Research Interests

As a member of the Medical Image Processing group my work concerns Image Analysis of the Eye Background (Retina Images). From the medical point of view we try to detect and analyze certain changes on the eye fundus, such as

• Detection of glaucoma disease,
• Changes due to diabetes, 
• Age estimation.

My research work involves image processing, feature extraction, dimension reduction, classification, evaluation and visualization. In cooperation with the Augenklinik Erlangen different image modalities are acquired in clinical practice. One of our main goals is to extract useful information from color fundus fotos. The images are relatively easy and cheap to gain and the modality is one of the most common ones in ophthalmology.

In addition to that we use laser scans from the HRT Lasertomograph, which acquires reflectance images as well as topographic images of the eye background. Multimodal image classification is an important step towards more robust and more accurate risk indicators.

Current projects

We focus our efforts to automatically analyse and classify retina images by appearanced based algorithms. Methods to segment physiological structures like the optic nerve head or the vessel tree are also used. In particular, I am interested in multimodal, segmentation-free image classification applied in the following projects:

• Retina Image Similarity, Retina content-based image retrieval (Retina CBIR) and visualization
• Retina Image Quality Assessment 
• CATEYE framework - computer automated detection of eye diseases

International Max-Planck Research School on Optics and Imaging (IMPRS)

Currently I am receiving a scholarship from the IMPRS.