Cape Town - 2026 ISMRM-ISMRT Annual Meeting and Exhibition - Validating Diffusion MRI: Reproducibility, Simulations, and Phantoms

Validating Diffusion MRI: Reproducibility, Simulations, and Phantoms

Digital Poster

Diffusion

Tuesday, 12 May 2026

Digital Posters Row E

09:15 - 10:10

Session Number: 464-02

No CME/CE Credit

Similar Sessions

This session focuses on methods for validating diffusion MRI techniques and ensuring reproducibility across scanners, sites, and studies. Contributions cover the use of histology, simulations, physical phantoms, and reproducibility studies to assess model performance, quantify variability, and build confidence in diffusion MRI measurements.

Skill Level: Intermediate

		464-02-001. A simulation framework for optimising diffusion MRI for microstructural sensitivity with experimental validation in phantoms Zhiyu Zheng, Karla Miller, Mohamed Tachrount, Fenglei Zhou, Kamila Szulc-Lerch, Sean Smart, Kamila Blachowiak, Benjamin Tendler, Michiel Cottaar Centre for Integrative Neuroimaging, FMRIB, Nuffield Department of Clinical Neurosciences, University of Oxford, United Kingdom Impact: We present an experimentally validated protocol optimisation framework combining Monte-Carlo simulations with estimation theory. It systematically identifies dMRI protocols that maximise sensitivity to target microstructural parameters, and is applicable to a broad range of tissue substrates and protocol constraints.
		464-02-002. Evaluation of the NEXI model of gray matter using Monte Carlo simulations in realistic substrates Rita Oliveira, Jasmine Nguyen-Duc, Ileana Jelescu Lausanne University Hospital (CHUV) and University of Lausanne (UNIL), Lausanne, Switzerland Impact: Validation of NEXI in realistic GM-like substrates with neurite beading, undulation, and varying permeability shows that NEXI estimates remain reliable under structural complexity. These results support its potential for in vivo GM microstructure mapping and studies of brain disorders.
		464-02-003. Statistical Pore Imaging Yang Yang, Adam Threlfall, Fenglei Zhou, Marco Palombo, Derek Jones, Salvatore Torquato, Leandro Beltrachini Cardiff University, Cardiff, United Kingdom Impact: This study establishes a mechanistic link between diffusion MRI (dMRI) signals and statistical microstructural descriptors, enabling non-invasive, model-free tissue reconstruction. By leveraging dMRI’s statistical information, it lays the foundation for in-vivo histology beyond the constraints of conventional biophysical models.
		464-02-004. Characterizing Fiber Bundles in an Anisotropic Diffusion Phantom: Comparison of Phantom Replicas Sofia Chavez, Israa Saber, Fergal Kerins PreOperative Performance, Toronto, Canada Impact: This study presents an image analysis pipeline that yields tolerance thresholds for equivalency of anisotropic synthetic fiber bundles in a novel diffusion phantom across replicas. This work enables cross-site result comparison and supports identification of site-specific variations for data harmonization.
		464-02-005. Framework for Interpreting Low b-value Diffusion Tensor Imaging by Inferring Flow-velocity Vector Fields in Oscillatory Flow Yoshitaka Bito, Hiroyuki Kameda, Tomohiro Otani, Noriyuki Fujima, Naoya Kinota, Daisuke Kato, Takaaki Fujii, Kinya Ishizaka, Kenji Hirata, Kohsuke Kudo Hokkaido University, Sappro, Japan Impact: A framework was developed to interpret low b-value DTI by inferring flow-velocity vector fields under oscillatory flow, enabling clearer insight into CSF dynamics that are central to brain clearance and have remained difficult to analyze.
		464-02-006. Impact of formalin fixation on biventricular parameters in cardiac diffusion tensor imaging: Insights from a swine model Leyi Zhu, Jing Xu, Huaying Zhang, Chen Cui, Peng Sun, Minjie Lu Fuwai Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China Impact: It is advisable to perform ex-vivo cardiac diffusion tensor imaging with fresh tissue samples, given the effects of formalin fixation.
		464-02-007. Green-dMRIPrep: diffusion MRI preprocessing pipeline with integrated QC, energy tracking, and carbon emissions auditing Muhammad Usman Akbar, Maëliss Jallais, Marco Palombo Cardiff University Brain Research Imaging Centre (CUBRIC), School of Psychology, Cardiff University, Cardiff, United Kingdom Impact: This GUI-driven pipeline simplifies dMRI preprocessing, embeds carbon auditing and automated quality control, and promotes reproducible, energy-aware neuroimaging practices bridging usability, transparency, and sustainability in research workflows.
