Tuesday, October 25, 2016


Time for today's PACS opus...I can report a day of incremental progress.

All three of us met with the Head of the Radiology Department first thing this morning and then with the residents. I will have a session with them tomorrow at 8, wherein I will run a PACS demo on the big screen in the Radiology Conference Room, and take questions as I go. I think this will prove valuable and will lay the foundation for some one-on-one time later on.

The rest of my day was spent mostly in the PACS/reading room, again working closely with Dr. B., the IT-savvy resident. (Honestly, I think that when I’m done here, he will be Korle Bu's resident-advocate for all things PACS!) And thanks to my PACS admins back home, I was able to solve the problem or loading the Merge client onto Windows 10 laptops, so we now have a bunch of very happy residents!

I made several work lists, which I modified as Dr. B requested. The IBMerge PACS is so powerful in this regard that just about any worklist you can think of can made, given enough time, energy, and willingness to work through the list of check-boxes and drop-down menus. I'll be quizzing the residents (and any attendings I can reach) about their needs for more of these.

We found a few possible, temporary (I would hope) work-arounds for the lack of reports. Clearly, what we ultimately need is a RIS, but short of that there are two ways to get a report associated with a particular study. First, one could copy the typed report and paste it into the comment field of the exam's order window. This drops all formatting, and the comment window shows only four lines at a time. The other possibility is to use the “snipping” tool to create a JPEG image of the report, upload it into PACS, and the use the QC editor to merge it into the main study as an extra series. A little more tedious I'm afraid.

I worked with Ben as well on uploading fluoroscopic and sonographic images. We have some hurdles here, in that when multiple studies are loaded from the modality onto a disc or flash drive, and they attempt to load the whole thing at once, all the exams get loaded under the first patient’s demographics. The only solution to this is to load the patients one at a time (my recommendation) or manually split them later (which might be easy to forget to do and would be even more tedious.)

I noted that virtually all CT’s have a huge number of slices, some up to 2500 or so. Even head CT’s are pushing 900 and more slices for a pre and post contrast exam. I spoke at length with Dr. Buckman about this, and I would like to get everyone's thoughts as well. (I read CT, but I'm no luminary.) Dr. B. feels that the thinnest slices should be used to avoid missing tiny abnormalities. While there apparently is no significant malpractice problem here in Ghana, the residents are VERY contentious, and sincerely wish to do no harm. This is admirable to the max. But when I asked for an example of something missed because thin sections weren't available, he showed me a 2cm lesion. In my experience, the thinnest sections (the Toshiba Aquilion One produces 0.5 mm slices) are useful for exquisite reformats and CT angiography. Most of us old folks don't have the stamina to peruse 600 slice sequences 50 or 70 times per day, and I don't think there is much that would be missed by using 5mm (or even 2.5mm) reformats. Add to this the multiplanar reformats created on the scanner and also sent to PACS, and we have a situation that will deplete the SAN very rapidly. Thus, some compromise is needed. I'm thinking that with the Merge PACS ability to create reformats and renderings in the viewer, perhaps only the thin sections should be sent and nothing else. Alternatively, they could create all the reformats on the CT and then lose the thins. But this is a waaaaaay above my pay-grade so I'm going to defer to those much wiser.

Stay tuned!

1 comment :

Stacey Gordon said...

Way to go folks!!!