Screening of adolescent idiopathic scoliosis using generative adversarial network (GAN) inversion method in chest radiographs

• Published in: PloS one 2023-05, Vol.18 (5), p.e0285489-e0285489
• Main Authors: Lee, Jun Soo, Shin, Keewon, Ryu, Seung Min, Jegal, Seong Gyu, Lee, Woojin, Yoon, Min A, Hong, Gil-Sun, Paik, Sanghyun, Kim, Namkug
• Format: Article
• Language: English
• Subjects: Ablation; Adolescent; Adolescents; Algorithms; Artificial neural networks; Automation; Biology and Life Sciences; Chest; Computer and Information Sciences; Datasets; Deep learning; Diagnosis; Diagnosis, Computer-Assisted - methods; Generative adversarial networks; Health aspects; Humans; Inversion; Kyphosis; Machine learning; Medicine and Health Sciences; Methods; Modelling; Multilayer perceptrons; Multilayers; Neural networks; Neural Networks, Computer; Physical Sciences; Physiological aspects; Radiographs; Radiography; Research and Analysis Methods; Scoliosis; Scoliosis - diagnostic imaging; Screening; Social Sciences; Teenagers; Young adults; Youth
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0285489

Conventional computer-aided diagnosis using convolutional neural networks (CNN) has limitations in detecting sensitive changes and determining accurate decision boundaries in spectral and structural diseases such as scoliosis. We devised a new method to detect and diagnose adolescent idiopathic scoliosis in chest X-rays (CXRs) employing the latent space's discriminative ability in the generative adversarial network (GAN) and a simple multi-layer perceptron (MLP) to screen adolescent idiopathic scoliosis CXRs. Our model was trained and validated in a two-step manner. First, we trained a GAN using CXRs with various scoliosis severities and utilized the trained network as a feature extractor using the GAN inversion method. Second, we classified each vector from the latent space using a simple MLP. The 2-layer MLP exhibited the best classification in the ablation study. With this model, the area under the receiver operating characteristic (AUROC) curves were 0.850 in the internal and 0.847 in the external datasets. Furthermore, when the sensitivity was fixed at 0.9, the model's specificity was 0.697 in the internal and 0.646 in the external datasets. We developed a classifier for Adolescent idiopathic scoliosis (AIS) through generative representation learning. Our model shows good AUROC under screening chest radiographs in both the internal and external datasets. Our model has learned the spectral severity of AIS, enabling it to generate normal images even when trained solely on scoliosis radiographs.