Prediction of Stable and Progressive Mild Cognitive Impairment Using Graph Kernels and Magnetic Resonance Imaging Features

• Grant number: 25/02513-2
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: May 01, 2025
• End date: April 30, 2026
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Ricardo José Ferrari
• Grantee: Davi Cerchiari Alves
• Host Institution: Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil

Abstract

Alzheimer's Disease (AD) is one of the leading causes of dementia in the elderly, characterized by progressive cognitive decline. Individuals with Mild Cognitive Impairment (MCI) can follow two distinct outcomes: remaining stable (sMCI) or progressing to AD (pMCI). Early prediction of this progression is crucial for more effective therapeutic interventions, enabling disease progression delay and improving patients' quality of life. This study proposes using Graph Kernels to model brain connectivity and classify individuals with sMCI and pMCI based on structural magnetic resonance imaging (3D MRI). To achieve this, the images will be segmented using the DKT atlas, allowing the extraction of specific anatomical brain regions, which will be represented as vertices in a graph. The connections between these vertices (edges) will be defined based on different distance metrics between extracted attributes, including texture, volumes of white and gray matter, cerebrospinal fluid, and statistical features of displacement vector fields. The study will involve implementing and optimizing Graph Kernel-based models, exploring different criteria for edge definition and selecting relevant vertex attributes. Model evaluation will be performed using well-established quantitative metrics from the literature, allowing the determination of their effectiveness in predicting MCI progression. This graph-based approach is expected to provide a more detailed analysis of brain alterations associated with AD, contributing to advancements in computational neuroimaging techniques and aiding in the early diagnosis of the disease.