AI-4 Spatial Multiomics Transfer (KI4Spatial Data)

Focus:	Life Science Technologies Transfer
Type of funding:	Project funding programmes
Funded institution:	Mannheim University of Applied Sciences

“AI-4-Spatial Multiomics Transfer” (KI4Spatial Data) is developing new AI methods to better understand tissue samples from patients with multiple sclerosis and myositis. The goal is to create personalized and resource-optimized diagnostic processes and therapies.

Goals

Today, tissue samples can be analyzed using innovative technologies, such as spatial gene analysis, metabolic measurements, and 3D microscopy. Each method provides different insights, but so far there have been no tools available to meaningfully integrate all of this data.

The KI4Spatial Data project is developing AI systems that integrate different data sources to identify spatial patterns in tissue. It is important to determine exactly where specific molecules are located within the tissue, as their spatial distribution can provide insights into disease processes. A particularly innovative aspect of the approach is training the AI using complex analyses and then applying this knowledge to a fast, cost-effective imaging technique - quantum cascade laser infrared imaging (QCL-IR). Based on these results, spatially resolved diagnostic processes in clinical settings can be accelerated and optimized in terms of resource use in the future, while also enabling personalized therapeutic interventions.

The project is a collaboration between the CeMOS Research and Transfer Center at Mannheim University of Applied Sciences, Mannheim University Medical Center, and the medical faculties at Heidelberg University.