New scanning method detects basal cell carcinoma early using AI

Researchers at FAU combine optical diagnosis method with AI-assisted image recognition.

Until now, diagnosing skin tumors has relied on detecting visible changes to the skin. Line-Field Confocal Optical Coherence Tomography (LC-OCT) is a well-established method used for diagnosis. Now, researchers at Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) led by Dr. Moritz Ronicke from the Department of Dermatology at Uniklinikum Erlangen have developed a new approach: In a feasibility study, the team introduced a procedure involving systematic screening of the faces of patients at a high risk of developing skin cancer. Using this procedure, researchers succeeded for the first time in detecting basal cell carcinoma even before any visible changes to the skin developed.

Basal cell carcinoma

Basal cell carcinoma is the most common type of skin cancer across the globe. It spreads locally, destroying the skin in its vicinity and thereby potentially damaging structures such as the nose or eyes if it occurs on the face. "Detecting basal cell carcinoma early allows for less invasive treatment options," claims Dr. Moritz Ronicke. "In some cases, all that is required is a cream."

LC-OCT delivers high resolution, three-dimensional images of the skin.

LC-OCT combines two techniques: optical coherence tomography (OCT), that indicates the thickness and extent of the tumor, and confocal microscopy ((LC: Line-field Confocal Microscopy), that takes clear images at the cellular level. The result: vertical, horizontal and three-dimensional images of the skin in real time with a resolution at micrometer scale, in other words a thousandth of a millimeter. By way of comparison, a human hair is approximately 50 micrometers thick.

The role of artificial intelligence

The AI delivers in real time a color-coded probability value for the presence of basal cell carcinoma. This allows the physician to recognize suspicious areas at a glance. It is still the physician who makes the final diagnosis.

Future applications

LC-OCT is already being used as standard to detect basal cell carcinoma in the early stages and monitor the success of non-invasive therapies.

At the current time, however, it is not possible to offer SUBSCAN routinely, as there is still a lack of data on the sensitivity of the method and it is still too time-consuming.

Source:

Friedrich-Alexander-Universität Erlangen-Nürnberg

Journal reference: