Ikonisys, a company specializing in the early and accurate detection of cancer, and the Department of Electronics, Information and Bioengineering of the Politecnico di Milano signed a research contract to initiate a project to bear on the challenging problem of identifying specific cells in complex tissues through designing, training and deployment of a deep learning model able to detect them.

Prof. Giacomo Boracchi will lead the research bringing his expertise in image processing and machine learning. In particular, the approach will aim to recognize clinically relevant cells of specific types in darkfield images of lung tissue acquired from the Ikoniscope20. Segmentation of single cell nuclei is a frequent challenge of microscopy image processing, often being the first step of many quantitative data analysis pipelines.

The ongoing collaboration will allow Ikonisys to further enhance its advanced image analysis capabilities and to accelerate the development of the Ikoniscope AI project with the implementation of deep learning models and novel state-of-art techniques into the instrument’s workflow, leading to the development of new applications or enhancement of the existing ones, especially in the most challenging fields such as circulating tumor cells.