Study on the improvement of image interpretation skill of non-radiology residents based on an image case database in gastroenterology department

Abstract

Section Background

Medical imaging plays a very important role in the diagnosis, treatment and prognosis of digestive diseases. This manuscript aimed to explore the value of a teaching model based on an image case database in improving the image interpretation skills of non-radiology residents training in gastroenterology department.

AbstractSection Methods

From October 2023 to March 2024, all the residents training in the gastroenterology department were selected as the research subjects. Using the implementation date of the teaching model based on an image case database as the cutoff point, the experimental group comprised 27 residents (January–March 2024), while the control group included 24 (October–December 2023). The experimental group learned by using a teaching mode based on an image case database, whereas the control group learned by using conventional teaching methods. The effects of the two teaching modes were evaluated through the examination of admission, mid-term and exit performance scores, and questionnaires were used to feedback the evaluation of the residents of the experimental group on the teaching mode of the case database.

AbstractSection Results

The scores of non-radiology residents in the experimental group and the control group showed no difference in the admission examination ( P = 0.728), but the scores in the experimental group were significantly higher than those in the control group at the mid-term (P = 0.006) and exit examination (P = 0.002), respectively. In the experimental group, 100% (27/27) of the residents believed that radiology was very important for the mastery of clinical knowledge in gastroenterology, 88.9% (24/27) of these residents believed that the teaching mode significantly improved their image interpretation skills and interest in using images to learn, and 81.5% (22/27) were very willing to accept the teaching mode of image case database in other clinical departments.

AbstractSection Conclusion

The teaching mode based on an image case database could improve the image interpretation skills of non-radiology residents training in the gastroenterology department, and could be popularized in more departments for residents training.

Background

Digestive system includes liver, gallbladder, pancreas, spleen, esophagus, stomach, intestine and other organs. Accurate diagnosis of diseases in various organs requires the aid of imaging, so it is extremely important for non-radiology residents to fully master the imaging characteristics of the digestive system for their clinical work [1, 2]. In the traditional residency training, the non-radiology residents usually receive training in the radiology department for one month, during which they concentrate on learning the the imaging characteristics of common diseases in the central nervous system, respiratory system, musculoskeletal system, circulatory system and digestive system [3, 4]. Clearly, the allotted time for learning about the imaging characteristics of the digestive system is constrained, making it challenging for residents to thoroughly grasp the necessary imaging knowledge within a limited period. Furthermore, they often find it difficult to seamlessly integrate this knowledge with practical clinical cases [5]. If the non-radiology residents acquire knowledge of the related imaging characteristics of disease during the training in gastroenterology department, it will improve their skills to interpret image and enhance their clinical post competence [6, 7].

The application of case-based learning (CBL) has evolved with the times and can be effectively integrated with teaching methods such as problem-based learning (PBL) and lecture-based learning (LBL) [8]. CBL guides participants to analyze and discuss through real or simulated cases, strengthening their problem-solving abilities and emphasizing the central role of participants. The authenticity of the cases further stimulates students’ interest and motivation [9]. PBL enhances participants’ comprehensive abilities and innovative thinking through questioning, discussion, inquiry, and collaborative problem-solving [10]. However, these two teaching methods are more suitable for participants with a wide range of basic knowledge and require substantial preparation time [9, 10]. In contrast, LBL focuses on building a systematic knowledge framework through lectures, which is beneficial for basic knowledge accumulation but may limit participants’ initiative and creativity [11]. Given that residents have busy clinical schedules while also needing to study, integrating the strengths of these teaching methods and utilizing their fragmented time effectively could enhance their learning outcomes.

Combining the advantages of the above teaching methods, we have summarized a teaching model based on five sections, including sub-specialty (S), problem-based learning (P), assessment (A), report (R), and reading skill (K), named as “SPARK” [12]. We have developed a corresponding software platform based on the SPARK teaching model with more than 50,000 cases. This case database is a commonwealth image case database collaboratively developed by more than 40 radiology departments in China, and is accessible through a dedicated mobile application (APP) named SPARK（Fig. 1). The SPARK platform includes five learning sections: “S”, “P”, “A”, “R”and “K” [13]. Anyone can install the APP on their mobile devices and engage in repeated learning anytime, anywhere. Meanwhile, teachers can observe students’ progress in real-time via the platform backend and assess their learning effectiveness, thereby achieving more personalized and efficient teaching management. In our previous research, we used the image case database in the practical teaching of clinical undergraduates to improve their imaging interpretation skills. As a result, the participants’ image interpretation skills have been greatly improved after using our teaching model [12, 13].

