Review on computational methods for the detection and classification of Parkinson's Disease

Computers in Biology and Medicine, Journal Year: 2025, Volume and Issue: 187, P. 109767 - 109767

Published: Feb. 11, 2025

Language: Английский

Deep Learning-Based Feature Extraction and Machine Learning Models for Parkinson's Disease Prediction Using DaTscan Image

Biomedical & Pharmacology Journal, Journal Year: 2025, Volume and Issue: 18(December Spl Edition), P. 161 - 177

Published: Jan. 20, 2025

Parkinson's disease (PD) is a chronic, non-fatal, and well-known progressive neurological disorder, the symptoms of which often overlap with other diseases. Effective treatment diseases also requires accurate early diagnosis, way that patients can lead healthy productive lives. The main PD signs are resting tremors, muscular rigidity, akinesia, postural instability, non-motor signs. Clinician-filled dynamics have traditionally been an essential approach to monitoring evaluating Disease using checklists. Accurate timely diagnosis (PD), chronic ailment, be difficult due its overlapping those disorders. therapy improvement in quality life for depend on detection. To improve classification performance, this study investigates transfer learning, uses pre-trained models extract features from massive datasets. Transfer learning improves generalization permits domain adaptation, especially small or resource-constrained datasets, while lowering training time, resource needs, overfitting concerns. This work aims design assess general paradigm reliable prognosis Parkinson’s based DaTscan images consider feature extraction performance variety ML algorithms. explore use deep order accuracy. sample made up 594 68 participants, 43 26 healthy. Out four algorithms employed; Random Forest, Neural Network, Logistic Regression, Gradient Boosting models, learning-based were applied. Four indices accuracy, namely Area Under Curve (AUC), Classification Accuracy (CA), F1 Score, Precision, Recall Matthews Correlation Coefficient (MCC) used evaluate machine task such as Boosting. networks outperformed showing robustness reliability AUC 0.996, CA 0.973, MCC 0.946. performed competitively, coming second 0.995 0.925. Forest worst, 0.986 0.905, whereas Regression had 0.991 0.926. These results demonstrate how well neural perform high-precision tasks point gradient boosting more computationally effective option.

Language: Английский

A new network structure for Parkinson's handwriting image recognition

Medical Engineering & Physics, Journal Year: 2025, Volume and Issue: unknown, P. 104333 - 104333

Published: April 1, 2025

Language: Английский