Content area
Computer-aided applications in medicine and biomedicine drive the advancement of diagnostics, treatment, and research by leveraging data processing, analysis, and innovation. Computer technologies are used in medicine in imaging, diagnosis, storing and processing information, and staff management. The review will organize these progressions by significant features, which are medical imaging, diagnostic systems, data management, and their impact on daily clinical work. It explores diverse computer applications in medicine, including simulations, modeling, data visualization, and advanced data processing. Pattern classification techniques, decision support systems, and the integration of supercomputers—particularly in drug development—are discussed, alongside natural user interfaces and the application of Computer Methods and Programs in Biomedicine, with attention to human–computer interface integration. Modern computer-based imaging modalities like CT and MRI are also examined in detail, as are clinical applications of CAD for improving medical procedures. Lastly, the review projects future trends in cost-effective, high-quality telemedicine, remote consultations, integrated health records, computer-based learning, and disease management. Overall, this comprehensive discussion highlights the multifaceted impact of computing on the continued evolution of healthcare.
Introduction
Computers have become integral to modern society, particularly in fields such as engineering, education, and healthcare. Still, there are ethical questions that come with the quick introduction of computing technologies into healthcare. AI-enabled decision support systems can unwittingly mirror the biases found in previous medical records and result in unfair health care. Furthermore, there are worries about privacy when medical information is not protected or shared without the patient’s permission, and since AI models are not always clear, holding the system responsible for errors can be tough. In addition, when AI is used, it can be challenging to identify the root cause of issues, which concerns some when wrong decisions happen. Since these models are not always clear, healthcare professionals cannot always understand the reasoning behind major decisions. In medicine, computing technologies, including software systems, embedded hardware, and artificial intelligence (AI), enhance data storage, processing, and communication. AI, in particular, has allowed for the automation of diagnostic tools and decision support systems, helping clinicians make more accurate, timely decisions. The Internet further amplifies collaboration, allowing doctors from around the world to consult in real time, share research findings, and access the latest treatment protocols [1]. Computer technology has lifted patient care to new heights, yet its rapid spread forces us to think hard about the ethical and social fallout. Because most files are now stored and sent online, keeping sensitive health information private and secure sits at the top of the worry list. While helpful, electronic health records and AI diagnostic tools open doors to possible data breaches or snooping. Strong encryption, regular audits, and strict rules like HIPAA are still the first line of defense [2]. The trouble is that tech keeps sprinting ahead of the rules, making it tough for hospitals and clinics to stay in step [3]. However, while advancements in medical computing have greatly benefited healthcare, challenges remain, such as ensuring data privacy, improving system interoperability, and overcoming the high costs associated with new technologies [4, 5]. Medical professionals discuss medical matters via medical forums and share their knowledge by blogging, writing papers, and contributing to online medical publications. The Internet and easy access to computers have enabled the rapid distribution of updates and information on new treatment approaches in the medical industry, making it easily accessible to the public. There are several studies on computer-assisted methods in healthcare, but most concentrate on just a few areas or use technology that is not modern enough. Such prior works usually do not address integration issues, ethics, or provide a wide look at the system level. Instead, this review tries to close these gaps by bringing together a framework that covers important domains such as imaging, diagnosis, decision systems, and human–computer interaction, and it also discusses issues with using AI in the real world, its reliability, and future potential. Because of this, our review is useful for researchers and those working in the field of biomedicine.
Doctors can quickly exchange images and information, enabling prompt decision-making. They can quickly seek guidance and share expertise online, facilitating convenient collaboration [6]. Computer-aided surgery, also referred to as computer-assisted surgery, or CAS, is a subdivision of medicine that is evolving at a very high rate. Even though computer-assisted surgery has many advantages, it has faced several problems during its application in regular surgery. Such issues as software mistakes, faulty data, and insufficient experience from surgeons have caused problems during procedures. An example is that sometimes, errors in a robotic system can take more time and cause more problems during a surgery. Besides, the high price of robotic surgery systems has prevented many areas from using them, especially where resources are limited. It employs, incorporates, or makes use of sophisticated computer systems. Simulators and robotic devices are applied to boost the accuracy with which surgical procedures are performed. In CAS, a virtual model of the patient is created, allowing physicians to devise plans for strategies and rehearse the procedures before executing them. The impact of these systems is noticeable in areas such as neurosurgery and orthopedics. As with any advancement, robotic systems have made surgery more precise and less invasive; however, they are accompanied by expensive costs, extensive training requirements, and skepticism surrounding the dependability of AI technologies in surgery. In the future, the application of AI technologies in computer-assisted surgery (CAS) will likely center around increasing system flexibility and support for decision-making during surgery. In addition, some minor malfunctions during robotic surgery have resulted in the operation continuing for longer and increased risks to patients. Their occurrence proves that more reliable management and control devices are necessary in surgical computing systems [7]. Computer intelligence is used in both situations, highlighting the applications of computers in medicine. MRI and CT imaging technologies have transformed diagnostic medicine by enabling non-invasive examination of the body’s interior. For example, soft tissues are augmented with high-resolution scans through MRI, and advanced computational techniques of CT deliver intricate cross-sectional views of organs and bones. Alongside their achievements, however, these technologies encounter challenges, such as their elevated costs and the specialized equipment required [8, 9]. Recent leaps in image segmentation, mostly driven by deep learning techniques, have noticeably sharpened the accuracy with which doctors spot unusual regions. AI-powered models that carve medical scans into meaningful sections now smooth the review process, appearing in successful reports on both lung cancer screening and brain tumor mapping [10, 11]. Computers are integral to creating three-dimensional medical images, with specialized imaging devices such as CT and MRI scanners relying on high-performance computing hardware and software systems. Many modern medical devices, including patient monitoring systems and diagnostic tools, operate using embedded computing circuits and AI algorithms to provide real-time analysis and decision support [12].
Many modern medical devices function according to pre-established instructions. The majority of medical equipment is essentially comprised of computer-based circuitry and logic. The operation of hospital bed beeping systems, X-ray devices, emergency alarm systems, and other medical appliances relies on computer logic. Initially, computers found their way into essential medicine. Over time, as both computer technology and medical research progressed, the bond between computers and medicine strengthened. Recently, the use of computer technology in medical information has become a significant focus of scientific research, capturing widespread attention from society. To appreciate the depth of modern innovations in healthcare, it is essential to understand how computing technology evolved and laid the groundwork for today’s systems.
Research Questions
We aimed to map out the recent advancements in computers used in medicine and biomedicine with this scoping review. Considering how interdisciplinary and quickly the field is advancing, the following research questions (RQs) have been developed to direct the systematic review:
RQ1
What are the major areas in medicine and biomedicine where computer-aided technologies are implemented, and what mechanisms do they use?
RQ2
What are the constraints, moral problems, and difficulties in using computer-aided systems in healthcare today?
RQ3
How have recent inventions in artificial intelligence, human–computer interaction, and high-performance computing affected accuracy in diagnostics, the pace of medical work, and patient success?
RQ4
Which areas of computer-aided methods need more research, and what possible benefits can they bring to clinical and biomedical research?
With these research questions as a guide, we categorize studies, study ongoing trends, and gather insights from many areas.
Methodology
This review captures the most recent changes in healthcare using medical imaging, EHR, and AI. A strict procedure is used that is easy to understand, includes clear guidelines for selecting research, and designates the databases to utilize. The study approach guarantees scientific rigor and useful clinical thoughts.
Scope and Delimitation
In the review, attention is given to new medical imaging, EHR, and AI tools and how they influence patient care. The literature covered for this research was published after 2010, reflecting recent advancements. Most studies cover North America and Europe since they are highly active in research and use the latest technology. Even so, academic papers from Asia were added to the review when they offered practical information about how AI is used in rural or neglected regions of the world. Because of this, the conclusions remain meaningful for today’s medical treatment.
Inclusion Criteria
Only studies that met the requirements and methodology were allowed to be included in the review. Only papers that had been reviewed by other experts were used to ensure quality. The research had to concentrate on medical imaging, electronic health records, or AI in healthcare. The objective for medical imaging was to help with accurate diagnoses and better clinical choices, and EHR research aimed to improve how care is given and protect patients. Those in AI research should examine both diagnostic algorithms and decision support systems. The studies need to be clinically relevant outside of research, and, because of access problems, only English language studies were considered.
Exclusion Criteria
Only articles published from 2020 to 2025 that qualify as relevant and properly scientific were included in this review. Researchers use only scientific journals with a good reputation that publish work after peer review. These studies should be about medical imaging technologies (MRI, PET, CT, X-ray), EHR, or AI in healthcare and how they are used in clinical work. In the field of medical imaging, studies ought to focus on the technologies used for diagnostic purposes in making clinical decisions. EHR studies need to assess the role of EHR systems in designing better ways of teamwork and safety in healthcare. AI studies must consider diagnostic algorithms and systems that aid in making decisions. Outcomes from empirical research and randomized trials are the main focus of the study review. For accessibility, the review uses research only in the English language.
Database Usage and Search Strategy
To guarantee that all included studies were both comprehensive and relevant, the search was carried out using important academic databases. Among these, I used PubMed to search for medical imaging and EHR-related studies, IEEE Xplore for AI, machine learning, and systems that aid decision-making, and Google Scholar and ScienceDirect to cover many different research topics in healthcare technology. Some of the words used in the search included “Medical Imaging,” “MRI,” “PET,” and “AI in healthcare,” among others. A thorough search was accomplished by combining the keywords with AND and OR. This strategy gave attention to both simple overviews and detailed descriptions of how these technologies are used in practice.
Study Selection Process
The articles found were then reviewed by title and abstract to check their relevance. We selected and analyzed full-text papers by assessing their quality using a given checklist. Some of the important aspects assessed were study design, number of participants, clinical importance, and how well the paper was written. To ensure the review findings are accurate, only studies with results that matter clinically and are well-published in reputed journals were selected.
Data Extraction and Synthesis
The following details were noted from every eligible study: its goals, approach used, focus on imaging or algorithms, and the results on effectiveness, safety, and accuracy. Notes were taken about the study area and place of care to understand where technology is being used. Putting data into groups based on areas like medical imaging, EHR, and AI made it easier to merge conclusions. This approach allowed identifying patterns in the literature regarding AI’s ability to diagnose accurately, how EHR leads to more effective care, and what medical imaging brings for early diagnosis.
Quality Assessment
Every study was examined using the same checklist, which looked at its design, methods, number of participants, the importance of its findings, and whether it could be used clinically. By using these approaches, the studies could be depended on by real healthcare professionals. Research studies with top scientific value were selected for analysis.
Now, using this method, reviewers explain in detail how the review was conducted, ensuring their results are verified by research and are useful in modern clinical care.
Brief History of Computing
The development of a computerized cytoanalyzer in 1954 advanced medical technology by allowing slide evaluation of bulk cells for cancer signs [13]. Decision-making and reliability in the healthcare sector have been greatly improved with the aid of AI in recent years. A case in point is the AI boosted analytics rolled out in 2023, which augments decision-making and delivery in healthcare [6]. In the early 1960s, hospitals employed computers for accounting and administration [14]. Integration of real-time data acquisition with machine learning algorithms is two of the latest innovations made in the ECG analysis, which resulted in accurate diagnosis as well as personalized treatment planning for heart monitoring [15]. In 1964, the IBM System/360 and DEC PDP-8 minicomputers were introduced, altering medical informatics. There has been a significant advancement in the medical image analysis field owing to AI and computer vision. Olveres et al. mention that the application of AI still further increases the accuracy of medical imaging diagnosis [9]. In 1966, Massachusetts General Hospital introduced MUMPS as a healthcare programming language. France established IMIA in 1968 [16]. AI technologies are now commonplace in the processing of clinical data, enhancing the efficiency of clinical decision-making. Zewail and Saber address the impact of AI on the processing of clinical data and the delivery of healthcare services [17]. With machine learning, artificial intelligence, and medical imaging science have developed, modernized, and transformed for the better. Newer works, such as Barragán-Montero et al., discuss AI’s application in medical imaging and its improvements in efficiency and accuracy [8]. Early in the 1980s, medical informatics was defined in 1977, and ACMI was created in 1984. The 1980s networking boom led to the 1987 Health Level Seven, Inc. (HL7) clinical data exchange standard. In 1988, IBM supported MUMPS [18]. Medical imaging today sees the use of MRI, CT, and PET scans to deliver more effective and progressive ways to diagnose conditions. This part of the guideline discusses how these imaging modes are used and reviews the new computer tools that provide clearer, truer images, accurate diagnosis, and effective treatment planning. Wald et al. published an extensive analysis on the application of AI in medical imaging, particularly in inexpensive and portable MRI technologies [19]. The modern digital tools, including those using AI for remote consultations, have contributed to the expansion of telemedicine. As Shah and Khang state, the Internet of medical things (IoMT) marks a new era in the digital evolution of healthcare services [5]. AI chips now sitting on most smartphones let apps spot health warnings the user might miss. Recent tests show these tools help track long-term issues in real-time and even power quick video checkups with doctors, proving just how handy mobile health tech can be [20]. The incorporation of artificial intelligence into imaging technology has greatly enhanced the diagnostic accuracy and quality of images produced by multidetector CT scanners. Barragán-Montero et al. analyze AI's applications in CT imaging and diagnostic improvements [8]. In summary, AI has transformed the handling of clinical data and automated medical imaging procedures, resulting in a higher accuracy rate of medical diagnoses and improved operational efficiency. With the progression of AI technologies, incorporation into clinical work will continue to improve the quality of care by augmenting judgment power, minimizing human errors, and offering tailored treatment. Addressing issues such as data quality, opacity of the algorithms, and the reliability of the AI algorithms by clinicians poses a significant challenge to the potential of AI in medicine (Fig. 1).
[See PDF for image]
Fig. 1
An approximate chronology of milestone computing breakthroughs in medicine from earlier machines like the IBM 650 to modern counterparts such as electronic health records (EHR). This helps in understanding the impacts these developments have on the modern healthcare system
Kendall et al. [21] investigated the use of computer technology in analyzing the identification of anomalies in mammograms. Encarnacao [22] extended the scope of computer-aided design by categorizing a computer's characteristics to identify mammogram abnormalities: calcification, speculation, roughness, and form. Several more characteristics have been established after this first categorization. The USA has authorized the ImageChecker CAD system, developed by R2 Technology, Inc. The Food and Drug Administration (FDA) approved radiographic mammography in June 1998 and digital mammography in 2001 [23, 24]. By 2001, over 130 units were used, mainly at educational institutions. However, in 2005 to 2020, it was documented that over 5000 CAD units were actively being used across various healthcare settings. CAD devices detect and locate lesions by applying a marker or symbol to the region of concern. This does not imply malignancy, but rather signifies a suspicious region that requires evaluation by a radiologist [25]. Grid CAD is an evolved version of CAD software that has emerged in recent years. Grid CAD enables users to compare images against a database, display images from the database, and execute several CAD algorithms on a group of images [26]. Such early innovation in medical computing paved the way for contemporary medical systems, which now offer real-time monitoring of patients, predictive analytics, and automated decision-making. With computing becoming more sophisticated, its impact extended beyond diagnostics and imaging into data management, clinical documentation, and patient records, marking a pivotal evolution in healthcare delivery (Fig. 2).
[See PDF for image]
Fig. 2
The process of selecting studies that were used as per the PRISMA guidelines
Medical Records and Clinical Trials
Machine learning and AI are found more and more in medical imaging, diagnostics, data handling, and clinical procedures, aiding doctors by providing immediate analysis, valuable predictions, and extra help making decisions. The traditional medical record was based on a paper system and constitutes the first step of documentation, with healthcare systems undergoing digitization leading to AI-powered clinical trials that aid in data gathering, data analysis, and overall enhancement of patient satisfaction. The attention to automation in medical records is necessary due to the crucial role of electronic records in facilitating clinical trials. These trials involve gathering and analyzing data from specific patient interactions to evaluate the effectiveness and safety of new treatments and to understand disease processes that are not well comprehended. Medical researchers have limitations due to inefficient techniques for gathering the necessary data for clinical trials. These techniques often include manually recording information onto datasheets, which are then transferred into computer databases for statistical analysis. The method is time-consuming, filled with possibilities for mistakes, and contributes to the already expensive nature of randomized prospective research methods. Utilizing electronic health records (EHRs) provides several benefits to those doing clinical research.