On the basis of previous research, this study compares the teaching effectiveness of a case-based teaching method and conventional teaching method in the gastroenterology department, aiming at providing new ideas and methods for non-radiology residents to learn the imaging characteristics of diseases and improve their image interpretation skills.

[IMAGE OMITTED: SEE PDF]

Materials and methods

Participants

From October 2023 to March 2024, all of the non-radiology residents training in the gastroenterology department, Hangzhou First People’s Hospital affiliated to West Lake University School of Medicine were selected as the research subjects. This study was approved by the hospital ethics committee (ethical review approval No. KY-20230905-0201-01, Clinical trial number: not applicable), and all the participating residents signed an informed consent. Using the PASS software (version 15) for sample size estimation with a power of 0.8, a significance level (α) of 0.05, an mean difference (δ) of 10.0, and a standard deviation (σ) of 12.0, it was determined that at least 46 subjects need to be included. Meanwhile, using the implementation date of the teaching model based on an image case database as the cutoff point (January 2024), the residents training from January to March 2024 were all taken as the experimental group, while those training from October to December 2023 were all taken as the control group. The number of residents in the experimental group and the control group were 27 and 24 respectively. There was no statistical difference in gender, grade and academic degree between the residents in the experimental group and the control group (P > 0.05), as shown in Table 1.

[IMAGE OMITTED: SEE PDF]

Inclusion and exclusion criteria

Inclusion criteria: (1) All the residents were non-medical imaging residents, including internal medicine, surgery, pediatrics, obstetrics and gynecology, anesthesiology, otorhinolaryngology, etc.; (2)The study time in gastroenterology department was one month; (3) All the residents had no other experience being taught medical imaging, except for university study of medical imaging and clinical teaching materials; (4)The assessment at each stage was finished; (5) The completion rate of the case database learning task in the experimental group exceeded 90%. Exclusion criteria: (1) The residents majored in medical imaging including radiology, nuclear medicine and ultrasound; (2) The residents majored in gastroenterology; (3) The completion rate of the case database learning task is less than 90%; (4) The residents failed to complete any stage of the assessment (Fig. 2).

[IMAGE OMITTED: SEE PDF]

Study design

Teaching methods of the experimental group

The teaching mode based on an image case database was implemented in the experimental group of non-radiology residents. According to the “Standardized Training Contents and Standards for Residents”, non-radiology residents who trained in the gastroenterology department for one month from January to March 2024 were included in the experimental group. The diseases that residents training in gastroenterology department need to master were peptic ulcer, posthepatitic cirrhosis, primary liver cancer, esophageal cancer, colon cancer, gastric cancer, acute and chronic pancreatitis and inflammatory bowel disease [14]. For each disease, 20 cases were selected from the case database as the learning task, assigned one week in advance. The question set was composed of the following two sections on the APP: (1) Assessment(A): A total of 180 questions were set to interpret the optimal imaging findings of the diseases (Fig. 3). Students were required to select the best answer based on the provided images. The images used in these cases were specifically chosen from the central slice of each lesion sequence, including CT and MRI images, and all cases were confirmed by pathological examination. (2) Reading Skill (K): For each disease, three videos were set, amounting to a total of 27 videos, with each video lasting 2 to 3 min. Each video focused on a real clinical case to show the imaging diagnosis process for the disease, primarily interpreting the imaging characteristics while also introducing the patient’s general information, medical history, physical examination findings, laboratory test results, and prognosis (Fig. 4).

Residents were required to complete their tasks on the APP, and they had the option to review and repeatedly practice the questions they answered incorrectly. The SPARK platform allows for the real-time monitoring of each student’s completion of learning tasks. The residents were required to learn for at least 30 min every day. The backend statistics revealed that, on average, students actually spent between 30 and 45 min per day logged into the APP for learning.

[IMAGE OMITTED: SEE PDF]

Teaching methods of the control group

Routine teaching mode was implemented in the non-radiology residents in the control group. According to the needs of the residents, the imaging characteristics of the above diseases was explained by the teachers of the gastroenterology department for 20 to 30 min per day, and residents further studied it through online learning platforms and books, especially the educational resources on the WeChat official account (WeChat is a widely used APP in China, primarily for daily conversations. It supports sending text, images, and voice messages, and also provides learning materials, online courses). Similarly, residents were required to actively engage in study for at least 30 min every day.

