Affiliations: [a] Department of Computer Science, University of Cyprus, P.O. Box 20537, CY-1678 Nicosia, Cyprus. E-mail: csfranks@ucy.ac.cy, pattichi@ucy.ac.cy, schizas@ucy.ac.cy | [b] Department of Electron Microscopy, The Cyprus Institute of Neurology and Genetics, P.O. Box 23462, Nicosia, Cyprus
Correspondence:
[*]
Corresponding author: Constantinos S. Pattichis, Department of Computer Science, University of Cyprus, P.O. Box 20537, CY-1678 Nicosia, Cyprus. Tel.: +357 2 892244; Fax: +357 2 339062; E-mail: pattichi@ucy.ac.cy.
Abstract: The Biopsy Analysis Support System (BASS), previously used for image analysis of immunohistochemically stained sections of breast carcinoma, has been extended to include indexing and content-based retrieval of biopsy slide images from a database of 57 captured cases. Images from histopathological biopsy slides are described and these are accessed in terms of the properties of either individual nuclei or groups of cell nuclei present in the slide. Visual similarity of cases is specified in terms of a diagnostic index, commonly known as the H-score, which incorporates the heterogeneity of nuclear staining intensity, as well as the percentage of nuclei staining at specific intensities. The system provides a platform that can be exploited in telepathology and teleconsultation, but further research is needed to explore its full potential and accuracy in a diagnostic clinical environment.
Keywords: breast cancer biopsy, content-based retrieval of biopsy slide images, digital medical libraries, optical imaging
