Posture detection for physical therapy application

The recent Covid-19 pandemic has greatly affected the way people interact and has caused widespread social disruptions. A survey conducted by the world physiotherapy member organization showed that over 70% of physiotherapy services were impacted by the pandemic, primarily due to the harsh lockdowns...

Full description

Saved in:
Bibliographic Details
Main Author: Yeo, Nigel
Other Authors: Pui Tze Sian
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2023
Subjects:
Online Access:https://hdl.handle.net/10356/166826
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:The recent Covid-19 pandemic has greatly affected the way people interact and has caused widespread social disruptions. A survey conducted by the world physiotherapy member organization showed that over 70% of physiotherapy services were impacted by the pandemic, primarily due to the harsh lockdowns imposed by many countries and the fear of patients contracting the virus. This has led to a shortage of physiotherapists and has made it difficult for many patients to access timely and consistent physiotherapy treatment. To address these challenges, there is a need for a cost-effective and remote physiotherapy solution for patients who are unable to access in-person treatment. This project aimed to develop such a solution by creating a real-time posture detector using the Body Detection package for a physiotherapy mobile application. Data was collected and used to optimize a pre-trained machine learning model which was utilized to classify poses and developed pose estimation algorithms for physiotherapy applications. The application was developed using Flutter SDK and Dart programming language and leverages the Body Detection package to detect the spatial locations of key body points as well as their geometrical angles. The development of pose estimation algorithms represents a major shift in the assessment and analysis of human movement. By utilizing advanced computer vision, these algorithms can track human motion in real-time through video footage captured by common, low-cost devices such as smartphones, tablets, and laptops. This makes it possible to provide remote and affordable physiotherapy treatment to patients who are unable to access in-person treatment. The results of the tests conducted on the application demonstrate its effectiveness in accurately detecting the posture of the user. This project has successfully developed a solution that can provide remote and affordable physiotherapy treatment and has the potential to positively impact the provision of physiotherapy services.