Notícies Notícies

Torna a la pàgina principal
Vés enrere

Thesis Defence: Mohamed Oualid Benkarim

Thesis Defence: Mohamed Oualid Benkarim

10.10.2018

Mohamed Oualid Benkarim

Multi-atlas Segmentation and Analysis of the Fetal Brain in Ventriculomegaly

Supervisors: Dr. Gerard Sanromà (Center for Neurodegenerative Diseases) & Dr. Gemma Piella.

14/09/2018

 

Thesis Brief Description:

In my thesis I developed automatic segmentation approaches that can be applied to brains at different life stages, including the gestational period, and investigated in utero brain development in ventriculomegaly. In the first part, we proposed two novel segmentation frameworks that approach the segmentation problem from two different but complementary perspectives. The first approach attempts to mitigate systematic errors caused by registration failures, while the second approach focuses on reducing the impact of interpolation errors, which are inherent to most traditional segmentation approaches. In the second part of my thesis, we analyzed the relationship of ventricular dilation with alterations in cortical folding. First, we studied this association using either lateral ventricular volume or diagnosis as a descriptor of ventriculomegaly. Then statistical analysis and sparse linear regression were used to assess the relationship from two different approaches. In our second study, we proposed a novel approach to jointly analyze cortical and ventricular shapes based on their growth patterns, which allowed us to find fine-scaled associations between both shapes.

 

 

Experience as a PhD Student:

I many times heard that the PhD is a long-distance race. Indeed, it is! Now I look back with satisfaction and tenderness. During these 4 years at UPF I have really enjoyed research, the open-endedness of its challenges, and the gratification of my little achievements. The interdisciplinary character of my PhD, bridging between machine learning and in utero neurodevelopment, gave me the opportunity to get involved in an exciting research line and work closely with an inspiring team of researchers. I am very grateful to my supervisors who supported me and showed me how to unfold new paths to my research goals I could have never thought of by myself. Finally, I cannot imagine traveling this journey with better companions than my fellow researchers and colleagues, who made this a positive learning endeavor and a pleasant life experience.

 

Categories: