The AiCOLO project associates computer scientists, specialized in data and image analysis, with pathologists and biostatisticians, specialized in digital pathology and colon cancer. This project focuses on two methodological aspects of WSI imaging, at the convergence of image analysis and data-mining, with specific objectives:
Features extraction and coarse image classification for determining basic regions of interest and biological objects (e.g. tumor, stromal structures, vessels, muscle, immune infiltrate, normal tissue) using deep learning frameworks that have proven their efficiency in such problems with the aim to use these data to define patients prognosis. The task will be to develop a new tool to assist pathologists by giving an automatic estimation of the prognosis using digitized WSI.
Use of deep neural network to predict colon cancer genetic features and describe the tissue features built by the neural network to make this prediction.
More details here…