Large Margin and Local Structure Preservation Sparse Representation Classifier for Alzheimer's Magnetic Resonance...

Area of Science:

• Neuroimaging
• Machine Learning
• Medical Diagnostics

Background:

• Alzheimer's disease (AD) is a progressive dementia characterized by brain atrophy.
• Early diagnosis of AD is crucial for effective management and treatment.
• Machine learning and MRI show promise for AD detection, but traditional methods have limitations.

Purpose of the Study:

• To develop an improved sparse representation classifier (SRC) for enhanced Alzheimer's disease diagnosis.
• To address the limitations of traditional SRC in capturing global and local image structures.

Main Methods:

• Introduction of a Large Margin and Local Structure Preservation Sparse Representation Classifier (LMLS-SRC).
• Incorporation of a large margin term to ensure compactness of representation coefficients within classes and separation between classes.
• Integration of a local structure preservation term to maintain the manifold structure of the data.
• Application of the ℓ -norm to enhance model sparsity and robustness.

Main Results:

• The LMLS-SRC algorithm demonstrated effective classification performance on the KAGGLE Alzheimer's dataset.
• The method successfully differentiated between non-AD, moderate AD, mild AD, and very mild AD stages.
• Experimental results indicate superior classification accuracy compared to traditional SRC methods.

Conclusions:

• The LMLS-SRC algorithm offers a promising approach for the early and accurate diagnosis of Alzheimer's disease using brain MRI.
• The method's ability to preserve local structure and enforce large margins improves diagnostic capabilities.
• This advancement holds significant clinical value and social importance for managing AD progression.