A051 and A060 were severely motion corrupted datasets.  Only whole tumour volume and ADC measurements are present at baseline for these patients.

Of the remaining patients, you should be able to find at least one baseline measurement for Ktrans, ve, vp, enhancing fraction and IAUC.  

Visit 1 and visit 2 are both baseline visits, and are prior to treatment.  Where only one of these were successfully acquired, this can be taken as baseline data (e.g. A010 visit 1 is not present, but A010 visit 2 can be used as baseline; similarly for A029).

Some patients had multiple tumours outlined, and for reference sake these are followed through to the final result, even if a given tumour was excluded from further analysis.  The tumour number is recorded as t1, t2, … .  For example, A013 had four tumours outlined.  t1 and t2 did not produce reliable data, and so NaN values are recorded for visit 1 through to visit 5.  t3 and t4 do have data for all parameters.

I hope this is helpful.