Introduction:

• The goal here is to show concrete examples of good and bad data quality (which can be assessed in many many different ways, using different types of visualizations, and looking at different aspects of the data).
• Apparently, there is a lot of "visual expertise" that one acquires with lots of experience with poor/fair/good/excellent MRI data.

Video walkthroughs:

•  20260223 - In this video, we visually inspect raw T1 and T2 data for a single subject (across many T1 and T2 scans), commenting on image quality, usability, and so on. 

Criteria for looking at slices / volumes of raw or processed fMRI volumes:

• Ghosts
• Overall SNR
• Distortion quality (brain asymmetry)
• Blurriness beyond the nominal voxel size
• Any egregious imaging artifacts
• Gray/white contrast is nice or not?
• Signal dropout

Other notes:

• If you "early mask" your data, you may not realize how bad stuff was.