AI Powered Mobile Analysis of Scoliosis among Children in Qinghai- Tibetan Plateau of China

Adolescent idiopathic scoliosis (AIS) represents the foremost spinal disorder in children. The conventional clinical benchmark for AIS severity assessment relies on the Cobb angle measured from coronal X-ray images. However, in remote areas with constrained medical resources, the absence of X-ray facilities and specialized expertise poses significant challenges. In order to address this gap, our study introduces an accessible radiation-free optical assessment method embodied within the mobile application, AlignProCARE. This application facilitates the acquisition of RGBD images of patients' bare backs through smartphones, subsequently analyzed by our artificial intelligence algorithm, ScolioNet, to determine AIS severity. Tested on 80 patients across the Qinghai-Tibet Plateau, AlignProCARE demonstrated exceptional performance, achieving sensitivity and specificity rates exceeding 0.974, indicating its potential as an accessible tool for AIS evaluation in resource-limited regions.