Diagnostic Accuracy of Artificial Intelligence Models for the Detection of Ankle and Foot Fractures: A Systematic Review and Meta-analysis
Archives of Academic Emergency Medicine,
Vol. 14 No. 1 (2026),
1 October 2025
,
Page e57
https://doi.org/10.22037/aaem.v14i1.3163
Abstract
Introduction: Foot and ankle fractures are common injuries that may be missed on radiographs, particularly when subtle or complex fracture patterns are present. Artificial intelligence (AI) has emerged as a potential adjunct to radiographic interpretation, but reported diagnostic performance varies across studies and AI models. Therefore, a comprehensive meta-analysis is warranted to provide a more robust understanding of the diagnostic performance and applicability of AI models in this field. Methods: In this systematic review and meta-analysis, we searched MEDLINE and PubMed Central (PMC) via PubMed, Scopus, and Web of Science from inception to September 2026. Diagnostic accuracy measures were pooled using a bivariate random-effects model, and subgroup analyses were performed to explore potential sources of heterogeneity. Results: The pooled sensitivity was 0.90 (95% confidence interval (CI): 0.86–0.92), while the pooled specificity was 0.88 (95% CI: 0.82–0.93). The pooled positive likelihood ratio (PLR) was 7.6 (95% CI: 4.8–12), and the pooled negative likelihood ratio (NLR) was 0.12 (95% CI: 0.09–0.16). The area under the summary receiver operating characteristic (SROC) curve was 0.94. Conclusion: AI models demonstrate strong diagnostic performance for detecting foot and ankle fractures on radiographs. Diagnostic performance differed by anatomical region, with higher sensitivity for foot fractures and higher specificity for ankle fractures. These findings support AI as a promising adjunct to radiographic interpretation, although prospective, multicenter studies with external validation are needed to confirm its clinical utility and real-world effectiveness.
- Ankle Fractures
- Artificial Intelligence
- Deep Learning
- Diagnosis
- Foot Fractures
- Foot Injuries
- Meta-analysis
How to Cite
References
1. Cheng CT, Hsu CP, Ooyang CH, Chou CY, Lin NY, Lin JY, et al. Evaluation of ensemble strategy on the development of multiple view ankle fracture detection algorithm. Br J Radiol. 2023;96(1145):20220924.
2. Lau H, Dreyer MA. Cuboid Stress Fractures. StatPearls. Treasure Island (FL): StatPearls Publishing, StatPearls Pub-lishing LLC.;2026.
3. De Boer AS, Schepers T, Panneman MJ, Van Beeck EF, Van Lieshout EM. Health care consumption and costs due to foot and ankle injuries in the Netherlands, 1986-2010. BMC Musculoskelet Disord. 2014;15:128.
4. Davidovitch RI, Walsh M, Spitzer A, Egol KA. Functional outcome after operatively treated ankle fractures in the el-derly. Foot Ankle Int. 2009;30(8):728-33.
5. Beckenkamp PR, Lin CW, Chagpar S, Herbert RD, van der Ploeg HP, Moseley AM. Prognosis of physical function follow-ing ankle fracture: a systematic review with meta-analysis. J Orthop Sports Phys Ther. 2014;44(11):841-51, b2.
6. Mehta SS, Rees K, Cutler L, Mangwani J. Understanding risks and complications in the management of ankle frac-tures. Indian J Orthop. 2014;48(5):445-52.
7. Brandser EA, Braksiek RJ, El-Khoury GY, Saltzman CL, Marsh JL, Clark WA, et al. Missed fractures on emergency room ankle radiographs: An analysis of 433 patients. Emerg Radiol. 1997;4(5):295-302.
8. Young KW, Park YU, Kim JS, Cho HK, Choo HS, Park JH. Misdiagnosis of Talar Body or Neck Fractures as Ankle Sprains in Low Energy Traumas. Clin Orthop Surg. 2016;8(3):303-9.
9. Judd DB, Kim DH. Foot fractures frequently misdiagnosed as ankle sprains. Am Fam Physician. 2002;66(5):785-94.
10. Wei CJ, Tsai WC, Tiu CM, Wu HT, Chiou HJ, Chang CY. Systematic analysis of missed extremity fractures in emer-gency radiology. Acta Radiol. 2006;47(7):710-7.
11. Gitto S, Serpi F, Albano D, Risoleo G, Fusco S, Messina C, et al. AI applications in musculoskeletal imaging: a narrative review. Eur Radiol Exp. 2024;8(1):22.
