I was looking into x-ray DICOMS with photometric interpretation MONOCHROME1 and had a few questions:

1) I have a DICOM file with window width = 4096. It seems like the pixel array gets inverted by doing new array = 2^12 - pixel_array (from my understanding, this gives a p-value), before the pixels get scaled from 0-255 according to the appropriate window center and width values and the image gets displayed on screen. Does the value 2^12 correspond to the window width being 4096, the no. of bits stored = 12, or something else entirely?

2) I have another file which does not have window width set as a DICOM attribute. How would the pixel array be converted here? Just for reference, bits allocated=16, bits stored=15, LUT Descriptor=15309\7364\16 and rescale type=P-VALUES. new array = 2^15 - pixel_array does not seem to work for me. Also in RadiAnt, the window length and window width values for this scan are 17722 and 30073 respectively, but using them in any way doesn't seem to help, and I anyway cannot access them through the dicom attributes list

Also, I've tried using the maximum value in the array since P-Value = maximum value - output value, but that doesn't seem to work for me either :(

These are general DICOM technical questions not specifically related to RadiAnt.
Please check the DICOM documentation: http://dicom.nema.org/dicom/2013/output/chtml/part03/sect_C.11.html