RSNA 2010: Two SPECT/CT’s

The one piece of equipment I would really like to have is a SPECT/CT scanner.  For some reason, our hospital doesn’t want to part with the $1 Million or so that one would cost, but I’m hoping that the claim of having an extra CT in the house will help my case.  We’ll see.  Maybe I’ll just buy the darn thing myself and park it in a trailer…

My criteria for the scanner remain fairly simple:  I need an excellent SPECT camera integrated to an excellent, DIAGNOSTIC quality CT.  That’s it.  Until recently, only the Siemens Symbia T series fit the bill, but as I have reported earlier, GE finally listened to me and put a proper CT, the BrightSpeed 16, in the package, creating the Discovery NM/CT 670.  The 670 comes in only one main flavor, with the 16 slice CT, while the Symbia can be had with 2, 6, or 16 slice versions.  Beyond that, the hardware differences between the two are relatively minor.  The GE has semiautomatic collimator change, while the Siemens will do this all by itself.  There are various automatic protocols for scanning and QC, and even center of rotation correction.  Both have 70cm apertures and both can handle 500 pound bar-be-cue addicts as we tend to have back home.  The Symbia has been around for a few years, and it is mated to the older Emotion CT platform.  Although this is still being sold, the upgrade paths for Emotion are not as robust as those for the newer BrightSpeed, and some of the newer software toys are not going to be available.  CARE, one of the dose-reduction initiatives will go Emotional, if you will, but SAFIRE, the iterative-reconstruction software, won’t.  I didn’t get dose figures, but I’ll have to assume we won’t clear that 2.4 mSv threshold.  I could be wrong. 

GE touts the new version 3 of the Xceleris computer for acquisition, processing, and viewing, although the viewer is rather busy and doesn’t do some stuff that the AW workstation does do such as edge/threshold detection of a lesion.  There is, however, rigid registration of an old to a new study, and red contour lines are superimposed to confirm the match.  It is possible to superimpose two molecular imaging studies (PET or SPECT) over one CT with a triple color map.  There is some new cardiac software, which allows CT acquisition for just the rest or a stress nuclear perfusion which will then apply it to the stress or rest images respectively.  “Cardiac Morphing” gives a more accurate “splash” display, which can show the various slices of a perfusion study corrected for time.  In other words, I can see all the slices as they would appear in end diastole or end systole, which gives a bit different view than the conventional summated slices. 

Siemens is still showing the eSoft software I’ve been using with my Biograph PET/CT.  There is now a remote client, symbia.net, which (finally!) allows viewing on any computer.  I did not see anything about how syngo.via might be used for processing or viewing the Symbia images.

Images.  Here’s where I’m going to get myself in trouble.  I’ve just gone through an ACR accreditation at one of our places, and one of the older cameras didn’t do well with bone SPECT.  Thus, I wanted to see how each of these new toys would do in that space.  I have to say both produced rather similar images.  And they were not great.  Now, when you superimpose the blobs over the CT acquired contemporaneously, they don’t look so bad.  But viewing a bone SPECT on each workstation from the sample data that is always present at a trade show was, well, disappointing.  The vertebral bodies were muddled, and not clearly demarcated.  Disc spaces were faintly seen, if at all.  One team tried to refilter the data, but to no avail. 

Since I don’t have funding at the moment, I have the luxury of sitting back and watching for improvement.  To be scrupulously fair, I’ll probably have to see both of these battleships out in the field, with live data rather than the canned-RSNA images.  Now that I’ve dissed both machines, I’ll bet the next bone SPECT on display will be incredible.  I’m hoping to see clearly, without having to use much imagination, some sharper images.  But then we are talking about Unclear Medicine, aren’t we? 

More to come from sunny, but extremely cold and windy Chicago.

RSNA 2010: Merge

You might think that the folks at Merge do nothing but play all day.  Their booth here at RSNA has an entire section devoted to a video arcade (see above), and a wall of candy bins.  The major attraction, however, was this fine Tesla electric roadster in Merge Orange. (Yes, that is a stock color for Tesla.) 

Merge actually owns the car, and will use it for advertising purposes.  I certainly got a jolt out of sitting in it, mainly because one has to free-fall into the seat which is about 1cm off the ground.  Getting out is a bit of a production for someone my age as well.  The Tesla’s body is designed by Lotus, but mine isn’t.

