Recent trends in smartphone-based optical imaging biosensors for genetic testing: A review

Abstract

Genetic testing plays an important role in human health management and disease prevention. However, traditional genetic testing methods have been unable to fulfill the current diagnostic needs owing to several limitations, including high cost, complexity, and difficulty in performing. Smartphones have multiple inherent advantages, such as easy portability, ubiquity, fast processing speed, and excellent imaging capabilities, and have enormous potential in realizing rapid on-site genetic testing. The present review documents the research progress of smartphone-based optical imaging biosensors in the fields of colorimetry, fluorescence, and microscopic imaging for genetic testing. Furthermore, the review describes their potential applications in diagnostics, which range from infectious diseases to hereditary diseases and cancers. Finally, the challenges and perspectives of smartphone-based optical imaging biosensors regarding genetic testing are discussed.

INTRODUCTION

With indepth studies on the molecular biology mechanism of diseases as well as the smooth implementation of the Human Genome Project, the relationship between genes and diseases is becoming increasingly clear.¹ A gene comprises DNA or RNA sequences that carry genetic information. Genes, also known as genetic factors, are the basic units that record and transmit the genetic information of life. Genetic testing, which refers to the analysis of molecular sequences that carry genetic information, that is, the analysis of DNA or RNA sequences, is a prerequisite for understanding the relationship between diseases and genes. Owing to the rapid development of molecular biology theory and technology, many genetic testing methods were developed, including polymerase chain reaction (PCR), nucleic acid hybridization probe technique, direct sequencing.^2,3 The basic principle of all genetic testing methods is amplifying and transforming the DNA or RNA sequence information into optical signals or electrical signals that can be recognized by specific devices through some innovative strategies, and determining the sequence information of genetic molecules via the optical or electrical signals. As a basic technique in life sciences, genetic testing is widely used in tumor screening, pathogen detection, biomedical research, agricultural breeding, judicial identification, food safety, and other fields.⁴ However, with rapid developments in the genetic detection technology, detection equipment requires higher performance and sophistication. Traditional laboratory analysis and testing instruments are bulky, complex, and costly and rely on well-trained professionals, which restricts their application in easy-to-use, point-of-care testing (POCT). A portable instrument for the rapid and visual detection of genes is urgently required to achieve onsite real-time detection for genetic testing.^5,6

Recent advances in the smartphone technology offer great potential for manufacturing portable instruments, as smartphone-based analytical sensors may be more accessible than traditional laboratory analytical equipment, with the advantages of small size, low cost, convenient portability, easy-to-operate.^7–10 Smartphones are integrated with many components that can be used for biological detection, including abundant built-in sensors (such as cameras) for detecting biological signals, powerful processors and memory for analyzing data, and high-resolution screens for displaying results. Considering these advantages and the widespread use of smartphones in recent years, various true portable smartphone-based biosensors have been developed for numerous applications, which include food safety and biosecurity, medical diagnostics, environmental science, and others.^11–13

This review discusses the recent progress in the smartphone-based portable biosensor technology designed to detect DNA and RNA, and highlights long-standing limitations and related challenges of this technology. These biosensors with optical imaging devices are always promising as inexpensive and hypersensitive diagnostic tools in the absence of good analytical instruments or sufficient resources. Herein, three types of smartphone-based optical imaging biosensors are discussed according to the detection technique: smartphone-based colorimetric biosensor,¹⁴ smartphone-based fluorescence biosensor,¹⁵ and smartphone-based microscopic imaging biosensor (see Figure 1). The targets of these detection techniques include the different disease biomarkers of infectious diseases,¹⁶ hereditary diseases,¹⁷ and cancers.¹⁸ It must be noted that these three types do not fully cover all the existing types of smartphone-based biosensors, other types of smartphone-based biosensors such as electrochemical sensors and multimode-sensing sensors are also applied to the field of genetic testing; however, the review mainly considers the common optical biosensing applications of these three types as per the existing literature. In addition to optical sensing and imaging biosensors, non-optical biosensors (electrochemical biosensors, electrochemiluminescence sensors, and others) are also being developed for genetic testing, although they are not the focus of this review.

FIGURE 1. Summary of the representative achievements of smartphone-based optical imaging biosensor in three different applications. The applications cover detections of infectious diseases, hereditary diseases, and cancers.16–18 Reproduced with permission from Ref. 19 Copyright 2018, Elsevier. Reproduced with permission from Ref. 20 Copyright 2022, Elsevier. Reproduced with permission from Ref. 21 Copyright 2021, American Institute of Chemical Engineers. Reproduced with permission from Ref. 22 Copyright 2017, Springer, Nature. Reproduced with permission from Ref. 23 Copyright 2011, Elsevier. Reproduced with permission from Ref. 24 Copyright 2017, American Chemical Society. Reproduced with permission from Ref. 25 Copyright 2020, Elsevier. Reproduced with permission from Ref. 26 Copyright 2018, Elsevier. Reproduced with permission from Ref. 27 Copyright 2019, American Chemical Society.

