Table 8 Quality metrics description.

From: A longitudinal neuroimaging dataset on arithmetic processing in school children

Image Type Metric Description
Both Entropy-focus criterion (efc) Indication of blurring and ghosting caused by head motion. Lower values are better24.
Signal-to-noise ratio (snr) A measure of quality of signal sensitive to the MRI system and parameters. Higher values are better25.
T1-weighted Coefficient of joint variation (cjv) A measure of the performance of INU correction. Higher values indicate greater amounts of head motion and large INU artifacts. Lower values are better26.
Contrast-to-noise ratio (cnr) A measure of the contrast between white matter and gray matter. Higher values are better25.
Intensity non-uniformity median (inu_med) Location and spread of the bias field from INU correction. Values closer to 1.0 are better27.
White-matter to maximum intensity ratio (wm2max) Median intensity of white matter over the 95th percentile of full intensity distribution. Values between 0.6 and 0.8 are best22.
T2-weighted Mean framewise displacement (fd_mean) Measure of head movement across data acquisition calculated by realignment estimates. Lower values are better28.
Ghost-to-signal ratio (gsr_y) The intensity of Nyquist ghost signal in the y-direction due to suboptimal EPI sequence calibrations. Lower values are better29.
Normalized temporal derivative of RMS variance (dvars_std) Standardized measure of the intensity change across volumes. Lower values are better30.
Median temporal signal-to-noise ratio (tsnr) Median value of signal across time over temporal standard deviation. Higher values are better22.
  1. Image quality metric used to assess T1-weighted and T2-weighted data quality and their description, including information on how to interpret each of the values.