The Microstructural Features of the Diffusion-Simulated Connectivity (DiSCo) Dataset

Jonathan Rafael-Patino* (Invited author), Gabriel Girard (Invited author), Raphaël Truffet (Invited author), Marco Pizzolato (Invited author), Jean-Philippe Thiran (Invited author), Emmanuel Caruyer (Invited author)

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

We present a detailed description of the structural characteristics of the MICCAI 2021 Diffusion Simulated Connectivity (DiSCo) Challenge synthetic dataset. The DiSCo dataset are one of a kind numerical phantoms for the simulation of the diffusion-weighted images (DWIs) via Monte-Carlo diffusion simulations. The microscopic and macroscopic complexity of the synthetic substrates allows the evaluation of processing pipelines for the estimation of the quantitative structural connectivity. The diffusion-weighted signal in each voxel of the DWIs is obtained from Monte-Carlo simulations of particle dynamics within a substrate of an unprecedented size of 1 mm3, allowing for an image matrix size up to 40×40×40 voxels (isotropic voxel sizes of 25 µm). In this paper, we provide a characterization of the microstructural properties of the DiSCo dataset, which is composed of three numerical phantoms with comparable microstructure. We report the ground-truth tissue volume fractions (“intra-axonal”, “extra-axonal”, “myelin”), the fibre density, the bundle density and the fibre orientation distributions (FODs). We believe that this characterization will be beneficial for validating quantitative structural connectivity processing pipelines, and that could eventually find use in microstructural modelling based on machine learning approaches.
Original languageEnglish
Title of host publicationComputational Diffusion MRI
PublisherSpringer
Publication date2021
Pages159–170
ISBN (Print)978-3-030-87614-2
DOIs
Publication statusPublished - 2021
Event12th International Workshop on Computational Diffusion MRI - Strasbourg, France
Duration: 1 Oct 20211 Oct 2021
Conference number: 12

Conference

Conference12th International Workshop on Computational Diffusion MRI
Number12
Country/TerritoryFrance
CityStrasbourg
Period01/10/202101/10/2021
SeriesLecture Notes in Computer Science
Volume13006
ISSN0302-9743

Keywords

  • Monte-Carlo simulations
  • DW-MRI
  • Phantoms
  • Tractography
  • Microstructure

Fingerprint

Dive into the research topics of 'The Microstructural Features of the Diffusion-Simulated Connectivity (DiSCo) Dataset'. Together they form a unique fingerprint.

Cite this