Hola Ferran,
There is little to be gained from analysing a SINGLE dark frame.
It is far easier to analyse a SERIES of dark frames, before you integrate them to provide you with a MasterDark.
Ideally, when you look at such a 'series analysis' (as is produced by this tool in AIP4WIN) you should see a table containing statistical data for each image in the series. The statistical analysis should provide you with the Minimum ADU, the Maximum ADU, the Mean ADU, the Median ADU and the Standard Deviation of the ADU values.
Any dark frame in the series that significantly deviates from the others is worth excluding from further integration. I will often graph the data from each of the columns in such a statistical analysis table, which then allows me to take note of images that may be in some way 'different' from others - and then I can later open that individual frame, on screen, to decide 'why' it was different, and whether it should be totally excluded from all further integration.
If you only have a single Dark Frame, then it can be useful to subtract a Bias/Offset frame (if one of these is available), and to then apply an Auto-STF to help 'look' for anomalies, or outlying, data.
Knoing the MAX ADU vakue of a single dark frame can tell you whether you have any 'stuck' pixels, in the same way that a MIN ADU value might help tell you if you have any 'dead' pixels. With that information, you can possibly create a 'faulty pixel' map, allowing you to repair the pixel later in the process.
HTH, and apologies if this is not clear 'in English'
