Classification of Alzheimer's Disease in MRI based on Dictionary Learning and Heavy Tailed Modelling

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Medical Image Analysis

Background:

• Diagnosing brain diseases from anatomical images is complex and time-consuming for experts.
• Automated computational tools can enhance the speed and accuracy of brain disease diagnosis.
• Structural Magnetic Resonance Imaging (sMRI) is a key modality for visualizing brain structure.

Purpose of the Study:

• To develop and evaluate a computational model for classifying structural Magnetic Resonance Imaging (sMRI) scans.
• To assess the effectiveness of dictionary learning for feature extraction in brain MRI classification.
• To improve the accuracy and efficiency of automated brain disease diagnosis.

Main Methods:

• Utilized a Support Vector Machine (SVM) classifier for sMRI scan classification.
• Employed dictionary learning to create a feature space for analysis.
• Compared classification performance using learned dictionaries versus a predefined dictionary.
• Tested the framework on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset.

Main Results:

• Features learned via dictionary learning demonstrated superior performance in classification tasks.
• The proposed framework showed promising results in classifying sMRI scans.
• Learned features improved the model's ability to distinguish between different conditions.

Conclusions:

• Dictionary learning is an effective approach for extracting relevant features from brain MRI data.
• Automated classification models using learned features can aid in the diagnosis of brain diseases.
• The developed framework offers a potential improvement over traditional visual inspection methods for sMRI analysis.