Skin cancer diagnosis: researchers develop 3D imaging of skin structure

Contents

Researchers at Leibniz Universität Hannover (LUH) and Rostock University Medical Center have developed a procedure that uses three-dimensional imaging to reliably detect malignant skin changes at an early stage. The team led by Prof. Dr. Bernhard Roth from the Hannover Centre for Optical Technologies and Prof. Dr. Steffen Emmert from Rostock University Medical Centre received the Helmholtz Prize 2024 for the development of the skin cancer scanner.

The prize is endowed with 20,000 euros and is awarded for precision measurements in basic research and applied metrology. The award ceremony took place as part of the 24th International Metrology Confederation (IMEKO) in Hamburg, which has been meeting since 1958.

Looking under the skin without a scalpel

According to LUH, the interdisciplinary team from medicine and physics has developed a system "for three-dimensional imaging of the skin structure, including determination of the so-called depth of invasion and the benignity or malignancy of skin lesions". Previously, in everyday medical practice, it was necessary to surgically remove a tissue sample and examine it histologically in order to diagnose black skin cancer (melanoma) in particular. The new system enables non-invasive melanoma diagnostics – quickly and by laser scan from the outside.

The system combines four different measurement methods: optical coherence tomography (OCT), Raman spectroscopy (RS), photoacoustic tomography (PAT) and high-frequency ultrasound imaging (US). The essentially laser-based system enables measurements below the legally prescribed exposure limits for human skin. It has been tested in preclinical studies on several dozen patients. As soon as the database is large enough, AI software will be trained to diagnose skin changes in real time.

Clinical trials and certification planned

The next step will be clinical studies lasting several years. The aim is to certify the system as a medical device so that it can be used in practice in the foreseeable future. "If detected at an early stage, more than 90 percent of those affected survive the first five years after the start of treatment," says the Cluster of Excellence "PhoenixD: Photonics, Optics, Engineering - Innovation across Disciplines" at LUH, which is researching the digital optics of the future.

AI systems already widespread

Skin scan apps can already help with the early detection of skin cancer. Among other things, pattern recognition algorithms are used to analyze photos of skin spots and warn the user in case of suspicion. Powerful AI systems are trained with thousands of images of different types of skin cancer and harmless moles.

(mack)