SMARTPHONE-BASED OPTICAL IMAGING BIOSENSORS

Smartphone-based optical biosensors fully utilize the many advantages of smartphones. In this section, three different methods of smartphone-based optical imaging biosensors, including smartphone-based colorimetric detection, smartphone-based fluorescence detection, and smartphone-based microscopic imaging detection, are discussed.

Smartphone-based colorimetric detection

Colorimetry is a method that utilizes biological or chemical reactions to measure changes in the absorbance or reflectance of analytical reagent complexes. By measuring the amount of light passing through a solution at a specific wavelength, the concentration of the solution can be determined. Of all analytical instruments that exploit smartphones as a detector, colorimetric method-based instruments are the most promising. Colorimetry is most frequently used in the analytical and biochemical fields. The combination of colorimetric methods with smartphone-based imaging devices has enabled the development of POCT systems (Figure 2A). In addition, the quantification of color-based reactions can be easily completed without any extra instrument, using the powerful image processing and computational abilities of smartphones. Colorimetry is a widely utilized technique in smartphone-based genetic testing due to its low cost and convenience. In some studies, the content of biomarkers in reactants is analyzed by comparing changes in color.^25,28,30 With the development of technology, there are a variety of color digitization software based on mobile phones to distinguish color intensity. Such software is generally based on Red-Green-Blue (RGB) analysis, which utilizes color decomposition to sort colors into its R, G, and B components, has been widely used in color processing. Color decomposition is quantified into its R, G, and B components, and appropriate channels are selected for quantitative analysis of color intensity.^19,31 Nguyen et al. developed a smartphone-integrated gene analyzer based on the colorimetric method.³¹ By drawing the ratio of green/red channel intensity and reaction time, a quantitative loop-mediated isothermal amplification (LAMP) spectrum could be obtained, and 101 copies/μL of Escherichia coli O157:H7 could be detected within 60 minutes. More simply and quickly, researchers use gray value and H (Hue) value for quantitative analysis.^27,32

FIGURE 2. (A) Schematic illustration of smartphone-based colorimetric portable sensor for cell-free DNA detection. Reproduced with permission from Ref. 28 Copyright 2022, American Chemical Society. (B) Diagram of smartphone-based FRET (fluorescence resonance energy transfer) emission to detect multiplexed proteolytic activity assay. Reproduced with permission from Ref. 29 Copyright 2014, American Chemical Society. (C) Schematic representation of the device and detection principle of multi-module smartphone microscopy device. Reproduced with permission from Ref. 22 Copyright 2017, Springer Nature.

The use of single-wavelength light-emitting diode (LED) as an excitation light source is no longer sufficient for scenarios that require the detection of multiple analytes with different absorbances. Luo et al. designed a smartphone-based colorimetric device with multi-wavelength for the quantitative analysis of different liquids. The device was fabricated using plastic three-dimensional (3D) printing and six LEDs with various wavelengths as light sources mounted on a turntable. Six different kinds of common analytes selected and tested to verify the performance of the device demonstrated that their highly stable and accurate device met the conditions of POCT.³³

Although the approach of the smartphone-based colorimetric detection has been developed in various fields owing to their ease of operation, portability, and low cost, fluctuations in ambient light and the uncontrollable factor of field environment are major limitations for the application of smartphone-based colorimetric biosensor for POCT.³⁴ In other applications, external light sources are employed to eliminate detection errors caused by changes in natural light.^35,36 Additionally, a smartphone detection platform based on machine learning has been proposed. It utilizes machine learning classifiers to form a more robust and adaptive platform that can detect glucose independent of various light exposures.³⁷ These methods are effective at inhibiting the effects of ambient light and can be used in the field of POCT.

Smartphone-based fluorescence detection

Fluorescence detection analysis is a widely used and promising method in point-of-care (POC) detection due to its strong specificity, sensitivity, and high efficiency.³⁸ However, smartphone-based fluorescence biosensors are independent devices and require the support of other miniaturized systems, such as an excitation light source, filter module, imaging module, and portable power supply.^39,40 Previous studies have commonly relied on observing fluorescence intensity to achieve gene detection and quantitative analysis.^41–43 Rajendran et al. developed an independent convective PCR (cPCR) device connected to a smartphone, which uses natural convective heating to rapidly detect nucleic acid and realizes multiple amplification of DNA through simple hardware control and wireless communication through Bluetooth technology.⁴¹ A detection limit of 2.8 × 10³ lambda DNA fragments was obtained when a smartphone-based fluorescent reader was used for quantification. The device is simple and ingenious, but can only detect a single target at a time.