I finally had the opportunity to speak with Michael Ferro, Chairman and CEO of Merrick Ventures, which owns Merge.  I came away from this 5 minute face-to-face meeting exhausted.  This fellow has so much energy he could probably charge up the Tesla just by looking at it.  Once he realized who I was, though, he perked up even more, and (most embarrassingly) made a point to bow in my direction, claiming not to be “worthy” to speak to Doctor Dalai.  The guy knows how to deal with pretentious folks like me!  I just wish I’d had a camera to capture that moment.  And as if I weren’t flattered enough by that point, he went on to say that “the only things I read in this space are your blog and AuntMinnie.com, and your blog has a lot more influence!”  From your lips to God’s ears, Mr. Ferro.

We spent a few more moments discussing the state of Merge and healthcare.  Michael’s dedication to reshaping the landscape is clear, and he noted that he does not draw a salary from Merge.  The company took on a tremendous debt to buy AMICAS, and all possible funds are directed to paying that back, although much is still devoted to R&D.  Mr. Ferro is not so much interested in leveling the healthcare playing field, as in boosting the position of Merge customers on that field.  As a Merge customer, I’m OK with that.  The goal of making Merge a Billion Dollar company is still front and center, and I do think it will happen.

As the boss had to head home for a few moments to see his kid’s ballgame (that’s the kind of CEO I want to be when I grow up), I was handed off to Justin Dearborn, formerly CEO of Merge, and now President of Merge, as well as Jeff Surges, the new CEO of Merge, and former President of Sales at Allscripts.  I’m not sure if moving from CEO to President is a promotion or a demotion, but both titles sounds rather impressive.  I had met Mr. Dearborn earlier this year on my visit to Merge HQ.  He is as quiet and deliberate as Mr. Ferro is flamboyant. Mr. Surges also seems a bit more restrained than Mr. Ferro, but here everyone is on the same page, and that page is to bring Merge to the forefront of healthcare IT.  We discussed, among other things, how we radiologists could be helped though the maze of Meaningful Ruse, I mean Meaningful Use, and this is something Merge will leverage in its favor, as well as that of its needy customers (like me).  There is a significant amount of reward money to be had (on the order of $44,000 per doc) from the government for implementing a meaningfully usable EMR, and there are penalties for those who don’t take the cash.  Only the government could come up with something like that, but I digress.

As a legacy AMICAS customer, the one thing I want even more than a $44,000 handout, or even a Tesla (well, OK, I’d take the Tesla instead, but..) is confirmation that the upgrade roadmap I was given earlier this year will stay in place.  The President and CEO assured me that it would, and I was able to see signs elsewhere in the booth that this is so.  More on this momentarily.

After the meeting with the execs, I had a chance to wander the booth and see what’s new.

My first stop was to see my pride and joy, the Halo Viewer in AMICAS 6.0/MR4.  It should be clearly noted that MR4 is the first to be released under the Merge banner, and it shows significant incremental progress as compared to the older version (MR1) that I have on my test server.  To be fair, most of these were present in intervening service releases that came out, or were supposed to come out, before the Merge-r, but they are new to me.  There is 64-bit support for 3D functions, and the whole thing is now Windows 7 compatible.  I would be happier if it was Mac compatible, but I’ve learned not to ask for that one anymore from most any company, although Intelerad recently took the plunge into Mac-dom.  There are a number of other improvements, such as a nicer interface for the customization of the fast-right-click sequence, and a smoother implementation of the multi-step hanging protocols.  There is a History button that allows rapid access to the last 25 exams (or more if you set it that way).  An outside CD can be loaded in the viewer on the workstation without uploading to PACS.  The latter is possible, although this may need a PACSGear appliance.  The multidimensional RealTime Worklist is much as it was.  Keep in mind, you can have more than one open and visible at the same time.

There was a little hushed talk about what version 6.1 might or might not have, and not much was said because most details are still in flux.  That probably applies to version 7, the, ummmmmm, Merged PACS supposedly blending the best of AMICAS, Emageon, and Fusion/Matrix software.  Of course, no one knows just how that patois might look just yet, and I’m hoping for an invite to the new and improved Advisory Committee so I can offer my biased opinions.  It does appear that 6.1 will use the ECM, or Enterprise Content Management, the old Emageon VNA, as its back-end.  I expect Merge will be able to get this integration accomplished a bit faster than some of their counterparts managed with a similar move.  It helps that there are already 300 ECM’s out there, integrated to various other products. 

