Advanced MRI based Alzheimer's diagnosis through ensemble learning techniques.

Area of Science:

• Neuroimaging
• Artificial Intelligence
• Medical Diagnostics

Background:

• Alzheimer's Disease (AD) is a progressive neurodegenerative disorder impacting cognition and daily function.
• Early diagnosis of AD is crucial for effective management and treatment strategies.
• Advancements in MRI technology and machine learning offer new avenues for AD detection.

Purpose of the Study:

• To evaluate the efficacy of deep learning models, specifically Convolutional Neural Networks (CNNs), in diagnosing and staging Alzheimer's Disease using brain MRI data.
• To compare the diagnostic performance of individual pre-trained models (ResNet50, InceptionResNetv2) and a custom-trained CNN against a combined model approach.

Main Methods:

• Utilized brain MRI datasets to train and validate deep learning models.
• Employed Convolutional Neural Networks (CNNs), including ResNet50 and InceptionResNetv2, for pattern recognition in neuroimaging.
• Developed and tested a hybrid model combining multiple CNN architectures for enhanced diagnostic accuracy.

Main Results:

• The custom-trained CNN achieved 90.76% accuracy, ResNet50 reached 90.27%, and InceptionResNetv2 achieved 86.84% accuracy individually.
• The collaborative model, integrating all three architectures, demonstrated a superior accuracy rate of 94.27% in diagnosing Alzheimer's Disease stages.
• Combined model approach significantly outperformed individual models in identifying Alzheimer's-related patterns in MRI scans.

Conclusions:

• Deep learning models, particularly when combined, show significant promise for accurate and automated diagnosis of Alzheimer's Disease from MRI data.
• The integrated model approach offers enhanced diagnostic capabilities for staging Alzheimer's Disease, potentially improving patient outcomes.
• This study highlights the potential of AI in revolutionizing early detection and management of neurodegenerative conditions like Alzheimer's.