LADNET: An MRI-based deep learning approach for Alzheimer's disease detection.

Area of Science:

• Neurology
• Medical Imaging
• Artificial Intelligence

Background:

• Alzheimer's disease (AD) is the most prevalent form of dementia, characterized by progressive cognitive decline.
• Early diagnosis of AD is crucial for improving patient survival rates and management.
• Deep learning (DL) shows promise in analyzing complex medical images for disease identification.

Purpose of the Study:

• To design and evaluate the Lightweight Alzheimer's Disease Net (LADNET) model for accurate AD detection using MRI scans.
• To classify different dementia severities: Mild Dementia (MD), Moderate Dementia (MDD), and Very Mild Dementia (VMD), alongside Non-Demented (ND) individuals.
• To assess LADNET's performance against existing methods on a large, publicly available dataset.

Main Methods:

• Development of the LADNET model, a lightweight deep learning architecture.
• Utilizing magnetic resonance imaging (MRI) scans for AD detection and classification.
• Performance evaluation on a Kaggle dataset comprising over six thousand images across four categories.

Main Results:

• LADNET achieved 99.4% accuracy and 99% AUC, outperforming existing methods.
• The model demonstrated a lightweight design with approximately 1.2 million parameters.
• Achieved fast inference speed of 4.2 ms per image.

Conclusions:

• LADNET shows strong potential for clinical deployment in early Alzheimer's disease detection.
• The model's efficiency and accuracy can aid in extracting biomarker information from MRIs.
• LADNET may contribute to reducing patient burden and diagnostic costs in AD management.