Immediately after leaving academia I began to work at Nucleai, a cancer research company that utilizes deep learning to gain insights into biopsy images. During my initial months, I worked on developing a deep learning-based image analysis pipeline for analyzing multiplex immunofluorescence microscopy images. Later on, I became the main developer in charge of spatial feature calculation within our software team. spatial features are quantitative values derived from analyzing the results of our deep-learning models. These can be fairly basic values like ‘density of tumor cells’ to more complex features such as measures of cell clustering.
The position I hold involves both the development of the software branch in charge of this feature calculation as well as the research and integration of new features we (or our clients) wish to calculate.