An efficient diagnosis system for Parkinson's disease using kernel-based extreme learning machine with subtractive...

Area of Science:

• Biomedical Engineering
• Machine Learning for Healthcare
• Neurological Disorder Diagnostics

Background:

• Parkinson's disease (PD) diagnosis remains a challenge, necessitating advanced computational methods.
• Existing diagnostic approaches often lack the required sensitivity and specificity.
• The development of accurate and efficient diagnostic tools is crucial for timely intervention.

Purpose of the Study:

• To introduce a novel hybrid method, SCFW-KELM, for enhanced Parkinson's disease diagnosis.
• To evaluate the impact of kernel functions on the performance of the kernel-based extreme learning machine (KELM) classifier.
• To demonstrate the superiority of the proposed method over existing diagnostic approaches.

Main Methods:

• Integration of subtractive clustering features weighting (SCFW) for data preprocessing to reduce feature variance.
• Application of a fast classifier, kernel-based extreme learning machine (KELM), for PD diagnosis.
• Rigorous evaluation using a PD dataset, assessing classification accuracy, sensitivity, specificity, AUC, f-measure, and kappa statistics via 10-fold cross-validation.

Main Results:

• The SCFW-KELM method achieved exceptional performance metrics: 99.49% classification accuracy, 100% sensitivity, 99.39% specificity, 99.69% AUC, 0.9964 f-measure, and 0.9867 kappa value.
• Demonstrated significant outperformance compared to Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and standard Extreme Learning Machine (ELM) based methods.
• Achieved the highest classification results reported to date for PD diagnosis in the literature.

Conclusions:

• The proposed SCFW-KELM method offers a powerful and effective tool for the diagnosis of Parkinson's disease.
• The hybrid approach significantly enhances diagnostic accuracy and outperforms existing methods.
• SCFW-KELM presents a promising new candidate for clinical application in PD detection.