The New Language of Anatomy

Summer 2012

Michael I. Miller, Hershel and Ruth Seder Professor of Biomedical Engineering and director of the Whiting School’s Center for Imaging Science. Photo Credit: Peter Howard

Back in 2003, a pair of economists estimated that roughly 2 billion medical images were being created annually by hospitals worldwide. In the nine years since then, that pace may have doubled. The vast majority of those billions of images are glanced at briefly by a radiologist and then placed in the patient’s records, never to be studied again.

But what if those images could be fed into computer systems that would slowly be trained in the language of human anatomy, much as the software that underlies Apple’s Siri system has slowly learned the rules of human language by processing millions of spoken utterances? If computers studied the lexicon of the body-if they learned the thousands of tiny anatomical signs that mark various diseases across thousands of individuals-they could offer revolutionary new tools for early diagnosis and treatment.

That, at least, is the dream of Michael I. Miller, MS ’79, PhD ’84, the Hershel and Ruth Seder Professor of Biomedical Engineering and director of the Whiting School’s Center for Imaging Science. In the late 1990s, Miller had an insight that has transformed the mathematical analysis of human anatomy. Fifteen years later, that insight is beginning to bear fruit. With the help of physicians and other biomedical scientists from throughout Hopkins, Miller and his engineering colleagues have built computing systems that are studying small-scale anatomical features associated with heart failure, Alzheimer’s disease, schizophrenia, and a long list of other conditions.
What does the work of painter Pablo Picasso and linguist Noam Chomsky share in common with Miller’s work? Watch this video to find out.

At age 57, Miller is an imposingly tall man with warm eyes. He first came to Hopkins as a graduate student in 1976, and he has spent the bulk of his adult life here. (From 1983 to 1998, he worked as a researcher and faculty member at Washington University in St. Louis.) As he sits on a bench in front of Clark Hall, where his center is housed, he gestures at the quad and recalls how small the Hopkins biomedical engineering program (then one of the few such programs in the world) was when he arrived during the Ford administration.

“I believe there were only four doctoral students who started with me,” he says. “Today there are 20 or 30 every year. Back then, there was no undergraduate program in biomedical engineering. Now we have 400 undergraduates, and I think it’s one of the healthiest undergraduate majors in all of Hopkins.” But ask Miller if he foresaw any of this spectacular growth back in 1976, and he says no: He’s as surprised as anyone by the field’s explosion. In any case, he says, that’s not the kind of topic that he tends to dwell on. As he tells it, he’s been extremely fortunate to find and solve one interesting intellectual problem after another, and he’s spent 36 years keeping his head down, looking for new solutions.