Artificial Intelligence-Guided Assessment of Femoral Neck Fractures in Radiographs: A Systematic Review and Multilevel Meta-Analysis

Abstract

Artificial Intelligence (AI) is a dynamic area of computer science that is constantly expanding its practical benefits in various fields. The aim of this study was to analyze AI-guided radiological assessment of femoral neck fractures by performing a systematic review and multilevel meta-analysis of primary studies. The study protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) on May 21, 2024 [CRD42024541055]. The updated Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines were strictly followed. A systematic literature search of PubMed, Web of Science, Ovid (Med), and Epistemonikos databases was conducted until May 31, 2024. Critical appraisal using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool showed that the overall quality of the included studies was moderate. In addition, publication bias was presented in funnel plots. A frequentist multilevel meta-analysis was performed using a random effects model with inverse variance and restricted maximum likelihood heterogeneity estimator with Hartung-Knapp adjustment. The accuracy between AI-based and human assessment of femoral neck fractures, sensitivity and specificity with 95% confidence intervals (CIs) were calculated. Study heterogeneity was assessed using the Higgins test I² (low heterogeneity <25%, moderate heterogeneity: 25%–75%, and high heterogeneity >75%). Finally, 11 studies with a total of 21,163 radiographs were included for meta-analysis. The results of the study quality assessment using the QUADAS-2 tool are presented in Table 2. The funnel plots indicated a moderate publication bias. The AI showed excellent accuracy in assessment of femoral neck fractures (Accuracy = 0.91, 95% CI 0.83 to 0.96; I² = 99%; p < 0.01). The AI showed good sensitivity in assessment of femoral neck fractures (Sensitivity = 0.87, 95% CI 0.77 to 0.93; I² = 98%; p < 0.01). The AI showed excellent specificity in assessment of femoral neck fractures (Specificity = 0.91, 95% CI 0.77 to 0.97; I² = 97%; p < 0.01). AI-guided radiological assessment of femoral neck fractures showed excellent accuracy and specificity as well as good sensitivity. The use of AI as a faster and more reliable assessment tool and as an aid in radiological routine seems justified.

Keywords

artificial intelligence / deep learning / femoral neck fractures / hip fractures / meta-analysis / multilevel meta-analysis / neural network / radiographs

References

[1] Iqbal JD, Vinay R. Are we ready for artificial intelligence in medicine? Swiss Med Wkly. 2022; 152: w30179. https://doi.org/10.4414/smw.2022.w30179

[2] Langerhuizen DWG, Janssen SJ, Mallee WH, van den Bekerom MPJ, Ring D, Kerkhoffs GMMJ, et al. What are the applications and limitations of artificial intelligence for fracture detection and classification in Orthopaedic trauma imaging? A systematic review. Clin Orthop Relat Res. 2019; 477(11): 2482-2491. https://doi.org/10.1097/CORR.0000000000000848

[3] Larson N, Nguyen C, Do B, Kaul A, Larson A, Wang S, et al. Artificial intelligence system for automatic quantitative analysis and radiology reporting of leg length radiographs. J Digit Imaging. 2022; 35(6): 1494-1505. https://doi.org/10.1007/s10278-022-00671-2

[4] 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. https://doi.org/10.1148/radiol.211785

[5] Zhang X, Yang Y, Shen YW, Zhang KR, Jiang ZK, Ma LT, et al. Diagnostic accuracy and potential covariates of artificial intelligence for diagnosing orthopedic fractures: a systematic literature review and meta-analysis. Eur Radiol. 2022; 32(10): 7196-7216. https://doi.org/10.1007/s00330-022-08956-4

[6] Lex JR, Di Michele J, Koucheki R, Pincus D, Whyne C, Ravi B. Artificial intelligence for hip fracture detection and outcome prediction: a systematic review and meta-analysis. JAMA Netw Open. 2023; 6(3):e233391. https://doi.org/10.1001/jamanetworkopen.2023.3391

[7] Jung J, Dai J, Liu B, Wu Q. Artificial intelligence in fracture detection with different image modalities and data types: a systematic review and meta-analysis. PLOS Digit Health. 2024; 3(1):e0000438. https://doi.org/10.1371/journal.pdig.0000438

[8] Stöckle U, Lucke M, Haas NP. The femoral neck fracture. Dtsch Arztebl. 2005; 102(49):A-3426/B-2894/C-2710.

[9] Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Rev Esp Cardiol (Engl ed). 2021; 74(9): 790-799. English, Spanish. https://doi.org/10.1016/j.rec.2021.07.010

[10] Prill R, Karlsson J, Ayeni OR, Becker R. Author guidelines for conducting systematic reviews and meta-analyses. Knee Surg Sports Traumatol Arthrosc. 2021; 29(9): 2739-2744. https://doi.org/10.1007/s00167-021-06631-7

[11] Campbell JM, Klugar M, Ding S, Carmody DP, Hakonsen SJ, Jadotte YT, et al. Diagnostic test accuracy: methods for systematic review and meta-analysis. Int J Evid Based Healthc. 2015; 13(3): 154-162. https://doi.org/10.1097/XEB.0000000000000061

[12] Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002; 21: 1539-1558.

