Fighting with DICOM
At work this last week I’ve been working with DICOM support for a Seimens Isocentric C-Arm. DICOM is a command standard for data transfer for medical devices. It’s pretty extensive, and covers every possible type of equipment and situation.
Unfortunately, it’s a standard written by committee, not by engineers. So it’s big, and bloated, and almost impossible to understand. For an example, check the standard here. It’s in about 16 parts, with each part ranging from 300-1000 pages. It took me 2 hours just to find a list of “Field Types”. It’s only 1 table, only half a page, but it’s buried down in the “Data Structure and Semantics” section.
How do things like this become so widespread? Everyone I’ve talked to about DICOM agrees it’s awful. It’s only partially supported, and the way the standards are written it’s almost impossible to support everyone’s hardware. Alot of the “specifications” are pretty generic, and open to interpretation. So how did it become “the standard” for medical communications? Was everyone simply too lazy to come up with another, or is this a standard designed by PHB’s for PHB’s? And then forced upon engineers to make it work?
