PhoenixD member Prof. Dr. Bernhard Roth and his team are part of the iToBoS project, which the European Union funds with twelve million euros.
Three scientists from Leibniz University Hannover (LUH) are part of the new international project iToBoS, which aims to reduce the mortality rate of skin cancer by improving diagnostics from the earliest stage on. The project partners are working on linking imaging technologies with all relevant individual patient information for the first time. The EU is funding the interdisciplinary project with twelve million euros over the next four years within the Horizon 2020 research and innovation program.
The name explains the overall goal: the acronym iToBoS stands for "Intelligent Total Body Scanner for Early Detection of Melanoma". And this is precisely the kind of artificial intelligence-based skin cancer scanner the iToBoS team is striving to develop. The idea of analyzing the skin with optical technologies such as imaging, optical coherence tomography or laser analytics to detect skin lesions is widely investigated today. In practice, however, only visual diagnostics based on dermatoscopic devices placed directly on the suspicious skin areas are established so far. Moreover, these systems' contact mode of operation can falsify diagnoses and make it challenging to detect changes of skin lesions over time, particularly when patients have dozens of them on their bodies. The iToBoS researchers now aim to increase diagnostic precision with the help of non-contact imaging, which captures the entire skin surface simultaneously, in combination with individual patient data.
Linking imaging and patient data
The iToBoS full-body scanner will be equipped with optical imaging and artificial intelligence (AI) and incorporates information from all relevant sources, i.e. dermatoscopic images and patient records, into the diagnosis. For this purpose, the new system will also include specific characteristics such as age, gender, previous illnesses as well as the location and size of the skin malformation into account. The scanner independently evaluates the entire recorded skin surface, determines the suspicious skin areas, and combines and analyses all data. In addition, the iToBoS partners take care of the handling and protection of sensitive patient data because this is of paramount importance for the approval and acceptance of the new system on the market.
The AI in the planned skin cancer scanner will not only diagnose skin cancer but will also transparently explain which information was used to assess the clinical diagnosis for the first time. This approach will then allow a more comprehensive diagnosis tailored to the individual patient. Furthermore, through regular examinations, the system will reliably detect and document changes in the skin lesions over time. This functionality could enable the detection of skin cancer and potentially other skin diseases much earlier in the future.
The earlier melanoma skin cancer is detected, the higher the chances for curing: if detected at an early stage, more than 90 per cent of the patient survive the first five years after treatment. If the cancer is diagnosed at a later phase when it has usually formed metastases and spread, the survival rate decreases significantly to only about a quarter. With more than 144,000 new cases in 2018 in Europe alone, melanoma skin cancer is among the most prevalent cancer types. A more reliable diagnostic system can significantly reduce the number of skin cancer deaths, relieve the burden on healthcare systems, and thus unfold high socio-economic impact.
LUH focuses on illumination and imaging system
To develop the new skin cancer scanner, a total of 19 partner organisations from Europe, Asia and Australia are cooperating, including five research and university institutions, ten companies, three hospitals and one patient organisation. The coordination of iToBoS lies in the hands of Prof. Dr. Rafael Garcia from the University of Girona (Spain).
"In Hanover, we are focusing on the development of the illumination and imaging system, which uses tunable optics for imaging in particular," says Prof. Dr. Bernhard Roth, head of the Hanover Centre for Optical Technologies (HOT) at LUH. Roth has performed research on the optical detection of skin diseases for many years. This experience is now most valuable for the new project. He is supported by two scientists at the HOT, Lennart Jütte and Gaurav Sharma, who are both working on their PhD thesis in medical optics.
Roth also heads the Precision Metrology working group in the Cluster of Excellence "PhoenixD: Photonics, Optics, Engineering - Innovation across Disciplines". The Cluster's research focuses on digital optics of the future. Systems such as the one developed on iToBoS might one day be realized using additive manufacturing, i.e. 3D printing, as developed by the Cluster. As optical technologies are one of the key technologies of our time, they impact numerous applications from industry and agriculture to life sciences and provide new solutions which were not available before.
"The biggest challenge of the project is to reconcile imaging, mechanical design and artificial intelligence," says Lennart Jütte. "The international team has a very hands-on approach to problem-solving. We are happy to work in such an environment. This means that the results can be used worldwide very quickly," adds his colleague Gaurav Sharma.
Prof. Dr. Bernhard Roth (centre) is standing at an optical table with Lennart Jütte (left) and Gaurav Sharma (right) in the HOT test lab. Using white-light LED sources, the scientists record high-resolution 3D images of the skin lesions without contact. The photos are then automatically classified using artificial intelligence algorithms. The compact system consists exclusively of low-cost standard components. A test will run in a clinic is planned for the near future.
The scientists are available to answer questions. In case of interest, please contact:
Published by Sonja Smalian at Leibniz University Hannover.