This panel shows a mosaic enhancing the background around the head. Artifacts usually unveil themselves in the air surrounding the head, where no signal sources are present.
This panel shows a mosaic of the brain. This mosaic is the most suitable to screen head-motion intensity inhomogeneities, global/local noise, signal leakage (for example, from the eyeballs and across the phase-encoding axis), etc.
The hat-mask calculated internally by MRIQC. Some metrics will use this mask, for instance, to find out artifacts and estimate the spread of gaussian noise added to the signal. This mask leaves out the air around the face to avoid measuring noise sourcing from the eyeballs and their movement.
Brain tissue segmentation, as internally extracted by MRIQC. Defects on this segmentation, as well as noisy tissue labels could indicate problematic aspects of the image quality-wise.