Alzheimer's disease image classification based on enhanced residual attention network

Area of Science:

• Artificial Intelligence in Medicine
• Neuroscience
• Medical Imaging Analysis

Background:

• Alzheimer's Disease (AD) diagnosis faces challenges with traditional methods, including subjectivity, high costs, and misdiagnosis rates.
• Urgent need for accurate and early detection of Alzheimer's Disease to enable timely intervention.
• Limitations of current diagnostic approaches hinder effective early management of Alzheimer's Disease.

Purpose of the Study:

• To develop and evaluate a deep learning model for the early and accurate detection of Alzheimer's Disease.
• To improve the feature representation and classification accuracy in medical image analysis for Alzheimer's Disease detection.
• To provide a more objective and efficient diagnostic tool for Alzheimer's Disease.

Main Methods:

• Proposed an Enhanced Residual Attention Network (ERAN), a deep learning model for medical image classification.
• Integrated residual learning, attention mechanisms, and soft thresholding to enhance model performance.
• Utilized a dataset of medical images for training and testing the ERAN model's diagnostic capabilities.

Main Results:

• The ERAN model achieved a high accuracy of 99.36% in detecting Alzheimer's Disease.
• Demonstrated a low loss rate of 0.0264, indicating robust model performance.
• Experimental results confirmed the model's excellent performance on the Alzheimer's Disease test dataset.

Conclusions:

• The Enhanced Residual Attention Network (ERAN) shows significant potential for the early diagnosis of Alzheimer's Disease.
• The model's high accuracy and efficiency support its application in clinical settings for Alzheimer's Disease detection.
• This AI-driven approach offers a valuable tool to aid clinicians in the early diagnosis and treatment planning for Alzheimer's Disease.