Biomedical Computer Vision (BMCV)

Prof. Dr. Karl Rohr

We develop advanced image analysis methods for computer-based quantification of biomedical images with focus on deep learning methods for cell microscopy image data. A wide spectrum of novel methods has been introduced and successfully employed in different biomedical applications. The Biomedical Computer Vision Group achieved top-ranking results in international competitions on particle tracking and cell tracking. Our work is important to bridge the gap between experimental image data and quantitative modeling.

Research Strategy

The research group Biomedical Computer Vision (BMCV) develops methods and algorithms for computer-based analysis of biological and medical images. A main aim is to derive quantitative information about cellular and subcellular structures from microscopy image data. We have developed a wide spectrum of novel advanced image analysis methods for cell segmentation and tracking, particle detection and tracking, non-rigid image registration, vessel segmentation, and landmark localization comprising deep learning methods, model-based methods, and probabilistic methods. Top-ranking results in international competitions on particle and cell tracking were achieved. The developed methods have been successfully used in different applications to study virus infection, cell migration and division, and the genome architecture in cooperation with biomedical partners, particularly with research groups at BioQuant. Our work is important to bridge the gap between experimental image data and quantitative modeling.