Mohammed Chekroun

As a member of the Morpheo team at Inria Grenoble, I developed a robust pipeline for both rigid and non-rigid alignment of 3D CT scans. The primary goal was to automate the annotation of bones in foot CT scans using a single template patient, enabling the accurate generation of bone masks and meshes. I focused on the following key steps: • Conducting a comprehensive review of state-of-the-art segmentation methods in medical imaging • Testing and evaluating various 3D segmentation software solutions • Designing and implementing 3D segmentation algorithms for bone structures • Analyzing and optimizing segmentation results to ensure high precision Throughout this project, I utilized a range of technologies and tools, including OpenCV, SimpleITK, Open3D, 3D Slicer, and MeshLab to achieve advanced image processing and visualization.

Development and evaluation of automatic segmentation tool for thoracic structures on a 0.35 T MR-linac radiation therapy system.

Industrial project: Application of machine learning and deep learning models to create a real time analyzer of metal contents in flotation cells.