Posture Abnormality Classification Using Computer Vision–Based Gait Analysis: A Clinical Decision Support Framework

Abstract

The spinal deformities that comprise Adolescent Idiopathic Scoliosis (AIS), kyphosis, and lordosis have posed a major challenge to orthopedic medical facilities, and their longitudinal assessment is a traditionally language that uses ionizing radiography. The suggested paper is proposing a new system of classifying these abnormal posture by using Computer Vision based gait analysis in a Clinical Decision Support System(CDSS). We look at the development of dynamic analysis of motion on the surface statistics to the current literature to develop a systematic review of the analysis. We will consider the flaws of the existing standards of diagnosis, namely the risks of multiple radiation exposures in adolescent and aging ages. Moreover, we discuss the combination of the concepts of Artificial Intelligence (AI) and Explainable AI (XAI) to improve the diagnostic accuracy, yet still winning the clinical trust. It is suggested that vision-based systems that are non-invasive are a viable alternative to mass screening and continuous monitoring of spinal pathologies and have a reasonable cost.