Nov 19 2013

“Towards A Universal Name Space” – Dec 6, 2013 Workshop at MIT

Published by under workshops

We will be meeting at the offices of the W3C at MIT.

The success of the World Wide Web was due to an amazingly simple set of initial conditions.  In Weaving the Web Tim Berners-Lee wrote,

What was often difficult for people to understand about the
design of the web was that there was nothing else beyond URLs, HTTP, and
HTML.  There was no central computer “controlling” the web, no single
network on which these protocols worked, not even an organization anywhere
that “ran” the Web. The web was not a physical “thing” that existed
in a certain “place.” It was a “space” in which information could exist.”
Today’s health IT landscape has taken a wildly different approach.  The HITECH funding supported a top-down, “Castles and Drawbridges” approach, focusing on enterprise-centric information systems that are then supposed to be interfaced.  Whereas Google has given us immense power to find information quickly and in context, healthcare has stuck with primitive categorization scheme akin to libraries using card catalogs organized by card catalogs using the Dewey Decimal System, everything in its place and a place for everything.  At it’s root, medical nomenclature is based on assumptions of Aristotle’s law of the excluded middle – things can be either A or Not A, but not both.  We roll these assumptions up in to Boolean Lattices, layer upon layer of hierarchical decomposition which appear to give us precision and consistency. In reality, however are being dragged into a morass of exploding complexity, and our attempts to improve the situation are like trying to get out of a hole by digging it deeper. The hospital, Peter Drucker said, is the most complex organization in our society.  Over and above the complexities of modern medicine and discoveries in the life sciences, we have overlaid immense technical, economic, and political complexities in recent years.
It is questionable, whether we have the intellectual paraphernalia to deal with this level of complexity.
Taking lessons learned from the web, this invitation-only workshop will look at alternatives technology to get out out of the complexity morass.  We will examine Linked Data, using RDF and Semantic Web technology to provide a flexible model for linking health information at the patient level.  Rather than locking patient information into information “castles” and trying to establish a network of “drawbridges” to interface everything, an alternative approach would be to look at Health IT as creating an “information space” within which health information could exist.  This would become a large-scale, fine-grained network of information, giving us great freedom to rethink privacy, security, and provenance, in addition to supporting hierarchical structures as appropriate. (Here is a conversation between Ward Cunningham, creator of the Wiki, and Tom Munnecke, one of the architects of the VA VistA EHR system, comparing how their designs started from simple initial conditions.
A critical first step in this architecture is to define a URI (Universal Resource Identifier) or similar framework that is capable of naming all health information at .  What would this URI look like, and how could it facilitate advanced health IT?
This workshop is a continuation of one at MIT in April, 2013, as well as “RDF as A Universal Health Exchange Language” in Encinitas, CA. June 2013.
We will be meeting at MIT, courtesy of W3C, at Stata Center,  Friday, Dec 6 from 10AM to 2 PM
We are looking for “unencumbered thinkers” to participate in this workshop.  If you are interested in attending, please email [email protected]
Share

No responses yet

Sep 15 2013

NTakes: Divergence in Health Information Technology

Health care still exists in an iPhone versus Android world. We speak of our health care “system,” but that term implies rationality that is lacking when it comes to exchanging data.

Here is a personal example. A friend facing a major surgical procedure decided to secure a second opinion. He requested and received relevant portions of his medical record — on paper, at 75 cents a page — along with a CD containing imaging studies that had been done. His initial care was provided at a major teaching hospital, so he went to a similar medical center for the second opinion. But, when he presented the CD to the consulting physician, the response was, “We can’t use that.” It was in a format not compatible with the IT system at the second hospital.

“They knew where the CD was coming from. Why didn’t they warn me about that in the first place?” my friend wondered.

But, why should that have been necessary? Shouldn’t medical systems speak the same language?

