Living lab for the transfer of digital health applications and AI into healthcare
Artificial intelligence for skin tumor diagnostics
Visual medical inspection is one of the most important pillars in today’s skin cancer diagnostics: To make a diagnosis, medical professionals examine skin lesions visually. Current artificial intelligence (AI) systems achieve the accuracy of a specialized dermatologist. However, there is still room for improvement. In contrast to already existing approaches, the Intelligente Diagnostik (Intelligent Diagnostics) joint project aims to significantly increase the amount of medical data collected in each measurement to improve diagnostic quality. Intelligente Diagnostik is using quantitative multi- and hyperspectral imaging to determine physical and chemical properties of potential skin changes. This measurement method operates optically and therefore non-invasive. Based on these extensive data sets, new AI models will then be developed and trained. This will allow medical digital diagnostics to be used even more accurately, resulting in improved early detection of skin cancer and thus better medical prevention for the general population, thereby giving the medium-sized healthcare industry in Baden-Württemberg an innovative edge.
In addition to leading the consortium, the FZI is responsible for the development of the data and model management platform, which manages the newly created data sets and controls the training of the AI models.
Vice Division Manager
Division: Software Engineering
Applied Artificial Intelligence
In this research focus, the FZI prioritizes the topics of Artificial Intelligence (AI) as well as human and AI engineering. In addition, the FZI deals with questions on dedicated AI hardware and predictive AI.
Large data sets combined with artificial intelligence can offer significant added value in digital medical diagnostics. Artificial intelligence is already being used in clinical settings today. Examples include the classification of tens of thousands of digital photographs to provide diagnosis based on a new image. In order to improve the quality of diagnosis, we need even more data of skin lesions to train AI models. This means: As part of the funding, we will specifically build up data sets. We will also develop a platform to provide data and model management capabilities. Quantitative imaging techniques will be used to collect data, which will significantly increase in the amount of data per image. Our goal is to take digital medical diagnostics of skin cancer to a new level through quantitative imaging techniques and improved AI models.
A strong alliance for progress
Five institutes of the Innovationsallianz Baden-Württemberg (innBW) are involved in the project.
The innBW is a unique alliance in Germany of 13 non-university application-oriented research institutes supported by the Ministry of Economic Affairs, Labour and Tourism.