This page presents my research on automatic medical image segmentation.
Open-Source Software Development📦
</>💻💾As a contributor of the Spinal Cord Toolbox (SCT), I develop open-source tools for the processing and quantitative analysis of spinal cord MRI. My contributions span atlas development, deep learning, quantitative MRI, and automated image analysis.
Atlas-Based Segmentation🧩
AMU7T: Quantitative 7T Spinal Cord Template
June 2023
We developed AMU7T, a high-resolution quantitative spinal cord template at 7 Tesla with refined white and gray matter parcellations.
The template is fully integrated into the SCT and is compatible with the PAM50 template anatomical space.
Conference: ISMRM 2023 (Toronto, Canada)
Title: AMU7T: A 3D QT1 and T2star-Weighted High-Resolution In Vivo Template with Refined White and Gray Matter Parcellation Dedicated to 7T Spinal Cord MR Analyses
DOI: https://doi.org/10.58530/2023/0569
Deep Learning-Based Segmentation🔬
2D Multi-Class Spinal Cord and Gray Matter Segmentation at 7T
October 2021
We developed a deep learning model for the simultaneous segmentation of the spinal cord and gray matter from 7T T2*-weighted MRI.
The model is available in the SCT
Title: 2D Multi-Class Model for Gray and White Matter Segmentation of the Cervical Spinal Cord at 7T.
DOI: https://doi.org/arXiv:2110.06516
Multiple Sclerosis Lesion Segmentation on MP2RAGE MRI
January 2024
We developed a deep learning-based tool for automatic spinal cord multiple sclerosis lesion segmentation from MP2RAGE MRI.
The model has been integrated into the SCT.
Title: Automatic spinal cord multiple sclerosis lesion segmentation on MP2RAGE MRI.
Multiple Sclerosis and Related Disorders. DOI: https://doi.org/10.1016/j.msard.2026.107250
Contrast-Agnostic Gray Matter Segmentation
April 2025
We developed a deep learning framework capable of segmenting spinal cord gray matter across multiple MRI contrasts, magnetic field strengths (1.5T, 3T and 7T), anatomical regions (cervical, thoracic and lumbar), and neurological pathologies.
The model is integrated into the SCT.
Title: Automatic Spinal Cord Gray Matter Segmentation Across Multiple Contrasts, Magnetic Fields, Regions and Pathologies.
DOI: Coming soon. ISMRM Abstract
