Detection, classification, and characterization of proximal humerus fractures on plain radiographs.

The purpose of this study was to develop a convolutional neural network (CNN) for fracture detection, classification, and identification of greater tuberosity displacement ≥ 1 cm, neck-shaft angle (NSA) ≤ 100°, shaft translation, and articular fracture involvement, on plain radiographs. The CNN was trained and tested on radiographs sourced from 11 hospitals in Australia and externally validated on radiographs from the Netherlands. Each radiograph was paired with corresponding CT scans to serve as the reference standard based on dual independent evaluation by trained researchers and attending orthopaedic surgeons. Presence of a fracture, classification (non- to minimally displaced; two-part, multipart, and glenohumeral dislocation), and four characteristics were determined on 2D and 3D CT scans and subsequently allocated to each series of radiographs. Fracture characteristics included greater tuberosity displacement ≥ 1 cm, NSA ≤ 100°, shaft translation (0% to < 75%, 75% to 95%, > 95%), and the extent of articular involvement (0% to < 15%, 15% to 35%, or > 35%). For detection and classification, the algorithm was trained on 1,709 radiographs (n = 803), tested on 567 radiographs (n = 244), and subsequently externally validated on 535 radiographs (n = 227). For characterization, healthy shoulders and glenohumeral dislocation were excluded. The overall accuracy for fracture detection was 94% (area under the receiver operating characteristic curve (AUC) = 0.98) and for classification 78% (AUC 0.68 to 0.93). Accuracy to detect greater tuberosity fracture displacement ≥ 1 cm was 35.0% (AUC 0.57). The CNN did not recognize NSAs ≤ 100° (AUC 0.42), nor fractures with ≥ 75% shaft translation (AUC 0.51 to 0.53), or with ≥ 15% articular involvement (AUC 0.48 to 0.49). For all objectives, the model's performance on the external dataset showed similar accuracy levels. CNNs proficiently rule out proximal humerus fractures on plain radiographs. Despite rigorous training methodology based on CT imaging with multi-rater consensus to serve as the reference standard, artificial intelligence-driven classification is insufficient for clinical implementation. The CNN exhibited poor diagnostic ability to detect greater tuberosity displacement ≥ 1 cm and failed to identify NSAs ≤ 100°, shaft translations, or articular fractures.

Informations de copyright

© 2024 The British Editorial Society of Bone & Joint Surgery.

Déclaration de conflit d'intérêts

R. W. A. Spek received payments of an amount between USD 10,000 and USD 100,000 from the Flinders Foundation (Adelaide, Australia) for the purpose of this study. During the study period, R. W. A. Spek received payments of an amount between USD 10,000 and USD 100,000 from Prins Bernhard Cultuurfonds (Amsterdam, the Netherlands), Stichting Zabawas (The Hague, The Netherlands), and with an amount of less than USD 10,000 from Michael van Vloten Foundation (Rotterdam, The Netherlands) and Anna Fonds NOREF (Mijdrecht, the Netherlands), all of which were unrelated to this specific study. B. Jadav provided paid consultations for Johnson & Johnson, unrelated to the current study. G. I. Bain received a Flinders Foundation grant for this study, paid to Flinders University, as well as royalties or licenses from Fusetec, stock or stock options in Fusetec, and speaker payments or honoraria from Depuy Synthes and Medartis, none of which are related to this study. G. I. Bain also holds fiduciary roles in the Australian Hand Surgery Society, the Shoulder and Elbow Society of Australia, the Asia Pacific Wrist Association, the International Federation for the Societies for Surgery of the Hand, and the Journal of Wrist Surgery. J. L. Jaarsma is an unpaid executive of the Australian Orthopaedic Association. B. Jadav received a one-off consultation payment and payment for teaching courses from Johnson & Johnson, unrelated to this study. M. P. J. van den Bekerom receives fellowship support from Smith & Nephew, which contributed to this study.