Effective Alzheimer's disease detection using enhanced Xception blending with snapshot ensemble

Area of Science:

• Neuroscience
• Medical Imaging
• Artificial Intelligence

Background:

• Alzheimer's disease (AD) is a progressive neurodegenerative disorder causing dementia and cognitive decline.
• Early detection of AD is crucial for timely intervention and management, though challenging with traditional methods.
• Deep learning (DL) shows promise in enhancing AD detection through analysis of brain imaging data.

Purpose of the Study:

• To propose and evaluate an ensemble deep learning methodology for detecting Alzheimer's disease from brain MRI scans.
• To improve diagnostic accuracy and prediction of disease progression compared to conventional techniques.

Main Methods:

• An enhanced Xception architecture was utilized to generate multiple model snapshots for diverse MRI feature analysis.
• A decision-level fusion strategy combined decision scores with a Random Forest (RF) meta-learner via a blending algorithm.

Main Results:

• The proposed ensemble methodology achieved a high accuracy of 99.14% in categorizing Alzheimer's disease into four distinct groups.
• Experimental findings confirmed the efficacy of the ensemble technique in AD detection from MRI data.

Conclusions:

• The developed ensemble DL methodology offers a highly accurate approach for early Alzheimer's disease detection.
• This technique has the potential to support medical practitioners in providing individualized care plans for AD patients.
• Future work will focus on improving the model's generalization across diverse datasets.