Research at CIR is roughly grouped around 3 research lines. The borders are very transparent, and usually people collaborate across the topics.
Machine Learning & Neuroimaging
The interface between machine learning and neuroimaging is an exciting and rather controversial research field. We are working on novel methods that allow to capture and understand cognitive processes from neuroimaging data. The current research focuses manifold learning, functional atlas building, and multi-variate pattern analysis in fMRI data.
Computer Aided Diagnosis and Quantification
Computer aided diagnosis is the most clinically oriented part of our research. Projects include the automatic quantification of rheumatoid arthritis, computer based lung disease diagnosis the development of new ways to assess osteoporosis.
Computer Vision and Pattern Recognition
The basis of a large part of our work is rooted in the computer vision and pattern recognition community. Learning models from data, applying them during analysis, and devising methods to perform increasingly un-supervised learning from medical image data are at the center of this line. Projects include the development of a medical image search engine and computational anatomy. Although not tightly connected to purely clinical motivation, it serves as a fruitful experiment bed for new ideas.