CIBM | Center for Biomedical Imaging

CIBM MRI HUG UNIGE Clinical MR Imaging Section

Oscar Dabrowski completed his BSc at HEPIA (EIG) in 2013 and MSc in Computer Science in 2017 at the University of Geneva (UNIGE). During his studies, he explored various classical image/signal processing techniques. His Master’s thesis was related to the identification of pharmaceutical packages using local image descriptors. It entailed using the SIFT algorithm to compute local descriptors and leveraging the RANSAC algorithm to find robust keypoints allowing the computation of homographies for image alignment.

In 2023, he obtained his PhD in computer science at the University of Geneva, under the supervision of Prof. François Lazeyras, Prof. Bastien Chopard, Dr. Sébastien Courvoisier and Dr. Jean-Luc Falcone. In his PhD thesis, he focused on motion estimation and correction of MRI brain images in k-space, exploiting deep learning techniques for motion quantification in k-space and a model-based approach leveraging the non-uniform fast Fourier transform (NUFFT) for hallucination-free image reconstruction. He also had the opportunity to study MRI physics and acquire hands-on experience on a 3T MRI scanner. Alongside his PhD, he was a teaching assistant for computer programming courses for BSc students at UNIGE.

Dr. Dabrowski joined the CIBM MRI UNIGE section on February 1st, 2024, as a postdoctoral researcher. His research interests encompass signal/image processing, machine learning/deep learning techniques, for improving image quality in MRI. He has a strong interest in teaching and is participating in teaching activities as part of the “Advanced MRI techniques in neuroimaging” course.

Keywords: MRI, artifact correction, Artificial Intelligence, deep learning