On a much larger and incredibly expensive scale, health information technology has developed along dozens of proprietary lines, making the iPhone-Android divergence seem paltry in comparison. Numerous projects are being undertaken to have these disparate systems become “integrated” — a Rube Goldberg undertaking that by nature is overly complex and destined to be fraught with problems. We need only to look at the decades-long effort to “integrate” Defense Department and Veterans Affairs electronic health records. VA’s VistA is an open system built for clinicians that has proven itself over the years, gaining for VA the title of best health care in the U.S. DoD has gone the commercial, let’s-plug-pipes-and-wires-together approach and is still struggling to have a system its users actually want to use. (See Tom Munnecke’s open letter to Secretary of Defense Chuck Hagel).

Just sitting here at my iMac, I can have Amazon and Facebook connect or instantly send to Twitter an interesting article from almost any website. It’s all in the magic of the URL — a semantic approach to data.

Healthcare IT systems could use some of that magic. But, it will take a different sort of magic for the proprietary firms with mega dollars on the line to be willing to evolve to a more rational, web-like approach to health information.

Share

No responses yet

Jun 26 2013

Report on Workshop “RDF as a Universal Health Language” Encinitas, June 25-26, 2013

Published by under workshops

These are the videos from the workshop..

Share

One response so far

Jun 12 2013

Universal Health Exchange Language Workshop – Encinitas June 25-26, 2013

Published by under workshops

The President’s Council of Advisors on Science and Technology issued an influential 2010 report on Health IT calling for the creation of a Universal Health Exchange Language.  This was discussed at a recent workshop in San Francisco RDF as a Universal Health Exchange Language and previously at the New Health Project meeting at MIT and the New Health Project Science of the Individual workshop in San Diego.  I also discussed these topics in a video interview with Stanford Computer Science/Medical Informatics professor Gio Wiederhold.

We will be holding a workshop at my home in North San Diego County June 24th – 25th to examine both the theoretical and practical foundations of implementing a Universal Health Exchange Language.

If we are designing a language, we also need to address the pragmatic issues surrounding the language.  Who or what is the speech community who will be using it?  What will they be using it to express? Does this language supplement or replace spoken language?

It also brings up deeper questions relating to the foundations of the medical informatics involved:   At one end of the spectrum is a highly structured coding system, akin to a Dewey Decimal System indexing books in a card catalog in a library.  Books that don’t fit well into any given category fall out of the sieve into “miscellaneous” or the 000 category.  The other end of the spectrum is Google, treating “everything as miscellaneous,” and using no predefined structure to the data involved.

Schedule (draft; specific times for hangouts will be posted later this week)

Tues, 2-5PM (PDT)  “Beyond the Boolean Lattice” will examine the theoretical foundations of this, looking towards ways of sense-making of medical information that don’t depend on Boolean categorizations.  This might include Category Theory, Algebraic Group Theory, or other other approaches.  This will also examine how RDF/Semantic Web technologies might be used as a foundation to “bootstrap” current health IT technology to innovative ways of dealing with health information, clinical decision support, and research.  Tentative speakers:  David Ellerman, Erick Von Schweber, Ben Grosof (Hangout), John Mattison, Peter Norvig (Hangout)

5:00PM refreshments/appetizers, followed by Dinner.  Hosted by Conant and Associates, a physician-led health IT consulting group specializing in Nuance speech recognition technology.  Please contact Tom if you have any dietary restrictions.

Weds, 9 AM – 3 PM (PDT) “What is the Universe of a Universal Health Exchange Language?” This will be a broader look at the practical applications of a Universal Health Exchange Language.  We will look at language and speech communities from an anthropological perspective, with lessons learned from the design of the VA’s VistA.  We will look at Health IT and communications from an “information space” perspective, not simply a collection of interfaced enterprises.  We will look at issues of privacy, provenance, and complextity  from the perspective of a universally addressable information space, and what a universal health exchange language might facilitate this broader “universe” of communication.  Lunch and refreshments will be included.  David Booth (Hangout), Adrian Gropper (Hangout), Emory Fry, Tom Munnecke, Christophe Lambert (Hangout), Reed Conant, Conor Dowling (possible Hangout) + TBD

 Remote Access:

The primary purpose of this workshop is a small, face-to-face “salon” style discussion.  However, we recognize that there are many who can contribute to the discussion who can’t physically make it to the meeting.  Therefore, we will be setting up a Google Hangout for remote participants to listen in.  We will have some scheduled remote presentations via the Hangout, and may open the Hangout for general remote user comments, but that will be decided as the workshop progresses.

