A convolutional neural network to detect scoliosis treatment in radiographs

Abstract

The aim of this work is to propose a classification algorithm to automatically detect treatment for scoliosis (brace, implant or no treatment) in postero-anterior radiographs. Such automatic labelling of radiographs could represent a step towards global automatic radiological analysis. Methods Seven hundred and ninety-six frontal radiographies of adolescents were collected (84 patients wearing a brace, 325 with a spinal implant and 387 reference images with no treatment). The dataset was augmented to a total of 2096 images. A classification model was built, composed by a forward convolutional neural network (CNN) followed by a discriminant analysis; the output was a probability for a given image to contain a brace, a spinal implant or none. The model was validated with a stratified tenfold cross-validation procedure. Performance was estimated by calculating the average accuracy. Results 98.3% of the radiographs were correctly classified as either reference, brace or implant, excluding 2.0% unclassified images. 99.7% of brace radiographs were correctly detected, while most of the errors occurred in the reference group (i.e. 2.1% of reference images were wrongly classified). Conclusion The proposed classification model, the originality of which is the coupling of a CNN with discriminant analysis, can be used to automatically label radiographs for the presence of scoliosis treatment. This information is usually missing from DICOM metadata, so such method could facilitate the use of large databases. Furthermore, the same model architecture could potentially be applied for other radiograph classifications, such as sex and presence of scoliotic deformity.

Current and emerging artificial intelligence applications for pediatric musculoskeletal radiology

“…Other authors have assessed the ability of AI to predict scoliosis progression [74], assess the Risser stage [75], detect evidence of scoliosis treatment from radiographs [76] and to automate three-dimensional (3-D) spine reconstructions from biplanar images [77].…”

Using deep transfer learning to detect scoliosis and spondylolisthesis from x-ray images

“…Wang et al [ 15 ] designed a deep learning model to differentiate between progressive (P) and non-progressive (NP) classes at first clinic visit. Vergari et al [ 23 ] combined CNN with discriminate analysis to determine the type of scoliosis treatment appearing the X-ray image (i.e., brace, spinal implant, or neither). Although their study did not aim to diagnose scoliosis, the authors claim that their work will facilitate the processing of large databases for such research purposes.…”

Artificial Intelligence Accurately Detects Traumatic Thoracolumbar Fractures on Sagittal Radiographs

“…The recent explosion of using labeled data, namely ‘big data’, has brought upon the era of artificial intelligence (AI) into the field of medical diagnostics and imaging, which has particularly benefitted from the application of AI based innovations [ 18 ]. Regarding imaging of the spine, promising results in the assessment of degenerative disorders [ 19 ], adult deformities [ 20 ] and adolescent idiopathic scoliosis [ 21 ], as well as in the detection of primary and secondary bone tumors [ 22 , 23 ], and vertebral fractures [ 24 ] have been recently published. Deep learning (DL) is a machine learning method that uses an algorithmic structure most commonly based on neural networks, such as convolutional neural networks.…”