Microfluidic chips that utilize the high sensitivity and quantitative ability of fluorescence signals have been shown to have great potential for simultaneously multiple targets detection.^24,44,45 Chen et al. developed a platform for multipath analysis of disease-specific DNA sequences.²⁴ LAMP reagents were pre-deposited into different channels of microfluidic chips. When exposed to the target nucleic acid sequence of test samples, fluorescent products were produced. Software applications running on the smartphone's microprocessor will display and automatically analyze. The system is capable of detecting multiple nucleic acid targets simultaneously, which enables the identification of co-infection of multiple pathogen strains, and generate positive/negative tests for the presence of specific pathogens by combining experimental controls and repetitions. Zhou et al. reported a cost-effective and portable smartphone-based fluorogenic sensor to detect different foodborne pathogens,⁴⁴ including E. coli O157:H7. This sensor consists of using a microfluidic chip combined with a LAMP device that can maintain the temperature at 65°C. Besides, an android application was used to obtain the RGB values from the fluorescence signals. These results show that the smartphone-based fluorogenic sensor can be successfully applied to the POCT of foodborne pathogens with a detection limit of 2.8 × 10⁻⁵ ng/μL.

Fluorescence resonance energy transfer (FRET) is a type of fluorescence analysis that involves energy transfer between two fluorescent groups that are in close proximity. In this phenomenon, energy is transferred from the donor fluorophore to the acceptor fluorophore within a few Angstroms. Recently, smartphone-based FRET optical sensors have become popular in biotechnology and analytical testing due to their inherent photostability.^46–50 Selecting a FRET donor is crucial for an excellent smartphone-based optical sensing sensor. For FRET biosensors, organic fluorescent dyes are the most commonly used donors; however, most organic dyes have narrow excitation spectra, broad emission spectra, and susceptibility to photobleaching and photolysis, limiting their widespread application. Quantum dots (QDs), a novel type of fluorescent donor, have become the preferred choice of researchers owing to their narrow excitation fluorescence half-width, relatively broad wavelength-tunable emission spectrum, and excellent optical stability. Petryayeva and Algar developed a smartphone-based FRET platform combined with QDs and performed multiplexed proteolytic activity assay. In this study, three kinds of emitting QDs (QD450, QD540, and QD625) were conjugated with peptides that assembled with different acceptors (A647, QSY9, and QSY35), respectively, to generate FRET pairs.²⁹ QDs have some disadvantages, such as requiring multiple synthesis steps, having unstable structures, difficulty in storage, relatively high toxicity, and low fluorescence yield, suggesting the need of further improvements. Smartphone-based FRET sensors based on carbon QDs have drawn much attention owing to their exceptional fluorescence performance. Furthermore, Ren et al. devised a new strategy based on carbon QDs to detect amino-modified cDNA. The single excitation wavelength of 365 nm changed the imaging of the nanoprobe system from bright blue to colorless due to effective triggering by target bacteria, resulting in a fluorescent response to Vibrio parahaemolyticus. Ren et al. employed a paper sensor with a smartphone to detect V. parahaemolyticus. Furthermore, the study results showed that the detection limits of V. parahaemolyticus in pure culture and artificially contaminated solution were 8.9 and 6.1 × 10 cfu/mL, respectively.⁵¹

Smartphone-based microscopic imaging detection

With the high-resolution imaging capability of semiconductor imaging sensors, instead of traditional computers, smartphone-based optical image sensing, particularly optical microscopic imaging applications, has become the most common method for biological analysis (Figure 2C). Smartphone-based optical microscopes can detect the size, shape, brightness, color, and other morphological characteristics of biological samples using different external accessories, such as optical components, mechanical construction, and LED light sources.^22,52–54 These devices can be used for the detection of causative agents, cancer cells, DNA, RNA, viruses, and other substances.^55–58 Wei et al. developed a compact and portable smartphone-based microscopic imaging biosensor with high-power excitation and thin-film interference filter for detecting various bacteria and viruses. The performance of this device was tested by utilizing fluorescent particles with a diameter of 100 nm and fluorescently labeled human cytomegalovirus.⁵⁹ A highly sensitive detection was achieved by designing an additional thin-film filter to reduce background noise. Wei et al. subsequently proposed an improved method to increase the sensitivity of the smartphone-based microscopic imaging biosensor using a thin metal-based film with surface-enhanced fluorescence, which could realize single 50-nm fluorescent particle imaging and single QD imaging.⁶⁰ Chung et al. recently reported a noteworthy advance in smartphone-based microscopic imaging biosensors. They combined smartphone-based microscopic imaging with microfluidic paper analytic devices (μPADs).⁶¹ Noroviruses were detected directly by μPADs fabricated using ColorQube wax printer. The quantitative analysis of the intact norovirus samples obtained from different fields was performed using a smartphone-based fluorescence microscope and self-developed phone application. This biosensor enabled the detection at the level of individual virus particles. Compared with other methods of microscopic bioimaging, fluorescent bioimaging offers higher sensitivity and specificity, enabling better detection of biological targets (even at the nanoscale level).

