Improving the interpretability of the image quality metrics computed by MRIQC

[Remi-Gau]

Added as an issue for book keeping

Source: https://brainhack.ch/

Team Leaders:

Céline Provins and Mikkel Schöttner

MRIQC (Esteban et al. 2017) is a tool to help researchers perform quality control (QC) of their structural and functional MRI data. This tool not only outputs visual reports that can be manually rated but also automatically extracts a set of image quality metrics (IQMs). A question that comes often is : “How should we interpret the IQMs ? Which IQMs are more important ?”

In this project, we aspire to answer those questions with the help of the movement-related artefacts (MR-ART) dataset (Nárai et al. 2022). This dataset contains structural brain MRI images collected from 148 healthy adults which includes both motion-free and motion-affected data acquired from the same participants. Furthermore, the quality of the images has been rated by two expert raters. After running MRIQC on the dataset, the goal is to perform dimensionality reduction of the IQMs and to compare the IQMs to the manual ratings in the quest of improving the interpretability of the IQMs. Additionally, we can also get our hands dirty and rate the images ourselves. We would then compare our own manual quality ratings and the expert ones. Lastly, our method and findings will be gathered in a jupyter notebook that will contribute to the Nipreps QC book .

Our plan is not set in stone and we can adapt the analysis to the interests of the participants. Any ideas of analysis leveraging the manual quality ratings and IQMs extracted from this dataset is welcome !

References

Esteban, Oscar, Daniel Birman, Marie Schaer, Oluwasanmi O. Koyejo, Russell A. Poldrack, and Krzysztof J. Gorgolewski. 2017. “MRIQC: Advancing the Automatic Prediction of Image Quality in MRI from Unseen Sites.” Edited by Boris C Bernhardt. PLOS ONE 12 (9): e0184661–e0184661. https://doi.org/10.1371/journal.pone.0184661.
Nárai, Ádám, Petra Hermann, Tibor Auer, Péter Kemenczky, János Szalma, István Homolya, Eszter Somogyi, Pál Vakli, Béla Weiss, and Zoltán Vidnyánszky. 2022. “Movement-Related Artefacts (MR-ART) Dataset of Matched Motion-Corrupted and Clean Structural MRI Brain Scans.” Scientific Data 9 (1): 630. https://doi.org/10.1038/s41597-022-01694-8.