Neuroimagery for the assessment of patients suffering from addiction

Details

Supervisors

Macq, Benoît

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] en sciences informatiques, à finalité spécialisée

Abstract

Diffusion-Weighted Magnetic Resonance Imaging (DW-MRI) has emerged as a critical tool in visualizing and characterizing the microstructure of brain tissues non-invasively. This thesis explores the enhancement of a microstructure fingerprinting framework, initially developed by PhD. Rensonnet G., to improve the speed and efficiency of parameter estimation using genetic algorithms. The approach involves the creation of a dictionary of DW-MRI signals through Monte Carlo simulations, facilitating the identification of tissue properties such as axonal diameter, density, and orientation by matching these signals to real DW-MRI data. By employing genetic algorithms, the framework can explore the state space more efficiently, reducing the computational burden by 28% while maintaining the same accuracy of the microstructural parameter estimation. This work further examines the robustness of the proposed method through extensive simulations and comparative analyses, highlighting its potential for clinical applications in neuroimaging.