Enhancing Rehabilitation Outcomes with DynTherapy: An AI-Driven Personalized Approach

The increasing demand for accessible and personalized rehabilitation services has led to the development of Dyn-Therapy, an AI-driven application designed to deliver effective physical therapy at home. By leveraging machine learning and a diverse exercise library, DynTherapy provides real-time feedback and tailored guidance, overcoming the limitations of traditional at-home therapy methods. This paper details DynTherapy's methodology, including the creation of a novel dataset, data preprocessing, and the implementation of a k-nearest neighbors (k-NN) classifier for pose classification. Our results demonstrate the model's high accuracy, achieving approximately 98% in classifying physical therapy exercises, highlighting its potential to enhance rehabilitation outcomes and overall physical well-being. DynTherapy represents a cost-effective, accessible, and personalized solution for home-based rehabilitation, promising more effective therapy for patients.