Search the Elsmar Cove!
**Search ALL of** with DuckDuckGo including content not in the forum - Search results with No ads.

In what class - A, B or C - should PACS workstation be?



We develop from scratch our own DICOM Pacs diagnostic workstation. In which class (in terms of 62304) should it be? As the software is diagnosis aid, it does not control any therapy or diagnostic hardware, so in my opinion it should be in class A.
The diagnostic PACS workstation is medical device class IIa or IIb.
What is your opinion on the problem? Are there any more detailed guidelines about classification? Any examples?

Best regs,


This will be workstation mainly for radiologists and other qualified persons, where they will watch images and write diagnosis. In the past the images were hanged on lightbox, now they are displayed.
There are some basic tools like contrast adjustment, zoom, basic measurement, etc.
The images may be of different types - classic radiology, computed tomography, magnetic resonans, etc.
It is obvious, that wrong diagnosis may lead to death, but the radiologist's diagnosis is not the only basis for treatment. When you compare the software control in nuclear treatment device to software for "watching images" it seems, they shoud fall in different classes!
In my opinion, a radiology screen will be a "primary" diagnosis tool. On many occasions the image shown will be the only thing needed or used. Incorrect diagnosis could lead to failure to treat (a tumour not being seen) or unwarranted treatment (un-needed surgery).

The safety classification is a risk management task and in Clause 4.3 it states that the probability of failure of the software system is 100%. As such, and with the importance of radiology images, I can see no other option than to consider such a system as Class C. (Death or serious injury is possible).


Quite Involved in Discussions
Hello adam,

I've been working for a long time on PACS.
My opinion, with the data you provide here, is that your software is class B.
If I interpret well your description, your soft is only a viewer, with a few tools like contrast, light, or powerpoint tools (I caricature).
Your soft displays images, it doesn't give any additionnal information like statistical data or expert system advice.
It is always the responsibility of the medic to interpret correctly the images.

For FDA, your soft is a PACS (category LLZ) and is class II
For CE mark, your soft is class IIa, or even class I. There is a very good document about CE classification with a chapter about PACS here:

Given this, I think your soft is class B.
Last argument: class I or II devices for FDA and EC cannot be class C for IEC 62364. This would be a discrepancy between regulations and standards.

You're in good shape, you don't have to do everything in IEC 62304.

More info on software classification in EC, visit my blog:




Thank you for your detailed answer and useful links.
I must say, that classification in general is rather ambiguous matter and influenced a lot by "marketing" factors (at least at my market). We can see continuous process of increasing the class level of PACS products. If you are not in IIb class you are out of most deals. On the other hand, CR scanners are IIa class (last year most of them were class I...) while they are directly connected with X-ray devices (Class IIb) and make sophisticated processing of raw data. I think in near future, if only the 1st CR scanner will be classified as IIb, the rest will follow up.

I hope my message does not break the frame of this discussion.



Quite Involved in Discussions
classification in general is rather ambiguous matter and influenced a lot by "marketing" factors (at least at my market)
Making high classification a marketing argument is a funny way of dealing with regulations! :nope:
That's really interesting! Thank you for the info.
Top Bottom