December 1999

A virtual necessity

The Virtual Human will be a massive, but immensely valuable, undertaking

The sky’s the limit when you think of useful applications of a “Virtual Human”— a computational model of a human being. Imagine an integrated, functioning simulation of the processes of all the bodily organs, united into a model of a living, breathing person.

Recently ORNL organized a multiagency meeting of leaders to discuss the merits and challenges of the Virtual Human program. The group, which included officials of the National Institutes of Health, DOE, the National Science Foundation, the National Academy of Sciences, the Department of Defense, numerous universities and the private sector, not only concluded that the Virtual Human program was worth doing, but also attached an urgency to the task.

"The time is right for physics and biology to get married."
“This could be man’s largest and most complex undertaking since World War II,” said one attendee.

To temper the enthusiasm, ask this question: Where does one start?

“To build a house, you dig a hole and lay some blocks. To create a virtual human, the issues are far more complex,” says the Life Sciences Division’s Clay Easterly.

Easterly is spearheading ORNL’s drive to attract interest and funding for a nascent Virtual Human program. Easterly says that ORNL was involved early on in human modeling and is well situated to lead an international effort to create a computerized, Web-based model.

ORNL built “phantoms,” physical and mathematical models, shortly after World War II to estimate radiation doses received by survivors of the Japan atomic bomb blasts. As strides were made in health physics, the phantoms evolved into a “Reference Man” and the mathematical models became more sophisticated and complex.

Easterly believes that ORNL can use its historic expertise to pull an international effort into an integrated human model. The unabated progress of computer technology makes fantastic ideas such as this one seem doable.

“We recently completed a roadmapping exercise,” he says. “We asked what the risk elements are: Do we have sufficient data in open literature? Who are the good players? “Such an exercise would take many years and much money. We can only do elementary judgments, but the responses that we’ve been getting have been positive. It’s a good sign that we’re not crazy.

“The time is right for physics and biology to get married,” Easterly declares.

The processes that allow us to live can be expressed succinctly in terms of science, Easterly says. For instance, the transfer of oxygen from the air to the bloodstream through the lungs, propelled by the beating heart, is “not super hard” to describe in terms of physics and chemistry.

“On the other hand, modeling the liver would be very difficult. The liver is a big, motionless chemical plant,” he says.

ORNL’s Virtual Human work has been supported initially by internal program development and LDRD funding and, to a smaller extent, by the U.S. Marine Corps, which is interested in computational models for the development of nonlethal weapons. Tests of kinetic weapons being developed have been confined to gelatin models, which are not that effective for nonlethal weapons studies.

In that instance, a Virtual Human’s application would be similar to ORNL’s computer simulations and animations of car crashes. But Easterly believes the eventual uses of a Virtual Human are as undefined now as the Internet’s impact was in the early ’90s.

“One thing I’d like to do with Virtual Human is provide a focus for development, where we could factor in all of the contributing elements to living. I’m thinking of things like diurnals (day-night cycles) and physiological variations over a lifetime.

Eventual uses of the Virtual Human are as undefined now as the Internet's impact was in the early "90s.
“A personal model could store your medical records—health experiences, CAT scans, information from biomedical sensors. Such a model would allow physicians to scan back and forward to refer to past treatments and how they have worked. One could also factor in genetics. These records could be compressed into a functioning model.

“In 10 to 20 years we might have terabytes of data storage on a three-by-five card,” Easterly says, envisioning a computer model of oneself filed away in the doctor’s office. Current major undertakings, such as the Human Genome project, could be factored into the Virtual Human.

“The Human Genome project wants to sequence information. The data is all very similar in nature. Our vision is to stimulate the integration of that and similar models that have been or are being developed. A lot of work is being done already, in normal body processes and with diseases like emphysema.

“Western science breaks things down into pieces and studies the fool out of them. We do that well; just look at the journals or a typical dissertation. With the computational toolboxes that are increasingly available to us, I think we have a responsibility to take these pieces and figure out how these models work together.”

It will take much planning, work and funding, Easterly acknowledges, before Virtual Human can even crawl. When the model is up and running, however, the benefits will prove invaluable.—B.C.


[ ORNL Reporter Home | ORNL Home | Comments ]