Traditional Paper-Based Medical Records
Traditional paper-based record systems include recording patient health care information using physical storage media such as films, paper, or discs and storing this information in physical storage facilities to be retrieved when required. The paper-based medical record is insufficient in addressing the requirements of modern medicine. During the nineteenth century, a personalized "lab notebook" emerged as a tool for physicians to document their observations and plans, enabling them to remember crucial information when encountering the same patient on successive occasions. Without any regulatory prerequisites and lacking assumptions about the record's role in fostering communication among various care providers, the pages of the record contained minimal data or test results [27]. To address evolving healthcare and medical needs, the record that effectively served doctors a century ago has encountered notable challenges. Practitioners may become frustrated when seeking additional information about a patient related to treatment. To improve clinical efficiency, practitioners want more dependable solutions that help access crucial information while treating patients. The electronic health record may increase patient access to patient-specific information, significantly improving healthcare quality and patient satisfaction.
Electronic Health Records
Electronic health record (EHR) systems consolidate patient information from various healthcare providers, streamlining interaction and management. Security, real-time interoperability of patient information privacy, blockchain, and cloud computing improves patient care, as well as aids in managing chronic illnesses while minimizing mistakes during care delivery [28]. The concept of an electronic health record was first introduced in the 1960s, and improvements continued in successive decades [29, 30]. EHRs maintain a patient's medical history, including diagnoses, treatments, and test results. Several healthcare organizations are creating integrated clinical workstations. These clinical entry points provide access to various medical services. Computational technologies assist clinical tasks like delivering test results and allowing doctors to write instructions. Telemedicine applications and transcribed reports are facilitated, with electronic clinical data tracking hospital patients and managing resources for administrative and financial objectives [31].
The Health Information Technology for Economic and Clinical Health (HITECH) Act was ratified in 2009 in USA to promote the swift adoption and utilization of electronic health record systems. This legislation empowered Medicare and Medicaid to encourage healthcare providers who can demonstrate meaningful use of certified EHR systems that meet specific objectives [32, 33–34]. Before the Act, multiple studies indicated minimal adoption rates of EHR and other health information systems Worldwide. Yoon et al. [35] found that only 1.5% of acute care hospitals had a comprehensive EHR. In comparison, an additional 7.6% had a basic EHR. Similarly, Abramson et al. [32] reported that only 11.9% of hospitals in the USA had implemented an EHR system before this period. Following the implementation of the Act, many studies conducted in different areas of the USA showed a rise in adoption rates. Emani et al. [36] reported a significant 336% increase in the proportion of doctors who reported having an electronic health record [37] a system that fulfills the requirements for a basic system between 2006 and 2015. However, many worldwide studies have documented a rise in the adoption and utilization of electronic health record systems, both with and without government interventions. This suggests that the increasing adoption of EHR systems may be attributed to their perceived advantages and the advancement of healthcare service delivery rather than solely dependent on federal initiatives or interventions [30, 32, 38, 39]. Even so, using electronic health records (EHR) on a wide scale has encountered various challenges. Such problems are due to expensive setup, challenges in letting systems from different hospitals share information, and some professionals who feel uncomfortable using the systems. Besides, a lack of infrastructure at some rural and small healthcare facilities prevents many people in need from benefiting from EHR systems.
Despite a noticeable increase in global adoption rates, the advancement has been slow, and many reasons have been identified as either constraining or facilitating the adoption of electronic health records in medical services. Non-profit hospitals and hospitals with a public insurance rate over 40% are more likely to implement an electronic health record. This correlation may be attributed to non-profit hospitals' reduced tax obligations, allowing them to allocate more financial resources toward health information systems [40]. The capacity of an electronic health record system to enhance the quality, productivity, and efficiency of the healthcare service system is well established. Electronic healthcare systems have shown a gradual and consistent growth in global adoption. However, the level of popularity does not align with the speed at which these systems are being adopted. Presently, an electronic health record system is regarded as a method to enhance healthcare quality, productivity, and security, as well as gather data and examine the prevalence of diseases. In summary, the adoption of electronic health records (EHR) has brought about a major change in the healthcare sector, improving the precision and accessibility of information and aiding in clinical decisions [5]. While there are obstacles in implementing EHR systems, they continue to boost the quality of healthcare offered globally and increase convenience as well as compliance with healthcare regulations. Though EHRs changed the landscape with how patient information is stored, the next step was to utilize real-time data to track patient health—this opened the door for computer-based patient monitoring. There is a good chance that EHR implementation will improve further with increased attention to system interoperability and UI/UX design.
Electronic Health Data Encryption
To ensure privacy and data protection, clinical information transmitted over the Internet must be encrypted. Additionally, reliable identity verification methods should be in place to control access for purposes like research or surveillance [41]. EHR systems make patient care smoother by placing complete, up-to-the-minute files at doctors' fingertips whenever they need them. That easy access helps providers choose the right treatments, cut down on mix-ups, and work side by side with nurses, specialists, and pharmacists without digging through paper stacks. Yet, the digital shelves still wobble under real problems. Hackers target health data, leaks can harm a patient's trust, and juggling so many logins, passwords, and vendor rules taxes staff time instead of saving it. Until stronger safeguards are built, the promise of flawless, secure records will lag behind the technology itself (Table 1).
Table 1. Medical records with merits and drawbacks
Records | Key aspects | Merits | Drawbacks | References |
|---|---|---|---|---|
Clinical trials and EHR | Utilization of digital records | Streamlined data capture | High initial overhead | [30] |
Paper-based issues | Legacy documentation systems | technical familiarity | Data latency | [30] |
EHR introduction | Integration of patient diagnoses, a unified repository | Real-time clinical decision support | Elevated cost of setup | [31] |
HITECH & adoption | EHR incentives | Adoption rates | compliance challenges | [34, 35–36] |
global uptake factors | EHR acceptance | Improve safety | Implementation barriers | [42] |
Encryption & data security | Data security | Security compliance | Security management | [43] |
HIPAA implications | Privacy regulations | Data protection | Compliance burden | [43] |
Computer-Based Patient Monitoring (CBPM)
The early computation models of patients represented only the data collection framework for the patients' information. Modern AI algorithms process immense streams of real-time bioelectric data, further enhancing patient monitoring and decision-making in critical care. Computers' capability to obtain and analyze bioelectric data has led to another application—specifically, tracking physiological events. The computer plays a crucial role in the experiment by constantly monitoring and processing the incoming data in real time [42]. Two often-used phrases in physiological monitoring are "online" and "real time." "Online" refers to the direct connection of the data source to the computer without any intermediate data recording. Real time refers to a computational setup where the values of parameters that change over time are created at their current pace. These attributes are important for proficiently monitoring configuration [43, 44]. It is important to highlight that online monitoring systems allow for examining physiological specimens, enabling the study of control systems responsible for maintaining equilibrium. Stark [45] has done pioneering work in this field. Recently, Laferrière-Langlois et al. [46] have used physiological monitoring methods in clinical settings to monitor patients during surgical anesthesia, while Weil [47] has used these techniques to monitor patients during circulatory shock. Boyd [48] has outlined four primary purposes of a physiological monitoring system. The following purposes are included: data gathering, data recording and output, data modification, and provision of warning devices if discrepancies exceed pre-defined limitations. Data acquisition depends on sensors that are predominantly free from artifacts, do not interfere with the physiological system under examination, and exhibit a prolonged lifespan. Biological engineers have played a crucial role in creating the necessary tools and equipment for monitoring various activities, such as developing more efficient sensors. Figure 3 showed computer-based patient monitoring.
[See PDF for image]
Fig. 3
The operation of computer-based patient monitoring (CBPM) systems in which bioelectric data is captured and analyzed continuously in real time. This enables timely life-saving interventions during critical care and other areas where they can greatly improve patient outcomes
In the future, there will be significant advancements in the physician's capacity to provide medical treatment to the sick via monitoring tools and procedures. CBPM proves particularly advantageous for many diseases, allowing patients to be monitored remotely and reducing the need for frequent hospital visits. In acute care settings, CBPM ensures prompt detection of deteriorating patient conditions, contributing to improved outcomes and patient safety as shown in Fig. 3. Integrating CBPM into electronic health records enhances clinical decision support systems, fostering data-informed care strategies. Challenges such as data security and interoperability need addressing. However, the potential of CBPM, augmented by artificial intelligence, is undeniable, ushering in a new era of personalized, proactive, and data-driven patient care. In addition to these tracking capabilities, computers are now essential to the diagnostic processes, improving precision and decision-making through computer-aided diagnosis systems. In addition to these monitoring functions, computers have now become an inseparable part of the diagnostic process, improving precision as well as computer-aided decision-making through diagnostic systems. Computer diagnosis through the evaluation of intricate images has led to the invention of CAD systems that assist clinicians in complex imaging data interpretation.
Computer-Aided Diagnosis
The contemporary systems for computer-aided diagnosis (CAD) signal a shift from the MYCIN-based frameworks for decision support in medicine to the automated recognition of complexities in medical imaging, with unrivaled precision. Computer-aided diagnosis (CAD) captures varying degrees of advanced software implementation into medical imaging equipment that automates finding and distinguishing diverse forms of abnormalities in medical images. CAD systems can utilize algorithms that employ AI to enable easier identification of complex problems by radiologists, like tumors or fractures, thereby improving accuracy [49]. CAD technologies have been confirmed to improve the precision and accuracy of diagnosis, detecting even the slightest abnormalities in certain imaging, like CT scans for cancerous growths. It has been discovered recently that AI-enhanced CAD systems, by utilizing extensive databases containing labeled medical images, can significantly minimize the number of incorrect positive results [50, 51]. Nonetheless, there remain difficulties about misleadingly positive results, especially with intricate imaging cases [52]. Employing AI and machine learning algorithms, CAD systems are incorporated to detect and screen for breast cancer via mammograms systematically. The system not only detects breast cancer, but it also supports advanced stage detection where the patient exhibits secondary disease symptoms. A successful application of deep learning algorithms, which is an aspect of machine learning, is in medical diagnostics, as it enables ease of pattern recognition in large datasets [53]. The algorithms help in the easier detection of diseases, and in the case of cancer diagnosis, the precision attained through the application of deep learning networks is astounding [54, 55]. Reported claims suggest that the algorithms can attain an equal or even greater accuracy level compared to human experts [56]. These technologies help radiologists by analyzing images and detecting abnormalities with great precision. Nevertheless, despite how much technology has improved efficiency in CAD systems, there are some challenges that it faces. For example, CAD systems may reduce human error, but, at the same time, increase the chances of inaccurate assumptions that require additional procedures that are not needed. Furthermore, these systems perform poorly if the training data provided is substandard, along with the algorithms used [57]. CAD is now a regular component of clinical operations. Further research in diagnostic radiology and medical imaging has focused on exploring this technique [58]. In clinical settings, it has been noticed that bias in diagnosis may arise because there is not enough diverse training information for a wide variety of patients. In addition, false positives may raise a patient’s anxiety and lead to higher healthcare spending, which stresses the importance of thorough testing in clinics. With CAD systems, further developments will most likely be aimed at reducing bias, increasing accuracy in algorithms, and enhancing AI integration in real-time clinical decision-making [59, 60]. These algorithms must utilize a feedback system based on both successful and not-so-successful predictive data streams. These intelligent units receive and analyze computerized multimedia data, employing data mining techniques, AI, and machine learning on extensive and intricate clinical information [61]. While the computer's significant impact on disease diagnostic efforts is not expected future, it is crucial to conduct additional research to better understand its potential significance and enhance its applications. Apart from their role in diagnostics, computers are used in other fields of medicine such as education, in simulations, for assisting with decisions, and in the modeling of surgeries. AI programs already prove they can lift the accuracy bar, especially in radiology and pathology. By scanning images pixel by pixel, these tools spot tiny clues that busy doctors might miss, return results in record time, and flag problems early. Yet the magic only works if the training data is clean, rich, and balanced. If that pool leans toward one age group or skin tone, the model may stumble with everyone else. On top of that, every new version of the software needs real-world checks and fresh, well-organized files to keep learning.
Computer Applications in Medicine
Electronic devices equipped with processing units have become vital to our daily lives. Computers, cellphones, and other devices that provide portable Internet connectivity are essential to current business, education, and sometimes even interpersonal connections. As a vital component of modern progress, the healthcare sector is subject to the same technological advancements as other industries. Any computer techniques showing technological and scientific potential are promptly developed and used in medicine. It is difficult to list all potential uses of computers in modern medicine since practically every facet of applied informatics is used in actual medical solutions. The importance of computer techniques in medicine is evolving at the same pace as computer science. Our work mainly focuses on various dimensions. Our primary areas include image recognition and processing, pattern classification, user interfaces, and visualization. This article will provide a concise overview of specific uses of computer methods (CM) in medicine. Telemedicine opens the door to healthcare for people living far from big hospitals—whether in small towns or remote mountain camps—who would otherwise waste half a day just getting to a specialist. By letting doctors meet patients on a smartphone screen, it eases traffic at clinics and emergency rooms. Yet the system still hinges on fast Internet, and that is often shaky or absent in many country neighborhoods. A video chat can tell you a lot, but it cannot feel a lump, check a heartbeat, or test reflexes, so serious questions still need an in-person visit. Finally, staring at a screen instead of sitting beside the doctor can make the human bond a little weaker (Fig. 4).
[See PDF for image]
Fig. 4
The image shown emphasizes the important computer applications in medicine such as simulations, data processing, visualization and pattern classification, and decision support systems, which assist in the practice of medicine and patient recovery. These technologies improve diagnosis, treatment, and the efficiency of healthcare services
Simulations and Modeling
Preoperative planning, enhanced by predicting complications and improving efficiency, is assisted by patient-specific 3D models to ensure better surgical outcomes. These models aid in surgical simulations and modeling [62, 63]. The first stage of medical practice involves acquiring knowledge and skills throughout the study. Simulation-based medical education has been more popular in the last 20 years [64, 65]. CM enables the simulation of surgical procedures, eliminating the potential risks to patients' well-being. Huang et al. [66] used a computer program for preoperative surgical planning and educational purposes. They combine a patient-specific fracture with a generic full-length 3D bone model to create a customized bone model. Combining these two elements creates a composite three-dimensional model that accurately depicts the broken bone. Furthermore, Luria [67] documents a case of intricate wrist deformity resulting from an accident, which was effectively remedied by a slow corrective process after computer modeling. The simulation facilitated a suitable operational strategy by precisely computing three-dimensional deformity using computer-generated bone models.
Data Visualization
The visualization of data in medical imaging assists healthcare professionals in interpreting intricate information by combining different imaging techniques to provide a holistic view, thereby enhancing the precision of the diagnoses and improving decision-making processes [68]. Recent research shows that adding smart image analysis software to scanners helps spot lung and breast tumors much earlier than before. Doctors are also using the same technology to watch how well treatment is working and to see how tumors react to different drugs [69]. In a study, Vellido [70] demonstrated a method for visualizing medical data that effectively integrates information from several sources into a combined and comprehensible appearance. The system enables users to investigate and control volumetric datasets, visualize dataset analysis in specific areas, integrate 2D and 3D imaging techniques, and provide outcomes of vector-based computer simulations.