		464-02-008. Whole-brain diffusion MRI assessment of microglial morphology in physiological and humanized mouse models. Akari Takekoshi, Fumiko Seki, Junichi Hata, Ikumi Katano Tokyo Metropolitan University, Tokyo, Japan Impact: This study demonstrates that diffusion MRI enables non-invasive, whole-brain assessment of microglial morphology and can distinguish subtle cellular differences across physiological and humanized mouse models, offering a potential imaging biomarker for studying neuroimmune mechanisms in health and disease.
		464-02-009. Post-mortem diffusion MRI reveals downstream tract damage increases with injury-to-death interval after spinal cord injury Nikolai Lesack, Sarah Morris, Taylor Swift-LaPointe, Andrew Yung, Kirsten Bale, Shana George, Andrew Bauman, Piotr Kozlowski, Zahra Samadi-Bahrami, Caron Fournier, Pushwant Mattu, Kevin Dong, Femke Streijger, G. R. Wayne Moore, Adam Velenosi, Veronica Hirsch-Reinshagen, Brian Kwon, Cornelia Laule University of British Columbia, Vancouver, Canada Impact: Advanced diffusion MRI detects progressive axonal degeneration after spinal cord injury, offering insights into the timescale of Wallerian degeneration. These findings can inform clinical assessment and staging of secondary injury, and support the development of biomarkers for evaluating treatments.
		464-02-010. Robustness of quantitative diffusion metrics from eight models in musculoskeletal system: a phantom study Jingyu Zhong, Yang Song, Yue Xing, Yangfan Hu, Wenjie Lu, Weiwu Yao Tongren hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China Impact: Quantitative diffusion metrics, as promising tools for providing functional information in musculoskeletal system, should be interpreted with caution because they are fragile to acquisition parameters but not scan-rescans, which calls for standardization of scan protocols before clinical implementation.
		464-02-011. Repeatability and Reproducibility of Fractional Anisotropy Across Commonly Used Research Protocols in Diffusion Tensor MRI Matheus T, Damon Carter, Ryan Willoughby, Jane Allendorfer, Virendra Mishra University of Alabama at Birmingham, Birmingham, United States of America Impact: Given the importance of repeatability and reproducibility of fractional anisotropy (FA) measures across widely used dMRI protocols, we propose caution in pooling high-resolution and low-resolution dMRI data, while further demonstrating the benefits of acquiring data at higher resolutions.
		464-02-012. Evaluating Phase Correction Methods for Complex Denoising of dMRI Data Across Vendors Francesco D'Antonio, Olivier Mougin, Steen Moeller, Essa Yacoub, Shaun Warrington, Paul Morgan, Stam Sotiropoulos University of Nottingham, Nottingham, United Kingdom Impact: Using decorrelated kernels as a phase correction step prior to denoising high-resolution complex dMRI data significantly reduces noise-floor bias across vendor data without introducing filtering artefacts, compared to other phase correction approaches commonly used in reconstruction and denoising pipelines.
		464-02-013. A Large-Format 3D-Printed Diffusion Phantom for Cross-Platform Muscle Microstructure Validation David Berry, Lin Huang, Jascha Gaarder, Yajun Ma, Vitaly Galinsky, Lawrence Frank, Shaochen Chen, Samuel Ward University of California, San Diego, United States of America Impact: This large-format 3D-printed diffusion phantom enables standardized cross-platform validation of muscle-mimicking microstructure, advancing quantitative dMRI calibration across scanners and supporting development of clinically translatable diffusion biomarkers of muscle pathology and regeneration.
		464-02-014. Enhancing cellular tumor conspicuity with ultra-high b-value multishell DWI: a simulation study Gérémy Michaud, Philippe Dionne, Laura Natalia Beltran, Michèle Desjardins, Maxime Descoteaux, Louis Gagnon Université Laval, Québec, Canada Impact: Realistic Monte Carlo simulations indicate that adding ultra-high b-value (b = 8000 s/mm²) shells to multishell DWI increases diffusion-model sensitivity to tumor cellularity when analyzed with RSI, guiding acquisition optimization and motivating in-vivo validation for improved glioblastoma delineation.
		464-02-015. Adding Soy Lecithin to Water Reduces the ADC to Tissue-Equivalent Values: The Effects of Concentration and Temperature Victor Fritz, Siri Raupach, Fritz Schick University of Tuebingen, Tuebingen, Germany Impact: This work introduces a calibration-model for soy lecithin solutions to account for temperature-related ADC shifts in diffusion phantoms. This model may help researchers to improve consistency of DWI-measurements across scanners and protocols, supporting more reliable comparisons under varying scanning conditions.