Citation

Gurucharan, J. and Padmapriya, M. and Grace, M. Rishi and Krishna, S. Barath and Elamaran, V. and Chinnaiyan, Senthilpari (2025) Machine Learning Networks' Potential for Parkinson's Disease Diagnosis Without Coding Experience. In: 2025 Multimedia University Engineering Conference, MECON 2025, 21 July 2025 - 23 July 2025, Cyberjaya, Malaysia.

Text 445.pdf - Published Version Restricted to Repository staff only Download (1MB)

Abstract

Millions of people worldwide suffer from Parkinson's disease (PD), a chronic brain ailment. It happens when dopamine, a neurotransmitter that regulates movement, is produced by dying or injured brain cells. This results in PD, which impairs posture, balance, and movement. Early identification is essential to reduce the disease's course and enhance PD sufferers' quality of life. Google's Teachable Machine (GTM) automatically provides diagnostic performance and builds a classification network when images are uploaded. GTM uses a pre-trained model (MobileNetV2) for image classification, leveraging transfer learning. Four different handwriting datasets, such as Parkinson’s Drawings (PD), Augmented PD, HandPD, and NewHandPD, are used in this study. We implemented 13 teachable machine deep learning (DL) models to assess diverse hand-drawn patterns, including spirals, meanders, circles, waves, and combinations, in the original samples from the distinct datasets. DL model 5 achieves 99.59% accuracy, 99.19% precision, 100% sensitivity, and 99.60% F1-score in the classification of PD patients.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	Artificial Intelligence, Deep Learning, Machine Learning, Parkinson’s disease, Teachable Machine.
Subjects:	Q Science > QA Mathematics > QA71-90 Instruments and machines
Divisions:	Faculty of Artificial Intelligence & Engineering (FAIE)
Depositing User:	Ms Suzilawati Abu Samah
Date Deposited:	19 Mar 2026 01:59
Last Modified:	19 Mar 2026 01:59
URII:	http://shdl.mmu.edu.my/id/eprint/15607

Downloads

Edit (login required)