When I drop a point dose tool on the displayed graticule in Portal Dosimetry and I read back the center pixel of the image in python I get a similar but not exact match (same image looked at in two applications).
The isocenter is reported at 0,0,0 in the image header. Am I missing a shift of some kind or an SID correction?
RI.1.2.246.352.62.1.4625589395255983245.9103590929697145013.dcm (1.5 MB)
I haven’t messed with portal dosimetry yet, so I don’t have any knowledgeable answers there. My guess is that there is an additional calibration/normalization from PD on top of the raw data. I’d test more points to see if it’s a normalization thing (maybe the PD cal was off?) or it changes across the field (flatness correction?). You could look at the image in an independent DICOM viewer (e.g. imageJ) to see if the values are the same (this reminds me I should make a rescaled_array property for dicom images). That way you’ll know if it’s PD that’s different. Let me know how it turns out.