How Do You Like My Bodacious Tatas?

I guess nothing is sacred anymore. I have despised Ford Motor Company ever since it made the decision not to recall the exploding Pintos because it was cheaper to pay awards in wrongful-death lawsuits. I was very disappointed to see the results of Ford’s buying spree of luxury nameplates; Ford presently owns Jaguar, Land Rover, and Volvo. But now, Ford will sell the first two to India’s Tata Group, although this will be more of a partnership than a complete sale.

Tata is probably going to be the world’s largest car manufacturer. It is actually a very large operation:

The Tata Group is one of India’s oldest, largest and most respected business conglomerates. The Group’s businesses are spread over seven business sectors. It comprises 98 companies and operates in six continents. It employs some 2,89,500 people and collectively has a shareholder base of over 2.9 million and market capitalisation of $66.9 billion as on February 21, 2008.

While the British (automotive) Crown Jewels are certainly a nice acquisition, Tata’s greatest achievement is probably going to be the introduction of the Nano, a US$2500 “people’s car”:

There is a much larger draw for a US$2,500 mini-car than a US$75,000 luxury SUV. Nothing like covering both ends of the market.

I’m still stuck on Japanese quality, having been very pleased with my Lexuses and Toyotas for the past 20+ years. It will be interesting to see what Tata does with its new toys. So, if anyone out there wants to keep us updated on this topic, please feel free to send in photos of your bodacious Tatas. Thank you.

Stacks of Tiles

I like some degree of flexibility in my PACS display, but I like it simple. This has been a very difficult concept to convey to Agfa in the process of my occasionally-welcome attempts to help with Impax 6.x development. (I do have to tell you that Impax has been behaving quite a bit better these days, with crashes occuring only very rarely.) I have had a really hard time convincing them that any series should be deployable in any viewport. In other words, if I want to put the same CT series on the screen in 10 places, that should be possible. This idea just hasn’t yet penetrated. The only way to do anything like this now is to create a “clone window”, which can be set up to display multiple instances of the same series. However, I cannot link a “regular” series to one in the clone window, and there are some glitches in the implementation of window and level adjustments, rendering the clones limited at best.

Why is this important to me? An article that I rediscovered recently from the American Journal of Roentgenology by Drs. Kim and Lee, et. al., gives one possible answer. These radiologists from the Department of Radiology of Seoul National University looked at scans of hepatocellular carcinoma in two different modes. First, they used the old “tile” mode, showing a dozen slices from the scan on screen at once. Frankly, nobody does it this way anymore. Secondly, they used “multisynchronized stack” mode to view the multiphasic dynamic CT. While they didn’t include any pictures in the article, I’m assuming the stack display looked something like this:

The key here is that you have multiple phases (with different series) of the same study, synchronized to the same level. A dynamic study of this type means that the patient was scanned as IV contrast was injected, and then at several fixed delay times thereafter. Thus, if the reader does stumble upon a lesion, he/she can see how the contrast material enters and exits the lesion, which yields a better depiction of the contents of the mass. With this display, you have all phases of the study up at the same time, and synchronization allows you to scroll through the study and view the lesion on all of the phases at once. This is much easier than trying to view “tiles” of all the series, and even better than viewing one stacked series at a time.

The flexible display lends itself to other approaches. I like to view abdominal and chest studies in this 4-up configuration with soft-tissue, bone, and lung windows, as well as the scout, all on-screen at once.

I do the same thing with head CT’s, adding in a “blood” and a very narrow windowed viewport to aid in detecting subtle CVA’s.

I guess the point of all this is that every rad reads just a little bit differently (although no one uses tile-mode much anymore). For my 30 rads, I need something that will adapt to our various ways of doing things. Many, if not most, other systems have the any-series-any-viewport approach these days. Most can actually handle the linked scrolling of the images, although our older installation of Centricity slows down markedly when windows are linked. Agfa still has its own version…send in the clone windows! Maybe we’ll see the “any viewport” thing in Impax 7.x?

A New Cellphone Accessory

Now here’s a good idea… I should probably patent this, as I came up with it all by myself, but since a search revealed that others have mentioned it (but not actually patented it or manufactured it), I won’t bother. I’d just like to see someone actually make this thing.

We’ve all seen folks wandering around with Bluetooth headsets jammed in their ears, seemingly talking into space. I own one myself, but rarely use it because I can’t stand to have something on my ear for more than a few minutes at a time.