Interoperability is the name of the game for the future, and Merge wants to be a big player in that arena.  I never understood why until today, when a former Click colleague of Mr. Ferro’s told me a story to which many can relate.  It seems that around the time Michael sold off Click, for a very large sum, he began to experience headaches, and the local docs couldn’t find the reason.  He carried his jacket of films from place to place, and in many instances, he was told that he had to have a repeat study.  There had to be a better way, and the purchase of Merge was the first step in the quest to find it.  


Ultimately, the idea is not dissimilar to my desire to fix the portable patient mess, and Merge attacks it in a different manner than lifeIMAGE.  By leveraging various items from their acquired toy chests, Merge has assembled iConnect, “The Next Big Thing In Interoperability”.  (At least they don’t claim it to be a big f’n deal, although it has potential to be just that.)  The concept is simple, really.  All we want to do is access any image on any device at any time.  Simple?  No, it’s not, not at all.  But the iConnect portfolio offers one path toward the greater goal.  

  • iConnect Access is a web-based platform-agnostic viewer, using AJAX for display.  It works very nicely on the iPad with a limited tool-set, and even better on a PC or a Mac with additional tools.  It has 510K approval for diagnostic reads.
  • iConnect Share is a drag-and-drop solution for studies on CD.
  • iConnect VNA is the ECM from Emageon as noted above.
  • iConnect Exchange is a bundle of the above modules.
  • iConnect DR uses the various components to establish a disaster recovery plan.
  • iConnect Kiosk is the free-standing registration portal I mentioned earlier.  
  • iConnect PHR allows storage of images from the other ‘ologies, Pathology, Ophthalmology, etc.

As an aside, another vendor asked me if the Merge Kiosk would accept a patient-entered CD.  At this point, the answer is no.  But I asked my friend Luc, director of that part of the project if that was something they might add.  He considered it for a moment, and thought it might not be a bad idea, but thought the logistics of trying to get a patient to deal with a balky disk would be too hard to overcome.  I tend to agree.

I’m becoming more and more comfortable living under Merge management.  I had my fears, as you all know, but I think they have been proven mostly baseless. This is a somewhat different healthcare journey than I had signed up to take, but it is worthy none the less.

And now, if you’ll excuse me, it is time that iConnect with a friend from Perth for dinner here in frigid Chicago.

RSNA 2010: Carestream and Sectra

Being a Personality in the PACS field (I doubt this is a goal to which many folks aspire) has significant advantages, one of which is the opportunity to be friends with some great people.  As Doctor Dalai, I’m in mild demand to come have a look at this product or that.  So far this trip, I’ve demo’ed Sectra and Carestream, the former at the request of Mike Cannavo, the One and Only PACSMan, and the latter at the request of Garn, our once and former AMICAS salesman.  (Actually, I go way back with Garn’s family…his father and sister sold my group its first Image Data PhotoPhone system back in 1989.)

Let’s start with Carestream.  Personally, the first thing that comes to mind when I hear that name is some sort of urological appliance, but I guess I’m the only one who has that problem.  In case you didn’t know, Carestream was once Kodak’s health imaging division, since bought out by Onex, a big venture-cap firm.  The software is written by Algotec, an Israeli firm with which I also have history, as they wrote the viewing software for our old Elscint CT’s.    
The latest and greatest version of Carestream’s RIS/PACS 11.1, bears only slight resemblance to the Algotec product I remember from SCAR (now SiiM) 2003.  This is a .NET program, which is self-deployed via web browser, like most of what’s out there these days.  
The Archive Explorer is Carestream’s flavor of worklist, and it is folder-based, somewhat like the Synapse approach.  The worklist disappears when a study is brought up on the viewer.  While the worklist is adequate, it doesn’t give the real-time color feedback as I have come to know and love with AMICAS.  
The viewer itself has some similarities to the newer versions of Microsoft Office, with tabs and tools much like Word or Excel.  The toolbar and right-click menu can be customized.  When the viewer is active, a “mini archive” appears on the left-most screen, which includes priors although multiple filters can be applied to this display.  The thumbnail images can be dragged to a miniature representation of the monitor set-up for display.  
Display protocols are automatically launched based on study attributes.  Advanced visualization can be included in the arrangement.  
Carestream has included one rather spectacular function, that of automatic registration of a study of any cross-sectional modality to one of the same or any other modality.  Thus, an old and a new CT could be registered to one another, facilitating comparison.  A PET and a CT, or a CT and an MRI could be similarly linked.  The algorithm doesn’t deform the comparison, but it will convert it to an MPR which might allow for a little closer match.  This is something I’ve only seen in advanced visualization programs, and it would be very nice to have.  
The system can bring up multiple old comparisons, but there is no clear flag or signal to let you instantly realize which is the current exam and which is the old one, and that could be problematic.
There is a rather complete vessel segmentation program that will identify the major blood vessels all by itself, and this works for coronaries as well.
There is a native report generator which is templated based on the study, and there is speech recognition with AutoText (precanned reports).
Carestream has a solution for enterprise viewing called SuperPACS, which essentially places a layer on top of disparate systems at multiple sites.  An “agent” wraps around the local system and builds an index, and Carestream’s unified/integrated reporting allows easier distribution of the dictation back to where it belongs.
The Cloud is big at Carestream.  The company owns ten data centers around the world, and is the largest supplier of cloud solutions.  
There is the now-obligatory iPad app with a zero-foot-print viewer.  It is patient based, and displays in HTML 5, since Flash won’t work on the ‘Pad.  
Sectra, made in Sweden, was once the supplier of Philips PACS.  Philips sadly dumped them for Stentor iSite, and Sectra had to go it alone in this country under its own name.  It recovered from the experience well, and to its great credit, never once (at least publicly) acted the part of jilted lover toward Philips.

