Towards clinic-ready relaxometry with MR-STAT: multi-site repeatability, motion detection and integration in the clinical data workflow

Motivation: Clinical adoption of transient-state based relaxometry methods necessitates the assessment of repeatability and robustness of the quantitative MRI method. Goal(s): Present a multi-site repeatability, automatic motion detection and DICOM-based reconstruction for the MR-STAT framework Approach: MR-STAT is a quantitative imaging framework that reconstruction T1/T2 directly from time-domain data. Repeatability was investigated in healthy volunteers at two different medical centers. A histogram-based automatic motion framework was developed and a DICOM-based reconstruction pipeline was validated. Results: Repeatability experiments demonstrated intra-subject of less than 1.2% for T1/T2. Motion was detected automatically in a subject with motion and DICOM-based reconstruction were equivalent to a raw-data pipeline. Impact: This work shows that quantitative MRI (MR-STAT) yields repeatable results and can be incorporated in a clinical data workflow by using a DICOM-based reconstruction and automatic motion detection.