12. Sharma S. Artificial intelligence for fracture diagnosis in orthopedic X-rays: current developments and future poten-tial. SICOT J. 2023;9:21.
13. Kuo RYL, Harrison C, Curran TA, Jones B, Freethy A, Cussons D, et al. Artificial Intelligence in Fracture Detection: A Systematic Review and Meta-Analysis. Radiology. 2022;304(1):50-62.
14. Ruitenbeek HC, Sahil S, Kumar A, Kushawaha RK, Tana-mala S, Sathyamurthy S, et al. Cross-validation of an artificial intelligence tool for fracture classification and localization on conventional radiography in Dutch population. Insights Im-aging. 2025;16(1):150.
15. Bousson V, Attané G, Benoist N, Perronne L, Diallo A, Hadid-Beurrier L, et al. Artificial Intelligence for Detecting Acute Fractures in Patients Admitted to an Emergency De-partment: Real-Life Performance of Three Commercial Algo-rithms. Acad Radiol. 2023;30(10):2118-39.
16. Kim T, Goh TS, Lee JS, Lee JH, Kim H, Jung ID. Transfer learning-based ensemble convolutional neural network for accelerated diagnosis of foot fractures. Phys Eng Sci Med. 2023;46(1):265-77.
17. McInnes MDF, Moher D, Thombs BD, McGrath TA, Bos-suyt PM, Clifford T, et al. Preferred Reporting Items for a Systematic Review and Meta-analysis of Diagnostic Test Ac-curacy Studies: The PRISMA-DTA Statement. JAMA. 2018;319(4):388-96.
18. Mangwani J, Jithin SP, Akram N, Ravi L, Vaishya R. Ma-chine learning and deep learning for ankle fracture detection in radiographs: A systematic review with narrative synthesis on diagnostic performance and future clinical integration. Foot (Edinb). 2026;68:102263.
19. Prasetiyo GT, Wijaya JH. The diagnostic accuracy of artifi-cial intelligence in detecting foot and ankle fractures: a sys-tematic review and meta-analysis of diagnostic test accuracy. J Orthop Surg Res. 2026.
20. Husarek J, Fuchss AMC, Ruder TD, Mougiakakou S, Ex-adaktylos A, Wahedi K, et al. Artificial intelligence for frac-ture detection on computed tomography: a comprehensive systematic review and meta-analysis of diagnostic test accu-racy in non-commercial and commercial solutions. Emerg Ra-diol. 2026;33(2):351-75.
21. Nowroozi A, Salehi MA, Shobeiri P, Agahi S, Momtaz-manesh S, Kaviani P, et al. Artificial intelligence diagnostic accuracy in fracture detection from plain radiographs and comparing it with clinicians: a systematic review and meta-analysis. Clin Radiol. 2024;79(8):579-88.
22. Guermazi A, Tannoury C, Kompel AJ, Murakami AM, Du-carouge A, Gillibert A, et al. Improving Radiographic Fracture Recognition Performance and Efficiency Using Artificial Intel-ligence. Radiology. 2022;302(3):627-36.
23. Kim J-H, Mo Y-C, Choi S-M, Hyun Y, Lee JW. Detecting Ankle Fractures in Plain Radiographs Using Deep Learning with Accurately Labeled Datasets Aided by Computed To-mography: A Retrospective Observational Study. Appl Sci (Basel). 2021;11(19):8791.
24. Kitamura G, Chung CY, Moore BE, 2nd. Ankle Fracture Detection Utilizing a Convolutional Neural Network Ensemble Implemented with a Small Sample, De Novo Training, and Multiview Incorporation. J Digit Imaging. 2019;32(4):672-7.
25. Prijs J, Liao Z, To MS, Verjans J, Jutte PC, Stirler V, et al. Development and external validation of automated detection, classification, and localization of ankle fractures: inside the black box of a convolutional neural network (CNN). Eur J Trauma Emerg Surg. 2023;49(2):1057-69.
26. Taseh A, Shah A, Eftekhari M, Flaherty A, Ebrahimi A, Jones S, et al. Detecting Fifth Metatarsal Fractures on Radio-graphs Through the Lens of Smartphones: The FIXUS AI Al-gorithm. Cureus. 2025;17(8):e91284.
27. Wang Y, Li Y, Lin G, Zhang Q, Zhong J, Zhang Y, et al. Lower-extremity fatigue fracture detection and grading based on deep learning models of radiographs. Eur Radiol. 2023;33(1):555-65.
- Abstract Viewed: 64 times
- pdf Downloaded: 49 times
Download Statastics
Make a Submission
SJR
COPE
