The Research Unit at DESY focuses on the early stages of the imaging pipeline, developing  methods for advanced image reconstruction, including the optimization of measurements and the combination of classical methods with data-driven approaches. Our goal is to drag out a maximal amount of (quantitative) information from given or designed measurements. 

The Research Unit at MDC focuses on integrating heterogeneous imaging data across modalities, scales, and time. We develop concepts and algorithms for generic processing, stitching, fusion, and visualization of large, high-dimensional datasets. Our aim is to enable seamless analysis of complex imaging data without restrictions on the underlying modalities.

The Research Unit at DKFZ focuses on the downstream stages of the imaging pipeline, developing robust methods for automated image analysis and emphasizing rigorous validation. Our goal is to enable trustworthy and generalizable AI across scientific imaging domains. 

Inverse Problems: we focus on problems at the beginning of the imaging pipeline related to data acquisition and image formation. This is a crucial step in the pipeline, since information lost or not measured at this stage can hardly be recovered in later stages.

In the data acquisition, the change of some emitted signal interacting with the sample is measured. In order to produce meaningful images, an image reconstruction step is necessary. To achieve this task a mathematical model of the forward process leading to data generation has to be developed and then inverted to compute the image of interest. This process is called an inverse problem, whose stable solution relies on further a-priori knowledge or even training data about the type of images to be reconstructed. The development of algorithms for image reconstruction with highest quality is a major research topic for us.

The quality of the reconstructions is controlled via uncertainty quantification methods and feedback to the data acquisition is given by optimal experimental design techniques. Further important research questions investigated concern the compression of large scale data as well as problems at the edge to further steps in the pipeline such as denoising and registration.

Integrative imaging data sciences: we focus on integrating heterogeneous imaging data across modalities, scales, and time. They develop concepts and algorithms for generic processing, stitching, fusion, and visualization of large, high-dimensional datasets. 

The amount of image data, algorithms and visualization solutions is growing vastly. This results in the urgent demand for integration across multiple modalities and scales in space and time. We develop and provide HI solutions that can handle the very heterogeneous image data from the research areas of the Helmholtz Association without imposing restrictions on the respective image modalities. To lay the groundwork for the implementation of HI solutions, our team at MDC will focus on the following research topics:

• Develop concepts and algorithms for handling and generic processing of high-dimensional datasets.
• Develop algorithms for large, high-dimensional image data stitching, fusion and visualization

Our “Integrative Imaging Data Science” Group specializes on new concepts, mathematical approaches and representations for large scale image data. Our work includes the development of algorithms, computational resources and visualization solutions across scales in space, time and modalities. We bundle expertise in:

• concepts and algorithms for handling, generic processing and representation of high-dimensional datasets
• image analysis and visualization across scales
• frameworks for large data management, image analysis and data abstraction

Image Analysis & Validation (DKFZ): we focus on the downstream stages of the imaging pipeline – specifically, the development, validation, and deployment of advanced AI-based methods for automated image analysis. These phases are critical for extracting high-level, domain-relevant information from complex imaging data. Our work addresses algorithmic challenges in interpreting, quantifying, and validating image-derived information, with the overarching goal of enabling trustworthy, robust, and generalizable AI-driven analysis. 

We conduct interdisciplinary research in three core areas:

(1) automated image analysis, where we work to enhance the robustness and generalizability of AI methods across diverse and imperfect real-world datasets;

(2) human-machine interaction, which aims to integrate humans as active participants in AI development and deployment to ensure transparency, trust, and safety; and

(3) validation and benchmarking, where we lead efforts to develop and standardize validation practices and develop tools that enable reproducible and transparent assessments of algorithm performance.

By addressing these pivotal stages in the imaging pipeline, our mission is to advance the reliability and impact of AI-based image analysis across scientific and societal applications.