But for those so inclined, I have invented (via paste-up, anyway) the Bluetooth headset/camera. Just what everyone needs! Note the camera lens grafted onto the headset’s anterior tip. The concept here is that the camera points in the direction of your head (OK, not necessarily your gaze, but…) and lets you photograph or video whatever you are looking at. I think the bandwidth of Bluetooth is sufficient to handle this application, and otherwise it should work like any other cell-phone camera. Yes, it would add bulk to the headset, and drain the battery faster, but that’s minor compared to the incredible convenience this will offer.

Do I have any takers?

Promiscous Blogs Contribute To Science!

When you think about it, this blog, and blogging in general, represents something totally revolutionary in the history of information. I can sit here in front of my computer, post some occasionally-good ideas, and publish them to the world. Literally! I have had hits from all over the globe, from Russia, China, Vietnam, Israel, Dubai, and many others. (For some reason, the plurality of overseas hits seem to come from Belgium and Canada.) The power of the individual to reach anyone else in the world has never been so great.

But a blog is not a static, isolated thing. It changes (when the blogger gets around to it), other people leave comments, and perhaps most importantly, other blogs or websites can be linked to mine. This is a property unique to the internet, allowing instant access to the thoughts of others within the body of your own writings.

Allison McCook, writing on the blog portion of the magazine The Scientist, quotes writer Clive Thompson as calling blogging “highly promiscuous” for this reason.

It’s a basic concept. Thompson — a surprisingly dapper (for a writer), well-coiffed, quick-talking presenter — explained that he constantly feeds his blog, collisiondetection.net, because blogging is “highly promiscuous” — meaning, you blog and link to another blog, then that person links to you in a future post, and so on. You find out who’s linked to you (technorati.com ), check them out, and see other blogs by like-minded people, who might think about something you’d never considered before.

This leads to the concept of “crowdsourcing“, “The idea is to use the internet to get large numbers of people to help with a task. They may do it for money — usually not much — or out of interest or simply because it’s fun.”

So, let’s put this into practice. I get something upwards of 200 hits on this site per day. Not a lot, but still an amazing number to me. But only rarely, perhaps three or four times per week, will someone comment back, either by e-mail or through the comment field on the blog itself. So, I ask you to do your part for science. Post comments. Link to my blog. Ask questions. Help my blog become more promiscuous. Bet that got your attention!

SpaceNavigator…a tool for SpaceCadets and maybe PACS?

I can find inspiration anywhere, it seems. This morning I was perusing the Buy.com website, looking for bargains, when I stumbled upon this entry:

Is working in 3D frustrating? Feel like you can’t get the model in the right orientation or move the world to the right location? If only you could put your hand straight into the display and directly interact with your 3D design. 3Dconnexion SpaceNavigator lets you come as close to that experience as you can get without actually putting your hand in the display. Push, pull, tilt or twist the SpaceNavigator’s controller cap to interact with your 3D world or model with ease. SpaceNavigator PE supports over 100 popular 3D apps including Google Earth, Microsoft Virtual Earth, SolidWorks, Autodesk Inventor and lots more. Work in 3D as it was meant to be. SpaceNavigator PE is meant for non-commercial use and includes web based support.

Well, most of us do find working with 3D frustrating, as a regular ol’ mouse is geared for 2D control. There are various kludges that give us the third dimension, but in general, they aren’t very intuitive. You may remember my attempts to make a jog-shuttle device, the ShuttlePro2, work with the Amicas/Voxar 3D software. I really couldn’t get it going properly. But keep in mind, Darth Vader’s Codpiece was designed mainly for flipping video frames. It does work well for MPR viewing, but not so great for volume rendered images. It just doesn’t have a control for the third dimension!

The SpaceNavigator does have a Z-axis control, solving this deficiency:

Pressure sensing technology allows the controller cap to become a virtual extension of you. Push, pull, twist or tilt the cap a fraction of inch to simultaneously pan, zoom and rotate 3D imagery. Increase pressure to go fast or decrease pressure to make intricate adjustments. . . The key advantage of a SpaceNavigator over a mouse and keyboard is the ease of performing intricate adjustments to camera views and models with out repeatedly stopping to change directions, zoom, or rotate models. With a SpaceNavigator you can do all three at once.

