Subtracted EFC gives 81.33% sensitivity in MR-ART HM2 to HM1 transfer with a 95% training target.

This study tested whether an EFC threshold trained to reach at least 95% sensitivity for grade-3 scans in one MR-ART head-motion condition retained that sensitivity in the other condition, and whether subtracting each participant’s standard-scan EFC changed transfer results. The analysis used supplied MRIQC 0.16.1 metrics and expert grades from OpenNeuro ds004173 version 1.0.2: 436 scans from 148 participants, with 278 motion scans graded 2 or 3. For each target participant, the threshold was the (floor(P/20) + 1)-th increasing grade-3 training score after excluding that participant, where P is the number of positive training observations. Across-condition sensitivity for subtracted EFC was 61/75 (81.33%) from HM2 to HM1 and 121/122 (99.18%) from HM1 to HM2; the HM2-to-HM1 result was below the 95% training target. The 5,000-draw participant bootstrap gave pointwise limits of 0.718739–0.948718 for that sensitivity. Equal-condition mean sensitivity changed from 0.931803 for raw EFC to 0.902568 for subtracted EFC, while mean specificity changed from 0.226117 to 0.353621; both intervals for changes included zero. These results apply to this rule and sample. They do not establish performance in new participants, clinical accuracy, or a cause for the transfer difference. The source study used one scanner and protocol, and this analysis did not measure motion or recalculate metrics from images.

MR-ART · MRI quality · entropy focus criterion · threshold transfer · sensitivity · specificity · standard scan subtraction · participant bootstrap