Advanced Data Processing
Before being presented to a doctor, data must be prepared using specialized image-processing techniques. In [71] the authors suggest a novel approach to determine the size and function of the bone marrow by integrating the structural and functional information obtained from CT and PET scans. In [72] the authors show another use of image-processing techniques. They give a complete protocol for calculating the starting boundary of the liver, which is part of the operation for segmenting inner organs. Medical data may include irrelevant information for a specific diagnostic technique, such as image collection artifacts, which might manifest as high-frequency noise [73]. The existence of this noise might hinder or perhaps render the diagnostic process difficult or unachievable. The most effective approach is to use low-pass filtering to mitigate the impact of high-frequency noise on medical data. Filtering methods, such as Gaussian and median filters are essential in medical image-processing as they significantly improve image noise and retain vital elements. These filters are not only fundamental for improving image quality but are also seamlessly integrated into medical image-processing pipelines used in clinical settings. Soft tissue structures are preserved while stochastic noise is diminished through the use of Gaussian filtering in CT and MRI images. In the medical pipeline, this technique is considered to be part of the image normalization process, which takes place before other steps like segmentation or pattern recognition. Gaussian filters help in minimizing the impact of noise on the images, thus strengthening later diagnostic procedures, such as tumor detection or blood vessel segmentation [74, 75]. As clinical utilization advances, the concurrent development of computational techniques in biomedical research have paved new pathways in the modeling of anatomy, biomechanics, and rehabilitation technologies. Median filtering is extensively used in X-ray and ultrasound imaging due to its edge-preserving features, which facilitate the enhancement of bone and vascular structures in noisy images. In clinical pipelines, this filter is part of real-time processing systems where images are captured and processed during a surgical procedure or a diagnosis is done at the scene for immediate analysis [76, 77]. The application of these filters occurs at multiple points in the medical image analysis workflow and is integrated with other processes such as feature extraction, segmentation, and even diagnostic systems based on machine learning. As an example, Gaussian filtering is applied in the preprocessing phase before the training of deep learning models focused on anomaly detection, including tumors or cardiac abnormalities [78, 79]. Image processing determines a transformation that optimally aligns two images based on similarity criteria. While rigid and affine transformations only capture overall geometric variations between images, elastic systems can handle global and local variances [80]. Image processing is widely used in several areas of medical imaging, with a particular emphasis on creating deformable brain graphics. Deformable brain graphics are templates that may be customized to represent different individuals' anatomy accurately [81]. Once the data is acquired, the image data undergoes preprocessing, allowing for delineation of areas of interest [82]. The ROI comprises tissues segmented to identify tumors or significant anatomical structures. The next stage is identifying the tumor, which is often accomplished using pattern classification approaches.
Pattern Classification Techniques
Image processing aims to enhance specific characteristics to simplify their identification and interpretation by a doctor. Computer-aided systems, powered by ML and AI, have played important roles in finding diseases such as cancer. The following section explains how CAD, pattern classification, and decision support systems help to improve the accuracy of making diagnoses. The study [83] used a computerized tool to identify asthma patients who are potentially getting inadequate pharmacological treatment in community pharmacies based on their prescription data. The researchers aimed to delineate the distinctive attributes of tongue images in patients with various Chinese medicine syndromes associated with lung cancer [84]. Additionally, they were required to discover the fundamental patterns in the alterations of these tongue images using a digital analysis instrument capable of objectively describing the tongue characteristics in lung cancer patients with different syndromes. The research objective was to create and evaluate an open-source artificial intelligence software based on artificial neural networks. This software aims to assist nuclear medicine doctors in identifying coronary artery disease from myocardial perfusion SPECT [85] and aid in decision-making [86]. The researchers of the study [87] introduced an improved computer-assisted diagnostic method for thyroid disorders, which uses a fuzzy k-nearest neighbor (FKNN) classifier. In study [37], the researchers categorized medical diagnostic data related to breast cancer. Classifications have been constructed using k-nearest neighbor (KNN), neural fuzzy (NF), and quadratic classifier (QC) single model schemes, as well as their related ensemble models. Furthermore, a composite group model including all three techniques has been developed for further validation. In AI and pattern classification, sophisticated machine learning technologies integrated into software and medical devices are of great importance. Decision support systems (DSS) coupled with supercomputers are used in the enhancement of the diagnostic procedure; this includes the analysis of large data volumes through AI models and high-performance computing (HPC) hardware in drug development.
Decision Support Systems (DSS)
AI-integrated decision support systems (DSSs) assist practitioners in early disease detection by evaluating patient information and interpreting medical images, including CT and MRI scans. These systems sharpen the accuracy of diagnostics, minimize mistakes, and improve the treatment choices made, thereby enhancing the results for the patients [88]. However, some selections may be consulted using mainly created plans that provide possible solutions. Article [89] discusses how computer-driven methods might decrease the time a patient has to be on mechanical ventilation. Following the successful resolution of life-threatening medical conditions, it is advisable to transition patients away from artificial breathing promptly. Not doing so puts them at risk of experiencing higher levels of disease. The researcher anticipates that computerized weaning would surpass doctor weaning in units where limited resources constrain doctors and are unable from considering the possibility, early in a patient's treatment, that the patient may be capable of independent breathing. Another instance of a decision support system (DSS) is referenced [90]. The issue discussed in this article pertains to adolescent idiopathic scoliosis, a condition characterized by the intricate nature of its geometry, clinical assessment, and therapy. Computer technologies may be used to enhance the management of AIS. Support vector and fuzzy clustering classifiers can reorganize AIS spines that exhibit comparable curvature and curvature development. Applications using artificial neural networks and surface topography algorithms have shown high accuracy in calculating the actual and expected Cobb angle while minimizing exposure to radiation [91]. Rule-based algorithms have the potential to enhance the reliability of categorization. Fuzzy logic may compute the average of many rules derived from the literature and provide a measure of confidence in areas without explicit agreement, such as the degrees of fusion in AIS [92].
Application of Supercomputing Methods in Medicine
Certain computations conducted while exploring biological structural properties are too complex to be executed on conventional workstations. Protein structure prediction (PSP) is an unresolved issue with many practical implications in biology, medicine, and biochemistry [93]. The problem involves an ample search space, and analyzing each protein structure requires substantial computational time. To address this, it is crucial to utilize HP parallel computing platforms and establish efficient search procedures within the space of potential protein conformations [94].
Human–Computer Interfaces in Medicine
Over the last several years, many innovative computer systems using a non-image-based navigation system have been used in hospitals and clinics [95]. The article [96] uses image-processing techniques and open-source software libraries to create a gesture recognition and hand-tracking system. This system is based on the Kinect sensor. It allows surgeons to navigate through images intraoperatively without physically touching them. The navigation is done using a personal computer.
Drug Discovery
Contemporary computer-assisted drug creation methods use the latest advancements in computational quantum chemistry. One of the primary objectives of modern drug design is to comprehend the molecular mechanisms behind therapeutic action and investigate the complex chemical interactions that occur during drug transport [97].
Natural User Interfaces in Medicine
The development of natural user interfaces (NUIs) and gesture recognition technologies could improve the interactions between medical personnel and medical systems. However, their clinical applications are scarce. Particularly, NUIs enable less physical contact with the equipment, but issues such as sensor accuracy and system reaction time need to be overcome for integration into sensitive medical settings [98]. As an instance, the use of gesture recognition technologies in the operating theater can potentially reduce the touching of screens and therefore lessen the possibility of infections, but extensive clinical trials should be conducted to determine the actual value of these technologies on clinical outcomes [99]. The use of gesture recognition in hospitals is limited because sensors are blocked, there is too much noise, and lighting is inconsistent, which cause issues with its reliability and usability. Equipment in critical care must be super accurate and capable of handling mistakes well. Historically, medical diagnostic assistance systems needed reliability and the ability to produce advice to gain acceptance within the medical community quickly. Today, an equally crucial component is how such programs deliver facts to the user and the uniqueness of the interface. An extensive study on human–computer interaction has been undertaken over many years. Contemporary home and mobile computer systems now often have built-in cameras and inexpensive multimedia devices that may be connected via USB. Consequently, there is a significant need for applications that use such sensors. The articles [23] and [100] provide a novel method for recognizing human body positions and movement sequences. The foundation of our study relies on syntactic description using the Gesture Description Language (GDL). The method uses forward chaining to infer a schema based on rules established using a formal LALR language. The collection of regulations is called the GDL script. At the same time, the automated reasoning module equipped with a memory structure resembling a heap is known as a GDL interpreter [101] (Table 2).
Table 2. Computer applications in medicine
Application | Purpose | Techniques | Potential benefits | Challenges | Implementation | References |
|---|---|---|---|---|---|---|
Simulations and modeling | Surgical simulation | 3D bone modeling | Surgical enhancement | System complexity | Training collaboration | [57, 58, 59–60] |
Data visualization | Images visualization | Image fusion | Diagnostic confidence | Handling large data volumes | Adoption of specialized graphic algorithms | [61, 62–63] |
Advanced data processing | Preprocess medical images to extract clinically relevant information | CT–PET data fusion | More accurate diagnoses | Managing high-frequency noise | Pipeline integration in radiology | [64, 65, 66, 67, 68, 69–70] |
Pattern classification techniques | Automate classification of medical imaging | KNN, Fuzzy KNN, Neural Networks | Early diagnostics | Algorithmic complexity | Integration in clinical decision workflows | [39, 71, 72, 73, 74–75] |
Decision support systems (DSS) | Decision support | AI algorithms | clinical efficacy | Potential over-reliance on DSS | Integration with EHRs | [76, 77, 78–79] |
Supercomputing methods in medicine | Computation tasks | High-performance parallel computing | Feasibility of large-scale simulations | High infrastructure costs | HPC collaboration | [80, 81] |
Human–computer interfaces in medicine | Clinician interaction | Gesture navigation | Surgical efficacy | Sensor accuracy | Interface standardization | [82, 83] |
Drug discovery | Computational chemistry | Molecular modeling | Faster iteration in drug design | Simulations challenges | Software integration | [84] |
Natural user interfaces in medicine | Diagnostic interaction | Gesture Description Language (GDL) | User accessibility | Reliance on robust sensor input system latency | Prototype development in imaging laboratories | [26, 85, 86, 87–88] |
In addition, we have successfully developed a prototype of a virtual three-dimensional desktop capable of showcasing two and three-dimensional medical data. The desktop interface may be used traditionally using a keyboard or mouse, or gesture-based controls that do not need physical contact with any device. The user uses their right hand to manipulate the visualizations. The user can either translate the volumes by sending them to the rear of the desktop or bringing them to the front to display information. They may also rotate or adjust the location of a clipping plane. The final feature involves modifying the transfer function. The prototype includes three pre-defined functions: one for displaying bones and the vascular system, another for adding less dense tissues with high transparency to the visualization, and a third one for reconstructing the patient's skin [102, 103]. The user performs predetermined actions using their right hand to transition between translation, rotation, and clipping modes. GDL recognizes and identifies those gestures.
Computer Applications in Biomedicine
While the clinical uses of computers have greatly enhanced the care of patients, it is in biomedicine where the impact is even more pronounced in advanced computing technologies in research, device development, and diagnostic tests. Computer techniques and software have been extensively used in biomedical research and applications for several decades. However, the current limitations in hardware and software for data gathering, processing, analysis, imaging, and visualization pose significant obstacles that hinder the use of cutting-edge computer techniques and programs in biomedical procedures [104]. This drives researchers to create innovative, efficient computer techniques and programs to advance biomedical research and applications. Fortunately, significant advancements in computer technology over the last decade have led to the development of a wide range of sophisticated mathematical and computational approaches. These methods have been suggested to enhance current biomedical research and explore several possible new applications. This paper will provide current computational methods for biomedical research and applications.
In their work, Gao et al. [105] used micro-CT to examine ankle sprains in humans by comparing the morphology of mouse and human ankle joints using a mouse model. Li et al. [106] used ultrasound imaging to diagnose unilateral peripheral entrapment neuropathy, taking advantage of its lack of radiation danger in comparison to CT scans. They specifically focused on the multilayer side-to-side image contrast for this purpose. Jiang et al. [107] introduced a rapid 3D ultrasound projection imaging technique for evaluating scoliosis and a method that uses ultrasound to extract coronal images for automatically measuring the curvature of the spine in teenage idiopathic scoliosis [108]. Ouyang et al. [109] conducted a thorough study of ultrasound detection techniques for identifying breast micro-calcification. In addition, Shen et al. [110] used a combination of several ultrasonic parameters to conduct a quantitative analysis of non-alcoholic fatty liver in rats. Zhang et al. [111] comprehensively evaluated mathematical modeling and computer-assisted needle trajectory planning for liver tumor thermal ablation. Zhu et al. [112] used surface electromyography to investigate and track changes in the paretic muscles during stroke recovery. Liu et al. [113] used gait analysis data to investigate the application of 3D printing in creating customized ankle–foot orthoses for individuals recovering from stroke in biomedical research. Yao et al. [114] further suggested a less intrusive method for treating calcaneal fractures using the sinus tarsi approach in conjunction with 3D printing.
In their study on body motions, Li et al. [115] used the concepts of time-varying synergy and synchronous synergy theories to examine the crawling action of human hands and knees. Li et al. [116] also, examine the cartilage and meniscus of the knee during a typical Tai Chi exercise, namely the twist step and brush knee. Xiao et al. [117] analyzed the difference between the dominant and non-dominant hands by including 3D kinematic data during a reaching exercise. In addition, Wu et al. [118] examined more complex movements, such as the backward somersault landing in artistic gymnastics, using biomechanical and neuromuscular methodologies. Zheng et al. [119] developed a model that utilizes geometric solutions to correct gaze in eye-tracking for eye movement analysis. In their study on brain activities, Ma et al. [120] used visual graphic stimuli to identify the location of brain activity in an event-related potential-based speller. Jiang et al. [121] constructed a transfer learning model based on deep learning to estimate the age of the brain using magnetic resonance T1-weighted images.
Advanced Computer-Based Techniques in Medical Imaging
Imaging methods stand as some of the most powerful and striking computer-based instruments in biomedicine. These sophisticated imaging techniques grant unprecedented access to anatomical and functional information, which is essential for accurate diagnosis and treatment design. The development of AI systems stands on the shoulders of the earlier CT and MRI imaging accomplishments, as they further facilitate advanced resolution, automated interpretation of images, and disease identification, reinforcing their pivotal role in modern medicine. Medical imaging refers to the visual depiction of the structure and function of various tissues and organs in the human body. It is used in clinical settings and medical research to examine the precise physiology and anatomy of the body, both in normal and diseased conditions. AI is steadily finding its way into eye care, and early results are encouraging for conditions like diabetic retinopathy and glaucoma. Recent studies show that deep learning systems can spot these problems with surprising accuracy, often matching or even exceeding results from trained specialists [122]. Medical imaging modalities are used to see internal anatomical structures under the dermis and skeletal framework, facilitating the identification of irregularities and the provision of therapeutic interventions for various ailments [123]. Medical imaging research continues. Some biological imaging technologies include thermography, electrical source imaging (ESI), medical photography, digital mammography, magnetic resonance imaging [72], tactile imaging, medical optical imaging, SPECT, and ultrasonic impedance source (EIT) [124]. These techniques are crucial for depicting biological structures and processes. Imaging technologies are crucial for the diagnosis of abnormalities. Medical workers may visualize medical data to see patients' conditions [70, 125]. MEG, EEG, and ECG measure data. However, they do not generate images. Instead, they represent the data as graphs or maps showing time information. These representations may have lower accuracy in conveying detailed information. Thus, these technologies might be seen as constituting a sort of medical imaging on a restricted scale. Around 5 billion studies on medical imaging methods have been presented [126]. Several methodologies have been devised and may be elucidated by their underlying principles, use in medical laboratories, and advancements in imaging methods. Advanced medical imaging methods include CT, PET, MRI, SPECT, sonography, and digital mammography. Below are the benefits and uses of these technologies in diagnosing, managing, and treating many conditions, including cardiovascular disease, cancer, neurological disorders, and trauma. CT, MRI, and PET scanners give doctors crystal-clear pictures of the inside of the body, and those sharp images are key for spotting many illnesses. When scans show trouble early, care teams can jump in faster, which often makes a big difference in how well patients do. Still, the tools come with downsides that can't be ignored. A single CT scan sends a dose of ionizing radiation through the body, and over time, that extra exposure can nudge the odds of developing cancer. MRIs skip the radiation, but they cost a small fortune and need bulky magnets and cooling gear that hospitals in low-budget regions simply do not own. Because of the price tag, both programs stay out of reach for many clinics that want them (Fig. 5).
[See PDF for image]
Fig. 5
Advanced computer-based techniques in medical imaging. It depicts different imaging techniques such as computed tomography (CT) and its subdivisions like volumetric quantitative CT (vQCT), hrCT, magnetic resonance imaging (MRI), medical ultrasound, positron emission tomography (PET), single-photon emission computed tomography (SPECT), and even micro-CT (μCT). The techniques provide precise, accurate, and high-detail images which aid in medical diagnostic evaluation and treatment processes
Computed Tomography (CT)
Hounsfield developed the prototype of the CT scanner in the 1960s [127]. Computed tomography, sometimes called X-ray CT, is a diagnostic imaging technique used by archeologists, biologists, radiologists, and several other specialists to provide detailed cross-sectional pictures of scanned materials. Medical technicians use CT scanners, instruments that generate images to identify problems and implement therapeutic interventions. In this technique, computers generate tomographic images by processing X-rays emitted from various angles. The computer-based technique has seen significant advancements, generating high-resolution reconstructed pictures [128]. The pharmaceutical industry analyses and improves medicine production for quality [129].
