Mohamed Aziz Zeroual

Mohamed Aziz Zeroual

Postdoctoral Research Fellow — Athinoula A. Martinos Center for Biomedical Imaging

PhD in MRI with expertise in advanced pulse sequence design, iterative MRI reconstruction, and high-performance computing. My current research at the Martinos Center focuses on field shimming, SAR management, and MR sequence design. I develop GPU-accelerated algorithms and motion-compensated reconstruction methods to advance clinical MRI applications.

Research Interests
MRI Field Shimming SAR Management MR Sequence Design Iterative Reconstruction Motion Correction High Performance Computing GPU / CUDA Parallel Algorithms Applied Mathematics

Publications & Conferences

2025 Journal
Feasibility of Implementing Motion-Compensated Magnetic Resonance Imaging Reconstruction on Graphics Processing Units Using Compute Unified Device Architecture
M. A. Zeroual, N. Dudysheva, V. Gras, F. Mauconduit, K. Isaieva, P.-A. Vuissoz, F. Odille
Applied Sciences, 15(11), 5840
2025 Conference
Feasibility of Implementing the Generalized Reconstruction by Inversion of Coupled Systems Method on GPUs Using CUDA
M. A. Zeroual, N. Dudysheva, V. Gras, F. Mauconduit, K. Isaieva, P.-A. Vuissoz, F. Odille
IEEE EMBC 2025, Denmark
2024 Journal
Performance Study of an MRI Motion-Compensated Reconstruction Program on Intel CPUs, AMD EPYC CPUs, and NVIDIA GPUs
M. A. Zeroual, K. Isaieva, P.-A. Vuissoz, F. Odille
Applied Sciences, 14(21), 9663
2024 Conference
Motion-Compensated Reconstruction for Brain MRI at 7T: Comparison of Rigid and Non-Rigid Motion Models
M. A. Zeroual, F. Mauconduit, V. Gras, K. Isaieva, P.-A. Vuissoz, F. Odille
ISMRM Workshop on Motion Correction, Canada, 2024 — Abstract 55
2024 Conference
Performance Study and Optimization of an MRI Motion-Compensated Reconstruction Program
M. A. Zeroual, K. Isaieva, P.-A. Vuissoz, F. Odille
ISMRM Annual Meeting, Singapore, 2024 — Abstract 4639
2023 Journal
Feasibility of Online Non-Rigid Motion Correction for High-Resolution Supine Breast MRI
K. Isaieva, C. Meullenet, P.-A. Vuissoz, M. Fauvel, L. Nohava, E. Laistler, M. A. Zeroual, P. Henrot, J. Felblinger, F. Odille
Magnetic Resonance in Medicine, 90(5), 2130–2143
2023 Conference
High Performance Computing Applied to the Study of MRI Reconstruction Codes
M. A. Zeroual, K. Isaieva, P.-A. Vuissoz, F. Odille
SFRMBM Conference, Paris, 2023

PhD Thesis

2025 Defended
Advanced Reconstruction Methods for High-Resolution Magnetic Resonance Imaging
M. A. Zeroual
Supervised by Freddy Odille & Pierre-André Vuissoz
Université de Lorraine — École doctorale IAEM Lorraine — IADI Laboratory (Inserm U1254)

This thesis addresses motion-compensated MRI reconstruction at ultra-high field (7T). The GRICS method was profiled and implemented on GPU using CUDA, achieving up to 14× acceleration on NVIDIA A100. Pseudo-randomized k-space trajectories were designed for MPRAGE & MP2RAGE sequences to distribute motion effects uniformly. Rigid and non-rigid motion models were compared, and the GRICS approach was benchmarked against a deep learning method (MC-NET), demonstrating superior sharpness and robustness for in vivo brain imaging at 7T.

Jury: H. Ratiney (President), D. Atkinson & A. Bustin (Reviewers)

Experience

2025 – Present
Postdoctoral Research Fellow
Athinoula A. Martinos Center for Biomedical Imaging
Massachusetts General Hospital & Harvard Medical School, Boston, USA
Research on field shimming, SAR management, and MR sequence design for advanced clinical and research MRI applications.
Oct 2022 – Oct 2025
PhD in MRI Reconstruction & Sequence Design
Inserm — IADI Laboratory (U1254) & Université de Lorraine
Nancy, France
Performance profiling and GPU implementation (CUDA) of the GRICS motion-compensated reconstruction method. Design of pseudo-randomized Cartesian k-space scanning for 3D MPRAGE & MP2RAGE sequences. Comparison of rigid and non-rigid motion models for brain MRI at 7T. Applications to breast MRI at 3T and brain MRI at 7T. Supervised by Freddy Odille and Pierre-André Vuissoz.
Feb – Aug 2022
Master Thesis — HPC & Seismic Imaging
Jean Kuntzmann Laboratory — CNRS, Inria, Grenoble INP
Grenoble, France
Mathematical modeling of the seismic imaging problem. Roofline-based performance analysis of the SEM46 Full Waveform Inversion framework. Investigation of vectorization and memory-system optimization.
Sep 2021 – Feb 2022
Research Assistant — Mathematical Modeling
Grenoble Alpes University
Grenoble, France
Developed robust algorithms for Spherical Harmonics in weather modeling, computing Legendre polynomial roots with methods that outperform standard NumPy solutions by circumventing overflow limitations.
Jun – Aug 2021
Research Assistant — Number Theory & Probability
Fourier Institute — CNRS & Grenoble Alpes University
Grenoble, France
Exploration of Hawkins' Lucky Number Theorem and Wunderlich's probabilistic setting for prime number theory. Developed complete mathematical demonstrations of random sieves.

Education

2022 – 2025
Doctor of Philosophy
Université de Lorraine
Computer Science, Automation and Signal Processing. Thesis: Advanced image acquisition and reconstruction methods for high-resolution MRI.
2019 – 2022
Engineering Degree in Computer Science & Applied Mathematics
Grenoble INP — Ensimag
Mathematical Modeling, Imaging and Simulation.
2021 – 2022
Master in Industrial and Applied Mathematics
Grenoble Alpes University
2017 – 2019
CPGE — Grandes Écoles Preparatory Classes
CPGE Meknes
Mathematics and Physics.

Teaching

IFMEM · CHRU de Nancy · Sept 2024 – Feb 2025
Mathematics & MRI Physics

Instructor for Bachelor students covering mathematical analysis, functions, and the physics of Magnetic Resonance Imaging.

Bachelor level 60 students
IFMEM · CHRU de Nancy · Sept – Nov 2023
Mathematics, Geometry & MRI Physics

Instructor for Bachelor students in mathematical analysis, functions, geometry, and MRI physics fundamentals.

Bachelor level 60 students

Get in Touch

Current Affiliation
Athinoula A. Martinos Center
for Biomedical Imaging
Massachusetts General Hospital
Harvard Medical School
Boston, MA, USA

Links