Skin scanner and AI used to ‘score’ disease severity in diabetics

Home / Research Updates / Skin scanner and AI used to ‘score’ disease severity in diabetics

Skin scanner and AI used to ‘score’ disease severity in diabetics

By Paul McClure

Researchers have used high-resolution scans of the skin of diabetics combined with AI to determine disease severity


Depositphotos

Researchers have used a high-resolution, non-invasive technique to obtain images of the tiny blood vessels under the skin of diabetics and used an AI algorithm to formulate a ‘score’ that can be used to determine disease severity. Once it’s made portable, the technique could be used to monitor the effectiveness of treatment.

Microangiopathy, where the capillary walls become so thick and weak that they bleed, leak protein and slow blood flow, is one of the major complications of diabetes. It can affect many organs of the body, including the skin.

Researchers from the Technical University of Munich (TUM) have developed a method of obtaining detailed images of blood vessels beneath the skin of diabetics and used AI to quantitatively determine the severity of the condition.

Optoacoustic imaging uses light pulses to generate ultrasound inside tissue. Tiny expansions and contractions of the tissue surrounding molecules that strongly absorb light create signals that are recorded by sensors and converted into high-resolution images. The oxygen-carrying protein hemoglobin is one such light-absorbing molecule, and since it’s concentrated in the blood vessels, optoacoustic imaging produces detailed images of vessels that other non-invasive techniques can’t. Moreover, it’s quick and doesn’t use radiation.

Here, the researchers developed a particular method of optoacoustic imaging called RSOM, short for raster-scan optoacoustic mesoscopy, that can obtain data on different skin depths simultaneously to a depth of 1 mm.

“Other optical methods do not achieve the depth or the detail reached by RSOM,” said Angelos Karlas, lead author of the study.

The researchers used RSOM to take images of the skin on the legs of 75 diabetics and a control group of 40 and used an AI algorithm to identify clinically relevant characteristics associated with diabetes complications. They created a list of 32 particularly significant changes to the skin’s microvasculature, including the diameter of blood vessels and the number of branches they had.

They observed that the number of vessels and branches in the dermal layer was reduced in diabetics, but increased in the epidermis, closer to the skin’s surface. All of the 32 characteristics they’d identified were affected by the progression and severity of the disease. By compiling the 32 characteristics, they calculated a ‘microangiopathy score’, linking the condition of small blood vessels in the skin and diabetes severity.

“With RSOM, we can now quantitatively describe the effects of diabetes,” said Vasilis Ntziachristos, the study’s corresponding author. “With the emerging ability to make RSOM portable and cost-effective, these findings open up a new way for continuous monitoring of the status of those affected – more than 400 million people worldwide. In the future, with fast and painless examinations, it would take just a few minutes to determine whether therapies are having an effect, even at home environments.”

The study was published in the journal Nature Biomedical Engineering.

Source: TUM

Leave a Reply

Your email address will not be published.