AI-based tool for early detection of Alzheimer's disease

Area of Science:

• Neuroimaging
• Artificial Intelligence
• Medical Diagnostics

Background:

• Alzheimer's disease (AD) is a progressive neurodegenerative disorder where early detection is crucial for management.
• Current diagnostic methods can be invasive or lack sensitivity for early-stage cognitive impairment.
• The hippocampus is an early-affected brain region critical for memory and a potential biomarker for AD.

Purpose of the Study:

• To develop and validate a novel AI-driven approach for early detection of Alzheimer's disease.
• To leverage the hippocampus region and VGG16 model with transfer learning for improved classification accuracy.
• To create a practical tool for radiologists to aid in AD diagnosis and progression tracking.

Main Methods:

• Utilized the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset.
• Employed the VGG16 convolutional neural network model with transfer learning.
• Incorporated advanced image preprocessing and a progressive data augmentation technique focusing on the hippocampus.

Main Results:

• Achieved high accuracy in classifying patients into cognitively normal (CN), mild cognitive impairment (MCI), and AD categories.
• Reported testing accuracy of 98.17%, validation accuracy of 97.52%, and training accuracy of 99.62%.
• Demonstrated enhanced model generalization through progressive data augmentation.

Conclusions:

• The proposed hippocampus-focused VGG16 model with transfer learning offers a highly accurate method for early AD detection.
• The developed AI tool, accessible via a website, empowers radiologists with predictive capabilities and visualization tools.
• This approach significantly contributes to advancing early diagnosis and management strategies for Alzheimer's disease.