Extended Data Fig. 9: Comparable test-retest reliability of functional connectomes between DeepPrep and fMRIPrep.
To estimate whole-brain and cortical functional connectomes, we employed 300 volumetric ROIs encompassing both cortical and subcortical regions, as well as Schaefer’s 300 cortical surface ROIs, respectively. Test-retest reliability was quantified by measuring the similarity between functional connectomes derived from two independent segments randomly sampled from one or multiple sessions in two repeated measured datasets, namely the MSC and CoRR-HNU dataset. To elucidate the relationship between test-retest reliability and data amounts, we assessed reliability using one to five sessions. No significant differences in the test-retest reliability of functional connectomes were observed between DeepPrep and fMRIPrep (two-tailed paired-sample t-tests, FDR-corrected p-values > 0.05). Data are presented as mean ± s.t.d.
