p-Scan: An application for diagnosing Parkinson's disease
November 2019 | ML, Healthcare
Problem
People in India, especially in the rural areas are not able to get the basic health facilities and in many cases the diseases go undetected. With mental health being considered as a taboo in our society many people don't even want to get diagnosed because of the societal pressure. In these cases an app can help in bridging this gap taking healthcare to their doorsteps, thus improving the conditions significantly.
Goal
To create a tool that can detect Parkinson's analyzing a persons voice and help them connect their values and individual circumstances with their treatment choice.
Role
Speech Processing
Machine Learning
Advisor
Prof. Aswani Kumar Cherukuri
About
pScan is an android application to identify whether a user has Parkinson’s Disease. It uses a non-invasive technique of analysing the voice. From the voice, over 20 features are extracted. These features are used to diagnose the disease. It is also capable of indicating where the user lies on the UPDRS.
Application Overview
In addition to detecting Parkinson’s the app also suggested the nearby treatment centres where the patient can go and get him checked for confirmation following which he can continue to receive treatment for the same. Snapshots attached below:
How does it even work ?