Assessment and evaluation

The assessment included admission examination, mid-term examination and exit examination. The three examinations were administered as closed-book assessments on three distinct occasions: the day of admission, the Friday of the second week, and the final working day of each month. Three sets of questions with similar difficulty were constructed by the teaching secretary of radiology department with 20 years of clinical and teaching experience, all of which were composed of single-choice questions for image interpretation (Fig. 3), with 10 questions each time. Residents were required to select the correct answer from five options, which include one accurate imaging diagnosis and four potential differential diagnoses, based on a series of actual CT or MRI images provided. The examination scope did not exceed the diseases required to be mastered in the current month. The teaching secretary completed the question-setting, distribution, and analysis of the exams on QuestionStar (QuestionStar is a platform that supports online questionnaire design and data collection, suitable for questionnaire surveys, online exams, assessments, and voting ). The exam questions must adhere to the training syllabus for residents and be reviewed by the teaching director with 25 years of clinical and teaching experience for each exam. At the end of each assessment, the teaching secretary reviewed the correct rate and error rate of each question through the QuestionStar. To better understand the questions that frequently led to incorrect answers and to strengthen the review in follow-up studies, the teaching secretary explained to the residents on WeChat, focusing on the image characteristics of questions with an error rate of more than 30%. By monitoring the students’ learning progress through the SPARK platform, the teaching secretary reminded the students in the experimental group to complete their learning tasks promptly after the mid-term assessment, while also encouraging the students in the control group to study diligently.

[IMAGE OMITTED: SEE PDF]

On the day of graduation, we distributed an anonymous questionnaire to the residents in the experimental group to assess their demand for imaging knowledge and gather feedback on our teaching model, which was implemented via the QuestionStar platform. The questionnaire was developed for this study, including: the importance of radiology in mastering clinical knowledge of gastroenterology, the impact of case database teaching mode based on imaging interpretation skills and learning interest, and whether residents were willing to accept case database teaching mode in other departments (S 1). Likert scale was used to assess responses with 1–5 levels indicating “Strongly disagree”, “Disagree”, “Neutral”, “Agree”and “Strongly agree”, respectively.

Statistical analyses were performed using R software (version 4.1.0, https://www.r-project.org/) and SPSS software (version 25, IBM Corporation, NY, USA). Shapiro-Wilk test was used to test the normality of residents’ examination scores at the admission, mid-term and exit stages. Continuous variables with normal distribution were presented as mean ± standard deviation, and the comparison between the control group and the experimental group was represented by the independent samples t-test. Continuous data with non-normal distribution were represented by median (interquartile range), and the comparison between the two groups was carried out using the Mann-Whitney U test. The distribution differences of sex, grade and academic degree between the two groups were tested by the chi-squared test or the Fisher’s exact test. The intra-group comparisons of the admission, mid-term, and exit examination scores for both the experimental and control groups were conducted as follows: When the data passed both the normality test and the homogeneity of variance test, a one-way analysis of variance was used for comparison. Otherwise, the Kruskal-Wallis test was employed for analysis. For post hoc pairwise comparisons, the Bonferroni correction method was applied to adjust the significance level. A P value < 0.05 indicated a statistically significant difference.

Results

The scores of non-radiology residents in the experimental group and the control group showed no difference in the admission examination ( P = 0.728). However, the scores in the experimental group were significantly higher than those in the control group at the mid-term (P = 0.006) and exit examination (P = 0.002), as shown in Table 1; Fig. 5.

[IMAGE OMITTED: SEE PDF]

In both the experimental and control groups, there were overall differences in the residents’ scores across the three assessments (admission, mid-term, and exit examinations) (P < 0.001). Further pairwise comparisons revealed that all pairwise comparisons of the three assessment scores within the experimental group showed statistically significant differences. In the control group, no statistically significant difference was found between the admission and mid-term examination scores, while statistically significant differences were observed between the entrance and exit examination scores as well as between the mid-term and exit examination scores. The specific P-values are detailed in Table 2.

