Can fracture non-union be predicted using deep learning?

Abstract

Fracture non-union remains a significant clinical challenge despite considerable advances in diagnostic imaging and treatment modalities. Unpredictable healing, repeated interventions, and prolonged disability contribute to high patient morbidity and increased healthcare costs. Early and reliable prediction of non-union is therefore essential for timely intervention. This review discusses traditional radiographic assessment using the Radiologic Union Scale for the Tibia (RUST), its inherent limitations, and the emerging role of artificial intelligence (AI) and deep learning in fracture analysis. In addition, we review recent studies-including Bayesian classifiers and simulation models-that integrate AI for early prediction of non-union, and we provide an updated summary table of key studies.

Keywords

Deep learning / machine learning / bone / fracture union / non-union

References

[1] Stewart SK. Fracture Non-Union: A Review of Clinical Challenges and Future Research Needs. Malays Orthop J. 2019; 13: 1-10.

[2] Nandra R, Grover L, Porter K. Fracture non-union epidemiology and treatment. Trauma. 2016; 18: 3-11.

[3] Pountos I, Georgouli T, Pneumaticos S, Giannoudis PV. Fracture non-union: Can biomarkers predict outcome? Injury. 2013; 44: 1725-32.

[4] Leow JM, Clement ND, Tawonsawatruk T, Simpson CJ, Simpson AH. The radiographic union scale in tibial (RUST) fractures: Reliability of the outcome measure at an independent centre. Bone Joint Res. 2016; 5: 116-21.

[5] Christiano AV, Goch AM, Leucht P, Konda SR, Egol KA. Radiographic union score for tibia fractures predicts success with operative treatment of tibial nonunion. J Clin Orthop Trauma. 2019; 10: 650-4.

[6] Van Houten AH, Heesterbeek PJ, Heerwaarden RJ, Tienen TG, Wymenga AB. Medial Open Wedge High Tibial Osteotomy: Can Delayed or Nonunion Be Predicted? Clin Orthop Relat Res. 2014; 472: 1217-23.

[7] Christiano AV, Pean CA, Leucht P, Konda SR, Egol KA. Scoring of radiographic cortical healing with the radiographic humerus union measurement predicts union in humeral shaft fractures. Eur J Orthop Surg Traumatol. 2020; 30: 835-8.

[8] Misir A, Uzun E, Kizkapan TB, Yildiz KI, Onder M, Ozcamdalli M. Reliability of RUST and Modified RUST Scores for the Evaluation of Union in Humeral Shaft Fractures Treated with Different Techniques. Indian J Orthop. 2020; 54: 121-6.

[9] Fordyce W, Kennedy G, Allen JR, Abdelmonem M, Evans J, Evans JT, et al. Validation of the Radiographic Union Score for HUmeral fractures (RUSHU): A retrospective study in an independent centre. Shoulder Elbow. 2023; 15: 390-7.

[10] Leow JM, Oliver WM, Bell KR, Molyneux SG, Clement ND, Duckworth AD. The Radiographic Union Score for ulnar fractures (RUSU) predicts ulnar shaft nonunion. Eur J Orthop Surg Traumatol. 2023; 33: 3167-73.

[11] Guevel B, Gokaraju K, Mohamed F, Sorensen F, Gillott E, Domos P. A comparative study of 6-week and 12-week Radiographic Union Scores for HUmeral fractures (RUSHU) as a predictor of humeral shaft non-union. Shoulder Elbow. 2022; 14: 295-303.

[12] Gupta GK, Mandal S, Rani S, Kumar T, Prasad P, Chakraborty R. Early prediction of non union tibia from post operative follow up 3rd month radiographic union status of tibial fracture (RUST) score: A systematic review and meta analysis. Trends Clin Med Sci. 2023: 693-7.

[13] Porter SM, Dailey HL, Hollar KA, Klein K, Harty JA, Lujan TJ. Automated measurement of fracture callus in radiographs using portable software. J Orthop Res. 2016; 34: 1224-33.

[14] Khan TA, Ling SH. Review on Electrical Impedance Tomography: Artificial Intelligence Methods and its Applications. Algorithms. 2019; 12: 88.

[15] Stojadinovic A, Kyle Potter B, Eberhardt J, Shawen SB, Andersen RC, Forsberg JA, et al. Development of a prognostic naive bayesian classifier for successful treatment of nonunions. J Bone Joint Surg Am. 2011; 93: 187-94.

[16] Reumann MK, Braun BJ, Menger MM, Springer F, Jazewitsch J, Schwarz T, et al. Artificial intelligence and novel approaches for treatment of non-union in bone: From established standard methods in medicine up to novel fields of research. Unfallchirurgie (Heidelb). 2022; 125: 611-8. German.

[17] Takahashi S, Terai H, Hoshino M, Tsujio T, Kato M, Toyoda H, et al. Machine-learning-based approach for nonunion prediction following osteoporotic vertebral fractures. Eur Spine J. 2023; 32: 3788-96.

[18] Degenhart C, Engelhardt L, Niemeyer F, Erne F, Braun B, Gebhard F, et al. Computer-Based Mechanobiological Fracture Healing Model Predicts Non-Union of Surgically Treated Diaphyseal Femur Fractures. J Clin Med. 2023; 12: 3461.

[19] Kalmet PHS, Sanduleanu S, Primakov S, Wu G, Jochems A, Refaee T, et al. Deep learning in fracture detection: a narrative review. Acta Orthop. 2020; 91: 215-20.

[20] Chung SW, Han SS, Lee JW, Oh KS, Kim NR, Yoon JP, et al. Automated detection and classification of the proximal humerus fracture by using deep learning algorithm. Acta Orthop. 2018; 89: 468-73.

[21] Tanzi L, Vezzetti E, Moreno R, Moos S. X-Ray Bone Fracture Classification Using Deep Learning: A Baseline for Designing a Reliable Approach. Appl Sci. 2020; 10: 150.