Dealing with assessment of motion artefacts locally, and thinking about how to get a more integrated solution.

eddy gives a matrix (in a text file) of slice-wise outliers detected. Ways this could possibly be used within dwipreproc:

• Detect volumes with greater than N number of slices dropped out, and reject - possibly re-running eddy afterwards.

• Report the mean number of dropout slices across all volumes - may be an indicator for rejection of a subject.

• Check the condition number of the A2SH transform before and after volume rejection; warn if the rejection of those volumes is going to cause a reduction in the achievable lmax due to condition number.