Research

Theseus Medico

My research is funded by the Federal Ministry of Economics and Technology (BMWi) within the Theseus Medico project. The task of Medico is to explore new technologies for providing computational knowledge-based assistance in clinical situations. Medico's Use Case is the efficient handling of large patient databases at the example of lymph node cases, which usually involve several series of radiological images. The three principal challenges are:

• The structuring of various types of medical data with respect to their semantic relations.
• The development of methods for the automatic derivation of meta information. This involves:
  • Text mining and classification
  • Image segmentation and classification
  • Rule learning for inference in medical ontologies
• Efficient similarity search on medical images

Areas of research

I am currently active in the fields

• Similarity Search in Spatial and Multimedia Databases
  • Similarity measures for multi-instance and multi-represented data objects
  • Spatial indexing
  • Dimension reduction
• Image Search
  • Representation of medical image data
  • Efficient content-based image retrieval

Furthermore, I am also interested in graphmining and graph learning approaches.