EFC grade agreement is higher than all controls in two MR-ART samples with 140 and 119 participants

We examined whether earlier entropy focus criterion (EFC) grade agreement findings in the Movement-related artefacts (MR-ART) MRI dataset persisted under sample changes and single evaluation exclusions. Using MRIQC 0.16.1 tables for 148 participants, we compared grade-2/grade-3 ordering for raw head-motion scores, same-participant standard subtraction, and 20,000 generated reference controls per sample. In the full sample, complete-triplet sample (140 donors), and standard-grade-1 sample (119 donors), each own-subtraction mean exceeded all generated controls in both all-participant and same-standard-grade schemes. Own means were 0.828290771, 0.821818583, and 0.833473389, respectively. All 147 single evaluation exclusions also remained above every control in both fixed control banks. These results cover the specified samples and generated control banks; they do not cover all possible permutations or provide population confidence intervals. The analysis uses the same dataset on which EFC was selected, so it gives no independent biological validation. The conditional p-values do not correct for earlier metric selection, and the arithmetic does not establish reference exchangeability. Each participant has only one standard scan, so the data cannot estimate within-participant standard repeat error variance.

entropy focus criterion · MRI · MR-ART · reference subtraction · permutation controls · grade agreement · sensitivity analysis · MRIQC