Using computed tomography, bladder, kidney, bone, neck, and head cancer are detected. It also detects infections [130, 131]. CT scans may show distant lung, bone, liver, and brain metastases. Computed tomography (CT) significantly impacts the lung and brain [132, 133]. CT scans monitor tumor size changes after treatment better [134]. Bronchus cancer patients may have swollen lymph nodes and the belly. CT scans before surgery [135]. CT scans detect cardiovascular, congenital, and myocardial bypass grafts [136, 137]. Gastroenterologists mostly use CT to examine the liver or pancreas of patients. CT scans can detect tumors with a diameter of 1.5 to 2.0 cm. In addition, this method [138] may also be utilized to monitor biliary blockage caused by lesions. Technology plays a crucial role in studying potential anomalies inside the abdomen with a high level of accuracy (95%) [139].
One significant limitation of computed tomography is its inability to detect large masses in the gastrointestinal system during abdominal examinations. Furthermore, it does not detect some mucosal abnormalities. Compared to other methods, the CT scan is a valuable tool for accurately managing abdominal disorders, such as esophageal, stomach, and rectal carcinoma [85, 140, 141]. The computed tomography technique can visualize the middle column of the spine in dislocation-type fractures in thoracolumbar fractures. CT scans can detect lesions and offer non-surgical treatment for some conditions, such as unstable burst injuries [141].
Volumetric quantitative CT (vQCT) evaluated the forearm and lumbar mid-trabecular vertebrae bone mineral density using a transverse CT slice. BMD is typical in advanced spiral QCT [142]. The hip cortical and spine trabecular bone may be detected using this technique to predict fracture risk. CT is unrivaled in anatomical imaging, whereas PET scans visualize an area’s metabolism, enabling earlier detection of activity such as disease processes. Improving the brain's 3D structure requires 0.5 mm isotropic spatial resolution. For accurate images, QCT's 1.5–2 mm resolution is insufficient. QCT contains this flaw. Typically, determining the femur's precise cortical thickness is simpler than measuring the spine's thickness, particularly among older individuals. Studies have shown that women have accelerated growth, with smaller vertebrae and less bone mass. Additionally, women exhibit a slower expansion rate in cross-sectional areas than males [143]. Quantitative computed tomography (QCT) is a medical imaging technology that can accurately measure bone density [144].
High-resolution CT is an advanced imaging technique that uses a current CT scanner. It often involves a high radiation dosage, but yields detailed images of bone structures, such as the forearm bone. These images analyze the cortical and trabecular network and texture [145]. Multiple cross-sectional investigations have shown that CT scans provide superior imaging findings in differentiating fractured vertebral trabecular structures from nonfractured structures, compared to DXA assessments of bone mineral density (BMD) [146].
Microscopy is the term used to refer to the technique of using micro-CT with a spatial resolution ranging from 1 to 100 μm. Micro-CT can replace conventional methods for in vivo measurements in mice and rats. Initially, the micro-CT technology used synchrotron radiations to get applications of ultra-high resolution [147]. Currently, the prevalent approach for micro-CT using X-ray tubes is mainly used in research facilities affiliated with universities and specialized healthcare centers. In order to create three-dimensional bone structures, a micro-CT scanner is equipped with specialized software, such as FEM. Finite element modeling (FEM) is primarily used in software engineering. The objective is to facilitate the examination of fractured bone structures by providing information on the 3D structures of bones, explicitly comparing fractured and nonfractured bone structures. Volumetric QCT is used to create structural models, and computer programs assign elastic characteristics to the components based on the bone density at the corresponding locations [148].
Arlot et al. [149] conducted a study to analyze the three-dimensional microstructure of bone in postmenopausal women suffering from osteoporotic disease. These women had a 36-month treatment with strontium ranelate therapy, which was administered appropriately. In the last 20 years, significant advancements have been made in imaging technology for investigating osteoporosis bone disease. Although advancements have been made in these technologies, bone imaging still faces several hurdles, including spatial resolution, sample size, complexity, time, radiation exposure, and expense. Furthermore, superior bone imaging requires stringent criteria such as high precision, replicability, accessibility, and appropriate monitoring protocols [150]. In recent years, using CT scans has become widespread in emergency rooms, particularly in the USA. Computed tomography is increasingly used to diagnose acute and chronic disorders, including life-threatening conditions such as stroke, brain damage, severe trauma, heart disease, stomach discomfort, pulmonary embolism, severe chest pain, and kidney abnormalities [151].
3D ultrasound computed tomography (3D USCT) has excellent potential as a method for imaging breast cancer. The USCT system has many key features, including the ability to capture attenuation, reflection, data gathering speed, speed of sound volume, and good image quality in a simultaneous and repeatable manner. The 3D USCT system is a highly advanced equipment used for clinical applications. The whole volume of the breast may be captured in under 4 min [152]. Although CT and MRI scanners are advanced, people living in low-resource areas still cannot use them because of their cost, upkeep, and the limited number of well-trained medical workers. As a result of little use, people in rural and underserved areas are less likely to get correct diagnoses.
Positron Emission Tomography (PET)
Positron emission tomography (PET) is a nuclear medicine technology that provides clear images of the overall concentration of radioactively labeled elements in the body, allowing for functional analysis. A positron emission tomography (PET) scan detects metabolic processes that assist in the early detection of inflammation and tumors. Besides aiding in early detection, PET scans are used in specialized areas of medicine such as oncology to stage cancers, monitor the effectiveness of treatment, and observe the treatment responses because of cancer tissues' abnormal activity before any changes to the tissues’ structure are detectable using other imaging techniques. This technology can identify and analyze biological processes inside live organisms, making it particularly suitable for therapeutic applications [153]. A computer system creates three-dimensional images of radionuclides that produce positrons within the body. PET CT scanners use a CT X-ray scan performed on the patient's body inside the same machine and session to provide 3D images. PET and NMR are quantitative radiological procedures that provide information on the physiology and biochemistry of an organism, allowing for the assessment of normal or pathological conditions. NMR is only sensitive to the distribution of chemicals containing hydrogen. It does not provide high-resolution images of other substances. It can quantify the overall concentration of creatine phosphate [154] and ATP in specific brain regions. Both NMR and PET fulfilled their respective diagnostic roles [155]. The first PET system was created at Massachusetts General Hospital in 1953. Subsequently, several devices were introduced, including the PET scanner, tomographic positron camera, and other PET equipment [156]. In conclusion, PET, MRI, and CT scans are modern medical imaging techniques that are essential to clinical diagnosis. Every modality has particular advantages when assessing various physiological processes and anatomical components, particularly in the fields of oncology, neurology, and musculoskeletal imaging [157]. Furthermore, the application of AI and machine learning improves the accuracy and effectiveness of these imaging modalities, allowing for heightened identification of irregularities and better overall results for patients.
Magnetic Resonance Imaging
Magnetic resonance imaging (MRI) utilizes powerful magnets and radio waves to image soft tissues with high precision. It is especially helpful in the assessment of the brain, spinal cord, and muscles. MRI scans do not require surgical procedures and do not use ionizing radiation, which is critical for imaging soft tissues, including in the diagnosis of diseases that affect the nervous and musculoskeletal systems, imaging, and resolving conditions [158]. MRI is a mostly non-invasive imaging method used to see the structure and function of the body in both healthy and diseased states. Physicists Paul Lauterbur and Peter Mansfield developed an MRI method called echo-planar imaging (EPI) in the late 1970s [159]. An MRI scanner uses magnetic fields, electric fields, and radio waves to generate images of organs and the anatomical composition of the body. The Tesla is the SI unit used to quantify magnetic flux density, often known as magnetic field intensity. The most frequent findings seen by MRI are multiple sclerosis, central nervous system tumors, infections of the brain and spine, stroke, ligament and tendon injuries, muscle atrophy, bone tumors, and blockage of blood arteries [160]. MRI employs non-ionizing radiation, often considered more favorable in comparison to CT. MRI offers exceptional differentiation of soft tissues, allowing for a clear distinction between the brain's white and gray matter structure. Recent strides in MRI machines- sharper images and on-board artificial intelligence pushed the method deeper into everyday care. Research shows the tool now spots problems in eyes and skin with remarkable precision [161, 162]. MRI utilizes several methods, including functional MRI, MRA, perfusion-weighted imaging (PWI), susceptibility-weighted imaging, gradient echo, and diffusion-weighted imaging [132], and spin echo. The technology produces high-quality images without requiring the subject to move [163, 164]. MRI has several advantages, including its painless nature, excellent spatial resolution, non-invasiveness, and use of non-ionizing radiation. MRI is mainly used autonomously for the examination of soft tissues [165].
One significant use of whole-body MRI is to examine metastases of the skeleton. The MRI technique enables tumor visualization due to the high concentration of protons in the tumor matrix. Skeletal scintigraphy (bone scan) is less sensitive than this imaging approach when measuring metastases of the skeleton. Whole-body MRI is a superior method for identifying spine, pelvis, and femur abnormalities. This approach is often a primary diagnostic tool for assessing whole-body fat, soft tissue disorders, and polymyositis disease [166]. A microfluidic lab-on-a-chip (LOC) is a device used as a cutting-edge technology in medical laboratories. The devices facilitate the reaction between the sample (cell suspensions) and reagents. MRI is considered an optimal instrument for assessing the response to the locus of control. MRI captures the signals from the exhausted fluid leavings in the instrument. MRI, in conjunction with MRS, observes the movement of fluids, the separation of chemical reactions, and the dispersion of substances in a lab-on-a-chip (LOC) device. However, both MRI and MRS exhibit limited sensitivity. There is optimism that MRI will be functional for developing microfluidic LOCs, which may have significant applications in medical diagnostic libraries [141].
Multinuclear 3D solid-state MRI visualizes dental bone and the calcium phosphate constituents of bone materials. Additionally, it provides details on the bone's chemistry and texture [167]. Advanced neuroimaging methods, such as high-resolution MRI, may be used to study the myeloarchitectural patterns in the human brain cortex. The staining approach has shown the bands of myelination. High-resolution MRI imaging technology may also provide a high-quality image of the same band. While modern technology has been extensively used in the visual system, more enhanced approaches are necessary to study other brain regions. Overcoming the high signal-to-noise ratio difficulty in MRI machines, which is caused by higher image resolution, is difficult [168, 169]. FMRI and other forms of MRI have significantly contributed to advancing our knowledge about disorders, their etiology, and strategies for their management.
Single-Photon Emission Computed Tomography (SPECT)
SPECT is an imaging method that uses gamma rays to generate accurate three-dimensional reconstructions of things. In 1963, Kuhl and Edwards published the first findings on SPECT [170]. The progressive integration of computer-attached systems with spinning gamma cameras has resulted in a new modality. SPECT has emerged as a very effective medical imaging modality in research and therapeutic settings. SPECT tracks the three-dimensional data of an item by generating a sequence of thin sections from tomographic images. These crucial tomographic images might enhance the ability to discover profound and minuscule patient fractures [171, 172].
In SPECT, the detector directly captures the emitted gamma rays from radioactive tracers. The computer system processes the data obtained from the detector. It accurately represents the region where the radioactive tracers are introduced. The SPECT imaging approach is more cost-effective than the sole use of imaging tiny animals. It can monitor bone metabolism, cardiac diseases, and blood flow in the cerebrum [173]. SPECT has also been developed for neurochemical brain imaging. It uses robust imaging technology to assess and study neuropsychiatric illnesses. This approach is crucial for monitoring the pathophysiology and other complicated brain diseases. The SPECT/CT approach is promising for quantifying soft tissues, articular cartilage, and synovial structures. It also delivers promising findings in assessing osteochondral anomalies [174].
Medical Ultrasound
Ultrasonic imaging technology was used for brain imaging. Ultrasonography is a widely used imaging tool employed in diagnostic laboratories and clinics. It is devoid of any potential harm from radiation, somewhat more cost-effective, and more portable than other imaging methods such as MRI and CT [141]. This method is used across several disciplines. Ultrasound, a medical technique, employs high-frequency sound waves to diagnose the structure and organs of the body. The ultrasonic system operates at a high frequency. It is used by specialized technicians or physicians to visually examine the blood arteries, heart, kidney, liver, and other organs inside the body. A transducer is the most crucial element of ultrasound. An ultrasonic transducer can transform sound waves into electrical signals and vice versa, converting sound waves into an electrical signal [175]. A sonogram, often known as an ultrasonic image, is created by emitting pulses of sound waves into tissues using a specialized probe. Various tissues exhibit varying degrees of sound reflection. The operator detects and presents an image of the reflected sound waves, often called echoes. Ultrasound scanning is an efficient and dependable method extensively used for monitoring multiple pregnancies, placenta previa, normal pregnancy, and other anomalies throughout rest and pregnancy [176].
Ultrasound technology may also be effectively used to identify kidney masses. The ultrasound imaging approach correctly identified 86% of carcinomas and 98% of kidney cysts among 111 patients. Ultrasound is a secure, uncomplicated, cost-effective diagnostic technique for identifying intricate kidney masses [177]. Ultrasound screening enables the visualization of cartilage to assess instability, aberrant positioning of the femoral head inside the acetabulum, and infants' dysplasia development [178]. A high-powered, low-frequency ultrasound device may also be a non-invasive medication delivery method. Low-frequency ultrasound modality facilitates the efficient delivery of medicines and proteins with high molecular weight into human skin due to its capacity to enhance permeability. Insulin, erythropoietin, and interferon-gamma molecules may be efficiently and securely transported through the human epidermis [179]. Researchers [180, 181] in molecular biology have used a unique molecular imaging technology called molecular ultrasound to track changes in the rate of expression of molecular markers found on targets inside blood vessels. The contrast agents used in sophisticated molecular ultrasound imaging technologies primarily consist of micro- or nanosized particles with surface ligands, often called microbubbles. In the future, molecular ultrasound imaging technology will have a significant therapeutic impact by providing sensitive and accurate imaging of complex disorders at the molecular level.
Transvaginal ultrasound (TVS), an ultrasound imaging method, has significantly transformed our comprehension of pregnancy diagnosis and treatment. The TVS provides a comprehensive insight into early pregnancy complications. This examines both the location and viability of the pregnancy. The televisions detected early evidence of pregnancy viability by monitoring fetal cardiac activity during prenatal [182]. Transvaginal sonography (TVS) is a very sensitive method used to detect early miscarriage. This technique aims to identify blood flow and trophoblastic tissue in the intervillous space. Using color Doppler images makes it possible to determine the degree of expectant management accurately [183].
Functional ultrasound is a very effective method for imaging the brain, capable of detecting temporary changes in blood volume more precisely than other brain imaging techniques. The blood volume in tiny arteries may be assessed by functional ultrasonography, which employs plane-wave illumination with a high frame rate. Functional ultrasonography can identify the brain's active section [184]. Ultrasound biomicroscope (UBM), commonly called high-resolution ultrasound, is used for clinical imaging of the human eye. UBM utilizes frequencies of 35 MHz or higher to provide high-resolution images beyond the capabilities of traditional ophthalmic ultrasound methods. Ultrasonography bio-microscopy (UBM) may be used to diagnose ocular injuries and complicated hypotony [185]. The device may detect eye lens displacement, lens subluxation, zonular defect, iridodialysis, cataracts, and hyphemia. Hyphemia is a medical disorder characterized by blood in the front part of the eye due to an injury [186].
CT Colonography
Colorectal cancer is among the top five most often diagnosed types of cancer based on anatomical location. Early identification and elimination of polyps, precursors to cancer, may decrease its occurrence [187]. CADe can enhance precision in identifying polyps, minimize inconsistencies among observers, and minimize the time required for image interpretation [188]. Trilisky et al. [189] stated that optic colonoscopy and CT colonography detection rates are similar. Computer-aided design tools provide the reader with potential options in this imaging modality. The reader may use this information in three ways: examining the CAD candidates, evaluating the CAD results after an initial visual assessment, and analyzing the images alongside the CAD results. A significant limitation of using CAD for CT is the detection rates of false positives. However, it is noted that the radiologist rejects the majority of them. The authors [82] conduct a comprehensive examination of several forms of false positives (FPs) and highlight that the primary goal of computer-aided detection in this context is to minimize the rate of detection mistakes rather than errors related to lesion characterization. Manjunath et al. [154] discussed using neural networks to segment the colon in CT colonography. The complexity of this work arises from the existence of additional air-filled structures, regions with high density such as contrast fluid or bones, folded structures, and obstacles. The authors elucidated their process after examining existing studies that presented techniques for colon segmentation. The procedure involves extracting air packets, lungs, and fluid sequentially and performing many processing steps and Boolean operations on diverse structures. System performance is often assessed by calculating sensitivity, specificity, and accuracy.
