Page 102 - GTM-4-3

Global Translational Medicine CNNs for overfitting and generalizability in fracture detection

doi: 10.1371/journal.pdig.0000438 doi: 10.1016/j.patter.2020.100129
10. Li H, Li J, Guan X, Liang B, Lai Y, Luo X. Research on 20. Maleki F, Ovens K, Gupta R, Reinhold C, Spatz A,
Overfitting of Deep Learning. In: Proceedings of the Forghani R. Generalizability of machine learning models:
2019 15 International Conference on Computational Quantitative evaluation of three methodological pitfalls.
th
Intelligence and Security (CIS). IEEE; 2019. p. 78-81. Radiol Artif Intell. 2023;5:e220028.
doi: 10.1109/cis.2019.00025 doi: 10.1148/ryai.220028
11. Zhang H, Zhang L, Jiang Y. Overfitting and 21. Salehinejad H, Kitamura J, Ditkofsky N, et al. A real-world
Underfitting Analysis for Deep Learning Based End- demonstration of machine learning generalizability in the
to-end Communication Systems. In: Proceedings of detection of intracranial hemorrhage on head computerized
the 2019 11 International Conference on Wireless tomography. Sci Rep. 2021;11:17051.
th
Communications and Signal Processing (WCSP). IEEE; 2019.
p. 1-6. doi: 10.1038/s41598-021-95533-2
22. Zihni E, Madai VI, Livne M, et al. Opening the black box of
doi: 10.1109/wcsp.2019.8927876
artificial intelligence for clinical decision support: A study
12. Diogo P, Morais M, Calisto FM, et al. Weakly-supervised predicting stroke outcome. PLoS One. 2020;15:e0231166.
diagnosis and detection of breast cancer using deep
multiple instance learning. In: 2023 IEEE 20 International doi: 10.1371/journal.pone.0231166
th
Symposium on Biomedical Imaging (ISBI). IEEE; 2023. p. 1-4. 23. Yang G, Ye Q, Xia J. Unbox the black-box for the medical
doi: 10.1109/isbi53787.2023.10230448. explainable AI via multi-modal and multi-centre data
fusion: A mini-review, two showcases and beyond. Inf
13. Thomas RL, Uminsky D. Reliance on metrics is a fundamental Fusion. 2022;77:29-52.
challenge for AI. Patterns (N Y). 2022;3:100476.
doi: 10.1016/j.inffus.2021.07.016
doi: 10.1016/j.patter.2022.100476
24. Felder RM. Coming to terms with the black box problem:
14. Ektefaie Y, Shen A, Bykova D, Marin MG, Zitnik M, How to justify AI systems in health care. Hastings Cent
Farhat M. Evaluating generalizability of artificial intelligence Rep. 2021;51:38-45.
models for molecular datasets. Nat Mach Intell. 2024;
6:1512-1524. doi: 10.1002/hast.1248
doi: 10.1038/s42256-024-00931-6 25. Reyna MA, Nsoesie EO, Clifford GD. Rethinking algorithm
performance metrics for artificial intelligence in diagnostic
15. Foody GM. Challenges in the real world use of classification medicine. JAMA. 2022;328:329.
accuracy metrics: From recall and precision to the Matthews
correlation coefficient. PLoS One. 2023;18:e0291908. doi: 10.1001/jama.2022.10561
doi: 10.1371/journal.pone.0291908 26. Chowdhury R. Bone Fracture Detection Using CNN; 2024.
Available from: https://www.kaggle.com/code/27ituparna/
16. Husain G, Mayer J, Bekbolatova M, Vathappallil P, Matalia M, bonefracture-cnn [Last accessed on 2025 Jan 11].
Toma M. Machine learning for medical image classification.
Acad Med. 2024;1(4):1-18. 27. Chaskar P. Bone Fracture Detection - 97% Accuracy CNN;
2024. Available from: https://www.kaggle.com/code/
doi: 10.20935/AcadMed7444 prasadchaskar/bone-fracture-detection-97-accuracy-cnn
17. Buddhiraju A, Chen TLW, Subih MA, Seo HH, [Last accessed on 2025 Jan 11].
Esposito JG, Kwon YM. Validation and generalizability of 28. Chaddad A, Hu Y, Wu Y, et al. Generalizable and explainable
machine learning models for the prediction of discharge deep learning for medical image computing: An overview.
disposition following revision total knee arthroplasty. Curr Opin Biomed Eng. 2025;33(3):100567.
J Arthroplasty. 2023;38:S253-S258.
doi: 10.1016/j.cobme.2024.100567
doi: 10.1016/j.arth.2023.02.054
29. U.S. Food and Drug Administration. Considerations for
18. Sarker IH. Machine learning: Algorithms, real-world the Use of Artificial Intelligence to Support Regulatory
applications and research directions. SN Comput Sci. Decision-Making for Drug and Biological Products. Draft
2021;2:160. Guidance for Industry; 2025. Available from: https://www.
doi: 10.1007/s42979-021-00592-x fda.gov/media/184830/download [Last accessed on 2025
Mar 06].
19. Ho SY, Phua K, Wong L, Bin Goh WW. Extensions of
the external validation for checking learned model 30. Alam A, Al-Shamayleh AS, Thalji N, et al. Novel transfer
interpretability and generalizability. Patterns (N Y). 2020; learning based bone fracture detection using radiographic
1:100129. images. BMC Med Imaging. 2025;25:5.

Volume 4 Issue 3 (2025) 94 doi: 10.36922/gtm.8526