Introducing AnatomyGPT: A customized artificial intelligence application for anatomical sciences education
Abstract
Artificial intelligence (AI) technologies are poised to become an increasingly important part of education in the anatomical sciences. OpenAI has also introduced generative pretrained transformers (GPTs), which are customizable versions of the standard ChatGPT application. There is little research that has explored the potential of GPTs to serve as intelligent tutoring systems for learning the anatomical sciences. The objective of this study was to describe the design and explore the performance of AnatomyGPT, a customized artificial intelligence application intended for anatomical sciences education. The AnatomyGPT application was configured with GPT Builder by uploading open‐source textbooks as knowledge sources and by providing pedagogical instructions for how to interact with users. The performance of AnatomyGPT was compared with ChatGPT by evaluating the responses of both applications to prompts of the National Board of Medical Examiners (NBME) sample items with respect to accuracy, rationales, and citations. AnatomyGPT achieved high scores on the NBME sample items for Gross Anatomy, Embryology, Histology, and Neuroscience and scored comparably to ChatGPT. In addition, AnatomyGPT provided several citations in the responses that it generated, while ChatGPT provided none. Both GPTs provided rationales for all sample items. The customized AnatomyGPT application demonstrated preliminary potential as an intelligent tutoring system by generating responses with increased citations as compared with the standard ChatGPT application. The findings of this study suggest that instructors and students may wish to create their own custom GPTs for teaching and learning anatomy. Future research is needed to further develop and characterize the potential of GPTs for anatomy education.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Fields such as Biology, Business, and Medicine remain underexplored, largely due to challenges in adapting general-purpose GenAI models to specialized or niche contexts.
“In contrast, fields such as Biology (Collins et al., 2024), Business (Milić et al., 2024; Netland et al. 2025), and Medicine (Hudon et al., 2024) remain underexplored, largely due to challenges in adapting general-purpose GenAI models to specialized or niche contexts.”
From Pedagogical Applications of Generative AI in Higher Education: A Systematic Review of the Field (2025) · cites this paper as Collins et al., 2024
Specialized GPT tools, like AnatomyGPT, have emerged to overcome specific limitations associated with general-purpose platforms such as ChatGPT, Google Gemini, and Claude.
“However, specialized GPT tools, like AnatomyGPT, have emerged to overcome specific limitations associated with generalpurpose platforms such as ChatGPT, Google Gemini, and Claude.”
From Pedagogical Applications of Generative AI in Higher Education: A Systematic Review of the Field (2025) · cites this paper as Collins et al., 2024
Cites
Gemini: A Family of Highly Capable Multimodal Models
Gemini Team, Anil, R., Borgeaud, S., Alayrac, J.-B., Yu, J., Soricut, R., Schalkwyk, J., Dai, A. M., Hauth, A., Millican, K., et al. (2023) · arXiv · doi:10.48550/arxiv.2312.11805 · Cited by 990 · see citing works
Effectiveness of Intelligent Tutoring Systems: A Meta-Analytic Review
Kulik, J. A., & Fletcher, J. D. (2016) · Review of Educational Research · doi:10.3102/0034654315581420 · Cited by 930 · cited 1× · see citing works
“For example, a 2016 meta-analysis of 50 studies involving intelligent tutoring systems found an average effect of raising test scores 0.66 standard deviations above controls, representing an improvement from the 50th to 75th percentile (Kulik & Fletcher, 2016).”
InfluentialIntelligent tutoring systems and learning outcomes: A meta-analysis
Ma, W., Adesope, O. O., Nesbit, J. C., & Liu, Q. (2014) · Journal of Educational Psychology · doi:10.1037/a0037123 · Cited by 840 · cited 1× · see citing works
“An earlier meta-analysis involving 107 effect sizes found that intelligent tutoring systems improve student achievement with effect sizes ranging from 0.35 to 0.57 depending on the control comparison (Ma et al., 2014).”
InfluentialArtificial Intelligence for Health Professions Educators
Lomis, K., Jeffries, P., Palatta, A., Sage, M., Sheikh, J., Sheperis, C., & Whelan, A. (2021) · NAM Perspectives · doi:10.31478/202109a · Cited by 136 · see citing works
Sriwastwa, A., Ravi, P., Emmert, A., Chokshi, S., Kondor, S., Dhal, K., Patel, P., Chepelev, L. L., Rybicki, F. J., & Gupta, R. (2023) · 3D Printing in Medicine · doi:10.1186/s41205-023-00186-8 · Cited by 28 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Pedagogical Applications of Generative AI in Higher Education: A Systematic Review of the Field
Qian, Y. (2025) · TechTrends · doi:10.1007/s11528-025-01100-1 · Cited by 206 · see citing works
“In contrast, fields such as Biology (Collins et al., 2024), Business (Milić et al., 2024; Netland et al. 2025), and Medicine (Hudon et al., 2024) remain underexplored, largely due to challenges in adapting general-purpose GenAI models to specialized or niche contexts.”
Lin, Y., Luo, Z., Ye, Z., Zhong, N., Zhao, L., Zhang, L., Li, X., Chen, Z., & Chen, Y. (2025) · JMIR Medical Education · doi:10.2196/71125 · Cited by 31 · see citing works
“If the data is insufficient or skewed, it may lead to potential biases and limitations in practical applications, causing underperformance in less-represented areas [33,59, 73,82,85,86,111,117,122,126,140-143].”
Using aggregated AI detector outcomes to eliminate false-positives in STEM-student writing.
Hyatt, J.-P. K., Bienenstock, E. J., Firetto, C. M., Woods, E. R., & Comus, R. C. (2025) · Advances in Physiology Education · doi:10.1152/advan.00235.2024 · Cited by 4 · see citing works
- Authors
- Collins, B. R., Black, E. W., & Rarey, K. E.
- Affiliations
- University of Florida Health (US)
- Corresponding author
- Collins, Bradley R.
- Venue
- Clinical Anatomy
- Publisher
- Wiley
- Published
- 2024-05-09
- Volume
- 37
- Issue
- 6
- Pages
- 661-669
- ISSN
- 0897-3806 (ISSN-L)
- 1098-2353
- OpenAlex
- W4396772600
- Cited by
- 61
- References
- 5
- Citations by year
- 2026: 25 · 2025: 32 · 2024: 4
- FWCI
- 5.71
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Clinical Anatomy (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Physical Sciences › Engineering › Biomedical Engineering › Anatomy and Medical Technology
- Physical Sciences › Computer Science › Information Systems › Artificial Intelligence in Education
- Keywords
- generative pre-trained transformer, ChatGPT, intelligent tutoring systems, gross anatomy, embryology, histology, citation generation, response accuracy
- MeSH terms
- Anatomy, Artificial Intelligence, Computer-Assisted Instruction, Humans, Software
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
