Table 1.
Summary of IClinfMRI models and functions used to process the fMRI data.
| Function | Module | Software (Function) | |||||
|---|---|---|---|---|---|---|---|
| DICOM Import | Task fMRI | Resting-state fMRI | CVR Mapping | fMRI to PACS | |||
| Data input | DICOM to NIfTI | ✓ | dcm2nii | ||||
| Preprocessing | Slice timing | ✓ | ✓ | ✓ | AFNI (3dTshift) | ||
| Motion correction | ✓ | ✓ | ✓ | AFNI (3dvolreg) | |||
| Aligning EPI to T1 by coordinate-based registration | ✓ | ✓ | ✓ | AFNI (3dresample) | |||
| Aligning EPI to T1 by boundary-based registration | ✓ | ✓ | ✓ | AFNI(align_epi_anat.py) | |||
| Aligning EPI to T1 by intensity-based registration | ✓ | ✓ | ✓ | AFNI (align_epi_anat.py) | |||
| De-spiking & Detrending | ✓ | ✓ | ✓ | AFNI (3dDespike; 3dDetrend) | |||
| Nuisance Regression | ✓ | AFNI (3dBandpass) | |||||
| Segmentation | ✓ | ✓ | ✓ | SPM12 (Segmentation) | |||
| Bandpass filtering | ✓ | AFNI (3dBandpass) | |||||
| Smoothing | ✓ | ✓ | ✓ | AFNI (3dmerge) | |||
| Detection | HRF-based GLM | ✓ | AFNI (3dDeconvolve & 3dREMLfit) | ||||
| Seed selection on the graph | ✓ | In-house script | |||||
| Seed preview on the graph | ✓ | In-house script | |||||
| Pearson correlation | ✓ | AFNI (3dfim+) | |||||
| Seed-based GLM | ✓ | AFNI (3dDeconvolve & 3dREMLfit) | |||||
| RRF-based GLM with varying latency | ✓ | AFNI (3dDeconvolve) | |||||
| HRF-based GLM with varying latency | ✓ | AFNI (3dDeconvolve) & SPM12 (HRF) | |||||
| Display | Adjustable threshold, window, and color map | ✓ | ✓ | ✓ | In-house script | ||
| Fusion of both fMRI and CVR maps on anatomical images | ✓ | In-house script | |||||
| Data output | NIfTI to DICOM | ✓ | In-house script | ||||
DICOM, Digital Imaging and Communications in Medicine; GLM, general linear model; NIfTI, Neuroimaging Informatics Technology Initiative; CVR, cerebrovascular reactivity; HRF, hemodynamic response function; RRF, respiratory response function.