Related Experiment Video
Updated: Oct 15, 2026

A Magnetic Resonance Imaging Protocol for Stroke Onset Time Estimation in Permanent Cerebral Ischemia
Published on: September 16, 2017
Implementation and validation of a clinically compatible four-shell diffusion MRI protocol for ischemic stroke
Emmanuelle Gourieux1,2,3, Johan Pietras1, Aurélien Delphin1
1Univ. Grenoble Alpes, Inserm, US17, CNRS, UAR 3552, CHU Grenoble Alpes, IRMaGe, Grenoble, France.
Abstract:
Advanced multi-shell diffusion MRI enables microstructural modeling beyond conventional tensor metrics but remains underutilized in clinical stroke research. Within the multicenter RESSTORE clinical trial, we implemented a clinically compatible four-shell diffusion protocol optimized for 3T scanners (10-minute acquisition time) and developed a fully reproducible open-source processing pipeline, implemented within a BIDS-compatible framework, including preprocessing, diffusion tensor modeling, NODDI fitting, tractography, tractometry, and atlas-based extraction. The protocol was first evaluated in twelve healthy volunteers (HV), including a test-retest reliability assessment in six participants. It was subsequently applied in twelve patients 7 days after ischemic stroke. Microstructural metrics including FA, ODI and NDI were compared between stroke patients and HV to assess acquisition and processing pipeline feasibility and to evaluate the metric-specific ability to detect lesion-related microstructural alterations. Reliability metrics demonstrated stable parameter estimation for ODI, FA and NDI. Interhemispheric and between-group comparisons revealed significant differences for ODI and FA, whereas NDI, despite good reliability, did not show significant differences between HV and stroke patients in this dataset.

