Techn. Fakultät Willkommen am Institut für Informatik FAU-Logo

Medical Image Reconstruction

The medical image reconstruction group deals with the generation of slice images. These images are created from lower dimensional input data. For the case of CT, the observed projections are either 1-D or 2-D signals that are used to generate 2-D or 3-D slice data. In MR the signal is sampled in the Fourier space of the image and reconstructed using an inverse Fourier transform. Our work is closely related to other main fields in medical imaging, like image segmentation and registration as reconstruction quality heavily influences the subsequent processing.

CT Reconstruction


CONRAD is a software platform for simulation and reconstruction of flat-panel CT images.

Perfusion C-arm CT

Perfusion C-arm CT is a novel technology to measure capillary blood flow (CBF) with slowly rotating C-arm systems.

3-D Imaging of the heart chambers with C-arm CT

In this project, the focus of the 3-D reconstruction is on the left ventricle (LV). Regarding the long acquisition time of the C-arm system for the acquisition of the projection images (> 5s), the heart motion has to be considered.

3-D Imaging of coronary vasculature using C-arm CT

The focus of this project is the optimisation of the 3-D reconstruction of coronary vasculature towards a quantitative representation.

Region of interest reconstruction in C-arm CT

3D region of interest imaging reduces the radiation dose to the patients without compromising image quailty.

Multi-material beam hardening correction in CT

The aim of the project is to effectively reduce beam hardening artifacts in datasets consisting of multiple materials.

Spatial-temporal Total Variation Regularization (STTVR) for 4D-CT Reconstruction

4D-CT reconstruction based on compressed sensing.

Robust 2D/3D-Registration for Real-Time Patient Motion Compensation

Towards robust and real-time patient motion compensation for 2D/3D overlay applications

MR Reconstruction

Iterative Reconstruction for Cardiac Magnetic Resonance Imaging

The focus of this project is the accelleration of the data acquisition of coronary imaging with the use of iterative reconstruction and respiratory motion compensation.

Iterative Reconstruction Techniques for Radial Perfusion MRI

Radial k-space sampling trajectories are a promising alternative to Cartesian sampling, in particular for dynamic contrast-enhanced MRI.

Iterative Reconstruction Techniques for MR angiography and flow

Non-contrast enhanced MRA and flow acquisition are promising applications for iterative reconstruction.

Molecular Imaging

Attenuation Correction for Hybrid Imaging Systems

Attenuation maps for correcting PET data needs to be derived from MR information in current PET/MR hybrid systems

GPU based quantitative reconstruction in SPECT/CT

Graphical Processing Units (GPU)s with their tremendous parallelity concept and computational power are being used in this project to develop novel quantitative reconstruction algorithms.