[13] Açıcı K, Sümer E, Beyaz S. Comparison of different machine learning approaches to detect femoral neck fractures in x-ray images. Health Technol. 2011; 11: 643-653. https://doi.org/10.1007/s12553-021-00543-9

[14] Adams M, Chen W, Holcdorf D, McCusker MW, Howe PD, Gaillard F. Computer vs human: deep learning versus perceptual training for the detection of neck of femur fractures. J Med Imaging Radiat Oncol. 2019; 63(1): 27-32. https://doi.org/10.1111/1754-9485.12828

[15] Bae J, Yu S, Oh J, Kim TH, Chung JH, Byun H, et al. External validation of deep learning algorithm for detecting and visualizing femoral neck fracture including displaced and non-displaced fracture on plain X-ray. J Digit Imaging. 2021; 34(5): 1099-1109. https://doi.org/10.1007/s10278-021-00499-2

[16] Beyaz S, Açıcı K, Sümer E. Femoral neck fracture detection in X-ray images using deep learning and genetic algorithm approaches. Jt Dis Relat Surg. 2020; 31(2): 175-183. https://doi.org/10.5606/ehc.2020.72163

[17] Guy S, Jacquet C, Tsenkoff D, Argenson JN, Ollivier M. Deep learning for the radiographic diagnosis of proximal femur fractures: limitations and programming issues. Orthop Traumatol Surg Res. 2021; 107(2):102837. https://doi.org/10.1016/j.otsr.2021.102837

[18] Hsieh SL, Chiang JL, Chuang CH, Chen YY, Hsu CJ. A computer-assisted diagnostic method for accurate detection of early nondisplaced fractures of the femoral neck. Biomedicine. 2023; 11(11):3100. https://doi.org/10.3390/biomedicines11113100

[19] Jiménez-Sánchez A, Kazi A, Albarqouni S, Kirchhoff C, Biberthaler P, Navab N, et al. Precise proximal femur fracture classification for interactive training and surgical planning. Int J Comput Assist Radiol Surg. 2020; 15(5): 847-857. https://doi.org/10.1007/s11548-020-02150-x

[20] Krogue JD, Cheng KV, Hwang KM, Toogood P, Meinberg EG, Geiger EJ, et al. Automatic hip fracture identification and functional subclassification with deep learning. Radiol ArtifIntell. 2020; 2(2):e190023. https://doi.org/10.1148/ryai.2020190023

[21] Mutasa S, Varada S, Goel A, Wong TT, Rasiej MJ. Advanced deep learning techniques applied to automated femoral neck fracture detection and classification. J Digit Imaging. 2020; 33(5): 1209-1217. https://doi.org/10.1007/s10278-020-00364-8

[22] Yamada Y, Maki S, Kishida S, Nagai H, Arima J, Yamakawa N, et al. Automated classification of hip fractures using deep convolutional neural networks with orthopedic surgeon-level accuracy: ensemble decision-making with antero-posterior and lateral radiographs. Acta Orthop. 2020; 91(6): 699-704. https://doi.org/10.1080/17453674.2020.1803664

[23] Yu JS, Yu SM, Erdal BS, Demirer M, Gupta V, Bigelow M, et al. Detection and localisation of hip fractures on anteroposterior radiographs with artificial intelligence: proof of concept. Clin Radiol. 2020; 75(3): 237.e1-237.e9. https://doi.org/10.1016/j.crad.2019.10.022

[24] Cheng CT, Wang Y, Chen HW, Hsiao PM, Yeh CN, Hsieh CH, et al. A scalable physician-level deep learning algorithm detects universal trauma on pelvic radiographs. Nat Commun. 2021; 12(1): 1066. https://doi.org/10.1038/s41467-021-21311-3

[25] Cheng CT, Ho TY, Lee TY, Chang CC, Chou CC, Chen CC, et al. Application of a deep learning algorithm for detection and visualization of hip fractures on plain pelvic radiographs. Eur Radiol. 2019; 29(10): 5469-5477. https://doi.org/10.1007/s00330-019-06167-y

[26] Murphy EA, Ehrhardt B, Gregson CL, von Arx OA, Hartley A, Whitehouse MR, et al. Machine learning outperforms clinical experts in classification of hip fractures. Sci Rep. 2022; 12(1): 2058. https://doi.org/10.1038/s41598-022-06018-9

[27] Khan SK, Kalra S, Khanna A, Thiruvengada MM, Parker MJ. Timing of surgery for hip fractures: a systematic review of 52 published studies involving 291,413 patients. Injury. 2009; 40(7): 692-697. https://doi.org/10.1016/j.injury.2009.01.010

[28] Leicht H, Gaertner T, Günster C, Halder AM, Hoffmann R, Jeschke E, et al. Time to surgery and outcome in the treatment of proximal femoral fractures. Dtsch Arztebl Int. 2021; 118(26): 454-461. https://doi.org/10.3238/arztebl.m2021.0165

[29] Twinprai N, Boonrod A, Boonrod A, Chindaprasirt J, Sirithanaphol W, Chindaprasirt P, et al. Artificial intelligence (AI) vs. human in hip fracture detection. Heliyon. 2022; 8(11):e11266. https://doi.org/10.1016/j.heliyon.2022.e11266

RIGHTS & PERMISSIONS

2024 The Author(s). Orthopaedic Surgery published by Tianjin Hospital and John Wiley & Sons Australia, Ltd.