Spectro-Image Analysis with Vision Graph Neural Networks and Contrastive Learning for Parkinson’s Disease Detection

This study presents a novel framework that integrates Vision Graph Neural Networks (ViGs) with supervised contrastive learning for enhanced spectro-temporal image analysis of speech signals in Parkinson’s disease (PD) detection. The approach introduces a frequency band decomposition strategy that tr...

Full description

Saved in:
Bibliographic Details
Main Authors: Nuwan Madusanka, Hadi Sedigh Malekroodi, H. M. K. K. M. B. Herath, Chaminda Hewage, Myunggi Yi, Byeong-Il Lee
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Journal of Imaging
Subjects:
Online Access:https://www.mdpi.com/2313-433X/11/7/220
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849246323145965568
author Nuwan Madusanka
Hadi Sedigh Malekroodi
H. M. K. K. M. B. Herath
Chaminda Hewage
Myunggi Yi
Byeong-Il Lee
author_facet Nuwan Madusanka
Hadi Sedigh Malekroodi
H. M. K. K. M. B. Herath
Chaminda Hewage
Myunggi Yi
Byeong-Il Lee
author_sort Nuwan Madusanka
collection DOAJ
description This study presents a novel framework that integrates Vision Graph Neural Networks (ViGs) with supervised contrastive learning for enhanced spectro-temporal image analysis of speech signals in Parkinson’s disease (PD) detection. The approach introduces a frequency band decomposition strategy that transforms raw audio into three complementary spectral representations, capturing distinct PD-specific characteristics across low-frequency (0–2 kHz), mid-frequency (2–6 kHz), and high-frequency (6 kHz+) bands. The framework processes mel multi-band spectro-temporal representations through a ViG architecture that models complex graph-based relationships between spectral and temporal components, trained using a supervised contrastive objective that learns discriminative representations distinguishing PD-affected from healthy speech patterns. Comprehensive experimental validation on multi-institutional datasets from Italy, Colombia, and Spain demonstrates that the proposed ViG-contrastive framework achieves superior classification performance, with the ViG-M-GELU architecture achieving 91.78% test accuracy. The integration of graph neural networks with contrastive learning enables effective learning from limited labeled data while capturing complex spectro-temporal relationships that traditional Convolution Neural Network (CNN) approaches miss, representing a promising direction for developing more accurate and clinically viable speech-based diagnostic tools for PD.
format Article
id doaj-art-6b07bc21298c46f08e2ddcbf44de193b
institution Kabale University
issn 2313-433X
language English
publishDate 2025-07-01
publisher MDPI AG
record_format Article
series Journal of Imaging
spelling doaj-art-6b07bc21298c46f08e2ddcbf44de193b2025-08-20T03:58:31ZengMDPI AGJournal of Imaging2313-433X2025-07-0111722010.3390/jimaging11070220Spectro-Image Analysis with Vision Graph Neural Networks and Contrastive Learning for Parkinson’s Disease DetectionNuwan Madusanka0Hadi Sedigh Malekroodi1H. M. K. K. M. B. Herath2Chaminda Hewage3Myunggi Yi4Byeong-Il Lee5Digital Healthcare Research Center, Pukyong National University, Busan 48513, Republic of KoreaIndustry 4.0 Convergence Bionics Engineering, Pukyoung National University, Busan 48513, Republic of KoreaIndustry 4.0 Convergence Bionics Engineering, Pukyoung National University, Busan 48513, Republic of KoreaCardiff School of Technologies, Cardiff Metropolitan University, Cardiff CF23 6PS, UKDigital Healthcare Research Center, Pukyong National University, Busan 48513, Republic of KoreaDigital Healthcare Research Center, Pukyong National University, Busan 48513, Republic of KoreaThis study presents a novel framework that integrates Vision Graph Neural Networks (ViGs) with supervised contrastive learning for enhanced spectro-temporal image analysis of speech signals in Parkinson’s disease (PD) detection. The approach introduces a frequency band decomposition strategy that transforms raw audio into three complementary spectral representations, capturing distinct PD-specific characteristics across low-frequency (0–2 kHz), mid-frequency (2–6 kHz), and high-frequency (6 kHz+) bands. The framework processes mel multi-band spectro-temporal representations through a ViG architecture that models complex graph-based relationships between spectral and temporal components, trained using a supervised contrastive objective that learns discriminative representations distinguishing PD-affected from healthy speech patterns. Comprehensive experimental validation on multi-institutional datasets from Italy, Colombia, and Spain demonstrates that the proposed ViG-contrastive framework achieves superior classification performance, with the ViG-M-GELU architecture achieving 91.78% test accuracy. The integration of graph neural networks with contrastive learning enables effective learning from limited labeled data while capturing complex spectro-temporal relationships that traditional Convolution Neural Network (CNN) approaches miss, representing a promising direction for developing more accurate and clinically viable speech-based diagnostic tools for PD.https://www.mdpi.com/2313-433X/11/7/220Vision Graph Neural Networkssupervised contrastive learningParkinson’s diseasespeech analysisfrequency band decompositionspectro-temporal analysis
spellingShingle Nuwan Madusanka
Hadi Sedigh Malekroodi
H. M. K. K. M. B. Herath
Chaminda Hewage
Myunggi Yi
Byeong-Il Lee
Spectro-Image Analysis with Vision Graph Neural Networks and Contrastive Learning for Parkinson’s Disease Detection
Journal of Imaging
Vision Graph Neural Networks
supervised contrastive learning
Parkinson’s disease
speech analysis
frequency band decomposition
spectro-temporal analysis
title Spectro-Image Analysis with Vision Graph Neural Networks and Contrastive Learning for Parkinson’s Disease Detection
title_full Spectro-Image Analysis with Vision Graph Neural Networks and Contrastive Learning for Parkinson’s Disease Detection
title_fullStr Spectro-Image Analysis with Vision Graph Neural Networks and Contrastive Learning for Parkinson’s Disease Detection
title_full_unstemmed Spectro-Image Analysis with Vision Graph Neural Networks and Contrastive Learning for Parkinson’s Disease Detection
title_short Spectro-Image Analysis with Vision Graph Neural Networks and Contrastive Learning for Parkinson’s Disease Detection
title_sort spectro image analysis with vision graph neural networks and contrastive learning for parkinson s disease detection
topic Vision Graph Neural Networks
supervised contrastive learning
Parkinson’s disease
speech analysis
frequency band decomposition
spectro-temporal analysis
url https://www.mdpi.com/2313-433X/11/7/220
work_keys_str_mv AT nuwanmadusanka spectroimageanalysiswithvisiongraphneuralnetworksandcontrastivelearningforparkinsonsdiseasedetection
AT hadisedighmalekroodi spectroimageanalysiswithvisiongraphneuralnetworksandcontrastivelearningforparkinsonsdiseasedetection
AT hmkkmbherath spectroimageanalysiswithvisiongraphneuralnetworksandcontrastivelearningforparkinsonsdiseasedetection
AT chamindahewage spectroimageanalysiswithvisiongraphneuralnetworksandcontrastivelearningforparkinsonsdiseasedetection
AT myunggiyi spectroimageanalysiswithvisiongraphneuralnetworksandcontrastivelearningforparkinsonsdiseasedetection
AT byeongillee spectroimageanalysiswithvisiongraphneuralnetworksandcontrastivelearningforparkinsonsdiseasedetection