Phase Optics Microscopy Artificial Intelligence for Fertility

The developmental potential of the oocyte depends largely on the quality of its cytoplasm. Female fertility is at threat in our societies, as women tend to postpone childbearing. Millions of couples around the world use assisted reproduction (ART) to conceive. Subjective morphological criteria are common in clinics to select the "best oocytes". Although fluorescent probes would be ideal for this task, they are not compatible with clinical use due to their photo-toxicity. In this project, we will combine 1) an AI-based approach to learn and automatically reconstruct oocyte structures as displayed by fluorescent probes but only from non-invasive transmitted light images and 2) non-invasive phase microscopy to measure oocyte dry mass. Our project will allow both qualitative and quantitative phenotyping of the most promising oocytes, useful for both basic research and for clinical use.