APPLICATIONS IN GENETIC TESTING

Smartphone-based optical imaging biosensor, as an important tool in genetic testing, is used for the rapid detection of multiple biomarkers or portable identification of pathogens.^{28,40,43,62–64} In recent years, a trend toward “telemedicine,” in which medical testing equipment acts as a remote detection terminal and sends output result or diagnostic data directly to the medical institution or specialist for further investigation, has become increasingly apparent.^65–67 In contrast to traditional genetic testing devices that rely on expensive and complex equipment requiring technical expertise,^68–71 smartphones with the advantages of numerous different sensors, high-resolution cameras, small size, data processing ability, and connectivity have become dominant in resource-limited settings and found their way into the field of POCT.^{41,45,47,52,72} Various smartphone-based optical imaging detection platforms have been applied in many genetic testing scenarios.^46,73–77 According to the different principles of genetic testing applications, the major applications of smartphone-based optical imaging biosensors can be divided into the following three types: infectious disease, hereditary disease, and cancer.

Infectious diseases

The prevalence and spread of infectious diseases (acquired immune deficiency syndrome, influenza, severe acute respiratory syndrome [SARS], and others) have caused a huge impact on human society.^78–80 These diseases not only endanger human health but also cause panic in the society and chaos in people's daily lives. The outbreak of the novel coronavirus (coronavirus disease 2019 [COVID-19]) caused huge economic losses and public health problems.^66,67,81 Researchers have become increasingly aware that the rapid diagnosis of infectious diseases and accurate identification of their etiology are of great significance for preventing the early causative agent transmission and outbreaks of infectious diseases.^82,83 However, the traditional molecular detection equipment for infectious pathogens is often heavy and expensive, thereby limiting on-site detection application. To overcome this difficulty, simple, inexpensive, and portable testing platforms based on the smartphone optical imaging technology have been developed for diagnosing mainly infectious diseases.^{27,43,55,57,58,61,84–86}

Avian influenza, a type of respiratory infectious disease, causes serious harm to human health.^87,88 For detecting the avian influenza virus (AIV), Yeo et al. developed a smartphone-based rapid fluorescent diagnostic system (SRFDS) to detect H5N1 virus in human throat samples.⁸⁹ The system uses fluorescent coumarin-derived dendrimer-based bioconjugation and LED modules, they have also developed a highly sensitive and rapid diagnostic method for H9N2 viruses carried by poultry.⁹⁰ Their research was limited to the imaging performance of smartphones at the time. Kim et al. utilized near-infrared (NIR)-to-NIR up-conversion nanoparticles (UCNPs) to create a sensor that detects AIV nucleoproteins (NPs) from clinical samples within 20 minutes.⁹¹ Xia et al. have developed a smartphone-based nanomaterial colorimetric assay for highly sensitive and selective detection of avian influenza virus.⁹² The colorimetric reactions are based on on-chip gold nanoparticles and able to achieve visual detection of the virus. However, this method is slightly more costly and time-consuming. Jiang et al.’s study has more advantages in this aspect, they evaluated a fast multiplex detection method for tracing AIV DNA biomarkers based on the concept of colorimetric detection and developed a one-step quick unlabeled imaging array.¹⁶ They included three subtypes of AIV DNA biomarkers (H1N1, H7N9, and H5N1) as target models and designed a series of specific catalytic hairpin assembly amplification responses that were capable of efficient amplification. The amplification of DNA double-helix produces a 3′ overhang of the G-quadruplex structure, which can selectively and strongly incorporate a specific fluorescence probe thioflavin T (ThT) (Figure 3A). ThT emits fluorescence and colorimetric signals whose signal intensity is strong enough to be detected by smartphones. All fluorescent images of the array were captured and analyzed on Huawei Nova 4e (Figure 3B). This method could simultaneously identify and detect three DNA biomarker models in a small-volume solution (50 mL) within 20 minutes (Figure 3C,D). Combined with the output fluorescence images and grayscale analysis, the detection limits of the imaging array for H1N1, H7N9, and H5N1 were 136, 141, and 129 pM, respectively. In the study, the researchers also verified the mismatch recognition ability and anti-interference ability of the detection system.