[IMAGE OMITTED: SEE PDF]

27 questionnaires were sent out in the experimental group, and 27 were collected, with an effective questionnaire recovery rate of 100%. In the experimental group, 100% (27/27) of the non-radiology residents considered radiology to be very important (stongly agree) for the acquisition of clinical knowledge in gastroenterology. Among the residents, 88.9% (24/27) of these residents believed that this teaching mode significantly improved their image interpretation skills and interest in using images to learn (stongly agree), while 7.4% (2/27) and 3.7% (1/27) chose “agree”and “neutral”, respectively. A total of 81.5% (22/27) of these residents were very willing to accept the case database teaching mode by other departments, while 11.1% (3/27) chosing “agree”, 3.7% (1/27) “neutral” and 3.7% (1/27) “disagree”, as shown in Table 3.

[IMAGE OMITTED: SEE PDF]

QuestionStar platform can show the correct rate and error rate of each question. The teaching secretary analyzed the questions that led to incorrect answers in the admission, mid-term, and exit examinations of the two groups. The two groups of residents had similar types of these questions, including liver cancer, inflammatory bowel disease, and colorectal cancer, with an error rate more than 30%. After each mid-term and exit examination, the teaching secretary explained the questions that frequently led to incorrect answers in both the experimental group and the control group through the WeChat communication platform, with a focus on the image characteristics of these cases.

Discussion

The value of CBL teaching method in the acquisition of knowledge about the digestive system has been recognized by many scholars, and it has been applied to different teaching dimensions [15]. Shi et al. [16] showed that CBL multi-modal imaging teaching was more helpful to improve students’ image discrimination and clinical analysis ability, and helped to increase students’ learning efficiency, improve theoretical knowledge and practical ability. Zhu et al. [17] revealed that a CBL teaching method using a periodontal clinical database software could effectively improve dental residents’ achievement in the exam of the standardized residency training program. However, there is no data available on the use of image case-based teaching methods to promote acquisition of knowledge in non-radiology residents in the gastroenterology department.

In this study, the teaching model based on image case database improved the image interpretation skills of non-radiology residents training in the gastroenterology department. Initially, there were no statistically significant differences in the scores at the admission examination between the two experimental and control group. After a period of study, the median score of the mid-term exam in the experimental group (70 points) was higher than that of the admission exam (60 points), while the median scores of the two exams in the control group remained the same (both 60 points). The improvement in the mid-term exam scores of the experimental group was significantly greater than that of the control group.

Following a prompt from the teaching secretary of the gastroenterology department, the pace of learning tasks for both the experimental and control groups notably increased, with the experimental group experiencing a particularly significant acceleration. This group also benefits from real-time supervision facilitated by the case database management system’s backend. Ultimately, residents in both the experimental and control groups achieved higher scores in the final assessment compared to their initial admission and mid-term evaluations. Notably, the experimental group demonstrated a more significant improvement over the control group, indicating that an effective supervision and management mechanism was fundamental to enhancing the image interpretation skills of non-radiology residents [18,19,20]. The performance of the experimental group was higher than that of the control group, indicating that the case database based imaging teaching mode could improve the image interpretation skills of non-radiology residents in gastroenterology department. On one hand, because the actual cases of diseases that need to be mastered in the gastroenterology department were sent to the residents on the APP of the imaging case database, the residents learned the information provided by imaging in a task-driven and immersive mode, which was helpful for the non-radiology residents in the experimental group to understand and remember [21, 22]. On the other hand, the imaging case database offered a more comprehensive collection of typical cases, fostering the organic integration of theoretical knowledge, clinical practice, and imaging characteristics for non-radiology residents in the experimental group. This integration, in turn, enhanced their capacity to interpret images effectively [23, 24].

The teaching mode based on the image case database meets the requirements of modern image knowledge learning, including the imaging characteristics, diagnosis, and differential diagnosis of diseases. As imaging technology advances and diagnostic capabilities escalate, all the non-radiology residents in the experimental group believed that imaging was very important for mastering the clinical knowledge of gastroenterology. The case database was installed on the mobile phone of the residents in the experimental group, which could make full use of their fragmented time, and their learning was no longer limited by time and space, which was in line with the current learning habits of residents. Furthermore, 88.9% of the residents in this group believed that this teaching mode significantly improved their image interpretation skills and interest in using images to learn. It is conceivable that if the image case database is presented with the entire volume in a large medical-grade display in the future, clinical departments implement a similar teaching model based on such image case database, allowing non-radiology residents to learn 5–8 types of disease presentations closely related to their respective clinical fields and integrating this learning into their clinical practice, it could help enhance their overall skills to interpret imaging studies. Our survey indicated that 81.5% of the residents in the experimental group were willing to accept this teaching mode, thereby offering us new directions and innovative concepts to enrich our future clinical teaching endeavors.