Radiography
Cao et al. [190] developed computer-aided detection systems for identifying lung nodules. The image collection consisted of multi-projection chest radiography, with each patient having three projections. The dataset included a total of 59 patients. The images included both genuine and artificial nodules of varying sizes. Two methods were employed. Lung nodules were first improved and segmented using image-processing techniques. Before classification, segmentation calculated the candidates' attributes using dynamic programming to eliminate false positives. The second method, "fusion CAD," recorded nodules and minimized false positives (FPs) by linking nodule candidates in distinct projections. Their fusion CAD approach minimized false positives and dramatically improved lung nodule identification. Shi et al. [191] addressed the false positive nodule rate in several commercial computer-aided detection software versions for detecting chest radiography nodules. Their study using MTANNs minimized false positives while maintaining sensitivity. They start by creating MTANNs to identify specific false positives. The outputs of each network are combined by a three-layer traditional artificial neural network with different activation functions. Nodules are distinguished from non-nodules using this integrated network (Table 3).
Table 3. Advanced computer-based techniques in medical imaging
Technique | Applications | Advantages | Limitations | Advancements | References |
|---|---|---|---|---|---|
Computed tomography (CT) | Detection of tumors, organ abnormalities | trauma imaging | Limited sensitivity for certain mucosal or GI abnormalities | Volumetric quantitative CT (vQCT) for bone density | [111, 112, 115, 116, 117, 118, 119, 120, 121, 122–123, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135–136] |
Positron emission tomography (PET) | Oncology: tumor metabolism and staging Functional imagining | Metabolic imaging | Cost complexity | Molecular imaging | [137, 138, 139–140] |
Magnetic resonance imaging (MRI) | Diagnostic imaging, | Soft tissue imaging | Imaging limitation | Advance MRI | [65, 116, 125, 141, 142, 143, 144, 145, 146, 147, 148–149] |
Single-photon emission CT (SPECT) | Functional assessment | Cardiac imaging | Resolution limitation | SPECT integration | [150, 151, 152, 153–154] |
Medical ultrasound | Obstetrics (fetal monitoring) | Portable imaging | Operator-dependent image quality Limited penetration in bone | Functional ultrasound for brain perfusion (“fUS”) | [125, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165–166] |
CT colonography | Colonoscopy screening | 3D colonography | Colonoscopy limitation | Computer-aided detection (CAD) integration | [70, 138, 167, 168–169] |
Radiography (X-ray) | Chest imaging | Fast, widely available, and low-cost X-ray imaging | 2D projection only (overlapping structures) | Multi-projection chest radiography with computer-aided detection | [170, 171] |
Clinical Applications of Computer-Aided Design
Gu et al. [192] used deep learning techniques and image modalities in this anatomical region. A customized convolutional neural network (CNN) was used to construct computer-aided diagnostic software to distinguish benign and malignant lung nodules in CT scans. Researchers [193, 194–195] assessed several standard CADx methodologies and techniques. The deep learning model Stacked Denoising Auto-Encoder differentiated and categorized best. Their computer-aided diagnosis (CADx) system has been used to assess ultrasound breast imaging errors, showcasing its wide variety of applications. Hua [196] identified CT scans of lung nodules using convolutional neural networks (CNNs) and deep belief networks. The testing images are obtained from the Lung Image Database Consortium (LIDC), including a curated collection of nodules larger than 3 mm. The DBN and CNN models exhibited much higher specificity and sensitivity than techniques relying on manually derived features.
Shen et al. [197] recently used Multi-scale CNNs to classify lung nodules. This technique utilizes various scales of nodular patches for training. It obtains a notable classification accuracy of 86.4% without nodule segmentation. This study used radiological studies obtained from the LIDC-IDRI database. Additional benefits include the system's ability to withstand noise and effectively capture nodular variations and distinctive feature data. The CAD method was built using a parallel combination of three CNNs and integrating with a random forest classifier. Chen et al. [198] assessed the diagnostic accuracy of an artificial neural network ensemble-based computer-aided detection method for categorizing nodules on thin-section lung CT images. Three categories of nodules (likely benign, indeterminate, and likely malignant) were developed. Integrating many artificial neural networks (ANNs) and classifiers enhances the capacity to generalize and improves the dependability of the results. The authors state their intention to investigate the impact of including image modality registration, additional patient-related parameters, and temporal analysis information on the overall performance of the ensemble.
Suzuki et al. [199] trained MTANNs to differentiate between nodules and vessels. Generating many training samples enhances generalization capacity and increases classification performance. He et al. [200] used feature computation techniques to classify the delicacy of lung nodules. Their next step was implementing an algorithm for segmenting lung nodules, where an optimal threshold was chosen based on estimated gradients. Following the selection and computation of a set of characteristics, an Artificial Neural Network was trained. A high level of concordance was seen between the segmentations performed by the computer and those performed by the radiologists. The case subtlety ratings computed by the CAD system closely matched the expert scores. This approach enables the assessment of database complexity and facilitates comparisons across CAD systems. Tokisa et al. [201] devised a technique to remove images, eliminate related artifacts, and enhance lesions. Candidates are identified using a series of image-processing algorithms, and a method is implemented to minimize false positives. The assessment yielded the following findings: a sensitivity of 87.1% with 2.07 false positives per scan. Radiologists might use a specific scan or a temporal fusion to enhance the diagnostic information. Suji et al. [202] comprehensively evaluated CAD methods for lung cancer in CT scans. The scientific studies analyzed often highlight the significant potential of increasing radiologists' performance in clinical practice via high sensitivity, the capacity to identify nodules, and low cost. The presentation covers various designs, preprocessing algorithms, segmentation and detection approaches, and ways for removing false positives. CAD has also been used to analyze interstitial lung diseases (ILDs). Anthimopoulos et al. [203] developed and trained a Convolutional Neural Network (CNN) using a patch dataset from CT images obtained from two distinct annotated databases. Seven classes were created for tissue categorization, consisting of six patterns often seen in interstitial lung diseases (ILDs) and healthy tissue. Their network achieves a commendable classification accuracy of 85.5%. In order to optimize the network design, many configurations and training choices were experimented.
Emerging Technologies in Medicine: AI, Deep Learning, and Big Data
AI is reshaping the field of medicine by making advancements in diagnosis, therapy, and pharmaceutical innovation. Deep Learning AI systems scrutinize huge volumes of medical information, including images and historical clinical data, to uncover patterns that facilitate clinical decision-making. AI systems can now categorize skin cancers with accuracy that matches the best dermatologists. In addition, AI systems promote personalized medicine by tailoring treatments to the genetic profile of the patient, which enhances outcomes while decreasing unintentional damage [54, 204].
Machine learning (ML), a subtype of artificial intelligence (AI), improves predictive analytics by evaluating historical patient data to forecast various health outcomes, such as readmission to the hospital or the onset of chronic diseases. The application of ML in clinical decision-making has been documented, with evidence-based decision support cited as one of the ML-derived recommendations that was implemented successfully [205]. Also, unstructured clinical notes are increasingly being processed with ML models for better management using natural language processing (NLP) [206].
Analyzing big data helps us recognize patterns that enhance disease prevention and improve diagnostic accuracy. This is achieved by processing large volumes of information, such as health records and data from wearable devices. Big data is also important for managing healthcare costs more efficiently and predicting epidemics [207]. In addition, the application of AI in gene editing using CRISPR technologies has increased accuracy and is propelling the discovery of biomarkers for Alzheimer’s and cancer, while also offering prospective therapies for some genetic disorders [208]. Overall, the major impact of AI, ML, and Big Data Analytics in healthcare is the transformation of the industry through improved diagnostics, personalized therapies, and accelerated research. On the other hand, it is important to consider the concerns of privacy of information and bias in algorithms so that these technologies can be used safely and fairly in the future.
Future Applications of Computers in Biomedicine
Developing technologies like artificial intelligence (AI), machine learning (ML), and Big Data Analytics are transforming the healthcare sector by providing more precise diagnoses, custom treatment, and improving patient care. Medical imaging is one of the fields where AI systems, especially those that implement deep learning, are advancing. They can now detect anomalies, including tumors, fractures, and other conditions, far more accurately than traditional methods [54]. Machine learning, most notably through predictive models, enables clinicians to design personalized treatment plans due to the large and diverse datasets that incorporate genetics, environment, and lifestyle. It also helps foresee treatment success and allows for proactive measures. Big Data Analytics has aggregated health data from electronic health records (EHRs), wearable devices, and even clinical trials. This allows healthcare systems to detect trends among certain population groups, thus improving proactive care and overall treatment efficacy [209].
Undoubtedly, these technologies have tremendous potential to create value; however, they do have challenges. Privacy and security of data remain an issue as sensitive health information needs to be protected with strong safeguards during sharing and processing on different platforms [210]. Furthermore, the inaccurate predictions and outcomes resulting from bias in data are a consequence of the lack of diversity and incompleteness in the data used for training AI and ML models [211]. Moreover, integration of these technologies into the current healthcare systems is expensive and difficult, especially for smaller healthcare organizations that do not have easy access to these tools [212]. As we have seen, the combination of artificial intelligence, machine learning, and Big Data Analytics can dramatically transform healthcare. However, an appropriate focus on ethics, integration of different systems, data privacy, and relevant technologies will be critical for their effective implementation aimed at achieving equity, improving the quality of patient care, and enhancing the efficiency of healthcare systems. AI’s disruptive potential in disease diagnosis and patient care has already been researched, indicating that AI can enhance clinical decision-making [54]. In addition, AI being integrated into telemedicine systems is bound to facilitate real-time monitoring of patient data, creating scalable and economical healthcare systems, especially in marginalized areas [213]. Imaging is one of the most revolutionary fields in biomedicine, as it now leverages sophisticated computer technologies for in-depth analysis on human systems and their functions. Such developments are aimed at improving healthcare systems to be more responsive, efficient, and personalized (Fig. 6).
[See PDF for image]
Fig. 6
Illustration features telemedicine and biomedicine with computer technologies, which include remote consultations, integrated health records, computer-aided learning, patient and caregiver education, and disease management. Facilitating technologies focus on increasing access to healthcare and improving the overall efficiency of the healthcare systems
High-Quality, Low-Cost Telemedicine
The telemedicine tests administered in the mid-1990s utilized expensive dedicated connection lines as well as specialized equipment. Unlike in the past, the Internet is now integrated into medical practice, allowing ease of access for physicians and patients situated at considerable distances [214]. In the future, the Internet will be able to facilitate real-time high-quality audio and video streaming, including auscultation of the heart and lungs. Standard computer workstations will also be utilized on both ends of the connection, further boosting the cost-effectiveness and operational efficiency of telemedicine.
Additionally, patients in rural regions no longer have to undertake unwarranted journeys to major medical centers to telemedicine. Direct expert consultations will be available for primary care clinicians remotely, enabling bespoke interaction from the comfort of their offices. The ability to complete several processes during a single appointment will reduce the frequency of tedious visits, and patients will no longer have to navigate the inconveniences associated with them. These promising results of ongoing projects would suggest that further implementation of these programs will be economically beneficial and enhance overall patient care.
Telemedicine now includes remote consultations, real-time monitoring, and chronic disease management, powered by wearable devices and IoMT technologies. These tools enable healthcare providers to monitor patient vitals remotely, improving access and early intervention, especially in rural areas. Telemedicine software advancements are making healthcare more efficient, cost-effective, and patient-centric [215]. It includes remote monitoring of vital signs, chronic disease management, and digital health interventions. Through telemedicine, healthcare providers can track patients' health metrics in real time, adjust treatment plans, and even manage long-term care for chronic conditions. The integration of wearable devices and sensor technologies allows telemedicine to provide continuous care and ensure early detection of potential health issues, reducing the need for frequent hospital visits [216, 217]
Remote consultations via video calls and secure messaging allow doctors to provide care to patients in remote areas. These consultations incorporate tools like digital stethoscopes and ECG monitors for accurate diagnostics. By enabling real-time monitoring, remote consultations improve patient outcomes and expand healthcare access [218]. These interactions are particularly beneficial to patients from rural settings or those who are too ill to visit clinics. Healthcare providers can discuss and evaluate symptoms, analyze medical records, or provide recommendations or prescriptions. The introduction of remote consultation has helped to enhance the availability of healthcare, especially in areas that lack proper facilities, and was instrumental during the COVID-19 pandemic in making it possible to continue providing services without needing physical attendance at the healthcare facility [125, 219].
Integrating rapid and efficient electronic communication between medical practitioners enhances the availability of specialized služl and increases the level of satisfaction of patients [220]. For instance, teleconsulting available to specialists in a regional medical center can provide professional insight immediately, such as for a patient with skin lesions that are deemed unusual. Image data and relevant information transmission greatly increase the possibility of high-quality, timely diagnostic care, all while significantly minimizing patient wait times. AI is making real inroads in dermatology, especially in skin-cancer work; recent deep learning systems now spot disorders like melanoma with an accuracy that matches many experienced doctors [221]. The subsequent focus is on the need to further these peripheral consultancy methods to be incorporated into routine care and enable them to be used as defined clinical services free of charge.
Integrated Health Records
Our objective is for a future where individuals no longer have to deal with the inconvenience of having their medical records spread across several doctors' offices and many hospital record rooms. Instead, their data will be digitally linked over the Internet, resulting in each individual having a singular "virtual health record"—a comprehensive and unified summary of all the medical treatment they have had throughout their lifetime. Moreover, this record shall be protected, handled with reverence and secrecy, and disclosed to healthcare professionals only with the patient's consent or in cases of medical urgency based on clearly established and enforced standards [222]. Significant measures have been implemented lately to enhance the probability of this situation occurring. Establishing a national health information infrastructure and using standardized terminology like SNOMED-CT and privacy regulations such as HIPAA are helping to achieve this objective.
Computer-Based Learning
In the coming years, medical students undergoing their orthopedics rotation can visit their institution's electronic learning center [223]. Here, they can utilize the Internet to engage with a three-dimensional "virtual reality" model of the knee, accessible on computers provided by the National Institutes of Health. This innovative approach aims to prepare them for their initial observation of arthroscopic knee surgery. Through immersive technologies, students can virtually explore and understand the knee's anatomical structures and spatial connections from different perspectives. Moreover, they can manipulate the model using a simulated arthroscope, gaining a surgeon's viewpoint before experiencing the procedure firsthand [224]. While remote access to medical dissections is gradually becoming available through the experimental Next-generation Internet, it still has some limitations. Looking forward, complex procedures may soon be routinely taught to medical students using computer simulations or models, offering immediate feedback and eliminating the necessity for patients or live models.
Patient and Provider Education
Health science institutions have expanded their educational offerings to the online realm, providing postgraduate education, refresher courses, and independent study opportunities for health science students through the Internet. In the future, medical practitioners may prescribe tailored video instructional programs for patients, delivering them directly to their home television sets via the Internet. Healthcare facilities are utilizing video servers on the Internet to disseminate resources to patients and offer their staff continuous nursing and medical education. Additionally, healthcare providers now offer patient-centric versions of electronic medical records accessible online, complemented by personalized online information addressing the patient's specific health concerns [225].
Disease Management
Most homes in the USA now have access to high-speed Internet via DSL, cable, or satellite connections. Over 50 million households now have broadband connectivity. Soon, physicians will go from utilizing telephones to remotely addressing patient issues to using their visual perception via two-way video connections. The individuals who are physically weak or unwell will be provided with "home visits" via video connections, thus reducing needless trips to medical facilities or emergency rooms. Additionally, care managers will have significant additional resources for monitoring patients and prioritizing preventive measures rather than managing crises. Preliminary trials indicate that patients exhibit notable excitement when videoconferencing encounters are conducted in their homes by doctors and nurses with whom they are already acquainted. With the emergence of highly efficient, multi-core supercomputers, scientists can now simulate complex biological phenomena, such as protein folding or drug-receptor binding. The modeling work may be performed utilizing high-speed Internet and remote computers, namely the biologically oriented supercomputers in San Diego, California, and Pittsburgh, Pennsylvania [226].
