ISSI Institute of Control & Computation Engineering


The research in the area of medical imaging is focused on automatic analysis of digital images of cytological microscope material obtained with non-aspirant fine needle biopsy of breast cancer. The main goal of the research team is the design of fully automatic sifting diagnosis systems, which assists the expert deciding whether the digital image considered represents benign or malicious breast tumour. The research is focused on all steps of the automatic diagnosis procedure - image filtering, segmentation, morphological indices exploration and classification. Multifarious approaches are implemented based, e.g., on the spherical and elliptic Hough transform, the Grow-Cut method, evolutionary algorithms, decision trees, the sonar method, watershed segmentation, colour homograms, cellular automation, fuzzy clustering and type-2 fuzzy sets.