Hi Guiseppe.
Yes, Bernd is correct, PixInsight (specifically PixelMath) doesn't really care where the images have come from, or how they have arrived - they just need to be 'images' (i.e. a 2-dimensional array of pixel intensity values, stored in any valid numerical format, and with uo to three 'channels' per image).
So, if that doen;t help, or make sense, please do not hesitate to come back to us with more details of what it is that you are trying to do, and the PixelMath expression(s) that you are trying to use.