Excellence Strategy

Department of Nuclear medicine and Clinical Imaging

The research focus of the Department of Nuclear Medicine and Clinical Molecular Imaging is on theranostics, which is made possible by translational research, an interdisciplinary exchange of chemists, physicists, biologists and physicians, as well as the most advanced imaging technologies and methodology.

 

Current projects

Axial field of view of conventional whole-body (WB) PET scanners. The Biograph Vision Quadra total-body (TB) PET scanner enables simultaneous imaging with just a single bed position for the first time.
Total-Body PET

The workgroup led by Dr. Fabian Schmidt is doing pioneering work in the field of total-body (TB) positron emission tomography (PET) with the first TB-PET/CT scanner at a university hospital in Germany. The extended axial coverage of this new technology increases sensitivity by a factor of 20 compared to standard whole-body scanners. This enables enormous potential for dose reduction, ultra-late imaging and holistic assessment of whole-body interactions. The focus of this working group is to exploit the full potential of TB-PET imaging for better patient care by developing various innovative methods based on patient data, phantom experiments and Monte Carlo simulations.

 

Translational Immune Imaging and Theranostic

Complex molecular and cellular processes determine the success or failure of cancer therapies. Innovative molecular imaging can help to understand the biological processes in the tumor and its environment and significantly improve the control of cancer therapies. TH workgroup of PD Dr. Johannes Schwenck therefore develop new tracers for PET imaging, evaluate them in preclinical experiments and transfer the successful concepts into the first prospective clinical studies.

Their focus is on the non-invasive examination of immune cells that are directed against the tumor, the characterization of therapeutically usable surface markers for e.g. radiopeptide or antibody therapies (theranostics), and on research into cell stress reactions such as tumor senescence, which have a decisive influence on the success of many cancer therapies.

The case study presents a PET scan, cluster segmentations, and immunohistochemistry (IHC) for all examined tumor groups. A significant reduction in tumor volume was observed in the combined therapy group compared to all other groups. Additionally, the signal intensity and pattern of tracer uptake are visually discernible. The middle column shows the regions of interest (ROIs) generated by Gaussian Mixture Model (GMM), which are automatically adjusted by the algorithm to the tumor volume. The IHC in the right column illustrates CD206 expression patterns, showing increased uptake at the edges of the tumor

Machine Learning based analysis of Tumor-Microenvironment

The tumor microenvironment is a complex system that can be non-invasively studied in vivo using PET/MRI. In recent years, antibody fragments and small molecules have been developed to analyze the immune tumor microenvironment. Dr. Salvador Castaneda's research group focuses on implementing computer-assisted image analysis to evaluate the distribution patterns of these imaging tracers within the tumor.

 

Imaging of the Tumormicroenvironment

One of our current focuses is on the discovery of imaging biomarkers within the tumor microenvironment using PET/MRI. This involves evaluating clinical imaging signatures of the tumor microenvironment and translating them back to the animal tumor models. Identification of these key imaging signatures can help define tumor regions that are resistant to clinical treatment. By leveraging preclinical imaging and applying machine learning to translate data from the macro to the microscale across species, our research aims to uncover critical aspects found in the tumor microenvironment of patients that can be explained cellularly and molecularly through animal models, while aiming to improve patient survival rates and expand treatment options in the clinic.

 

 

 

In the above example, glioma regions were clustered using 5 clusters following Akaike’s Information Criterion. There were statistically different predominant clusters in low grade gliomas (LGG, Cluster 4) and high-grade gliomas (HGG, Cluster 5). This difference can be observed in two exemplary subjects in panel A and in all subjects in panel B. Moreover, Cluster 3 was commonly found among both grade classes. This simple approach demonstrates the diverse bioimaging signatures that can be found in different grades of glioma that can disclosed in multiparametric PET/MRI using solely signal intensity. The “late” [18F]FET-PET shown above is a frame of the last 20 min. at the end of the 40 min. dynamic scan. The MRI is a T2-weighted fluid attenuated inverse recovery acquisition.