FIGURE 3. (A) Catalytic hairpin assembly amplification and ThT fluorescence lighting of AIV DNA biomarker. (B) Schematic diagram of the AIV multichannel detection method for fluorescence array imaging and grayscale analysis. (C) Grayscale image. (D) Corresponding gray level analysis. C1–C8 represent a single target DNA or a mixture of target DNA. The concentration of each target was 50 nM. Reproduced with permission from Ref. 16 Copyright 2021, Elsevier.

Hepatitis B, which has a major impact on the human liver, is an infectious disease with high incidence.^93,94 For detecting the hepatitis B virus (HBV), some studies have proposed portable detection methods to overcome the limitations of expensive and bulky detection equipment.^95–99 Giavazzi et al. conducted an earlier study on the use of smartphones for hepatitis B diagnosis.¹⁰⁰ Jiang et al. used smartphones to image and analyze genotype-specific monoclonal antibody (mAb) functionalized lateral flow strips for HBV genotyping, with the development of technology.¹⁰¹ Draz et al. developed a nanoparticle-enabled smartphone (NES) system using convolutional neural network (CNN) technology for simple and sensitive HBV virus detection.¹⁰² Xie et al. designed a self-driven microfluidic chip that utilizes LAMP technology and cell phone camera function to perform multiple detections of HBV.¹⁰³ However, this method is somewhat complex in terms of sample preparation. Tao et al. used probe DNA to regulate the catalytic behavior of mxene-probe DNA-ag/Pt nanohybrids and established, for the first time, a CRISPR-Cas12a-based colorimetric biosensor for targeted HBV detection.¹⁰⁴ By comparison, Li et al. provided a highly sensitive, augmentation-free solution by designing a smartphone-based, single-molecule microscopic imaging HBV detection technology.¹⁰⁵ The technology functioned in the following three steps. First, pretreated serum sample was placed on a paper nucleic acid extraction card to bind HBV DNA to the card surface, and the viral DNA was extracted via buffer treatment. Second, the reporter probe and capture probe labeled with QD fluorescent microspheres were hybridized with the obtained viral DNA to form an HBV detection complex ([reporter probe]–[HBV DNA]–[capture probe]). Finally, the complex was immobilized and fluorescently imaged using a portable, smartphone-based monomolecular imager, thereby enabling the amplification-free detection of individual HBV target DNA and determination and quantification of HBV viral load from the fluorescence signal. It can reach high sensitivity as LoD of 100 aM verified by the synthesized HBV target DNA and 10⁴ copy/mL (≈2000 IU/mL) verified by the clinical samples. This technology not only performs the rapid and highly sensitive detection of HBV but also greatly reduces the detection cost, thereby realizing the advantages of telemedicine using smartphone real-time communication.

COVID-19, a respiratory infectious disease, is ravaging the world, and many researchers have made contributions to the development of the portable detection of SARS coronavirus 2 (SARS-CoV-2).⁸³ Zhang et al. reported the development and performance benchmarking of an inexpensive (approximately $0.30 per test) colorimetric detection method for rapid (within 30 minutes from sample to answer) testing of SARS-CoV-2 variants.¹⁰⁶ Fozouni et al. reported a non-amplification CRISPR-Cas13a detection method for direct detection of SARS-CoV-2 from nasal swab RNA, which can be read using a smartphone microscope.¹⁰⁷ Breshears et al. developed a highly sensitive and low-cost immuno-fluorescent particle assay by using a paper-based microfluidic chip and smartphone-based fluorescent microscope that can detect SARS-CoV-2 from clinical saline gargle samples.¹⁰⁸ These methods are all characterized by low cost and rapid detection. As more people join SARS-CoV-2-related research, new methods are gradually emerging. Nguyen et al. detected the fluorescence signal of amplified amplicons from RT-PCR by using the complementary metal-oxide-semiconductor (CMOS) camera of a smartphone. They achieved the detection of SARS-CoV-2 biomarkers.¹⁰⁹ Ma et al. developed a CRISPR-Cas12a-driven visual biosensor with smartphone readout. The nucleic acid of SARS-CoV-2 triggers arbitrary degradation of single-stranded DNA based on CRISPR-Cas12a and disperses gold nanoparticles, resulting in observable color changes.¹¹⁰ Song et al. reported a colorimetric DNAzyme reaction triggered by LAMP with clustered regularly interspaced short palindromic repeats (CRISPR [DAMPR]) assay.⁸¹ This method combines LAMP, CRISPR, and other technologies and the use of 3D printing technology to produce a portable nucleic acid detection method based on smartphones to rapidly detect SARS-CoV-2 and its variants within 1 h. Likewise, Ning et al. reported a CRISPR–Cas12a assay specifically designed to detect SARS-CoV-2 and its variants for diagnosis.²⁰ This assay identified the single nucleotide polymorphisms of VOC by detecting mutations in the spike gene CRISPR PAM motif. This method exhibited good results and can potentially be used to track SARS-CoV-2 emerging variants.