The explanation for the questions that frequently led to incorrect answers was beneficial in enhancing the level of image interpretation skills. The types of diseases that non-radiology residents in both the experimental and control groups frequently answered incorrectly were roughly similar, but the test scores of the experimental group were still higher than those of the control group, suggesting that the former had a better grasp on the imaging characteristics of common diseases than the latter. The reasons include: (1) Throughout the 1-month clinical rotation, residents were exposed to a limited spectrum of diseases, where the imaging presentations often exemplified the complexities of “the same disease with varying appearances” and “similar appearances across different diseases”, which required repeated practice and learning to master [25]. However, the case database had functions such as repeated practice and error review, which could not be compared with conventional learning methods. (2) The learning mindset of non-radiology residents in the experimental group reflected their greater interest and flexibility, which can stimulate their interest in learning, which may be another reason why the performance of the control group was not as good as that of the experimental group [26].

There were several limitations in this study that warrant attention: (1) As a single-center design conducted within a single hospital department, the study was constrained by a limited sample size. Furthermore, the inclusion of all residents training in the gastroenterology department may have compromised the randomness of the study to some extent. (2) This study was a comparative study of the time axis before and after the implementation of image case database teaching, rather than a comparative study of the same period. Our current findings lay a preliminary groundwork for future in-depth research. Moving forward, we plan to expand the collaborative imaging teaching model, which was jointly developed by the teaching management, radiology, and clinical departments, to more clinical units. Additionally, we aim to conduct a multi-center, prospective controlled study to further validate the effectiveness of this model.

An increasing number of studies indicate that large language models (LLMs) can enrich case-based learning processes in medical education. For instance, Kıyak [27] used LLMs (ChatGPT-4 and Claude 3) to generated Script Concordance Test (SCT) items on abdominal radiology, which were evaluated through an expert panel of 16 radiologists. The study showed that using LLMs to generate SCT items can significantly streamline the process, reducing the time and effort required. This, in turn, provides substantial support to medical educators, enabling them to focus more on other critical aspects of teaching and assessment. In the future, we will also integrate LLMs with medical imaging databases and apply them to clinical residency training.

Data availability

The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

References

Shang Y, Cao KF, Yue JY, Zhao SZ, Hao SH, Sun YZ, Cui QY, Guo HM, Tang CH. Comparative effectiveness of various teaching modes, including PBL, CBL, and CTTM in paediatric medical education with combined online and offline approaches. BMC Med Educ. 2025;25(1):8.

Clarke K, Bilal M, Sánchez-Luna SA, Dalessio S, Maranki JL, Siddique SM. Impact of COVID-19 pandemic on training: global perceptions of gastroenterology and hepatology fellows in the USA. Dig Dis Sci. 2021;66(10):3307–11.

Eltorai AEM, Parris DJ, Markovich D, Tarrant MJ, Madan R, Krupat E, Matalon SA. The Pre-Radiology internship: room for improvement?? Clin Imaging. 2023;94:85–92.

Cole S, Fetter J, Oliver MG. Soaring to new heights: strengthening outcomes and assessment in residency. Ann Fam Med. 2023;21(2):197–8.

Paisant A, Skehan S, Colombié M, David A, Aubé C. Development and validation of core entrustable professional activities for abdominal radiology. Insights Imaging. 2023;14(1):142.

Montane B, Abraham A, Bustamante S, Vachharajani T, Ayad S, Devarajan J, Thompson D, Lee R, Rampersad P, Gage B, Reznicek E, Luo C, Wardrop R. Implementing an interdisciplinary procedure curriculum. Cureus. 2023;15(9):e44851.

Karademos JE, Yarris LM, Steinberg E. Teaching residents to teach: A pilot study for an innovative online curriculum. AEM Educ Train. 2024;8(1):e10946.

Cen XY, Hua Y, Niu S, Yu T. Application of case-based learning in medical student education: a meta-analysis. Eur Rev Med Pharmacol Sci. 2021;25(8):3173–81.