Click here for a video of the device in operation.

I guess we’ve gone from Darth Vader’s codpiece to an excised BMW iDrive.

The SpaceNavigator is made by 3DConnexion, a division of Logitech. It lists for $59.99, but Buy.com sells it for $39.99. Having spent more than this on the jog-shuttle, I’m a little hesitant to invest more. But this one is pretty tempting…

eRad PACS

Mark Twain once received a telegram from his publisher stating:

NEED TWO-PAGE SHORT STORY TWO DAYS.

Twain responded,

NO CAN DO TWO PAGES TWO DAYS. CAN DO THIRTY PAGES TWO DAYS. NEED THIRTY DAYS TO DO TWO PAGES.

In the same vein, Dorothy Parker, an American poet and writer once said to then-Vice President Calvin Coolidge, “Mr. Coolidge, I’ve made a bet against a fellow who said it was impossible to get more than two words out of you.” His famous reply: “You lose.” (Ms. Parker stated upon learning of Coolidge’s death many years later, “How can they tell?”)

Those of you who have followed my attempts at writing over the years know that I am far from the Calvin Coolidge of blogging; I can yammer on and on for pages, and sometimes not say very much in the process. However, when it comes to PACS, I think Mr. Coolidge would have agreed with my minimalist philosophy, probably best outlined in my old post, Feature Fatigue and Lego PACS. There is such a thing as too many words, too many adjustments, and too many features, even if they are all relevant, useful, and powerful. And to pervert the spirit of Mark Twain somewhat, it is far more difficult to produce something with fewer features than a product with lots of them.

The folks from eRad gave me the keys to a web-demo, and then made the trip down to my average little town in the South from their average little town in the South to do a proper demonstration. In spite of my long lead-up, this is a powerful program with a bright future. However, you can guess where we’re going with this in the end.

eRad’s mission statement unifies one’s view of their approach:

The Mission of eRad® is to unify diagnostic imaging processes and operations by:

  • Applying unique technology which will present patient images, diagnostic reports and medical records to Physicians, Technologists and Clinicians from a single unified data base.
  • Delivering images and information both error free and with high availability across a unified infrastructure.
  • Enhancing patient care and reducing costs through improved immediate and long-term access to unified imaging information.
  • Improving productivity through unified workflow.
  • Delivering high value, scalable, cost effective unified diagnostic imaging solutions.

We invite you to look carefully at our product, its features and functions. This web site presents much of what a prospective user or institution may need to understand our solutions. In addition to a detailed description of the eRAD PACS® product line, you will find descriptions of the hardware components we use, demonstrations of our product, profiles of our staff and much, much more.

We recognize that a web site is no substitute for an informed account executive or service representative and invite you to contact us at any time by telephone or e-mail to open a dialog with eRAD. We are eager to assist you and look forward to the opportunity to do so.

Actually, these are lofty, admirable, and ultimately achievable goals. This is certainly in stark contrast to some companies out there with a mission statement of “Sure it works, sort of…”

One thing eRad can unify is an enterprise, or even a collection of disparate PACS systems. Uniting a (mostly) homogeneous enterprise shouldn’t be that hard, but it certainly can be. eRad’s approach. . .

is similar to most, with a server at each site feeding into the central server. Others call the peripheral computers spoke servers, and the effect is the same. At least they don’t call for a separate computer for each scanner, like some systems are wont to do.

Trying to tie several different PACS together is a larger pain in the neck (or lower), but eRad has a better plan for this than most:

This is the sort of thing the VA has done with ScImage. Basically, it involves grafting their system over “legacy PACS” such as GE, Siemens, Sectra, and Agfa. This would really come in handy in an environment such as ours where our nighttime readers have to contend with two worklists from our Agfa system (even though it’s one PACS, the two hospital worklists resident cannot be, well, united), one unified list from the Amicas dual hospital system, and others coming in on the group-owned Amicas. One single worklist would be a lifesaver. Of course, we would then have to come up with a way for the dictations to be routed to their proper site and so on. I’m not sure that one has been solved as yet. Still, the fact that eRad recognizes the need for such an umbrella is certainly important.

