Health Informatics Group

Computational Histology

Analysis of histology images of the human tissue biopsies remains the most reliable way of diagnosing and grading cancer. The current gold-standard for quantification and diagnostics based on prepared slides of human tissue is often a human expert looking through a microscope at the slides.

Histology of medulloepithelioma. Image: Wiki

Histology of medulloepithelioma. Image: Wiki

Our aim is to enhance and extend this expensive and time-consuming process by performing automated analysis of high-resolution images of tissue. This investigation involves designing a combination of algorithms, scanners, stains and biomarkers that allow fast and accurate identification of aspects of clinical interest.

Recent Publications

International Conference on Bioinformatics and Computational Biology (BICOB), Honolulu, United States, 18/03/19.