Abstract #0771

Cross-scanner and cross-protocol harmonisation of multi-shell diffusion MRI data: open challenge and evaluation results

Lipeng Ning^1,2, Elisenda Bonet-Carne³, Francesco Grussu³, Farshid Sepehrband⁴, Enrico Kaden³, Jelle Veraart⁵, Stefano B Blumberg³, Can Son Khoo³, Marco Palombo³, Jaume Coll-Font^1,6, Benoit Scherrer^1,6, Simon K Warfield^1,6, Suheyla Cetin Karayumak^1,2, Yogesh Rathi^1,2, Simon Koppers⁷, Leon Weninger⁷, Julia Ebert⁷, Dorit Merhof⁷, Daniel Moyer⁴, Maximilian Pietsch⁸, Daan Christiaens⁸, Rui Teixeira⁸, Jacques-Donald Tournier⁸, Andrey Zhylka⁹, Josien Pluim⁹, Greg Parker¹⁰, Umesh Rudrapatna¹⁰, John Evans¹⁰, Cyril Charron¹⁰, Derek K Jones¹⁰, and Chantal W.M. Tax¹⁰

¹Harvard Medical School, Boston, MA, United States, ²Brigham and Women's Hospital, Boston, MA, United States, ³University College London, London, United Kingdom, ⁴University of Southern California, Los Angeles, CA, United States, ⁵New York University, New York, NY, United States, ⁶Boston Children's Hospital, Boston, MA, United States, ⁷RWTH Aachen University, Aachen, Germany, ⁸King's College London, London, United Kingdom, ⁹Eindhoven University of Technology, Eindhoven, Netherlands, ¹⁰Cardiff University, Cardiff, United Kingdom

We present a comparison of twelve different methods that estimate mappings between scanners for multi-shell diffusion MRI data harmonisation. The methods are evaluated on a benchmark dataset acquired from 15 subjects using two scanners that both have a standard and a state-of-the-art protocol. The results show that cross-vendor harmonisation and spatial/angular resolution enhancement of multi-shell diffusion data sets can be performed reliably using suitable algorithms. The dataset is available upon request and can serve as a useful testbed for the development diffusion MRI harmonisation algorithms.

This abstract and the presentation materials are available to members only; a login is required.

Join Here