For me as a radiologist, the business-end of a PACS is its user interface. The web-demo as well as the live session gave me a good glimpse into eRad’s GUI. I’m not going to do the features justice, but I’ll try to give you a feel of what we have. Keep in mind that there are a lot of nice features. A LOT of them. Here is a screen-capture of their worklist:

and here is a snapshot of their viewer:

Clicking on the pictures above will expand them, and this is necessary to see what I’m going to talk about.

The worklist is customizable, and sortable on any column head. There is an indicator as to whether a study is locked, and there are various controls for batch reading, demographic editing, and for creating custom filters. ANY part of the DICOM header can be made a part of the worklist and subsequently a filter or search:

If one paid for PACS by the number of GUI features, eRad would be very expensive indeed. Here are just the extra modules that can be added when you first download the program:

These add some significant higher-level visualization to the package. Look at the viewer itself. The first thing that strikes me is the sheer number of buttons. The layout can be changed, by dragging the bars as you might with Microsoft Word. There are separate buttons for the major permutations of image and series layout. Most everyone else does this with popup menus and only one or two buttons. Many of the remaining functions have separate buttons as well. Notice that the buttons take up a lot of real estate that might otherwise go to the image viewports themselves.

There are good implementations of triangulation, linking, etc., and the ever-important mark-as-read-go-to-the-next-study button. When you click said button, there are various options as to how to leave the study status, or this can be determined from preset instruction. Window and level can be set from presets, slider controls, or the usual up/down/left/right drag. The series tray has a little indicator that tells if you have viewed all images. You don’t want to miss the lesion on the last slice, now do you? There are separate buttons to reset image info, window/level, zoom, etc. There is a very nice report view that allows importing of key images. There is a pretty good spine-labelling module.

When going through the demo, I hinted at my discomfort with the multiple buttons and controls. My friends from eRad reassured me that there is a default configuration supplied out of the box that will get most rads up and running. But if you want to get in and change things, there is a way to do that. Oh, boy, is there a way to do that… Here are just a few of the 14 tabbed control panels for customization:

Oh, and the right-click menu brings up something like this:

This is where I have to say, “Whoa!” There is really a tremendous amount of power here, and in some respects, there is too much power. I have called this approach to a GUI the “Lego PACS“, because of the high level of customization. As I’ve said before, this is to some degree a term of endearment, as Lego was my absolute favorite toy as a child, and it is still hard to resist when I go past a toy store or especially the Lego Store at the Mall of America. But I still prefer a simple, clean streamlined interface. Yes, the eRad default might give me that, but then there is the temptation that is hard to resist to go in and “tweak”. OK, that’s personal problem, and not eRad’s fault, but still. . .

This is a program with great potential, one to watch closely. I think personally that the GUI needs a little smoothing and seasoning, but it is certainly usable. eRad has addressed a number of issues that are close to my heart, and some that the “big boys” haven’t even solved as yet. I will expect to see great things from them in the future. In the meantime, if you ask me to say three words about eRad, I would say, “Pretty good.”

Microsoft Enters PACS Game

Microsoft has been dabbling in the Healthcare IT space, and the fruits of their labor (and acquisitions) will debut at HIMSS later this month. The press release notes” The package is called Amalga, which to me is all too close to the word “amalgam,” the mercury-containing metal that used to be used in dental fillings.

Microsoft Amalga: The new version of the product formerly known as Azyxxi, Amalga is part of a new software category called Unified Intelligence Systems that allows hospital enterprises to unlock the power of all their data sitting in isolated clinical, financial and administrative systems. Without replacing current systems, it offers an innovative way to capture, consolidate, store, access and quickly present data in meaningful ways for use by clinicians and executives of leading-edge institutions. Amalga is designed for hospitals and health systems that have invested in a diverse set of IT solutions.

Microsoft Amalga Hospital Information System (HIS): The new version of the product previously named Hospital 2000, Amalga HIS is a state-of-the-art, fully integrated hospital information system designed for developing and emerging markets. Amalga HIS is built around an electronic medical record (EMR) with complete patient and bed management, laboratory, pharmacy, radiology information system and picture archiving and communication system (RIS/PACS), pathology, financial accounting, materials management, and human resource systems.

Microsoft Amalga RIS/PACS: The new version of the product formerly known as GCS Amalga is now available as a stand-alone system as well as an integrated component of Amalga HIS. The integrated architecture means that a radiologist can use a single application to manipulate and study images and access the patient medical record. The workstation interface is optimized for radiologist workflow, including support for predefined templates, an intuitive report editor and voice recognition capabilities.

