Thursday, August 20, 2009


One of the joys of being the Dalai Lama of PACS is that people in the industry occasionally seek out my opinion on some product or other. Some of these folks are big names in the business, and I'm always honored to share what little wisdom I might possess.

I first spoke with Hamid Tabatabaie when he was CEO of Amicas. He was about to come down to my little operation in the Deep South to see how things really work in the boonies. Sadly for me, Hamid left Amicas just before his proposed trip, and I didn't get to meet him until a few years later at an Amicas RSNA party. Amicas is one of those companies that doesn't make too many enemies along the way, and I have always been impressed to see a number of former employees (and one former CEO!) at their various events. When I finally did get to speak with him in person, Hamid, the consumate enterpreneur, hinted at a new, revolutionary venture. He asked for my help when the time was right, and of course I agreed.

Hamid's new baby is called lifeIMAGE, and it's potential is almost literally unlimited. Here's the premise:
The U.S. healthcare system wastefully spends $10–15 billion each year on redundant radiological exams that stem from a provider's lack of access to patients' prior imaging exams. Though most diagnostic exams are already available in standard electronic formats, a void exists in the sharing of existing imaging data. By filling that void, the healthcare system can save billions of dollars, eliminate needless exposure to patient radiation, and significantly reduce medical errors. . .

Before long, patients in the U.S. will choose to trust their health records to one or more health record aggregators in much the same way that they currently manage their money online in banks. Accessing and exchanging health records will be modeled after some current banking methods including PINs, routing numbers, wiring instructions, and deposit or withdrawal transactions. In this model, a number of major and specialty aggregators will service the market.

Much remains to be done in extracting and organizing quality data from the originating systems and information silos within the walls of hospitals and other healthcare settings. lifeIMAGE can provide a highly specialized and valuable service by focusing on the imaging component of the medical record. . .

The lifeIMAGE product platform will more effectively connect patients and providers to medical imaging studies and reports through two core components. First, for hospitals and imaging centers, the lifeIMAGE Local Appliance (LILA) will provide a better way for enterprises to manage imaging exams introduced by patients on portable media such as a CD. Next, a secure, cloud-based environment, that will provide the basis for more convenient image sharing and storing between patients and referring physicians.

Trust me, lifeIMAGE is addressing problems that need to be solved. A huge number of exams have to be repeated on an emergency basis because the original, performed at St. Elsewhere, cannot be retrieved in a timely manner. In fact, a Brigham and Women's study from 2008 found that 32% of patients transported to their ER who DID have CD's with them still got additional exams. 40 of those CD's didn't have reports, and 27 of those that did have reports had discrepancies.

We all know about the joys of exams coming in on CD-ROM. Even if the disk does actually load (which doesn't always happen), what do you do with the data? Do you pull it up side-by-side with the PACS? Do you load it into PACS? HOW do you load it into PACS? This all begs the question of the "portable patient," the little old lady who has been to St. Elsewhere last week, Mayo the week before, and the Doc-In-The-Box down the street yesterday. This situation of course is truly the bane of my existence. You may recall that I tried years ago to get our big hospital to partner with all of the other hospitals in town on a city-wide PACS to solve some of these issues, but of course, the then-CIO looked at me like I had three heads.

lifeIMAGE acknowledges the various realitites in our current system, and has found ways to deal with them effectively. Here is the general overview of their operation:

Basically, lifeIMAGE becomes the conduit, and repository, for images that need to be, well, portable.

The first part of this daunting task is to simplify it, and lifeIMAGE does make it easy to upload, nominate, and share studies:
  • Let anyone upload from any desktop

  • Upload images from CD to an inbox with a universal, zero-download viewer
  • Allow all physicians to nominate exams for inclusion in RIS/PACS and worklist for secondary read

  • Equip the system gatekeeper(s) with means to electronically accept or reject nomination requests

  • When accepted, automatically generate an accession number, and if needed, an order number

  • Gives physicians in boxes so that they may share for clinical and personal reasons

  • Enables sharing for collaboration
The first step in all this is the LILA, the lifeIMAGE Local Appliance. This is a server box that facilitates the uploading of a study into PACS. Unlike the "fire and forget" approach, this system creates the proper trail into (and out of) the RIS/PACS for the exam handed to the doc on CD.

Here's how it works in practice...just view the slideshow to see how you would upload a new study and nominate it to be included in PACS. The gatekeeper allows normalization of patient demographics and so on, and creates an accompanying order in the RIS. You can also see the concept of the "inbox" which houses studies for the particular physician, or patient. The inbox can be searched, tagged, or shared.

What you see above is primarily LILA version 1.0, the Local Appliance itself. One interesting point here: the viewer is actually written in Adobe Flash, allowing it to work across different platforms.

LILA 1.2 includes the Enterprise Indexer:

  • Index every report and image dataset in RIS & PACS

  • Instantly locate and display all cases of interest by:
    o Relevance
    o Modality, procedure, keyword, tag, and synonyms
    o Natural language & medical lexicon based searches

  • Bookmark or save any search

  • Add personal tags

  • Use for decision support, research, or education

  • Interactive and customizable charts for refining search criteria

  • Research mode for clinical analysis

  • Dashboard elements for tracking operational efficiencies
    o Imaging business intelligence

  • Investigating trends and discovering anomalies
LILA 2.0 adds something very critical to the future: cloud distribution to referring physicians, and sharing and communication via LILA 3.0 adds an inbox for the individual patient, as well as searching across the LILA.

While the initial components are remarkable in their own right, is the cloud distribution that takes my fancy. The cloud is a virtually-unlimited space in which to store data, and should be able to handle just about any amount presented to it. (Cloud storage and computing are offered by a number of vendors.) Here is the opportunity to create a regional, national, or even a world-wide database. The possibility finally exists for a repository of every image of every patient, no matter where in the world they may be. Ultimately, I'm thinking lifeIMAGE's technology could prove to be the ultimate (vendor-neutral, of course) PACS database. In the meantime, simply streamlining the process of CD-ROM loading into PACS is a monumental achievement in and of itself.

I'm looking for great things to come from this, folks.


Anonymous said...

Sound very unique and innovative. Is the product commercialized yet and is anyone using it outside of Mass General? Neat stuff.

David Clunie said...

So this is a competent identity-reconciling CD importer coupled with a regional archive and index ?

Cool, if they do a good job of it, and provide features that your PACS is lacking (but should have anyway).

Why do you want YAFV (yet another ... viewer), when everything should be viewable, comparable and reportable in your local PACS ?

One of the things the Canadian Health Infoway project has worked on is providing a regional archive, but making access to images in it via the local PACS interface feasible and transparent. Easier said than done of course. The central indexing of reports and images is something IHE XDS registries provide.

I will be interested to see what Life Image's identity provisioning, authentication and access control model is, which is always a problem with these types of solutions, and has been challenging for the Canadians and other XDS-based approaches.

I will also be interested to see what Life Image's business model for the service is (e.g., who is paying to store and access what and when).

I do hope that Life Image is going to provide standard mechanisms for contributing to and accessing the system ... the LILA "appliance" does not sound promising in this regard, with yet another ... box (YAFB) installed locally.

Life sure would be a whole lot easier if the government just decided to build a single nationwide PACS database and archive.


Anonymous said...

Curious to see what their business model is as well. There are already a couple players in this space - AGMednet and Intelemage. I wonder how this is different. Agree with Clunie on the LiLa device. AGMedNet is the same way, though Intelemage is hardware agnostic and web based.