Artificial intelligence improves the accuracy of residents in the diagnosis of hip fractures: a multicenter study

Abstract

Background Less experienced clinicians sometimes make misdiagnosis of hip fractures. We developed computer-aided diagnosis (CAD) system for hip fractures on plain X-rays using a deep learning model trained on a large dataset. In this study, we examined whether the accuracy of the diagnosis of hip fracture of the residents could be improved by using this system. Methods A deep convolutional neural network approach was used for machine learning. Pytorch 1.3 and Fast.ai 1.0 were applied as frameworks, and an EfficientNet-B4 model (a pre-trained ImageNet model) was used. We handled the 5295 X-rays from the patients with femoral neck fracture or femoral trochanteric fracture from 2009 to 2019. We excluded cases in which the bilateral hips were not included within an image range, and cases of femoral shaft fracture and periprosthetic fracture. Finally, we included 5242 AP pelvic X-rays from 4851 cases. We divided these 5242 images into two images per image, and prepared 5242 images including fracture site and 5242 images without fracture site. Thus, a total of 10,484 images were used for machine learning. The accuracy, sensitivity, specificity, F-value, and area under the curve (AUC) were assessed. Gradient-weighted class activation mapping (Grad-CAM) was used to conceptualize the basis for the diagnosis of the fracture by the deep learning algorithm. Secondly, we conducted a controlled experiment with clinicians. Thirty-one residents;young doctors within 2 years of graduation from medical school who rotate through various specialties, were tested using 300 hip fracture images that were randomly extracted from the dataset. We evaluated the diagnostic accuracy with and without the use of the CAD system for each of the 300 images. Results The accuracy, sensitivity, specificity, F-value, and AUC were 96.1, 95.2, 96.9%, 0.961, and 0.99, respectively, with the correct diagnostic basis generated by Grad-CAM. In the controlled experiment, the diagnostic accuracy of the residents significantly improved when they used the CAD system. Conclusions We developed a newly CAD system with a deep learning algorithm from a relatively large dataset from multiple institutions. Our system achieved high diagnostic performance. Our system improved the diagnostic accuracy of residents for hip fractures. Level of evidence Level III, Foundational evidence, before-after study. Clinical relevance: high

Artificial intelligence and machine learning on diagnosis and classification of hip fracture: systematic review

“…In 14 studies, 5 studies used Grad-CAM for highlight important regions. The information on AI for all included studies is presented in Table 3 [ 1 , 8 , 16 , 20 , 21 ].…”

“…In the emergency room, clinicians spend a lot of time and are exposed to mental stress [ 1 ]. There are many things to check due to various images and laboratory tests, and fatigued clinicians (especially residents) are prone to misdiagnosis [ 2 ].…”

“…Early diagnosis and treatment of elderly patients with hip fracture are very important for the clinical course [ 4 ]. Delay in diagnosis or surgery causes complications such as pneumonia and psoa in these patients and increases morbidity and mortality rates [ 1 ]. This not only reduces the patient's quality of life, but also causes economic exhaustion.…”

Automated fracture screening using an object detection algorithm on whole-body trauma computed tomography

“…Furthermore, Sato et al developed a CNN classification model from a relatively large dataset of hip fractures on plain radiographs. The CNN model itself achieved a high diagnostic performance and improved the diagnostic performance of resident doctors (sensitivity of 0.834 without aid to 0.906 with aid p < 0.01; accuracy of 0.847 without aid to 0.912 with aid; p < 0.01) 22 . In a previous automatic localization and classification study of rib fractures (1079 patients) on CT images, the precision of five radiologists improved from 0.803 to 0.911, the sensitivity increased from 0.624 to 0.863, and the diagnosis time was reduced by an average of 73.9 s with artificial intelligence–assisted diagnosis 16 .…”

Imaging, post-processing and navigation: Surgical applications in pelvic fracture treatment

“…Mapping of fracture lines may also provide further information to systematically analyze injury patterns and adjust treatment strategies [22] . Further advancements of visualization may also be an important element of education and training of unexperienced surgeons (23).…”