Table 3.
Description of PCP QAP Measures
| Spatial Metrics | Description | References |
|---|---|---|
| Contrast-to-noise ratio (CNR) (sMRI only) | MGM intensity—MWM intensity/SDair intensity. Larger values reflect a better distinction between WM and GM. | Magnotta et al., 2006 |
| Artifactual voxel detection (Qi1) (sMRI only) | Voxels with intensity corrupted by artifacts/voxels in the background. Larger values reflect more artifacts which likely due to motion or image instability. | Mortamet et al., 2009 |
| Smoothness of Voxels (FWHM)a | Full width at half maximum of the spatial distribution of the image intensity values. Larger values reflect more spatial smoothing perhaps due to motion or technical differences. | Friedman et al., 2006 |
| Signal-to-noise ratio (SNR) | MGM intensity/SDair intensity. Larger values reflect less noise. | Magnotta et al., 2006 |
| Temporal Metrics (fMRI and DTI only) | Description | References |
| Ghost-to-Signal Ratio (GSR)a | M signal in the “ghost” image divided by the M signal within the brain. Larger values reflect more ghosting likely due to physiological noise, motion, or technical issues. | Giannelli et al., 2010 |
| Mean frame-wise displacement- Jenkinson (meanFD)b | Sum absolute displacement changes in the x, y, and z directions and rotational changes around them. Rotational changes are given distance values based on changes across the surface of a 50 mm radius sphere. Larger values reflect more movement. | Jenkinson et al. 2002 |
| Standardized DVARSb | Spatial SD of the data temporal derivative normalized by the temporal SD and autocorrelation. Larger values reflect larger frame-to-frame differences in signal intensity due to head motion or scanner instability. | Nichols, 2012 |
| Global Correlation (GCORR)b | M correlation of all combinations of voxels in a time series. Illustrates differences between data due to motion/physiological noise. Larger values reflect a greater degree of spatial correlation between slices, which may be due to head motion or “signal leakage” in simultaneous multi-slice acquisitions. | – |
Here, we provide a brief description of the Preprocessed Connectome Project Quality Assessment Protocol. These measures have been computed for all structural MRI (sMRI) and resting-state functional MRI (R-fMRI) datasets in PRIME-DE. The table was adopted from Di Martino et al. (2017).
a
For R-fMRI data, these metrics are computed on mean functional data
b
For R-fMRI, these metrics are computed on time series data. M, mean; GM, gray matter; WM, white matter; SD, standard deviation