NCA-EVA: An Innovative Ensemble-Based Approach for Alzheimer's Disease Detection from Magnetic Resonance Imaging

Area of Science:

• Neuroscience
• Computer Science
• Medical Imaging

Background:

• Alzheimer's disease (AD) is a progressive neurodegenerative disorder affecting over 55 million people globally, with diagnosis challenges, especially in early stages.
• The rising prevalence of AD necessitates advanced diagnostic tools to improve speed and accuracy.
• Artificial intelligence (AI) is crucial for developing sophisticated computer-aided diagnosis (CADx) systems in healthcare.

Purpose of the Study:

• To propose the NCA-Enhanced Voting Algorithm for Alzheimer's Classification (NCA-EVA) for rapid and accurate AD diagnosis.
• To evaluate the performance of NCA-EVA in classifying four distinct stages of Alzheimer's disease.
• To assess the computational efficiency of NCA-EVA compared to existing methods.

Main Methods:

• Development and implementation of the NCA-Enhanced Voting Algorithm for Alzheimer's Classification (NCA-EVA).
• Training of 66 models for four-class AD data and 6 models for two-class AD data.
• Performance evaluation based on classification accuracy and processing time.

Main Results:

• NCA-EVA achieved high accuracy in classifying Alzheimer's disease stages: 98.97% for four-class and 99.87% for binary classification.
• The algorithm demonstrated exceptional speed, with a processing time of 1.26 seconds.
• NCA-EVA is approximately 1200 times faster than comparable studies utilizing NCA-based feature selection.

Conclusions:

• The proposed NCA-EVA enables rapid and highly accurate diagnosis of Alzheimer's disease.
• This AI-driven approach offers a significant advancement in computer-aided diagnosis for neurodegenerative disorders.
• NCA-EVA shows potential for broader applications in healthcare data analysis.