Percorrer por autor "Cardoso, Jorge Filipe Pereira"
A mostrar 1 - 1 de 1
Resultados por página
Opções de ordenação
- Machine learning for shoulder rehabilitationPublication . Cardoso, Jorge Filipe Pereira; Rodrigues, Maria de Fátima Coutinho; Cunha, Bruno Miguel AlmeidaThis project addressed the critical challenge of monitoring and evaluating home-based shoulder rehabilitation exercises through the application of Machine Learning tools. The rising prevalence of stroke-related disabilities and the shortage of qualified therapists, exacerbated by the COVID-19 pandemic, underscore the need for innovative solutions. The research explores the feasibility of leveraging Machine Learning models to assess the quality of rehabilitation exercises performed by patients in a home-based setting. Utilizing MediaPipe Pose framework, the study captures and analyzes patients' movements during exercises, comparing them with recordings of healthcare professionals executing the same tasks. The proposed Machine Learning model aims to provide insightful progress reports, aiding therapists in delivering focused and intensive therapy remotely. This project involved designing and implementing Machine Learning models, developing an application to be used by patients and healthcare professionals, prescribing exercises, evaluating the quality of rehabilitation exercises, and testing the solution end-to-end. This project was conducted in a partnership with physiotherapists from Center for Rehabilitation Research from Politécnico do Porto.