Addressing the Research Questions
The findings from the review connected to the set research questions and support the discussion and evaluation of computer-aided tools in healthcare.
In answer to RQ1, this review covered the key application areas of computer-aided technologies such as medical imaging, decision helpers, surgery assistance, keeping electronic health records, along with patient monitoring. The article also included details discussing their AI-driven method for classification and information on their modeling and image-processing techniques (Sects. 6 to 8). The grouping of these about technologies describes the areas they operate in and how they connect with other clinical objects.
In answer to RQ2, the review discovered that there are a number of lasting issues. It was recognized that major concerns in ethics related to AI include the fact that decisions may not be obvious, there could be biased data in training, and that algorithms are not always clear. When it comes to deployment, there were problems with infrastructural shortcomings, costly equipment, and doctors not wanting to use new systems. These challenges are significant in less developed countries because digitizing healthcare has not been a priority due to restricted funds. Additionally, problems in medical systems and using technology heavily increase the possibility of clinical and legal issues (Sects. 1, 7, 8).
The literature gave detailed insights about how advances in technologies have helped to improve accuracy, efficiency, and patient results in diagnostics. Thanks to AI, doctors use CAD for fast and reliable cancer finding, use gesture controls for guiding operations during surgery, and benefit from supercomputing for speeding up new drug development. These tools bring more than just support; they are changing the main concepts in current medicine by introducing personalized, anticipatory, and preventative approaches (Sects. 6.5 to 6.9).
For RQ4, the research discussed various major gaps in the literature. It is also clear that AI is not built with data from many demographic groups, clinical tests are not standard enough, and interoperability between EHR systems is lacking. The reviewers underlined the absence of broad ethical standards to help use autonomous technology in critical care. This proves there is a high need for studies that look at multiple areas, such as clinical, computational, and ethical concerns, in the creation of cutting-edge healthcare solutions (Sect. 9).
All in all, this study provides answers to the main questions and highlights the positive areas as well as some of the problems related to digital transformation in healthcare. This article aims to support people who work in medicine and biology in using computers responsibly and successfully in their day-to-day activities.
Conclusion and Recommendations
There is no doubt that the new trends in the use of computers in biomedicine will change how we diagnose, treat, and manage diseases, after noting the exceptional milestones achieved with the integration of AI technologies into medicine. This comprehensive analysis examines computing technology's transformative impact on medicine. The historical viewpoint recounts the change from paper records to electronic health records (EHRs), highlighting the importance of paper records and clinical trials. From computer-based patient monitoring to medical decision-making, the paper covers numerous applications. Simulations, modeling, data visualization, and complicated computer processing are examples of computer applications in medicine. Supercomputing, DSS, and pattern classification techniques are used in medicine. Human–computer interfaces, the impact of computing on drug development, and the various applications of computer methods and programs in biomedicine are covered. The comprehensive study explores modern computer-based medical imaging techniques, highlighting their many applications. Clinical applications enhance operations. Paper anticipates telemedicine, remote consultations, integrated health records, computer learning, patient and provider education, and successful disease management. Using a thorough study, computing's impact on healthcare's changing environment may enhance medical treatment. Numerous recommendations for developing and using these technologies for healthcare systems and patients were developed: (1) encourage developing and using standard interoperability protocols to promote communication and integration among healthcare systems; (2) emphasize cybersecurity to safeguard patient data and medical records. (3) Teach healthcare professionals advanced computing technologies; (4) prioritize user-friendly interface development in medical technologies; (5) improve regulatory frameworks to ensure ethical and responsible use of medical computing technologies; encourage collaboration between computer scientists, pharmaceutical researchers, and healthcare professionals to speed up medication development; invest in telemedicine infrastructure development and expansion for low-cost, high-quality services; (6) engage patients and providers using computer-based teaching. Improved health outcomes and disease management may be linked to improving patient and healthcare practitioner access to information and services. The future of using computers in medicine will depend on developing machine learning, strong security for data, and compatibility among different systems. AI makes health services more organized, customized, and reliable, enhancing patient care and the results of treatments. Computer technology has lifted patient care to new heights, yet its rapid spread forces us to think hard about the ethical and social fallout. Because most files are now stored and sent online, keeping sensitive health information private and secure sits at the top of the worry list. While helpful, electronic health records and AI diagnostic tools open doors to possible data breaches or snooping. Strong encryption, regular audits, and strict rules like HIPAA are still the first line of defense. The trouble is that tech keeps sprinting ahead of the rules, making it tough for hospitals and clinics to stay in step.
Author Contributions
QML, Wahab Hussain, ZXL, ZLJ, and YHZ conceived the study and drafted the manuscript. Sarfraz Hussain, HYL, XYJ, HWW, and DDW prepared the figures. All authors read and approved the final manuscript. All authors have agreed to be accountable for all aspects of this work.
Funding
This work was supported by grants from the Foundation of Science and Technology Department of Henan Province, China (no. 212102310103), Natural Science Foundation of Education Department of Henan Province, China (no. 21A320004), Foundation of the National Health Commission of Henan Province, China (no. Wjlx2020017), Foundation of the Educational Administration Department of Henan University, China (no. HDXJJG2020-83), Foundation of Science & Technology Department of Kaifeng City, Henan Province, China (no. 2203015).
Data Availability
Data sharing does not apply to this article, as no new data were created or analyzed in this study.
Declarations
Conflict of Interest
The authors declare that there are no conflicts of interest.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
1. Bhardwaj, H; Singh, S; Sood, S. Singh, S. Principles and foundations of artificial intelligence and internet of things technology. Artificial Intelligence to Solve Pervasive Internet of Things Issues; 2021; Amsterdam, Elsevier: pp. 377-392.
2. Shyamalee, T; Jayasekara, S; de Silva, D et al. Automated tool support for glaucoma identification with explainability using fundus images. IEEE Access; 2024; 12, pp. 17290-17307.
3. Korteling, J; Brouwer, A-M; Crutzen, R. Human-versus artificial intelligence. Front. Robot. AI; 2021; 4, 622364.
4. Whig, P; Singh, S; Kaur, S. Computational science role in medical and healthcare-related approach. Handbook of Computational Intelligence in Healthcare; 2023; Berlin, Springer: pp. 245-272.
5. Shah, V; Khang, A. Internet of medical things (IoMT) driving the digital transformation of the healthcare sector. Data-Centric AI Solutions and Emerging Technologies in the Healthcare Ecosystem; 2023; Boca Raton, CRC Press: pp. 15-26.
6. Zewail, A; Saber, S. AI-powered analytics in healthcare: enhancing decision-making and efficiency. Int. J. Appl. Health Care Anal.; 2023; 8,
7. Chen, Y-W; Chen, C-H; Chang, C-H. Computer-assisted surgery in medical and dental applications. J. Med. Biol. Eng.; 2021; 18,
8. Barragán-Montero, A; García-García, F; Orts-Escolano, S. Artificial intelligence and machine learning for medical imaging: a technology review. Physica Med.; 2021; 83, pp. 242-256.
9. Olveres, J; Fernández-Caballero, A; Martínez-González, P. What is new in computer vision and artificial intelligence in medical image analysis applications. Quant. Imaging Med. Surg.; 2021; 11,
10. Wijesinghe, C.B., Meedeniya, D., Yogarajah, P.: Cardiac MRI segmentation of ventricular structures and myocardium using U-Net variants. In: 2025 5th International Conference on Advanced Research in Computing (ICARC), pp. 1–6. IEEE (2025)
11. Samarasinghe, D; Perera, S; Jayasekara, S. Brain tumour segmentation and edge detection using self-supervised learning. Int. J. Online Biomed. Eng. (iJOE); 2025; 21,
12. Lin, L; Wang, L; Zhang, H. High-speed three-dimensional photoacoustic computed tomography for preclinical research and clinical translation. Light Sci. Appl.; 2021; 12,
13. Mitra, S; Das, S; Nasipuri, M. Cytology image analysis techniques toward automation: systematically revisited. ACM Comput. Surv.; 2021; 54,
14. Berner, ES; Detmer, DE; Simborg, DJ. Will the wave finally break? A brief view of the adoption of electronic medical records in the United States. J. Am. Med. Inform. Assoc.; 2005; 12,
15. Chen, SW; Chen, YC; Lee, CH. Review of ECG detection and classification based on deep learning: coherent taxonomy, motivation, open challenges and recommendations. Biomed. Signal Process. Control; 2022; 74, 103493.
16. Flynn, FJ. Problems and benefits of using a computer for laboratory data processing. Comput. Progr. Biomed.; 1969; 3, pp. 62-68.
17. Henzler, D; Müller, M; Lauterbach, K. Healthcare professionals’ perspectives on artificial intelligence in patient care: a systematic review of hindering and facilitating factors on different levels. BMC Health Serv. Res.; 2025; 25,
18. Fitzmaurice, JM; Adams, K; Eisenberg, JM. Three decades of research on computer applications in health care: medical informatics support at the Agency for Healthcare Research and Quality. J. Am. Med. Inform. Assoc.; 2002; 9,
19. Wald, LL; Stiles, MK; Gulani, V. Low-cost and portable MRI. J. Magn. Reson. Imaging; 2020; 52,
20. Wanasinghe, T., Jayasekara, S., de Silva, D.: CNN-based optimization for lung sound classification with mobile accessibility. In: 2024 International Research Conference on Smart Computing and Systems Engineering (SCSE), pp. 1–5. IEEE (2024)
21. Kendall, EJ; Barnett, MG; Chytyk-Praznik, K. Automatic detection of anomalies in screening mammograms. BMC Med. Imaging; 2013; 13,
22. Encarnacao, JL; Lindner, R; Schlechtendahl, EG. Computer Aided Design: Fundamentals and System Architectures; 2012; Berlin, Springer Science & Business Media:
23. Yazdian, P; Liu, C; Hodgins, JK. Motionscript: natural language descriptions for expressive 3D human motions. ACM Trans. Graph.; 2023; 42,
24. Nields, MW. FDA & digital mammography: why has FDA required full field digital mammography systems to be regulated as potentially dangerous devices for more than 10 years?. Acad. Radiol.; 2010; 17,
25. Du-Crow, E.: Computer aided detection in mammography (PhD thesis). The University of Manchester (United Kingdom) (2022)
26. Lalegani Dezaki, M; Safari, J; Khosrozadeh, A. Influence of infill patterns generated by CAD and FDM 3D printer on surface roughness and tensile strength properties. Mater. Res. Express; 2021; 11,
27. Desai, RJ; Shah, ND; Patel, MR. Comparison of machine learning methods with traditional models for use of administrative claims with electronic medical records to predict heart failure outcomes. JAMA Netw. Open; 2020; 3,
28. Negro-Calduch, E; Martínez-Costa, C; Simó-Alfonso, E. Technological progress in electronic health record system optimization: systematic review of systematic literature reviews. Int. J. Med. Inform.; 2021; 152, 104507.
29. Slaveykov, K; Angelova, M; Hristova, I. Electronic health records–benefits, savings and costs. Health Inform. J.; 2013; 3,
30. Al Aswad, A.M.: Issues concerning the adoption and usage of electronic medical records in ministry of health hospitals in Saudi Arabia (PhD thesis). University of Sheffield, School of Health and Related Research (2025)
31. Forde-Johnston, C; Butcher, D; Aveyard, H. An integrative review exploring the impact of electronic health records (EHR) on the quality of nurse–patient interactions and communication. J. Adv. Nurs.; 2023; 79,
32. Abramson, EL; Kaushal, R; Jha, AK. Physician experiences transitioning between an older versus newer electronic health record for electronic prescribing. J. Gen. Intern. Med.; 2012; 81,
33. Gabriel, MH; O’Malley, AS; Blumenthal, D. Progress and challenges with the implementation and use of electronic health records among critical access hospitals. Health Aff.; 2013; 32,
34. Frimpong, JA; Asare, R; Opoku, SO. Adoption of electronic health record among substance use disorder treatment programs: nationwide cross-sectional survey study. JMIR Ment. Health; 2023; 25,
35. Yoon, D; Park, J; Kim, H. Adoption of electronic health records in Korean tertiary teaching and general hospitals. J. Med. Syst.; 2012; 81,
36. Emani, S; Patel, VR; Jha, AK. Physician beliefs about the meaningful use of the electronic health record: a follow-up study. J. Am. Med. Inform. Assoc.; 2017; 8,
37. Tariq, M; Rehman, A; Khan, MU. Medical image based breast cancer diagnosis: state of the art and future directions. Biomed. Pharmacother.; 2021; 167, 114095.
38. Woldemariam, MT; Jimma, W Informatics, C. Adoption of electronic health record systems to enhance the quality of healthcare in low-income countries: a systematic review. BMJ Health Care Inform.; 2023; [DOI: https://dx.doi.org/10.1136/bmjhci-2022-100704]
39. Somanchi, S; Chatterjee, S; Mukherjee, S. Come together: an empirical investigation of EHR adoption, mergers, and survival in the US healthcare system. Health Serv. Res.; 2023; 58,
40. Frimpong Antwi, M. Impact of electronic health record system (EHRs) on healthcare quality at Asamankese Government Hospital, Ghana. Glob. J. Community Med.; 2023; 1,
41. Bhatia, T; Singh, S; Sood, S. Towards a secure incremental proxy re-encryption for e-healthcare data sharing in mobile cloud computing. Secur. Commun. Netw.; 2020; 32,
42. Yew, H.T., Abdullah, J., Salleh, R.: IoT based real-time remote patient monitoring system. In: 2020 16th IEEE International Colloquium on Signal Processing & its Applications (CSPA). IEEE (2020)
43. Ahmed, A; Islam, SMR; Kwak, KS. IoT-based real-time patients vital physiological parameters monitoring system using smart wearable sensors. J. Ambient. Intell. Humaniz. Comput.; 2022; 35,
44. Santos, MA; Costa, AM; de Almeida, AT. Online heart monitoring systems on the internet of health things environments: a survey, a reference model and an outlook. Comput. Netw.; 2020; 53, pp. 222-239.
45. Stark, L; Holtzman, NA; Gornbein, J. Remote real-time diagnosis of clinical electrocardiograms by a digital computer system. Am. J. Cardiol.; 1965; 126,
46. Laferrière-Langlois, P; Paquette, S; Ménard, JF. Depth of anesthesia and nociception monitoring: current state and vision for 2050. Anesthesiology; 2024; 138,
47. Weil, MH; Shubin, H; Rosoff, L. Fluid repletion in circulatory shock: central venous pressure and other practical guides. J. Am. Med. Assoc.; 1965; 192,
48. Boyd, WA. An “ideal” monitoring system to aid patient care. IEEE Trans. Biomed. Electron.; 1965; 3,
49. Iqbal, S; Rehman, A; Khan, MU. On the analyses of medical images using traditional machine learning techniques and convolutional neural networks. Biomed. Signal Process. Control; 2023; 30,
50. Al Mohammad, B; Brennan, PC; Mello-Thoms, C. A review of lung cancer screening and the role of computer-aided detection. Clin. Radiol.; 2017; 72,
51. Arun Kumar, S; Sasikala, S. Review on deep learning-based CAD systems for breast cancer diagnosis. Technol. Cancer Res. Treat.; 2023; 22, 15330338231177977.
52. Mayo, RC; Gatsonis, C; Schmidt, RA. Reduction of false-positive markings on mammograms: a retrospective comparison study using an artificial intelligence-based CAD. J. Digit. Imaging; 2019; 32,
53. Ahsan, MM; Luna, SA; Siddique, Z. Machine-learning-based disease diagnosis: a comprehensive review. Healthcare (Basel); 2022; 10,
54. Esteva, A; Kuprel, B; Novoa, RA. Dermatologist-level classification of skin cancer with deep neural networks. Nature; 2017; 542,
55. Tran, KA; Pham, TA; Lee, BH. Deep learning in cancer diagnosis, prognosis and treatment selection. Genome Med.; 2021; 13,
56. Liu, X; Zhou, B; Hu, X. A comparison of deep learning performance against health-care professionals in detecting diseases from medical imaging: a systematic review and meta-analysis. Lancet Digit. Health.; 2019; 1,
57. Loizidou, K; Pattichis, CS; Kyriacou, E. Computer-aided breast cancer detection and classification in mammography: a comprehensive review. Biomed. Pharmacother.; 2023; 167, 106554.
