Authors :
B Swetha Sree; B Nithya Sree; B Nandini; H Shravani; Lakshmi M R
Volume/Issue :
Volume 8 - 2023, Issue 12 - December
Google Scholar :
http://tinyurl.com/4bs8jpxx
Scribd :
http://tinyurl.com/3akxrb59
DOI :
https://doi.org/10.5281/zenodo.10474047
Abstract :
This paper investigates using machine
learning to help diagnose Parkinson's disease early on. It
specifically looks at two aspects - hand movements and
vocal features. Unique datasets for both the symptoms
are explored. Specialized techniques extract the most
distinguishing hand motions and speech characteristics
as biomarkers. Unlike traditional methods relying on one
feature only, this multimodal approach combines both
hand movement and voice biomarkers into one
computational model. Overall, the study illustrates
promise for machine learning tools enabling earlier
intervention for medical purposes, not individual
diagnosis. The focus remains on aiding clinicians rather
than replacing specialized assessments.
Keywords :
Deep Learning, Feature Extraction, Neural Networks, Pattern Recognition, Multimodal Approach, Computational Models.
This paper investigates using machine
learning to help diagnose Parkinson's disease early on. It
specifically looks at two aspects - hand movements and
vocal features. Unique datasets for both the symptoms
are explored. Specialized techniques extract the most
distinguishing hand motions and speech characteristics
as biomarkers. Unlike traditional methods relying on one
feature only, this multimodal approach combines both
hand movement and voice biomarkers into one
computational model. Overall, the study illustrates
promise for machine learning tools enabling earlier
intervention for medical purposes, not individual
diagnosis. The focus remains on aiding clinicians rather
than replacing specialized assessments.
Keywords :
Deep Learning, Feature Extraction, Neural Networks, Pattern Recognition, Multimodal Approach, Computational Models.