Of course, I’m more interested in the PACS product. The Microsoft Amalga website emphasizes the interconnectivity of the Amalga software:

Most companies don’t integrate their PACS and RIS software, translating into a forced fit between the systems. This often imposes the manual matching of studies between the PACS and RIS.

The Integration of Microsoft Amalga RIS/PACS provides a powerful, truly seamless system that can deliver quick, high-quality data to any department, which, in turn, can help hospitals increase patient turnaround time and enhance the patient experience. Integration also improves data integrity between PACS and RIS, can reduce transcription errors and duplication of data entry, and optimizes report turnaround. The system fully supports paperless, integrated workflows and facilitates easy access to patient medical information and order, scheduling, and study information.

The radiologist workstation is designed to optimize radiologist workflow. This unified system provides access not only to standard image manipulation tools, but also to the patient Electronic Medical Record (EMR), without requiring the radiologist to log on to external systems. Available information includes all previous laboratory and radiology results, the patient’s medication profile, and clinical notes. In addition, all historical studies are stored online rather than archived, which means historical films can be reviewed anytime, as needed, helping providers to improve their patient care.

The “key benefits” are mostly what you would expect:

  • Automatic order management integrated from EMR and RIS.
  • Automatic scanning and attachment of hard copies to study orders.
  • Image manipulation tools include 3-D cursor location.
    Multiple language support provides patient demographic and screen label data in any Unicode language.
  • Integrated database ensures patient medical information is accessible directly from the PACS.
  • Template-driven options include reporting and voice recognition.
  • Intuitive Report Editor accepts written or dictated reporting.
  • Preference for radiologist worklist studies are customizable.
  • Instant Study assignment to radiologist at time of ordering.
  • Warning system provides real-time notifications to prevent radiologists from reporting a study that is being reported by another radiologist.
  • Online Historical studies make historical studies available for quick retrieval, regardless of study age.
  • Unlimited Study revisions save any or all key images, window-level settings, and image annotations.
  • CD creation for PACS studies, Reports, and Electronic Medical Record. Options include DICOM and JPEG images, reports, and all or selected portions of a patient’s EMR.

Frankly, this doesn’t sound that much different than most modern web-based systems, at least those with online priors.

Microsoft, like some other big companies that come to mind, sometimes innovates and creates its own product, but often will simply buy the expertise and the software lock, stock and barrel. Such is the case here. Amalga started life as Azyxxi, which according to the Wikipedia, was “is a unified health enterprise platform designed to retrieve and display patient information from many sources, including scanned documents, electrocardiograms, X-rays, MRI scans and other medical imaging procedures, lab results, dictated reports of surgery, as well as patient demographics and contact information. It was developed by doctors and researchers at the Washington Hospital Center emergency department in 1996, and in 2006 it was acquired by the Microsoft Health Solutions Group, as part of a plan to enter the fast-growing market for health care information technology.” Washington Hospital is part of the MedStar Health system.

As for the RIS/PACS module, this has been borrowed/purchased/liberated (and expanded) from

GC RRITS SDN BHD (GCR), is a Malaysian Company, which provide integrated information technology solutions to the healthcare industry. The business is built around its core product, Hospital 2000, was developed and will continue to develop locally, based upon a proven international standards.

The RIS/PACS module of Hospital 2000 is called….drum roll, please….. Amalga! Here is a link to the full brochure, and here are some screenshots:

Interesting look. No doubt Microsoft will add its own touch to the interface. But how well does it work? How will it be marketed? I guess we’ll have to wait and see. Anyone want to send in a report from HIMSS?

Addendum:

Based on Aunt Minnie discussions, the RIS/PACS and other parts of the program may have been written by Global Care Solutions from Bangkok in cooperation with Bumrungrad Hospital. GCR may therefore be GCS’s Malaysian reseller. However, in the statement above, which is no longer on the GCR website, they refer to local development, and local for a Malaysian company probably means Malaysia rather than Thailand.

The GCS website (http://www.hospital2000.com) now resolves to the Microsoft Amalga website.

In the end, it probably doesn’t matter to anyone. Microsoft needed software in this space and found a reasonably good, and presumably reasonably priced product. Let’s hope they learn the lesson from GE and avoid running it into the ground.