Project description
Despite advances in cancer therapy, there is still no reliable way to predict which treatment will work best for each patient, leading to trial-and-error approaches. To address this, we developed zAvatars—zebrafish-based patient-derived xenografts that enable real-time testing of multiple therapies at single-cell resolution, achieving ~90% predictive accuracy in clinical studies.
However, the analysis of the large imaging datasets generated is currently manual, limiting scalability and consistency. We propose to develop an AI-powered pipeline to automate tumor quantification and apoptotic cell detection using advanced deep learning, enabling fast, reproducible analysis and accelerating the clinical adoption of this precision oncology platform.