58. Najjar, RJD. Redefining radiology: a review of artificial intelligence integration in medical imaging. J. Clin. Med.; 2023; 13,
59. Barua, R; Mondal, JJ; Analysis, TIMI. Analysis, study of the current trends of CAD (computer-aided detection) in modern medical imaging. Int. J. Med. Inform.; 2023; 176, pp. 35-50.
60. Wang, S., Li, J., Liu, Y.: Chatcad: interactive computer-aided diagnosis on medical image using large language models. arXiv preprint http://arxiv.org/abs/2303.11332. (2023)
61. Sabbella, D.S., Singh, A.: Artificial intelligence in 3D CAD modelling. In: 2020 International Conference on Emerging Trends in Information Technology and Engineering (ic-ETITE), pp. 1–6. IEEE (2020)
62. Segaran, N; Arora, S; Gupta, A. Application of 3D printing in preoperative planning. J. Clin. Med.; 2021; 10,
63. Nasrallah, A; Alqadi, L; Issa, A. Integrating 3D printing technology in surgical planning and prosthetic development: current application and future prospects. J. Med. Eng. Technol.; 2024; 48,
64. Wang, S; Li, J; Zhang, Y. Exploration of simulation-based medical education for undergraduate students. Chin. J. Med. Educ.; 2021; 100,
65. Oo, Y. M., Nataraja, R.: The application of simulation-based medical education in low-and middle-income countries; the Myanmar experience. In: Seminars in Pediatric Surgery, pp. 1–5. Elsevier (2020)
66. Huang, J-H; Chen, C-H; Lin, C-J. Surgical treatment for both-column acetabular fractures using pre-operative virtual simulation and three-dimensional printing techniques. J. Orthop. Surg. Res.; 2020; 133,
67. Luria, S. Understanding the patterns of deformity of wrist fractures using computer analysis. Clin. Res. Radiol.; 2020; 16,
68. Zhou, L; Li, X; Wang, H. A review of three-dimensional medical image visualization. J. Vis.; 2022; 2022, pp. 1-15.
69. Chibi, NT; Benrabah, M; Talea, M. Corrections to “A novel approach based on machine learning, blockchain and decision process for securing smart grid”. IEEE Access; 2025; 13, pp. 6154-6154.
70. Vellido, A. The importance of interpretability and visualization in machine learning for applications in medicine and health care. Neural Comput. Appl.; 2020; 32,
71. Hofbauer, LC; Pautrat, K; Chergui, N. Novel approaches to target the microenvironment of bone metastasis. Curr. Opin. Oncol.; 2021; 18,
72. Siri, SK; Kumar, SP; Latte, MV. Threshold-based new segmentation model to separate the liver from CT scan images. IETE J. Res.; 2022; 68,
73. Bogacz, A; Szymański, M; Wojewódzki, M. High-frequency audiometry in the diagnosis of tinnitus. J. Otolaryngol.-Head Neck Surg.; 2023; 52,
74. Abuya, TK; Rimiru, RM; Okeyo, GO. An image denoising technique using wavelet-anisotropic Gaussian filter-based denoising convolutional neural network for CT images. Appl. Sci.; 2023; 13,
75. Rahul, Goyal, B.: Gaussian filtering based image integration for improved disease diagnosis and treatment planning. In: 2021 9th International Conference on Reliability, Infocom Technologies and Optimization (Trends and Future Directions) (ICRITO), pp. 1–6. IEEE (2021)
76. Verma, K; Singh, BK; Thoke, AS. An enhancement in adaptive median filter for edge preservation. Procedia Comput. Sci.; 2015; 48, pp. 29-36.
77. Bhatti, A; Chen, X; Kolios, MC. Region-based SVD processing of high-frequency ultrafast ultrasound to visualize cutaneous vascular networks. Ultrasonics; 2023; 129, 106907.
78. Ahmad, I; Alqurashi, F. Early cancer detection using deep learning and medical imaging: a survey. Crit. Rev. Oncol./Hematol.; 2024; 204, 104528.
79. Dang, K; Li, J; Zhang, Y. A deep learning framework integrating MRI image preprocessing methods for brain tumor segmentation and classification. IBRO Neurosci. Rep.; 2022; 13, pp. 523-532.
80. Min, C; Li, Y; Wang, H. Non-rigid registration for infrared and visible images via gaussian weighted shape context and enhanced affine transformation. Remote Sens.; 2020; 8, pp. 42562-42575.
81. Huang, Y; Li, X; Zhang, H. Difficulty-aware hierarchical convolutional neural networks for deformable registration of brain MR images. Med. Image Anal.; 2021; 67, 101817.
82. Hassan, C; Pickhardt, PJ; Kim, D. Computer-aided detection-assisted colonoscopy: classification and relevance of false positives. Gastroenterology; 2020; 92,
83. Gagné, M; Ducharme, F; Boulet, LP. A patient decision aid for mild asthma: navigating a new asthma treatment paradigm. Patient Educ. Couns.; 2022; 201, 106568.
84. Zhang, X; Li, Y; Wang, Z. The positive role of traditional Chinese medicine as an adjunctive therapy for cancer. Chin. J. Integr. Med.; 2021; 15,
85. Daffner, R; Alavi, A; Wolf, PL. CT of the esophagus. Part II. carcinoma. Radiology; 1979; 133,
86. Papandrianos, N; Papageorgiou, ES. Automatic diagnosis of coronary artery disease in SPECT myocardial perfusion imaging employing deep learning. Appl. Sci.; 2021; 11,
87. Nalinipriya, G; Suresh, S; Ramakrishnan, S. Biomedical data mining for improved clinical diagnosis. Artificial Intelligence in Data Mining; 2021; Amsterdam, Elsevier: pp. 155-176.
88. Elhaddad, M; Hamam, S. AI-driven clinical decision support systems: an ongoing pursuit of potential. Cureus; 2024; 16,
89. Kampolis, CF; Rello, J; Pelosi, P. Comparison of advanced closed-loop ventilation modes with pressure support ventilation for weaning from mechanical ventilation in adults: a systematic review and meta-analysis. Crit. Care; 2022; 68, pp. 1-9.
90. Rees, SE; Goldberger, AL; Badawi, O. Transparent decision support for mechanical ventilation using visualization of clinical preferences. JMIR Med. Inform.; 2022; 21,
91. Samadi, B; Asadi, A; Sadeghi, M. Development of machine learning algorithms to identify the Cobb angle in adolescents with idiopathic scoliosis based on lumbosacral joint efforts during gait (case study). J. Orthop. Res. Therapy; 2023; 5,
92. Liu, W; Chen, X; Li, Y. Practical moving target detection in maritime environments using fuzzy multi-sensor data fusion. Sensors; 2021; 23, pp. 1860-1878.
93. Mufassirin, MM; Newton, MH; Sattar, A. Artificial intelligence for template-free protein structure prediction: a comprehensive review. Artif. Intell. Rev.; 2023; 56,
94. Murugan, NA; Sivakumar, S; Pandian, S. A review on parallel virtual screening softwares for high-performance computers. Curr. Comput.-Aided Drug Des.; 2022; 15,
95. Li, X; Xu, Y; Neuroscience, C. Role of human-computer interaction healthcare system in the teaching of physiology and medicine. J. Med. Educ. Curric. Dev.; 2022; 2022, pp. 1-8.
96. Banerjee, T.: Computer Vision-Based Hand Tracking and 3D Reconstruction as a Human-Computer Input Modality with Clinical Application. The University of Western Ontario (Canada) (2023)
97. Sabe, VT; Patel, DK; Patel, AK. Current trends in computer aided drug design and a highlight of drugs discovered via computational techniques: a review. Eur. J. Med. Chem.; 2021; 224, 113705.
98. Gao, K; Li, Y; Wang, H. Challenges and solutions for vision-based hand gesture interpretation: a review. Comput. Vis. Image Underst.; 2024; 248, 104095.
99. Davis, F; Venkatesh, V. Toward preprototype user acceptance testing of new information systems: implications for software project management. IEEE Trans. Eng. Manag.; 2004; 51,
100. Elkess, G., Elmoushy, S., Atia, A.: Karate first kata performance analysis and evaluation with computer vision and machine learning. In: 2023 International Mobile, Intelligent, and Ubiquitous Computing Conference (MIUCC), pp. 1–6. IEEE (2023)
101. Martens, C; Scott, P; McCusker, J. Modeling game mechanics with Ceptre. Proc. ACM Comput. Sci. Educ.; 2023; 7,
102. Ogiela, MR; Ogiela, L; Tadeusiewicz, R. Natural user interfaces for exploring and modeling medical images and defining gesture description technology. Handbook of Human-Computer Interaction in Healthcare; 2015; Berlin, Springer: pp. 205-279.
103. Loaiza, CR; Smith, J; Johnson, M. Conference on human–computer interaction. ACM Trans. Comput.-Hum. Interact.; 2022; 29,
104. Borah, G; Nath, AC. Merging tech and medicine: biomedical engineering at the cutting edge. J. Med. Syst. Pract.; 2023; 114, pp. 1-10.
105. Gao, C; Li, Y; Wang, Z. Comparative anatomy of the mouse and human ankle joint using micro-CT: utility of a mouse model to study human ankle sprains. J. Orthop. Res.; 2019; 16, pp. 2959-2972.4030719
106. Li, X; Zhang, H; Wang, L. A method of ultrasound diagnosis for unilateral peripheral entrapment neuropathy based on multilevel side-to-side image contrast. Ultrasound Med. Biol.; 2019; 16,
107. Jiang, WW; Li, Y; Zhang, H. A fast 3-D ultrasound projection imaging method for scoliosis assessment. IEEE Trans. Biomed. Eng.; 2019; 16,
108. Jiang, W-W; Li, X; Zhang, H. An automatic measurement method of spinal curvature on ultrasound coronal images in adolescent idiopathic scoliosis. Ultrasound Med. Biol.; 2020; 17, pp. 776-788.4045027
109. Ouyang, Y; Li, X; Zhang, H. A review of ultrasound detection methods for breast microcalcification. Ultrasound Med. Biol.; 2019; 16,
110. Shen, Y; Li, X; Zhang, H. Quantitative analysis of non-alcoholic fatty liver in rats via combining multiple ultrasound parameters. Ultrasound Med. Biol.; 2019; 16,
111. Zhang, R; Li, X; Zhang, H. Computer-assisted needle trajectory planning and mathematical modeling for liver tumor thermal ablation: a review. Med. Phys.; 2019; 16,
112. Zhu, G; Li, X; Zhang, H. Examining and monitoring paretic muscle changes during stroke rehabilitation using surface electromyography: a pilot study. J Rehabil Med; 2020; 17, pp. 216-234.4044999
113. Liu, Z; Li, X; Zhang, H. Additive manufacturing of specific ankle-foot orthoses for persons after stroke: a preliminary study based on gait analysis data. J. Neuroeng. Rehabil.; 2019; 16,
114. Yao, L; Li, X; Zhang, H. Minimally invasive treatment of calcaneal fractures via the sinus tarsi approach based on a 3D printing technique. J. Orthop. Surg. Res.; 2019; 16,
115. Li, T; Li, X; Zhang, H. Human hands-and-knees crawling movement analysis based on time-varying synergy and synchronous synergy theories. J. Biomech.; 2019; 16,
116. Li, Y; Li, X; Zhang, H. Biomechanical analysis of the meniscus and cartilage of the knee during a typical Tai Chi movement-brush-knee and twist-step. J. Biomech.; 2019; 16,
117. Xiao, X; Li, X; Zhang, H. Comparison of dominant hand to non-dominant hand in conduction of reaching task from 3D kinematic data: trade-off between successful rate and movement efficiency. Hum. Mov. Sci.; 2019; 16,
118. Wu, C; Li, X; Zhang, H. Biomechanical and neuromuscular strategies on backward somersault landing in artistic gymnastics: a case study. J. Sports Sci.; 2019; 16,
119. Zheng, X; Li, X; Zhang, H. A model-based method with geometric solutions for gaze correction in eye-tracking. IEEE Trans. Biomed. Eng.; 2020; 17,
120. Ma, Z; Li, X; Zhang, H. Driving event-related potential-based speller by localized posterior activities: an offline study. Brain Res. Bull.; 2020; 17, pp. 789-801.4045028
121. Jiang, H; Li, X; Zhang, H. Transfer learning on T1-weighted images for brain age estimation. Neuroimage; 2019; 16, pp. 4382-4398.
122. Wang, J; Li, X; Zhang, H. VAE-driven multimodal fusion for early cardiac disease detection. IEEE Access; 2024; 12, pp. 90535-90551.
123. Halalli, B; Makandar, A. Computer aided diagnosis-medical image analysis techniques. Biomed. Res.; 2018; 85, pp. 85-109.
124. Kasban, H; El-Bendary, N; Khedr, W. A comparative study of medical imaging techniques. Int. J. Comput. Appl.; 2015; 4,
125. Haleem, A; Javaid, M; Singh, R. Telemedicine for healthcare: capabilities, features, barriers, and applications. Sens. Int.; 2021; 2, 100117.
126. Domingues, I; Cardoso, JS; Pereira, JM. Using deep learning techniques in medical imaging: a systematic review of applications on CT and PET. Comput. Struct. Biotechnol. J.; 2020; 53, pp. 4093-4160.
127. Ambrose, E; Gould, T; Uttley, D. Jamie Ambrose. Br. Med. J.; 2006; 332,
128. Stock, S.R. Trends in the micro-and nano computed tomography 2010–2012. In: Developments in X-ray Tomography VIII, pp. 1–10. SPIE (2012)
129. Hancock, BC; Mullarney, MP. X-ray microtomography of solid dosage forms. Pharm. Technol.; 2005; 29,
130. Suppiah, S; Subramaniam, R; Ranganathan, S. A review on the usage of bone single-photon emission computed tomography/computed tomography in detecting skeletal metastases in the post-COVID-19 era: is it time to ditch planar and single-photon emission computed tomography only gamma camera systems?. J. Nucl. Med. Technol.; 2023; 38,
131. Greco, A; Martelli, S; Piantadosi, C. Imaging techniques in veterinary medicine. Part II: computed tomography, magnetic resonance imaging, nuclear medicine. Vet. Sci.; 2023; 10, 100467.
132. Veronesi, G; Pastorino, U; De Koning, HJ. Recommendations for implementing lung cancer screening with low-dose computed tomography in Europe. Eur. Respir. J.; 2020; 12,
133. Siwicka-Gieroba, D; Gieroba, M; Krawczyk, M. Concentration of apoptotic factors in bronchoalveolar lavage fluid, as potential brain-lung oxygen relationship, correspond to the severity of brain injury. J. Clin. Med.; 2023; 22,
134. Fan, Y; Li, J; Zhang, H. Targeted tumor hypoxia dual-mode CT/MR imaging and enhanced radiation therapy using dendrimer-based nanosensitizers. Adv. Funct. Mater.; 2020; 30,
135. Pautrat, K; Chergui, N. SARS-CoV-2 infection may result in appendicular syndrome: chest CT scan before appendectomy. Virchows Arch.; 2020; 157,
136. Jepson, BM; Geva, T; Johnson, JN. Proposed competencies for the performance of cardiovascular computed tomography in pediatric and adult congenital heart disease. J. Am. Coll. Cardiol.; 2023; 82,
137. Kwiecinski, J; Nowak, J; Kubica, J. Bypass grafting and native coronary artery disease activity. Cardiol. J.; 2022; 15,
138. Khadka, S; Adhikari, S; Shrestha, S. Multidetector computed tomography evaluation of obstructive jaundice: a cross-sectional study from a tertiary hospital of Nepal. J. Nepal Med. Assoc.; 2023; 6,
139. Knochel, J; Moss, AA; McCort, JM. Diagnosis of abdominal abscesses with computed tomography, ultrasound, and 111In leukocyte scans. Radiology; 1980; 137,
140. Dixon, A; Thompson, JS; Williams, G. Pre-operative computed tomography of carcinoma of the rectum. Br. Med. J.; 1981; 54,
141. Hussain, S; Javaid, M; Haleem, A. Modern diagnostic imaging technique applications and risk factors in the medical field: a review. J. Med. Imaging Radiat. Sci.; 2022; 2022, pp. 1-10.
