MRIQC: BIDS-native quality metrics for structural, functional, and diffusion MRI

MRIQC is a quality-control tool for MRI studies that need an automated, repeatable check before downstream analysis. It extracts no-reference image quality metrics — IQMs — from structural T1w and T2w scans, as well as functional and diffusion MRI, and generates visual reports from those metrics. The intended user is a research or engineering team running BIDS-organized MRI data through a preprocessing pipeline. MRIQC is designed to sit alongside that pipeline, either as a standalone QC step or as part of a larger workflow, and to surface scans that warrant closer inspection because of acquisition or data-quality issues. Architecturally, MRIQC implements a nipype workflow composed of modular sub-workflows. Those sub-workflows rely on external toolboxes, including ANTs and AFNI, which is a practical consideration for deployment: a working MRIQC installation depends on getting those third-party tools in place and keeping their versions aligned with the workflow. Two design choices shape how it should be used. First, preprocessing is kept minimal. The workflows estimate IQMs on the original data or on minimally processed derivatives, rather than waiting for a fully processed output. That reduces the chance that a later processing step masks a problem in the source scan, but it also means the metrics reflect the input as provided, so teams need to be clear about which derivative they are feeding in. Second, the project follows the Brain Imaging Data Structure and adopts the BIDS-App standard, which helps it fit existing BIDS-based pipelines and containers. On readiness, the README points to ongoing vetting rather than a single