Volumetric Histogram-Based Alzheimer's Disease Detection Using Support Vector Machine.

Area of Science:

• Neurology
• Medical Imaging
• Artificial Intelligence

Background:

• Alzheimer's disease (AD) poses a significant global health challenge.
• Accurate and early diagnosis of AD is crucial for effective management.
• Magnetic Resonance Imaging (MRI) provides detailed brain structure information.

Purpose of the Study:

• To apply machine learning techniques for binary classification of Alzheimer's disease from MRI scans.
• To compare the performance of different supervised learning algorithms for AD detection.
• To identify optimal MRI slices for classification using histogram analysis.

Main Methods:

• Utilized supervised learning algorithms to train classifiers on MRI data.
• Employed histogram analysis on all image slices to identify informative regions.
• Selected specific MRI slices based on highest performance for further analysis.
• Applied majority voting and weighted voting ensemble methods for classification.

Main Results:

• Compared accuracies of various supervised learning classifiers.
• Identified specific MRI slices with the highest discriminative power.
• Achieved a maximum classification accuracy of 69.5% using majority voting.

Conclusions:

• Machine learning, particularly with ensemble methods like majority voting, shows promise in classifying Alzheimer's disease from MRI scans.
• The study highlights the potential of automated analysis of neuroimaging data for AD diagnosis.
• Further research with larger datasets and advanced feature extraction may improve diagnostic accuracy.