Detecting Ankle Fractures in Plain Radiographs Using Deep Learning with Accurately Labeled Datasets Aided by Computed Tomography: A Retrospective Observational Study

Abstract

Ankle fractures are common and, compared to other injuries, tend to be overlooked in the emergency department. We aim to develop a deep learning algorithm that can detect not only definite fractures but also obscure fractures. We collected the data of 1226 patients with suspected ankle fractures and performed both X-rays and CT scans. With anteroposterior (AP) and lateral ankle X-rays of 1040 patients with fractures and 186 normal patients, we developed a deep learning model. The training, validation, and test datasets were split in a 3/1/1 ratio. Data augmentation and under-sampling techniques were administered as part of the preprocessing. The Inception V3 model was utilized for the image classification. Performance of the model was validated using a confusion matrix and the area under the receiver operating characteristic curve (AUC-ROC). For the AP and lateral trials, the best accuracy and AUC values were 83%/0.91 in AP and 90%/0.95 in lateral. Additionally, the mean accuracy and AUC values were 83%/0.89 for the AP trials and 83%/0.9 for the lateral trials. The reliable dataset resulted in the CNN model providing higher accuracy than in past studies.

Multiclass datasets expand neural network utility: an example on ankle radiographs

“…In this study, we introduced a dataset of annotated ankle x-rays and demonstrated the effect of its multiclass labeling by performing network training on the task of automated fracture detection on distinct subsets of images, thus quantifying the influence of selected image features. By utilizing customized state-of-the-art preprocessing and augmentation methods on both the images themselves as well as the composition of the training data set, our study performed better compared to most contemporary ankle studies, including both convolutional neural networks [ 16 , 17 ] and traditional machine learning methods [ 18 ]. For convolutional neural networks, the prevalent training protocol incorporates random dataset augmentation as performed in our study as well as triplication of the medical image for use on imagenet-pretrained networks.…”

“…Kim and colleagues achieved a mean AUC of 0.89 on ap-views using the pretrained Inceptionv3-network architecture without layer customization on a similarly sized dataset of 1226 examinations and a comparable augmentation scheme [ 16 ]. The ratio of positive to negative images was almost inverted with mostly fractured ankles (85% against 15%), which for a binary classification task should result in a comparable sampling bias.…”

Development of convolutional neural network model for diagnosing osteochondral lesions of the talus using anteroposterior ankle radiographs

“…Regarding previous DL models for diagnosing ankle pathologies using ankle radiographs as input data, only models for diagnosing ankle fractures have been developed. [19][20][21] The accuracies were reported to be approximately 80% to 90%. To the best of our knowledge, this study is the first to demonstrate the possibility of diagnosing OLTs using DL algorithms trained using ankle radiographs.…”

Harnessing ResNet50 and SENet for enhanced ankle fracture identification

“…In light of recent advancements in deep learning for ankle fracture detection, studies [ 21 ] and [ 22 ] have demonstrated the efficacy of employing deep convolutional neural networks (DCNNs) with radiographic images. Study [ 21 ] reported a high sensitivity of 98.7% and specificity of 98.6% using Inception V3 and ResNet-50 on radiographs, emphasizing the potential of DCNNs to accurately identify fractures from multiple views.…”