Quantifying brain anomalies in children with bilateral lesions leading to cerebral palsy via diffusion Magnetic Resonance Imaging

Details

Supervisors

Macq, Benoît ; Dricot, Laurence

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] : ingénieur civil biomédical

Abstract

Context: Cerebral Palsy (CP) is caused by non-progressive disturbances of the brain that arise during fetal or infant development (Rosenbaum et al, 2006). The lesions created by these disturbances modify the cerebral structure, including major fiber tracts of the white matter. To predict CP with certainty based on conventional structural Magnetic Resonance Imaging (MRI) is difficult. Moreover, the extent to which the integrity of the fiber tracts is restorable with motor training is unknown (Bleyenheuft et al, 2020). This is the reason why some other techniques were investigated to improve diagnoses and therapies. Objective: Diffusion Weighted Imaging (DWI) is sensitive to water molecule diffusion. Applying adequate mathematical models to DWI allows the microstructural analysis of white matter tracts. This thesis investigates these models to measure and to quantify the damages caused by bilateral lesions on the Corticospinal Tract (CST), the most important tract for fine motor skill (Papadelis et al, 2019). Experiment: This work is composed of two parts: the CST extraction and the quantification of the brain modifications due to CP. To extract the CST was a major challenge of this thesis: due to the small brain size of the subjects and completely changed brain structure of children with CP, the reference atlas cannot be used to isolate the CST. By consequence, this thesis evaluated the efficiency of the tracking method for CST extraction. The second step of this work was to quantify the CST modifications due to CP, by analysing the metrics of the microstructural models: Diffusion Tensor Imaging (DTI) and Neurite Orientation and Dispersion Density Imaging (NODDI). Results: The statistical analysis showed that the diffusion metrics (Mean Diffusivity (MD) and Radial Diffusivity (RD)) reflect an increased freedom of diffusion in the white matter and may reflect a degree of demyelination in children with CP. Moreover, an increase of Fractional Anisotropy (FA) highlights the lower fiber density in these children. This result was confirmed by the ICVF metric. Conclusion: The microstructure models based on DWI provide relevant metrics to measure and quantify the modification of brain structure in children with bilateral lesions.