About the project
This project aims to build an AI-driven system that analyses live-cell microscopy videos showing how immune cells attack cancer cells. The videos are generated in a biology lab where each experiment can be precisely controlled. You will create machine-learning and computer-vision algorithms that can detect, track, and model these cell-to-cell interactions, revealing patterns that explain when and why immune cells succeed or fail.
Antibody-based treatments have transformed how certain cancers are treated, offering highly targeted ways to attack malignant cells. Yet, even with the same type of therapy, patient responses vary widely. In some cases, immune cells effectively eliminate the cancer cells, while in others, the same treatment has little effect. This inconsistency points to complex, dynamic interactions between immune cells and cancer cells that remain poorly understood. Traditional experimental techniques capture only snapshots of these interactions, missing the rich temporal and behavioural patterns that unfold over time.
This project will use advanced artificial intelligence to help uncover what drives these differences. You will work with time-lapse microscopy videos showing immune cells interacting with cancer cells, developing computational methods to analyse, model, and explain their behaviour. The project combines modern machine learning, video analysis, and explainable AI to identify subtle cues and temporal dynamics that influence treatment outcomes.
You will design algorithms that can detect meaningful behaviours, learn from complex visual data, and provide insights that go beyond what human observation can achieve. The work will be carried out in close collaboration with experimental biologists who generate the data, but the project will primarily focus on the computational side developing, testing, and refining AI models to interpret biological processes.
