The visible human

By Andrea Gawrylewski

In 1991, Vic Spitzer and David Whitlock at the University of Colorado at Boulder went hunting for cadavers - one male, one female. It took two years to find bodies that were considered normal - meaning of feasible proportions (no more than six feet tall, 20 inches wide, and 14 inches deep), and no history of cancer, operations, transplants, or implants. "A normal cadaver is an oxymoron, because everybody dies from something," says Michael Ackerman from the National Library of Medicine, which funded the work. Finally, the researchers found what they were looking for: a 38-year-old male who died from lethal injection in a Texas prison, and a 58-year-old female who died from congestive heart failure, both Caucasian.

PSC Computer scientists used a grahical ray casting technique passed over the...

The scientists went about the business of freezing, slicing, and imaging the two cadavers, then feeding the data into a computer that would reconstruct the anatomy and physiology of the human body. The authors spent one year on each frozen cadaver, cutting slices 0.33 mm and 1.0 mm thick (for the female and male, respectively). Each cadaver was transected into slices with a milling machine (yielding 5,180 and 1,878 slices for the female and male, respectively), with digital and film pictures taken of each slice. The data set for the male was released in 1994 at 15 gigabytes; the female data set, 48 gigabytes, was released in 1995.

In the mid-90s only a few groups had the computing tools capable of manipulating what is now called the Visible Human data sets. (Now, of course, Ackerman carries both the male and female data sets around on his iPod.) The Pittsburgh Supercomputing Center was one of the first licensees to use the data. Led by Art Wetzel, researchers developed for the University of Michigan Medical School a volume browser, or tool for exploring in three dimensions various computer constructions of the data. Wetzel continues to incorporate into the data set the 70 mm film data set that was never processed, and produces even higher resolution of anatomic detail.

Since the Visible Human data set was released, more than 3,000 institutions in 42 countries, most of them medical schools, have licensed it. In one of Ackerman's favorite applications, Rutgers University researchers animated Rocky 3000, the Visible Human running, jumping rope, and doing pushups, exposed muscles rippling with every movement. In another project a car company (which Ackerman declined to name) is using the Visible Human data set to simulate how the body responds during a crash. The next step: trying to catalogue human variation using existing literature, capturing differences in human organ size, and the effects of disease.

Interested in reading more?

Magaizne Cover

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!