A Personalized Computer-Aided Diagnosis System for Mild Cognitive Impairment (MCI) Using Structural MRI (sMRI)

Area of Science:

• Neuroscience
• Medical Imaging
• Artificial Intelligence

Background:

• Alzheimer's disease (AD) is a progressive neurodegenerative disorder affecting the central nervous system (CNS).
• Early diagnosis of AD is challenging due to a significant delay between neuropathological onset and symptom manifestation.
• Mild cognitive impairment (MCI) represents a critical stage for intervention in individuals at risk of AD conversion.

Purpose of the Study:

• To develop and evaluate a computer-aided diagnosis (CAD) system for early detection of Alzheimer's disease (AD) in patients with mild cognitive impairment (MCI).
• To visualize the impact of AD on individual cerebral cortical regions.
• To identify specific brain regions associated with cognitive decline in MCI patients at risk for AD.

Main Methods:

• A four-step CAD system was implemented: scan preprocessing and cortex extraction, cortex reconstruction and feature extraction, feature fusion, and a two-level diagnosis (regional then global).
• The system processed scans from MCI patients, enabling visualization of localized effects on the cerebral cortex.
• Shape-based features were extracted and fused for diagnostic analysis.

Main Results:

• The proposed CAD system demonstrated strong performance, achieving a maximum accuracy of 86.30%, specificity of 88.33%, and sensitivity of 84.88%.
• The system effectively visualized the impact of AD on specific cerebral cortical regions.
• Correlations were found between behavioral/cognitive deficits and specific brain regions involved in language, executive function, and memory.

Conclusions:

• The developed CAD system shows promise for the early detection of Alzheimer's disease in MCI patients.
• The system's ability to analyze regional cortical changes aids in understanding AD's progression.
• The identified brain regions highlight key areas for future research and therapeutic targeting in AD and MCI.