Heidelberg University
Medical Faculty Mannheim
University Hospital Mannheim
CKM receives DFG funding for Project "Prdiktion von Therapieansprechen und Outcome beim lokal fortgeschrittenen Rektum-Karzinom mittels Radiomics und Deep Learning: eine beispielhafte Anwendung fr eine allgemein verwendbare, Deep Learning basierte Prozessierungs-Pipeline fr die Bild-Klassifikation" read more.
Alena-Kathrin Schnurr presented on our research on AI in medical imaging @ Research Plus forum
read more
Dr. Mathias Davids wins the I. I. Rabi Award of the International Society for Magnetic Resonance in Medicine (ISMRM),
read more, article in local newspaper (Mannheimer Morgen)
Collaborate Research Projects

Image Guided Interventions

Group Leader: Alena-Kathrin Schnurr

Research Profile

Segmentation of Abdominal Organs in Medical Images

We investigate AI-based algorithms for fast and robust segmentation of anatomical structures in CT and MRI. Our main focus is the optimization of the preprocessing pipeline and the application to abdominal images.

Persons involved: Alena-Kathrin Schnurr, Christian Tönnes

Synthetic CT Images as Training Data for Neural Networks

There is often a lack of annotated medical image data to train deep learning algorithms. Our approach to obtaining training data is to synthesize realistic image data from the XCAT phantom via a CycleGAN network. The XCAT phantom allows the integration of anatomical variability in combination with pixel-precise annotation masks. The synthetic images can be used to train neural networks such as segmentation or registration networks.

Persons involved: Dominik Bauer, Tom Russ, Alena-Kathrin Schnurr

Registration of Abdominal Organs in Medical Images

We investigate (multimodal and interdimensional) volume registration techniques.

Persons involved: Gordian Kabelitz, Barbara Waldkirch (associated)

Alternative Trajectories for Cone Beam Computed Tomography

We investigate new trajectories for CBCT to improve imaging performance, reduce the associated dose, and avoid collisions with robots, humans, and stationary objects. For this purpose, we have access to a programmable interface of an Artis Zeego robot, which provides us with the unique capability to implement arbitrary source-detector trajectories on a clinical CBCT system.

Persons involved: Christian Tönnes, Tom Russ

Slices from the XCAT Phantom with the corresponding synthesized CT, CBTC and T1-weighted MRI images.
Gif of an alternative trajectory using a Artis Zeego system.

Group members

Student members

  • Hannah Braun
  • Eva Oelschlegel
  • Constantin Ulrich, B.Sc. 🏆 🏆
  • Jan Schlenker
  • I Sun


Recent Publications

  • T. Russ, S. Goerttler, A.-K. Schnurr, D. Bauer, S. Hatamikia, L. Schad, F. Zöllner and K. Chung; Synthesis of CT Images from Digital Body Phantoms Using CycleGANs, International Journal of Computer Assisted Radiology and Surgery, accepted for publication
  • Schnurr, K. Chung, T. Russ, L. Schad and F. Zöllner; Simulation-Based Deep Artifact Correction with Convolutional Neural Networks for Limited Angle Artifacts, Z Med Phys, 29 (2), pp.150-161 (2019)
  • K. Chung, L. Schad and F. Zöllner; Tomosynthesis implementation with adaptive online calibration on clinical C-arm systems, Int J Comput Assist Radiol Surg, 13 (10), pp.1481-1495 (2018)

List of Publications

Contact: Dr. Frank Zöllner last modified: 20.08.2020
to top of page