Hereditary diseases

Genetic diseases such as congenital foolishness, congenital deafness, hemophilia, and others are usually caused by changes in genetic factors or are controlled by pathogenic genes. At present, there is no effective treatment for most genetic diseases.^111,112 Therefore, the prenatal diagnosis and prevention of genetic diseases are of special significance. Genetic testing is an important tool in the prevention and diagnosis of genetic diseases.¹¹³ However, traditional diagnostic techniques rely on resource-intensive equipment and skilled technicians and are difficult to perform in resource-limited regions. To overcome these shortcomings, some researchers developed low-cost and portable detection equipment, which can be used in the genetic diagnosis of certain genetic diseases (congenital deafness, cystic fibrosis, and others).

For detecting mutated genes in hereditary deafness, Zhang et al. developed a smartphone-based microarray decoding platform (Figure 4A).²³ In terms of sample processing, symmetric PCR and magnetic separation technologies were used to obtain a large amount of single-strand DNA (ssDNA), which were bound to magnetic beads coated with streptavidin. Next, the attached ssDNA was hybridized with the marker array for genotyping. The processed sample data were then entered into a smartphone-based microarray decoding platform. Inside the system, white LED and customized aperture are used to produce uniform sample illumination. The comparison results can be easily captured via a macro-lens combined with a smartphone camera to record the genotyping result images (Figure 4B). The smartphone is equipped with a special genotyping microarray decoding software, which can decipher and analyze the recorded pictures and generate genotyping reports. With the wireless data transmission function of the smartphone, genotyping reports can be easily emailed to remote doctors or designated hospitals through the network. Using this technology, Zhang et al. collected and analyzed a total of 51 samples of patients with hereditary deafness. For a DNA detection limit of 1 ng, nine mutation genes of hereditary hearing loss were genotyped simultaneously, which were verified using fluorescence confocal imaging.

FIGURE 4. Smartphone-based microarray decoding system. (A) An imaging adapter comprising an LED lighting source, diaphragm, macro-lens, and smartphone. (B) Picture of the imaging system for image acquisition on the slide (left). The collected pictures are automatically analyzed using the self-compiled special software, and the genotyping report is produced (middle). The user interface of the mobile phone analysis software (right). With the wireless data transmission function of smartphones, genotyping reports can be easily emailed to remote doctors or designated hospitals via the network. Reproduced with permission from Ref. 23 Copyright 2011, Elsevier.

Cystic fibrosis is an autosomal recessive disease. In most patients with cystic fibrosis, phenylalanine at the 508 position (ΔF508) in the cystic fibrosis transmembrane conductance regulator protein (CFTR) gene is deleted, which is clinically manifested as a viscous secretion in the lung. Malhotra et al. developed a FRET-based assay for oligonucleotides that can be used as cystic fibrosis indicators to distinguish wild-type sequences from mutant ones.¹¹⁴ They used a paper substrate as a platform for nucleic acid detection and selected green QDs (gQDs) and Cy3 dye-labeled oligonucleotides as donor–receptor pairs (gQD–Cy3) in the FRET process. The hybridization of oligonucleotide complementary chains enables the molecular fluorophore to approach QD and emit fluorescence through FRET. The resultant fluorescence image is then captured by the smartphone camera for subsequent data processing, and the CFTR gene mutation (ΔF508) is identified using color contrast.