142. Grampp, S; Klaushofer, K; Pietschmann, P. Assessment of the skeletal status by peripheral quantitative computed tomography of the forearm: short-term precision in vivo and comparison to dual X-ray absorptiometry. Osteoporos Int.; 2020; 10,
143. Chirvi, S; Suryavanshi, S; Kulkarni, R. Trabecular bone mineral density correlations using QCT: central and peripheral human skeleton. J. Bone Miner. Metab.; 2020; 112, 104076.
144. Whittier, DE; Boutroy, S; Dempster, DW. Guidelines for the assessment of bone density and microarchitecture in vivo using high-resolution peripheral quantitative computed tomography. J. Bone Miner. Res.; 2020; 31, pp. 1607-1627.
145. Chan, TJ; Rajapakse, CS. A super-resolution diffusion model for recovering bone microstructure from CT images. Acad. Radiol.; 2023; 5,
146. Gruenewald, LD; Link, TM; Engelke, K. Dual-energy CT-based opportunistic volumetric bone mineral density assessment of the distal radius. Radiology; 2023; 308,
147. Goff, E; Allen, MR; van der Meulen, MC. Large-scale quantification of human osteocyte lacunar morphological biomarkers as assessed by ultra-high-resolution desktop micro-computed tomography. Bone; 2021; 152, 116094.
148. Benca, E; Amini, M; Pahr, DH. Effect of CT imaging on the accuracy of the finite element modelling in bone. Eur. J. Mech. R/Solids; 2020; 4, pp. 1-8.
149. Arlot, ME; Seeman, E; Delmas, PD. Histomorphometric and μCT analysis of bone biopsies from postmenopausal osteoporotic women treated with strontium ranelate. Bone; 2008; 23,
150. Mićić, M.J.: Influence of the choice of protocols on the accuracy of three-dimensional medical models, surgical guides and bone replacement (doctoral dissertation). University of Belgrade (2023)
151. Gumede, L; Badriparsad, N. CT in an emergency setting. Computed Tomography: Advanced Clinical Applications; 2023; Berlin, Springer: pp. 39-60.
152. Zhou, L; Li, X; Wang, H. A coupling, stabilizing, and shaping strategy for breast ultrasound computed tomography (USCT) with a ring array transducer. Ultrasonics; 2024; 138, 107212.
153. Tian, M; Hardy, J; Masters, CL. International nuclear medicine consensus on the clinical use of amyloid positron emission tomography in Alzheimer’s disease. J. Nucl. Med. Mol. Imaging.; 2023; 3,
154. Manjunath, K; Siddalingaswamy, P; Prabhu, G. Domain-based analysis of colon polyp in CT colonography using image-processing techniques. Asian Pac. J. Cancer Prev.; 2019; 20,
155. Mostafapour, S; Beiki, D; Shiran, M. Ultra-low dose CT scanning for PET/CT. J. Nucl. Med. Technol.; 2024; 51,
156. Zimmer, LJ. Recent applications of positron emission tomographic (PET) imaging in psychiatric drug discovery. Expert Opin. Drug Discov.; 2023; 18,
157. Juweid, ME; Cheson, BD. Positron-emission tomography and assessment of cancer therapy. N. Engl. J. Med.; 2006; 354,
158. Leskinen, S; Kiviniemi, V; Ruohonen, J. Applications of functional magnetic resonance imaging to the study of functional connectivity and activation in neurological disease: a scoping review of the literature. World Neurosurg.; 2024; 189, pp. 185-192.
159. Kumar, A. History of MRI. Indian J. Radiol. Imaging; 2014; 94,
160. Shahzad, K; Mati, W. Advances in magnetic resonance imaging (MRI). Advances in Medical and Surgical Engineering; 2020; Amsterdam, Elsevier: pp. 121-142.
161. Dasanayaka, S., Jayasekara, S., de Silva, D.: Interpretable machine learning for brain tumor analysis using MRI. In: 2022 International Conference on Advanced Research in Computing (ICARC), pp. 212–217. IEEE (2022)
162. Gamage, L; Jayasekara, S; de Silva, D. Melanoma skin cancer identification with explainability utilizing mask guided technique. Electronics; 2024; 13,
163. Wintermark, M; Reich, DS; Albers, GW. Comparison of admission perfusion computed tomography and qualitative diffusion-and perfusion-weighted magnetic resonance imaging in acute stroke patients. Stroke; 2002; 33,
164. Johnson, M; Smith, J; Williams, L. Actual applications of magnetic resonance imaging in dentomaxillofacial region. J. Dent. Res.; 2022; 38,
165. Dodin, G; Laroche, M; Bonvalot, S. Added-value of advanced magnetic resonance imaging to conventional morphologic analysis for the differentiation between benign and malignant non-fatty soft-tissue tumors. Eur. Radiol.; 2021; 31, pp. 1536-1547.
166. Nakanishi, K; Tanaka, Y; Kawashima, H. Whole-body MRI: detecting bone metastases from prostate cancer. Jpn. J. Clin. Oncol.; 2021; 51,
167. Dubey, R., Sinha, N.: Structure and dynamics of native biological materials by solid-state NMR spectroscopy. In: NMR Spectroscopy for Probing Functional Dynamics at Biological Interfaces, pp. 614–655. The Royal Society of Chemistry (2022)
168. Chen, Z; Li, J; Zhang, H. Deep learning for image enhancement and correction in magnetic resonance imaging—state-of-the-art and challenges. Comput. Struct. Biotechnol. J.; 2023; 36,
169. Arnold, TC; Wald, LL; Gulani, V. Low-field MRI: clinical promise and challenges. J. Magn. Reson. Imaging; 2023; 57,
170. Piez, CW, Jr.; Holman, BL. Single photon emission computed tomography. Crit. Rev. Biomed. Eng.; 1985; 9,
171. Papathanassiou, D; Karantanas, AH; Damilakis, J. Single-photon emission computed tomography combined with computed tomography (SPECT/CT) in bone diseases. Eur. J. Radiol.; 2009; 76,
172. Kapsoritakis, N; Gourtsoyiannis, N; Prassopoulos, P. Clinical impact of targeted single-photon emission computed tomography/computed tomography (SPECT/CT) bone scintigraphy on the assessment of bone metastasis in cancer patients. J. Nucl. Med.; 2021; 42,
173. Bocancia-Mateescu, L-A; Mateescu, GD; Mihai, C. Nanobodies as diagnostic and therapeutic tools for cardiovascular diseases (CVDs). Int. J. Mol. Sci.; 2023; 16,
174. Raff, U; Hendee, W. Clinical applications and protocols of single photon emission computed tomography. Tomographic Methods in Nuclear Medicine; 2020; Boca Raton, CRC Press: pp. 53-104.
175. Mlosek, RK; Migda, B; Migda, M. High-frequency ultrasound in the 21st century. Ultrasound Med. Biol.; 2021; 20,
176. Golshan, R; Nikbakht, N; Ghaffari, S. The diagnostic accuracy of sonography for detecting placenta accreta spectrum in pregnant women, with placenta previa. J. Ultrasound. Med.; 2023; 42,
177. Zeng, SE; Cao, Y; He, X. Ultrasound, CT, and MR imaging for evaluation of cystic renal masses. Radiographics; 2022; 41,
178. Litrenta, J; Dormans, JP; Schwend, RM. Ultrasound evaluation of pediatric orthopaedic patients. J. Pediatr. Orthop.; 2020; 28,
179. Mei, L; Zhang, Z Biology, M. Advances in biological application of and research on low-frequency ultrasound. Ultrasound Med. Biol.; 2021; 47,
180. Köse, G; Darguzyte, M; Kiessling, F. Molecular ultrasound imaging. Int. J. Mol. Sci.; 2020; 10,
181. Langeveld, SA; Meijlink, M; Kooiman, KJ. Phospholipid-coated targeted microbubbles for ultrasound molecular imaging and therapy. Curr. Pharm. Biotechnol.; 2021; 63, pp. 171-179.
182. Rahman, S. Use of trans-vaginal ultrasound to diagnose and treat infertility and its performance. Open J. Radiol.; 2023; 13,
183. Hu, H-J; Li, X; Zhang, H. Comparison of the application value of transvaginal ultrasound and transabdominal ultrasound in the diagnosis of ectopic pregnancy. J. Med. Imaging Radiat. Sci.; 2023; 11,
184. Edelman, BJ; Macé, E. Functional ultrasound brain imaging: bridging networks, neurons, and behavior. Curr. Opin. Neurobiol.; 2021; 18, 100286.
185. Alexander, JL; Donahue, SP; Capone, AL. A systematic review of ultrasound biomicroscopy use in pediatric ophthalmology. J. AAPOS; 2021; 35,
186. Biçer, Ö; Hoşal, MB. The diagnostic value of ultrasound biomicroscopy in anterior segment diseases. Turk. J. Ophthalmol.; 2023; 53,
187. Wong, MC; Chen, CJ; Chen, TH. Differences in incidence and mortality trends of colorectal cancer worldwide based on sex, age, and anatomic location. JAMA Oncol.; 2021; 19,
188. Sha, J; Li, X; Zhang, H. Computed tomography colonography versus colonoscopy for detection of colorectal cancer: a diagnostic performance study. BMC Med.; 2020; 20, pp. 1-8.
189. Trilisky, I; Ward, E; Dachman, AH. Errors in CT colonography. Abdom. Imaging; 2015; 40, pp. 2099-2111.
190. Cao, W; Li, J; Zhang, H. A comprehensive review of computer-aided diagnosis of pulmonary nodules based on computed tomography scans. J. Med. Syst.; 2020; 8, pp. 154007-154023.
191. Agrawal, T; Choudhary, P. Segmentation and classification on chest radiography: a systematic survey. Artif. Intell. Rev.; 2023; 39,
192. Gu, Y; Li, J; Zhang, H. A survey of computer-aided diagnosis of lung nodules from CT scans using deep learning. Pattern Recognit.; 2021; 137, 104806.
193. Rizzi, M; D'Aloia, M. Computer aided system for breast cancer diagnosis. Biomed. Eng. Appl. Basis Commun.; 2014; 26,
194. Drukker, K; Sennett, CA; Giger, ML. Automated method for improving system performance of computer-aided diagnosis in breast ultrasound. IEEE Trans. Med. Imaging; 2008; 28,
195. Oza, P; Patel, D; Patel, A. Computer-aided breast cancer diagnosis: comparative analysis of breast imaging modalities and mammogram repositories. J. Med. Syst.; 2023; 19,
196. Hua, K.-L., Li, J., Zhang, H.: Computer-aided classification of lung nodules on computed tomography images via deep learning technique. In: 2015 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp. 2015–2022. IEEE (2015)
197. Shen, W., Wu, G., Suk, H.I.: Multi-scale convolutional neural networks for lung nodule classification. In: Information Processing in Medical Imaging: 24th International Conference, IPMI 2015, Sabhal Mor Ostaig, Isle of Skye, UK, June 28-July 3, 2015, Proceedings, pp. 384–395. Springer (2015)
198. Chen, H; Li, J; Zhang, H. Neural network ensemble-based computer-aided diagnosis for differentiation of lung nodules on CT images: clinical evaluation. Acad. Radiol.; 2010; 17,
199. Suzuki, K. Computerized detection of lesions in diagnostic images with early deep learning models. Machine and Deep Learning in Oncology, Medical Physics and Radiology; 2022; Berlin, Springer: pp. 175-204.
200. Wang, Y; Li, J; Zhang, H. Recognition of lung nodules in computerized tomography lung images using a hybrid method with class imbalance reduction. J. Med. Imaging Radiat. Sci.; 2023; 14,
201. Tokisa, T; Li, J; Zhang, H. Detection of lung nodule on temporal subtraction images based on artificial neural network. J. Digit. Imaging; 2012; 12,
202. Suji, RJ; Abraham, T; Soman, KP. A survey and taxonomy of 2.5 D approaches for lung segmentation and nodule detection in CT images. Pattern Recogn.; 2023; 137, 107437.
203. Anthimopoulos, M; Christodoulidis, S; Ebner, L. Lung pattern classification for interstitial lung diseases using a deep convolutional neural network. IEEE Trans. Med. Imaging; 2016; 35,
204. Topol, EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat. Med.; 2019; 25,
205. Shergill, M; Singh, S; Sood, S. Machine learning used to study risk factors for chronic diseases: a scoping review. Can. J. Public Health; 2025; 116,
206. Joulin, A., Grave, E., Bojanowski, P.: Bag of tricks for efficient text classification. arXiv preprint http://arxiv.org/abs/1607.01759 (2016)
207. Jones, SS; Podgurski, A; Bakken, S. Health information technology: an updated systematic review with a focus on meaningful use. Ann. Intern. Med.; 2014; 160,
208. Karimi, M; Li, J; Zhang, H. CRISPR-Cas13a as a next-generation tool for rapid and precise plant RNA virus diagnostics. Plant Methods; 2025; 21,
209. Sheng, J; Khan, Z; Wang, X. COVID-19 pandemic in the new era of big data analytics: methodological innovations and future research directions. Br. J. Manag.; 2020; 32, pp. 1164-1183.
210. Farhud, DD; Zokaei, S. Ethical issues of artificial intelligence in medicine and healthcare. Iran. J. Public Health; 2021; 50,
211. Obermeyer, Z; Powers, B; Vogeli, C. Dissecting racial bias in an algorithm used to manage the health of populations. Science; 2019; 366,
212. Roppelt, JS; Kanbach, DK; Kraus, S. Artificial intelligence in healthcare institutions: a systematic literature review on influencing factors. Technol. Soc.; 2024; 76, 102443.
213. Osama, M; Javaid, M; Haleem, A. Internet of medical things and healthcare 40: trends, requirements, challenges, and research directions. Sensors (Basel); 2023; 23,
214. Alenoghena, CO; Ojo, SO; Akinwale, AO. Telemedicine: a survey of telecommunication technologies, developments, and challenges. J. Telemed. Telecare; 2023; 12,
215. Huang, C; Li, J; Zhang, H. Internet of medical things: a systematic review. Neurocomputing; 2023; 557, 126719.
216. Vudathaneni, VKP; Chava, S; Gurram, R. The impact of telemedicine and remote patient monitoring on healthcare delivery: a comprehensive evaluation. Cureus; 2024; 16,
217. Corbett, JA; Opladen, JM; Bisognano, JD. Telemedicine can revolutionize the treatment of chronic disease. Int. J. Cardiol. Hypertens.; 2020; 7, 100051.
218. Ftouni, R; El Baba, F; El Rassi, R. Challenges of telemedicine during the COVID-19 pandemic: a systematic review. BMC Med. Inform. Decis. Mak.; 2022; 22,
219. Smith, C; Jones, R; Williams, L. The effectiveness of remote consultations during the COVID-19 pandemic: a tool for modernising the National Health Service (NHS). Cureus; 2022; 14,
220. Cerchione, R; Esposito, E; De Santis, G. Blockchain’s coming to hospital to digitalize healthcare services: designing a distributed electronic health record ecosystem. Technol. Forecast. Soc. Change; 2023; 120, 102480.
221. Samarasinghe, D; Perera, S; Jayasekara, S. Brain tumour segmentation and edge detection using self-supervised learning. Int. J. Online Biomed. Eng. (iJOE); 2025; 21, pp. 127-141.
222. Mahajan, HB; Singh, S; Sood, S. Integration of healthcare 4.0 and blockchain into secure cloud-based electronic health records systems. J. Med. Syst.; 2023; 13,
223. Berthold, DP; Smith, J; Johnson, M. Head-mounted display virtual reality is effective in orthopaedic training: a systematic review. J. Orthop. Surg. Res.; 2022; 17,
224. Sattar, M; Rehman, A; Khan, MU. Motivating medical students using virtual reality based education. J. Med. Educ.; 2020; 15,
225. Freeley, M. Current postgraduate training programs and online courses in precision medicine. Eur. J. Med. Health Sci.; 2020; 20,
226. Schlick, T; Smith, J; Johnson, M. Biomolecular modeling and simulation: a prospering multidisciplinary field. J. Comput. Chem.; 2021; 50, pp. 267-301.
© The Author(s) 2025. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.