AI-Powered Skin Scanner Assesses Diabetes Severity by Measuring Microvascular Changes

To study these changes, researchers have now combined artificial intelligence (AI) with innovative high-resolution optoacoustic imaging technology. This method allows for the assessment of microvascular changes in the skin, providing insights into the severity of diabetes.

Optoacoustic imaging technology operates by using light pulses to generate ultrasound within tissue. These ultrasound waves, produced by the minor expansions and contractions of tissue around light-absorbing molecules like hemoglobin, are captured by sensors and transformed into images. Hemoglobin’s high concentration in blood vessels makes optoacoustic imaging particularly effective for creating detailed images of these vessels, surpassing other non-invasive techniques in clarity. Although the principles of optoacoustics have been known for over a century, their application in medicine is relatively recent. A research team at the Technical University of Munich (TUM, Munich, Germany) has been pioneering various optoacoustic imaging methods, including "Raster-Scan Optoacoustic Mesoscopy” (RSOM), to explore the impact of diabetes on human skin.

In their study, the TUM researchers used RSOM to image blood vessels in the legs of 75 diabetic patients and a control group. An AI algorithm was then employed to identify diabetes-specific characteristics from these images. The researchers compiled a list of 32 significant changes related to alterations in the skin’s microvasculature, such as variations in the number of vessel branches or their diameter. Each of these 32 features correlates with the progression and severity of diabetes. However, it is the combined analysis and scoring of these characteristics that reveal the relationship between the skin’s small blood vessel condition and diabetes severity. This approach marks a significant advancement from traditional methods like biopsies, which, while established, do not accurately represent living conditions, can deform blood vessels, are invasive, and unsuitable for extended observation.

In contrast, RSOM measurements are non-invasive, quick, and do not require radiation or contrast agents. Unlike other optical methods, RSOM achieves unmatched depth and detail. A single RSOM measurement can simultaneously gather data from different skin layers. This capability allowed the researchers to discover that diabetes impacts vessels in varying skin layers in distinct ways. For instance, the number of vessels and branches in the deeper dermal layer decreased in diabetics, while increasing nearer the skin’s surface in the epidermal layer.

"With RSOM, we can now quantitatively describe the effects of diabetes," said Vasilis Ntziachristos, Professor of Biological Imaging at TUM. "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."

Related Links:
TUM