Detecting Skin Cancer
An optical solution may detect melanoma without a biopsy.
As a native of Nigeria, ORNL biomedical engineer Justin Baba never gets sunburned and has no significant risk of contracting melanoma, the fatal skin cancer. But he is passionately committed to helping light-skinned people by developing an accurate, portable, noninvasive technique for detecting melanoma.
Aided by ORNL's Philip Boudreaux and Nance Ericson and an optical company, Baba has built the first clinical automated instrument with no moving parts for using polarized light to image subsurface features of moles. The instrument has been tested on nine melanoma suspects at M. D. Anderson Cancer Center in Houston, Texas. "The preliminary results are encouraging," he says. While noting that the instrument shows promise in distinguishing between cancerous and non-malignant moles, he cautions that a much larger patient sample size is needed.
"I am really excited about this technology," says Baba, a staff researcher in the Monolithic Systems Development Group of ORNL's Engineering Science and Technology Division. "M. D. Anderson researchers like our optical biopsy instrument but they want us to get funding to miniaturize it." What physicians would like, he explains, is a Star Trek-like wand they can wave over patients to spot tumors in their early, curable stage.
Melanoma is the most fatal of skin cancers. However, if detected early, this skin cancer almost always can be cured by surgical removal of the melanoma, a dark-pigmented, malignant tumor that arises from a melanocyte—an epidermal cell that synthesizes melanin, the dark pigment found in skin and hair.
"Melanocytes are located deep in the skin's bloodless epidermal layer," Baba says. "But right below that layer is the dermal layer, which contains blood vessels. As the melanoma grows, it stretches down into the blood vessels. Once in the blood the cancer spreads like wildfire. In just a few months, the patient can develop lesions in the kidney and liver and eventually die."
Today, if a patient is concerned about a mole, the physician assesses whether a more intensive examination is required by visually inspecting the suspect mole, taking into account size, shape, color and the rate and nature of its growth. Suspect lesions are biopsied—a piece of mole is excised and examined for cancer. This process is invasive, painful, expensive and time-consuming but increasingly necessary for liability reasons. Madeline Duvic, an M. D. Anderson clinician and one of Baba's collaborators, claims that in her clinical practice typically only two of about 200 skin biopsies she performs yearly test positive for melanoma.
What physicians need is a noninvasive, highly accurate technology to screen people at high risk of getting melanoma, such as blondes with fair complexions and redheads with freckles who are exposed to sunlight, have been sunburned, and do not always wear sunscreen. "If physicians can detect and treat melanoma in its early stages, they can save lives," Baba says.
Baba and Gerard Coté, his Texas A&M University Ph.D. adviser, advocate the use of polarized light for noninvasive biopsies of suspect moles. Polarized light photons passing through the skin are scattered from fibrous proteins, or collagen, in the dermal cells. Collagen fluoresces when exposed to blue polarized light. A camera images the mole's underlying structural features, including melanoma-induced disruptions in collagen.
Compared with normal cells, cancer cells are larger, possess denser nuclei, require more blood flow to provide nutrients needed by fast-growing tumors and scatter light differently. Multi-colored polarized-light images reveal details that indicate these differences.
"Our potential solution to skin cancer detection is an automated Mueller Matrix polarization imaging system," Baba says. "The Mueller Matrix algorithm is a powerful tool for analyzing the images by taking into account changes in the linear and circular components acquired by polarized light scattered from moles. Thirty-six images are combined into 16 images that reveal differences between cancerous moles and normal ones."
In time, Justin Baba could have a significant answer to skin cancer.
Web site provided by Oak Ridge National Laboratory's Communications and External Relations