Application of a deep learning algorithm for detection and visualization of hip fractures on plain pelvic radiographs

Abstract

Objective To identify the feasibility of using a deep convolutional neural network (DCNN) for the detection and localization of hip fractures on plain frontal pelvic radiographs (PXRs). Summary of background data Hip fracture is a leading worldwide health problem for the elderly. A missed diagnosis of hip fracture on radiography leads to a dismal prognosis. The application of a DCNN to PXRs can potentially improve the accuracy and efficiency of hip fracture diagnosis. Methods A DCNN was pretrained using 25,505 limb radiographs between January 2012 and December 2017. It was retrained using 3605 PXRs between August 2008 and December 2016. The accuracy, sensitivity, false-negative rate, and area under the receiver operating characteristic curve (AUC) were evaluated on 100 independent PXRs acquired during 2017. The authors also used the visualization algorithm gradient-weighted class activation mapping (Grad-CAM) to confirm the validity of the model. Results The algorithm achieved an accuracy of 91%, a sensitivity of 98%, a false-negative rate of 2%, and an AUC of 0.98 for identifying hip fractures. The visualization algorithm showed an accuracy of 95.9% for lesion identification. Conclusions A DCNN not only detected hip fractures on PXRs with a low false-negative rate but also had high accuracy for localizing fracture lesions. The DCNN might be an efficient and economical model to help clinicians make a diagnosis without interrupting the current clinical pathway. Key Points • Automated detection of hip fractures on frontal pelvic radiographs may facilitate emergent screening and evaluation efforts for primary physicians. • Good visualization of the fracture site by Grad-CAM enables the rapid integration of this tool into the current medical system. • The feasibility and efficiency of utilizing a deep neural network have been confirmed for the screening of hip fractures .

X-Ray Bone Fracture Classification Using Deep Learning: A Baseline for Designing a Reliable Approach

“…To be clearer, an initial network that classifies between fracture and no-fracture and a subsequent one that takes the images predicted to be fracture and classifies them into three types (A, B and C). Finally, class activation mapping or similar technologies should be used to see where the network is focusing [21,26,28]. A clear example is found in [30], where the author understood, with the help of CAM, that the network was learning the wrong features.…”

“…This subsection focuses on the work of Cheng et al [21]. Aim: The aim of this work is to use a CNN (pre-trained with a dataset of medical images) to classify and localize hip fractures in plain frontal pelvic radiographs (PXRs).…”

Pattern Recognition in Musculoskeletal Imaging Using Artificial Intelligence

“…Cheng et al reported a deep CNN pretrained on limb radiographs that achieved 91% accuracy, 98% sensitivity, 2% false-negative rate, and an AUC of 0.98 for hip fractures on frontal pelvis radiographs. 56 Like many other articles in the literature, the authors also used gradient-weighted class activation mapping (i.e., a saliency map) to confirm that the pixels considered class discriminative were indeed over the fracture site, demonstrating an accuracy of 96% for fracture site localization (►Fig. 7).…”

Artificial intelligence system can achieve comparable results to experts for bone age assessment of Chinese children with abnormal growth and development

“…Artificial intelligence (AI), which has high potential in reducing labor requirement and intra-and inter-observer variations, is gaining popularity in medical field, especially in radiology (Cheng et al, 2019;Hosny et al, 2018). Deep learning, one of the advanced AI techniques, which can automatically learn features from images, has become a hot spot in recent years.…”