Automated medical image quality assessment

Assessment of image quality is a prerequisite for further processing in most clinical situations and scientific analysis. In certain situations, when image quality has to be assessed in a short time interval or when data are analyzed by automated methods, image quality assessment has to be performed automatically. To this end, we develop machine learning based methods for identification, localization and quantification of image artifacts in medical images, specifically in whole body MRI.

Medical image translation with Generative Adversarial Networks

The task of translating medical images between different domains has numerous useful applications. One example is the correction and restoration of artifact-corrupted images. Another potential application is the generation of novel image data. We are working on the development and refinement of medical image translation frameworks and their applications. Specifically, we are currently adressing GAN-based motion correction, GAN-based PET attenuation correction and GAN-based image inpainting.

lmage Segmentation

In the context of large epidemiological studies, manual image analysis is often not feasible due to the overwhelming amount of data. Thus, automated analyses of whole body data is necessary. In order to perform automated analyses, a prerequisite is often the segmentation of target tissues and organs. A major aspect of our work is thus the adaptation and implementation of automated whole body segmentation in the context of large clinical and epidemiological studies using deep-learning approaches.