Page Last Updated: May 12, 2026

MRI Post-Processing QC🔗

BME-X🔗

We evaluate image enhancement effects and association between BME-X QI and manual QC ratings. Manual visual QC is performed for all T1/T2 structural data pre- and post-BME-X enhancement to evaluate:

  • If the manual QC ratings improve post- BME-X enhancement (paired one-sided t-test of manual QC ratings pre- vs post-enhancement)
  • Association between BME-X QI scores and manual QC ratings before and after image enhancement, including:
    • Spearman correlations between BME-X QI scores and manual QC scores pre- vs post-enhancement
    • linear regression models testing whether BME-X QI scores predicts enhanced and/or original manual QC ratings

Code: bmex_qc_analysis.r
README: BME-X QC Analysis

HBCD QC-FC Summary Analysis🔗

We evaluate whether scan quality, as quantified by BrainSwipes QC summary score, meaningfully changes functional connectivity (FC) estimates in HBCD resting-state fMRI data. The script loads BrainSwipes QC scores and Gordon FC matrices (.pconn.nii) for a selected session, then evaluates how functional connectivity differs between high- and low-quality runs across multiple QC thresholds (0.9–0.1). Summary measures include within-network FC, between-network FC, and network distinctness (within-network FC - between-network FC).

Outputs include:

  • Run-level QC + FC summary tables
  • QC-vs-FC scatter plots
  • High-QC vs Low-QC comparison tables
  • Cutoff-based summary plots

Code: hbcd_qc_fc_analysis.r
README: HBCD QC-FC Analysis

Per-Network QC Impact Analysis🔗

We also evaluate the association between scan quality (BrainSwipes QC scores) and network-level FC metrics (including network distinctness and within-network mean FC) to identify networks most affected by scan quality. The script evaluates the relationship between QC and FC metrics per-network, ranks the networks by effect size, and flags those that display FDR-significant QC effects. The scriopt computes and plots the following for each metric:

  • LM slope: estimated association between QC and the FC metric
  • Pearson correlation (r) between QC and the FC metric across runs
  • Also outputs both raw and FDR-corrected p-values (correction applied separately within each metric)

Code: per_network_qc_impact_analysis.r
README: Per-Network QC Impact Analysis