Our aim is to increase the statistical power of clinical studies by combining datasets from different MRI scanners by MRI data harmonisation.
Diffusion MRI is being used increasingly in studies of the brain and other parts of the body for its ability to provide quantitative measures that are sensitive to changes in tissue microstructure.
However, inter-scanner and inter-protocol differences are known to induce significant measurement variability, which in turn jeopardises the ability to obtain ‘truly quantitative measures’ and challenges the reliable combination of different datasets.
Even though careful harmonisation of acquisition parameters can reduce variability, inter-protocol differences become almost inevitable with improvements in hardware and sequence design over time, even within a site.
In this work, we present a benchmark diffusion MRI database of the same subjects acquired on three distinct scanners with different maximum gradient strength (40, 80, and 300 mT/m), and with ‘standard’ and ‘state-of-the-art’ protocols, where the latter have higher spatial and angular resolution.
The dataset serves as a useful testbed for method development in cross-scanner/cross-protocol diffusion MRI harmonisation and quality enhancement. Using the database, we compared the performance of five different methods for estimating mappings between the scanners and protocols.
The results show that cross-scanner harmonisation of single-shell diffusion data sets can reduce the variability between scanners, and highlight the promises and shortcomings of today's data harmonisation.
Key highlights
- A benchmarking database for diffusion MRI data harmonisation is presented.
- The same 14 healthy controls were scanned on three scanners with five acquisition protocols.
- Five harmonisation algorithms are compared for two tasks:
- Matched resolution scanner-to-scanner mapping
- Spatial and angular resolution enhancement.
- Cross-scanner harmonisation can reduce the variability between scanners and protocols.
Registration
To obtain the data, a Data Sharing Agreement (DSA) between Challenge and the University must be signed.
Key points about the DSA
- The DSA must be signed by the Principal Investigator leading the team and by all team members. Please fill out the fields [NAME] and [ADDRESS].
- No modifications will be made to the DSA.
- Signed DSAs can be sent to cdmri@cardiff.ac.uk.
- The University warrants that the data we provide were acquired ethically. However, we have no knowledge of what research is permitted ethically at others’ institutions. The DSA includes a clause to highlight that the receiving researchers’ use of the data may require local ethical approval, which is outside the scope of the University's knowledge. Regarding 'providing Cardiff with evidence…', in most cases a statement such as 'no additional local ethical approval was required for this project' would be sufficient.
For any question, please do not to hesitate to contact the organisers directly.
Funding
This project was made possible by the Wellcome Trust Strategic Award and Netherlands Organisation for Scientific Research Find out more about our funders.
Core team
View the paper
Acknowledgements
This database was initiated by the 2017 and 2018 MICCAI Computational Diffusion MRI committees (Chantal Tax, Francesco Grussu, Enrico Kaden, Lipeng Ning, Jelle Veraart, Elisenda Bonet Carne, and Farshid Sepehrband).