Automated analysis of cancer histopathology images
We explore image encoding and classification models to perform predictions on various types of tumor histopathology images. This data naturally fits the framework of “Multiple Instance Learning”. The questions we are interested in is tumor detection, tumor type classification, as well as prediction of therapy response.
Beyond mere classification, we are interested in generating insights about the visual information that is key for a given classification task. In a first step, simple conclusions can be drawn from analysing saliency maps, for example attention heatmaps in transfomer-based classifiers. We extend the analysis of the detected regions of interest by quantifying the cell populations and their morphological features and spatial distribution.
Date:
09/11/2022