Cancers

Cancer, called a “terminal disease” in some regions, remains one of the most fearful diseases with a high mortality rate,¹¹⁵ which can have a great impact on people's lives. The accurate and rapid diagnosis and staging of cancer are crucial in effective cancer treatment and survival rate improvement.^116,117 Differences in specific nucleic acid levels between patients with cancer and healthy individuals can be used as a potential biomarker.^118,119 The application and importance of the nucleic acid detection technology in cancer diagnosis are growing.^120–122 However, traditional nucleic acid detection methods are expensive and need costly equipment, which are not conducive to their popularization and application. Therefore, researchers have developed some low-cost and portable nucleic acid detection technologies based on smartphone optical imaging for detecting certain cancer biomarkers.^{28,42,48,62,123}

Prostate cancer is a malignant hyperplasia of the epithelial cells of the prostate gland; it is common in middle-aged and elderly men. Some reported research has made efforts in smartphone-based prostate cancer detection. Lv et al. coupled CuxS nanocrystals with a portable infrared thermal imager on a smartphone, and used rolling circle amplification (RCA) technology to form CuxS nanocrystal connections. They achieved visual quantitative detection of prostate cancer biomarkers.¹²⁴ Amin et al. used 3D printing to manufacture a smartphone device and applied magnetic focusing technology to achieve the separation and detection of various cancer cells, including prostate cancer cells.¹²⁵ Barbosa et al. proposed a flexible detection system (MCFphone) for portable colorimetric and fluorescence quantitative sandwich immunoassay detection of prostate-specific antigen (PSA) with a detection limit as low as 0.08 ng/mL.¹²⁶ Chen et al. introduced a highly sensitive and simple microbubble digital detection readout method, which only requires smartphone bright-field imaging. By combining with machine learning algorithms, it can monitor PSA after prostatectomy, with a detection limit of 2.1 fM (0.060 pg/mL).¹²⁷ Prostate cancer antigen 3 (PCA3) is an overexpressed tumor marker in prostate cancer. Wang et al. studied the detection of PCA3.²⁵ They developed a low-cost portable nucleic acid detection platform prepared via 3D printing. The device comprised a reverse transcription-LAMP (RT-LAMP) chip, thermal module, and imaging module (Figure 5A,B). The sample and RT-LAMP reagent were loaded on the PVA amplification pad on the chip, and the sealed chip was heated and incubated on the hot module of the device. The RT-LAMP-amplified products were injected into the colorimetric detection section through a channel on a pre-calcined dry paper (Figure 5C). A smartphone imaging analysis was used to detect colorimetric changes on the disc. First, RT-LAMP amplification primers for PCA3 were designed, and then the RNA samples of different cell lines (such as A549, MCF-7, LNCap, A549 + LNCap, and MCF-7 + LNCap) were added into RT-LAMP reagent for amplification (Figure 5D). RT-LAMP products on A549 and MCF-7 tablets did not lead to any color change, but the samples containing LNCap showed a color change from orange to bright yellow, indicating that the latter exhibited PCA3 gene expression and that the platform had good specificity. They then used different concentrations of LNCap RNA to verify the sensitivity of the system. The colorimetric results are shown in Figure 5E. When the sample concentration is less than 0.0001 pg/μL, the naked eye cannot identify color change, suggesting that it has a good detection limit.

FIGURE 5. Device, process, and results of PCA gene testing. (A) The detection device comprises a light-guiding unit and heating unit. (B) Device structure obtained via 3D printing. (C) Transfer of solution between the amplification and detection regions. (D) PCA3 gene specificity test results: samples 1–5 were A549, MCF-7, LNCap, A549 + LNCap, and MCF-7 + LNCap RNA. (E) Adding samples of different concentrations to verify the results of the sensitivity of the genetic detection device. Reproduced with permission from Ref. 25 Copyright 2020, Elsevier.

Breast cancer is one of the major diseases that affect women's health. The convenient and quick diagnosis method of breast cancer is the focus of extensive research.^48,128 Some researchers have made contributions to smartphone-based breast cancer diagnosis. Tewary et al. created a universal 3D printing adapter that connects a smartphone to the eyepiece of a traditional microscope to obtain microscopic images of stained tissue slices for automated Ki-67 quantification in breast cancer assessment.¹²⁹ Prasad et al. described a portable fluorescence microarray imaging system connected to a smartphone for detecting breast cancer gene expression in exon 11 (BRCA-1).¹²⁸ Tran et al. proposed a smartphone-based imaging platform (SIP) coupled with a magnetic fluorescent super-nanoparticle component, which are magnetic iron oxide nanoparticles densely surrounded by many bright luminescent semiconductor QDs corona.¹³⁰ The combination of SIP and magnetic iron oxide nanoparticles allows for selective separation, fluorescence immunolabeling, and counting of human epidermal growth factor receptor 2 (HER2)-positive breast cancer cells. Compared to other studies, the work of Joh et al. is more likely to be applied in actual breast cancer diagnosis. They developed EpiView-D4, a smartphone-based multimodal detection platform for the immediate evaluation of breast tumor cytology and HER2 molecular expression.¹³¹ D4 is a polymer-based brush-based immunodiagnostic chip that contains location-based capture antibody (cAb) and fluorescently labeled soluble detection antibody (dAb) antibody pairs, spots of immobile cAb, and excess “soluble” fluorescently labeled dAb are printed directly onto POEGMA-coated glass, resulting in the dissolution of dAb when droplets of sample cell lysate droplets are deposited on the chip. If HER2 is present, the antibody pair will bind to form an antibody “sandwich,” which can be detected using fluorescence imaging to confirm HER2 presence. EpiView is essentially a smartphone-based customized modular, multifunctional microscope that can achieve bright-field and fluorescent imaging for pathological tissue imaging and D4 detection and readout. In the study, puncture sampling was performed on human breast cancer cells, and the samples were directly introduced into EpiView for bright-field imaging to directly observe pathological tissues. They also lysed cancer cell samples, applied the lysed products to the D4 chip, and then performed fluorescence imaging using EpiView. The concentration of HER2 was determined by measuring the fluorescence intensity.

