One-in-a-million bug in Philips Mx8000 CT scanner?
If I understand correctly, the "Content time" (0008,0033), represented
as a string, should have an even length. If that is the case, then I
have a DICOM CT sample, in which the time is of an exact (round) second
(i.e. 150720.000000). Usually the length of this tag is 14 (6 digits
for the hour + 1x'.' + 6 digits for the microsecond + a leading 0). In
that specific DICOM file, the length is 11 (odd), and there are only 3
zeros leading the decimal point.
In case of the CTN implementation, DCM_OpenFile() returns
DCM_UNEVENELEMENTLENGTH for that DICOM file, as it violates the
If this is true, and it happens only on round seconds, the nice thing
is that it's literally a "one in a millon" bug!
Is this a known bug? How do you think I should treat it?
Re: One-in-a-million bug in Philips Mx8000 CT scanner?
Write a repair script that checks for odd length tags from that
modality. So a million images later... there won't be a problem! --
Radworks can do this...