X-rays and magnetic resonance imaging are powerful tools in medicine, but neither tells doctors everything they want to know. To get a better picture of the structural effects and chemical details of various diseases, researchers are adapting a technology long used by satellites to study the ground below.
It’s called “multispectral imaging,” and it uses a camera and certain wavelengths of visible and infrared light to take pictures of moles and other surface structures. It can even photograph internal structures like the brain, say, or a tumor, because some wavelengths can penetrate the body without harming it. By choosing wavelengths tuned to different constituents of biological tissues, such as water and fat, researchers can pick up otherwise invisible details, much as satellites can “see” the heat of a dense urban area using the right wavelengths of infrared light.
Engineers at several universities-many with funding from the National Institutes of Health-are working to develop these imaging tools for medical applications. Because they can operate in real time, multispectral imagers could, for example, help guide doctors during surgery. Alexander Gorbach, a research scientist at NIH’s Clinical Center in Bethesda, MD, is building a system to help neurosurgeons distinguish between cancerous and normal tissue in a brain exposed for surgery-based on the fact that light bounces off the enlarged nuclei of cancer cells differently than it does off normal cells’ nuclei. The imager could help brain surgeons know precisely where to cut.
The technology could also help doctors get medical data without putting patients under the knife. “That’s one of the advantages, that you don’t have to remove tissue to get information,” says Michael Feld, director of MIT’s Laser Biomedical Research Center. Feld is working on a system to determine the calcium content of plaques in arteries, which could tell heart specialists if they’re likely to cause problems. Gathering such chemical clues is possible because each type of molecule reflects and absorbs light in a particular pattern.
Taking advantage of that effect could have a big impact on the diagnosis of some common cancers. Sergio Fantini, a professor of electrical engineering at Tufts University in Medford, MA, uses different wavelengths of infrared light to determine oxygen levels in breast tissue-important because tumors use more oxygen than normal cells do. Currently, up to 90 percent of suspicious spots found during conventional mammography turn out to be benign, but, says Fantini, “We hope to discriminate between benign and malignant tumors based on the oxygen level”-possibly sparing hundreds of thousands of women each year from painful biopsies. And researchers at the University of Texas at Austin are midway through a human trial to see if multispectral imaging could replace current tests for cervical cancer. They hope to cut the rate of false positives by as much as 40 percent, which they estimate could save $625 million every year. If such trials prove the medical value of multispectral imaging, doctors could soon see their patients in a whole new light.
AI is here.
Own what happens next at EmTech Digital 2019.