Zeri F, Eperjesi F, Woods C, Bandlitz S, Kumar Bhootra A, Joshi MR, Nagra M, Schweizer H, Naroo SA. Evidence-based teaching in contact lenses education: teaching and learning strategies. Cont Lens Anterior Eye. 2023;46(2):101822.

Zhang S, Xu J, Wang H, Zhang D, Zhang Q, Zou L. Effects of problem-based learning in Chinese radiology education: A systematic review and meta-analysis. Med (Baltim). 2018;97(9):e0069.

Peng W, Xiong Y, Wei J, Chen X, Huai W, He S, Liu D, Tian X, Tang S, Chen Y. Flipped classroom improves student learning outcome in Chinese pharmacy education: A systematic review and meta-analysis. Front Pharmacol. 2022;13:936899.

Li Y, Gao C, Zhu X, Zhu J, Ding Z, Han Z. Application of SPARK teaching in acute abdomen radiography teaching for undergraduate medical students. BMC Med Educ. 2022;22(1):881.

Li Y, Han Z, Xia Q, Gao C, Wang C, Zhu X, Ding Z, Zhu J. Application of the process-based teaching based on SPARK case database in the practice teaching of radiology in the musculoskeletal system for undergraduate medical students. BMC Med Educ. 2024;24(1):688.

Zhang J, Han X, Yang Z, Wang Z, Zheng J, Yang Z, Zhu J. Radiology residency training in china: results from the first retrospective nationwide survey. Insights Imaging. 2021;12(1):25.

Manas A, Ismail PMS, Mohan R, Madiraju GS, Mulla M, Mulla M, Babaji P. The role of combined problem-based learning (PBL) and case-based learning (CBL) teaching methodologies in dental education. J Educ Health Promot. 2024;13:417.

Xuehui S, Haiyan W, Fan Y, Jinqiong Z, Wenbin W. Application and evaluation of case-based multimodal imaging in the ocular fundus disease teaching of postgraduate students. BMC Med Educ. 2025;25(1):390.

Zhu W, Feng R, Fu Y. Effects of periodontal clinical database software in resident training during COVID-19 pandemic: a prospective observational study. BMC Med Educ. 2022;22(1):234.

Grijpma JW, Ramdas S, Broeksma L, et al. Learning from the experts: stimulating student engagement in Small-group active learning. Perspect Med Educ. 2024;13(1):229–38.

Zhang M, Wang YY, Bai Y, Chen XD, Guo H. Investigation on the construction of teaching case base in medical genetics. Yi Chuan. 2024;46(1):78–87.

Cataldi ML, Kelly-Hedrick M, Nanavati J, Chisolm MS, Anne LW. Post-residency medical education fellowships: a scoping review. Med Educ Online. 2021;26(1):1920084.

Yang X, Guo L, Chen X, Zhang J, Xiang Q. Medical humanities in resident training: A pilot study of specific courses in a large comprehensive hospital in China. J Med Educ Curric Dev. 2024;11:23821205241307444.

Meyer HS, Larsen K, Samuel A, Berkley H, Harvey M, Norton C, Maggio LA. Teaching medical students how to teach: A scoping review. Teach Learn Med. 2022;34(4):379–91.

Fossey S, Ather S, Davies S, Dhillon PS, Malik N, Phillips M, Harden S. Impact of COVID-19 on radiology training: Royal College of Radiologists Junior Radiologists Forum national survey. Clin Radiol. 2021;76(7):549.e9-549.e15.

Gordon EB, Wingrove P, Branstetter Iv BF, Hughes MA. Evidence for an adverse impact of remote readouts on radiology resident productivity: implications for training and clinical practice. PLOS Digit Health. 2023;2(9):e0000332.

Fischetti C, Bhatter P, Frisch E, Sidhu A, Helmy M, Lungren M, Duhaime E. The evolving importance of artificial intelligence and radiology in medical trainee education. Acad Radiol. 2022;29(Suppl 5):S70–5.

Tu W, Hibbert R, Kontolemos M, Dang W, Wood T, Verma R, McInnes MDF. Diagnostic radiology residency assessment tools: A scoping review. Can Assoc Radiol J. 2021;72(4):651–60.

Kıyak YS, Emekli E. Using large Language models to generate script concordance test in medical education: ChatGPT and Claude. Span J Med Educ. 2024;6(1). https://doi.org/10.6018/edumed.636331.