Attire:  San Diego Casual.  The evening breeze may be a bit chilly, so bring a jacket if you want to be outside.

To Participate

Please register at http://www.eventbrite.com/event/7074105841

Shared Folder

I’ve set up a Shared Google Docs Folder for presentations, notes, and papers.  This is an open directory for anyone having the link.  (For now).  We will have this document available for shared notes

Video/Photography:

We will be photographing and/or video recording during the workshop, which will be published online for general Internet viewing.  If you wish to opt-out of this procedure, let me know and we will respect your wishes.

For more information, contact Tom Munnecke [email protected]

Share

One response so far

Jun 05 2013

Interview with Gio Wiederhold, Stanford Professor of Computer Science

Published by under Uncategorized

While I was in San Francisco to attend another conference,  I stopped by the home of Dr. Gio Wiederhold, professor of Computer Science at Stanford.  He was very helpful to me in my early days of the design of VistA, so it was fun to have some time to do this oral history interview of him:

Share

No responses yet

Apr 23 2013

Report on Semantic Health Workshop at MIT April 19-20, 2013

Published by under workshops

MIT Workshop on Semantic Web for Health Care

Tom Munnecke, Joanne Luciano, Eric Prud’hommeaux, Brian Ahier, Adrian Gropper, Alex Tam, David Booth Photo by David Booth

We held scheduled a workshop at MIT’s Stata Center, courtesy of the W3C consortium.  Unfortunately, events surrounding the Boston Marathon Bombing put the city on lockdown.  So we had a Google Hangout on Friday, Apr. 19, an then held another meeting at MIT on Saturday afternoon.  Here are the notes from the meetings:  2013 MIT Semantic Health Workshop 19-Apr-2103  and 2013 MIT Semantic Health Workshop 20-Apr-2013

Here are the photos from the event:

"I need my Space"

Share

No responses yet

Mar 27 2013

There’s Gotta Be a Better Way

Published by under opinion

100,000+ pages of federal legislation controls our health care system today.  We are increasing the number of disease categories 10-fold, moving towards a disease coding system (ICD-10) that names 150,000 categories, including categories such as W59.83D “Subsequent encounter of being crushed by other nonvenomous reptile.”  An enormous array of standards committees seek to define coding systems, exchange mechanisms, privacy and security, and medical knowledge bases.

One way to look at this is to marvel at the scale of it all.  Perhaps when we reach 200,000 pages of legislation and 1.5 million terms for how to be sick, we will finally get a handle on the US health care system.

Another way to look at it is that we are trying to get out of hole by digging it deeper.  What we see are the symptoms of the Humpty Dumpty syndrome: we have broken health care into so many pieces that even with “all the kings horses and all the kings men” we can never put it together again.

Continue Reading »

Share

One response so far

Mar 21 2013

Report on Workshop: “Science of the Individual” Nov 14, 2012 in Encinitas, CA

Published by under workshops

These are presentations made at the November 14, 2012 New Health Project Workshop “Science of the Individual” in Encinitas, Ca.
Futurist David Brin talking about privacy, transparency, and accountability:

Caregraf CTO Conor Dowling, talking about Semantic Web

Science Fiction author Vernor Vinge talking about accelerating change

Heather Wood Ion

Scenes from the workshop

Tom Munnecke, Heather Wood Ion, Vanessa Moeder, David Brin, Vernor Vinge, Nancy Tomich, Christophe Lambert

Share

One response so far