In 2020, amid the COVID-19 pandemic, researchers at the University of Michigan Medical School conducted a study on pulse oximetry measurement, focusing on white patients versus Black patients. A pulse oximeter measures the saturation of oxygen carried in a patient’s red blood cells. Since oxygen is among the most frequently administered medical therapies, pulse oximetry is a critical tool used to make clinical decisions.
The study found that, in two large cohorts, Black patients had nearly three times the frequency of occult hypoxemia that was not detected by pulse oximetry as white patients. The clinicians concluded that there is an ongoing need to understand and correct racial bias in pulse oximetry and other forms of medical technology.
Now, in response to those findings, researchers at Johns Hopkins University developed a new imaging technique that produces significantly sharper images for all people and—importantly—excels with darker skin tones. It produced much clearer images of arteries running through the forearms of all participants, compared with standard imaging methods where it was nearly impossible to distinguish the arteries in darker-skinned individuals.
“There were patients with darker skin tones who were basically being sent home to die because the sensor wasn't calibrated toward their skin tone,” said co-senior author Muyinatu "Bisi" Bell, associate professor of electrical and computer engineering at Johns Hopkins. “Now, we show not only there is a problem with current methods but, more importantly, what we can do to reduce this bias.”
Filtering out the noise
Medical imaging is often based on photoacoustic imaging, a method that combines ultrasound and light waves to render medical images. Body tissue absorbing this light expands, producing subtle sound waves that ultrasound devices turn into images of blood vessels, tumors and other internal structures. In people with darker skin tones, melanin absorbs more of this light, which yields cluttered or noisy signals for ultrasound machines.
For the study, published in Photoacoustics, Bell’s team tested 18 male volunteers with skin tones classified as light, intermediate, tan, brown and dark. Volunteer 1 had the lightest skin, while Volunteer 18 had the darkest.
The team tested their new imaging technique at laser wavelengths of 750 nm, 810 nm and 870 nm. They found that 870 nm worked best for all participants, but especially those with darker skin, including Volunteer 18. According to the study, the photoacoustic imaging taken of Volunteer 18 contained “strong clutter artifacts due to strong optical absorption at the skin surface, which compromises radial artery visualization.”
Essentially, the 870 nm wavelength was able to filter the unwanted signals from images of darker skin—similar to the way a camera filter sharpens a blurry picture—to provide more accurate details about the location and presence of internal biological structures.
The researchers are now working to apply the new findings to breast cancer imaging since blood vessels can accumulate in and around tumors. Bell believes the work will improve surgical navigation, as well as medical diagnostics.
It’s a research area that is well-known in forensic science, where researchers are working to improve bruise detection across diverse skin types. The detection, or non-detection, of bruises on a patient can have significant implications both medically and legally from a domestic violence and/or sexual assault standpoint. Previous studies have revealed unintended racial bias when documenting bruises on dark skin tones under regular light.
In a 2020 study published in the Journal of Forensic Sciences, forensic nursing experts at George Mason University used yellow, orange and red filters across 10 different wavelengths ranging from 350 to 535 nm to analyze bruises 21 times over the course of 4 weeks post-injury.
The findings showed the team was five times more likely to detect a bruise on study participants with an alternate light source than with white light. More specifically, alternate light at 415 and 450 nanometer wavelengths with a yellow filter resulted in the greatest bruise detection, regardless of skin color.
Lead study author and forensic nursing expert Katherine Scafide was surprised by the results, initially expecting that when it comes to darker skin, higher wavelengths would be more effective in visualizing bruising—which was shown in Bell’s photoacoustic imaging research—since melanin absorbs light across all wavelengths.
“We're aiming to mitigate, and ideally eliminate, bias in imaging technologies by considering a wider diversity of people, whether it's skin tones, breast densities, body mass indexes—these are currently outliers for standard imaging techniques,” Bell said. “Our goal is to maximize the capabilities of our imaging systems for a wider range of our patient population.”