Sectra started the demo with a view of its prototype Visualization Table, a huge screen embedded in glass or plastic, designed for, you guessed it, autopsies!  I tried to get Mike to lay his head down over the image of the spinning skull (the screen uses multi-touch type gestures to manipulate 3D images), but for some reason he declined.  

We then saw Sectra’s obligatory iPad app, which was a little more robust than some, and was able to stream the image for instant display gratification.  We also saw the latest app for Windows Phone, which was actually rather more powerful, having a limited but still pretty good subset of the main worklist available at your fingertips.
And on to PACS.  There were several Mikes present in the booth, and the Mike that ran the demo did a great job.  It was only after he was done and I was sliding off my stool (no comments) that one of the other Mikes told him who I was.  He did recover well, but that brief flash of panic made me wonder just what my reputation has become out there in the big bad world of PACS.
Sectra has done something either very good or very bad, depending on your point of view, placing a fourth monitor on the right of the usual three for a RIS display.  Of course, you don’t have to use it, and the workflow can be driven by either RIS or PACS.  
The left-most monitor usually contains the information window, which is mostly similar to the last incarnation of Sectra PACS I saw a few years ago.  Basically, it consists of customizable panels which include the list of worklists, the worklist itself, a patient history panel, and a document/order and report panel.  This is a very busy screen.  Worklists are highly customizable, and can search across an enterprise, even tapping other connected PACS.  A little eye icon appears on the worklist in front of patients that are being read by someone else.  The worklist does not use color other than that little green eye, and the background of the entire Information window is a little bright…how about darkening it a bit so as not to dazzle my green eyes?
Filters can be applied to worklists for searches, and there is a nice big banner that comes up to tell you that filters are in place.  This may not sound like a big deal, but when you have some entry in the far right field of a search window that you forgot about, and you can’t get any search results, you realize that such a warning might be a darn good idea.  Here that, Centricity?
Sectra does clearly mark the studies as to current and prior.  
The cursor wraps around the screen, which helps limit mouse movement.  (Believe me, it works.)  
Hanging protocols have improved somewhat from the old version.  There is an unusual combination of text and dragging windows around to create a hanging protocol.  It is still a bit harder than AMICAS’ version, but much better than Agfa’s (which doesn’t work at all).  Staged hanging protocols are available.
High-level visualization, including PET/CT fusion and comparison, is accomplished through outside programs, with Mirada 7D recommended for PET in particular.  
For those who are allowed to import CD’s, there is a “temporary merge” function that will let you treat two patients as one.  Very powerful option, but I do advise extreme caution when using it!
There is an optional Image Central which can be used to import non-DICOM studies into the PACS, thus serving the ‘ologies other than Radiology.
Sectra uses off-the-shelf graphics cards to drive the high-res color monitors on its workstations.  
So, here we have another tale of two systems, and this time both are worthy of praise.  Neither is perfect, but I would be content to use either one.  In my always humble opinion, both could use a few tweaks, and I’ve mentioned a few of them above, but both companies seem to be interested in satisfying their users, which they appear to have done.
I’m not really sure why neither has had great penetration into the US market.  These are worthy contenders, and perhaps with a bit more (OK, a LOT more) publicity, they could be right up there with some of the larGEr vendors.  
Stay tuned.  More to come.