CONCLUSION AND FUTURE OUTLOOK

In the past few years, the development of smartphone-based optical imaging biosensors has provided several portable, efficient, low-cost, and easy-to-operate devices for POCT application and human healthcare. Toward this, the present article summarized the most recent progress of three typical smartphone-based optical imaging biosensing systems. Although all aspects of the technology have not been completely covered, other types of smartphone-based biosensors, such as electrochemical-based sensors, also have a lot of research and applications in the field of genetic testing,^132–135 and these types as described in the second verse are smartphone-based biosensors that employ the optical imaging method in genetic testing.

Many developments have occurred within the past few decades in the field of smartphone-based POCT. The integration of biomedicine and individual disciplines has brought new detection methods that are more convenient, more economical, and easier to use than traditional laboratory-based detection methods. However, many challenges regarding the design and assembly of the smartphone-based optical imaging biosensors remain, which need to be overcome. Major challenges are as follows: (i) fluctuations in ambient light may have a significant impact on the analysis results of these biosensors. In this case, external light sources are used to eliminate detection errors caused by changes in natural light.^35,36 Alternatively, trained machine learning classifiers are used to form a more robust and adaptive platform.³⁷ (ii) The structure of these biosensors may not be suitable for different smartphones. Small three-dimensional adjustment setups can be designed and installed to improve the adaptability of these biosensors to different smartphones, which can resolve the difference of them manifested in the size, shape, and location of the lens module.

The future development of smartphone-based optical imaging biosensors for genetic testing will focus on the following aspects: (a) improving the performance of the smartphone camera that decides the sensitivity of the optical imaging biosensors; (b) combining the smartphone-based biosensor with other advanced biochemical analysis techniques, for example, surface-enhanced Raman scattering (SERS) technology,¹³⁶ surface plasmon resonance technology,¹³⁷ and others; Lee et al. reported a method for quantitative analysis of respiratory bacterial DNA paper-based SERS biosensor. It can be a potential way for POCT application by integrating the smartphone.¹³⁸ (c) Using the modern image processing technology to resolve low-image resolution. In this context, deep learning provides an effective tool to enhance the performance of smartphone-based optical imaging biosensor.¹³⁹ In recent years, the availability of massive data and high-speed computing systems has led to the rapid development of deep learning techniques. Algorithms can be developed from scratch by using publicly available datasets, or using transfer-learned pretrained networks such as AlexNet, Vgg-16, and UNet.¹⁴⁰ These techniques have been proven to improve the image resolution of various microscopic techniques and realize the automatic classification of diseases. Moreover, the application of deep learning in smartphone-based biosensing has led to the development of portable devices, which shows a promising future for the field of genetic testing.

In general, thanks to the widespread use of smartphones in people's daily lives, drastic advances in the smartphone-based biosensing technology, and thorough tapping of the potential capacity of smartphone-based biosensors, a revolutionary change can be expected in the existing fields of diagnostic and healthcare systems, biological population evolution, food safety, and public health.

AUTHOR CONTRIBUTIONS

This paper was written by Yue Wu, Haotian Zong, and Yunshan Zhang. Diming Zhang guided the structural arrangement for this paper. All revisions were discussed by Jing Ye, Zhongyuan Xu, and Wenjian Yang. All authors read and approved the final manuscript.

ACKNOWLEDGMENTS

The work was financially supported by the Research Project of Zhejiang Lab (grant number 2022MH0AC01) and Postdoctoral research project of Zhejiang Lab (grant number 113009-UA2007QJ).

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.