Skip to content

This project focuses on predicting Parkinson's clinical scores using voice features extracted from the ICEBERG Voice dataset. The dataset, which is not publicly available, includes clinical scores recorded from facial impressions and gait analysis.

Notifications You must be signed in to change notification settings

YSMN-HMT/Praat-Project

Repository files navigation

Project Overview

This project is based on the original description of software for voice feature extraction using Praat, a well-known speech analysis tool. Praat processes voice recordings and associated Textgrids, and by utilizing custom Praat scripts, we extract various voice features for analysis. Check this Link for a paper and software: https://github.com/t-haehnel/MSA-Speech-Analysis-Praat/tree/master

Each patient attended screenings over 3 years to track the progression or diagnosis of Parkinson's disease. Voice recording sessions were conducted at least 4 times per patient, including 29 distinct voice tasks. Praat is used to extract voice features, which are then analyzed through regression models to predict clinical outcomes.

About

This project focuses on predicting Parkinson's clinical scores using voice features extracted from the ICEBERG Voice dataset. The dataset, which is not publicly available, includes clinical scores recorded from facial impressions and gait analysis.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published