Search

This study aimed to develop and evaluate a convolutional neural network for identifying scaphoid fractures on radiographs. A dataset of 1918 wrist radiographs (600 patients) was taken from an orthopaedic referral centre between 2010 to 2020. A YOLOv3 and a MobileNetV3 convolutional neural network were trained for scaphoid detection and fracture classification, respectively. The diagnostic performance of the convolutional neural network was compared with the majority decision of four hand surgeons. The convolutional neural network achieved a sensitivity of 82% and specificity of 94%, with an area under the receiver operating characteristic of 92%, whereas the surgeons achieved a sensitivity of 76% and specificity of 96%. The comparison indicated that the convolutional neural network's performance was similar to the majority vote of surgeons. It further revealed that convolutional neural network could be used in identifying scaphoid fractures on radiographs reliably, and has potential to achieve the expert-level performance. Level of evidence: III [ABSTRACT FROM AUTHOR]

Objective: Studies exploring the epidemiological characteristics of scaphoid fracture nonunion are important to understand the causes and lead to effective prevention strategy. However, such knowledge is limited especially in China. This study aims to reveal the clinical and epidemiological features of scaphoid fracture nonunion in Chinese patients. Methods: This was a retrospective study, which was based on patients with scaphoid fracture nonunion treated in a Chinese national orthopedic referral center from 1 August 2009 to 1 August 2020. The basic demographic and clinical characteristics, including gender, age at diagnosis of scaphoid fracture nonunion, dominant hand, the location of fracture, the side, the causes of the injury, age of the injury, age of the treatment for scaphoid fracture nonunion, and the delay period, were retrieved from the medical records. The patients were divided into a non‐delay group and a delay group based on whether they had visited the doctor within 1 month after the injury. Scaphoid fracture was classified into the proximal, waist, and distal fractures according to the location. The t‐test or Mann–Whitney U test was used to compare the difference between groups when the dependent variable is continuous, while a chi‐square test was used when the dependent variable is categorical. Results: A total of 363 patients were included in this study. The mean age at the diagnosis of scaphoid fracture nonunion was 32 (ranging from 14 to 78). Waist fracture nonunion (76.5%) was most common. Among the 169 patients who sought medical care within 1 month after the injury (non‐delay group), more than half (90 cases) were misdiagnosed. 23 cases could not recall the injury, so the time of delay could not be determined. Among the 171 patients who did not seek medical care for acute injury (delay group), the average time was 52 months between the initial injury and the treatment. The proportion of waist nonunions was higher among patients in the delay group than that in the non‐delay group. Conclusions: Nonunion of scaphoid fracture is caused mainly by delayed treatment and misdiagnosis in China, suggesting that timely treatment and improved diagnosis could lower the prevalence. The fracture location may be a factor contributing to the delaying of seeking medical care. [ABSTRACT FROM AUTHOR]