Standards Sought for Hyperspectral Imaging

A team from the National Institute of Standards and Technology has begun gathering data they expect will lead to universal calibration standards for hyperspectral imaging, saying the lack of such standards has impeded this noninvasive imaging technique.

“The potential of the technology has been proven, but the problem is that researchers are simply lacking a way to assure consistent results between labs,” said NIST researcher David Allen. “Standards development has itself been hindered by a lack of human skin reflectance data, especially in the ultraviolet and short-wave infrared.”

The top image shows skin as normally viewed. At bottom are samples of the same images with enhanced contrast in false color via hyperspectral imaging to show the variability between subjects. Courtesy of NIST.

The researchers have begun investigating how human skin looks under various wavelengths of light. Hyperspectral imagers are extremely sensitive to many wavelengths, from UV to IR. This is unlike conventional imagers, which can see only a narrow portion of the electromagnetic spectrum.

The researchers have also been working to quantify the spectral variability within an individual, as well as between individuals, that exist due to inherent biological differences.

As part of the study, the researchers have been collecting data from 28 subjects. Images of normal tissue on a test area on each subject’s forearm have been taken, in addition to three reflectance measurements of that test area.

Healthy tissue must first be imaged and analyzed before the researchers can “delve into what diseased tissue looks like hyperspectrally,” said Dr. Catherine Cooksey, a researcher in NIST’s Sensor Science Division.

“Skin reflectance varies due to skin pigmentation, tissue density, lipid content and blood volume changes,” she said. “We need good data from a wide variety of sources.”

For more information, visit www.nist.gov.