Preimplantation genetic diagnosis (PGD) in IVF requires precise handling of embryos and micropipettes - a process traditionally done manually, making it slow, operator-dependent, and prone to error. This project delivered a computer vision workflow to automatically detect and track blastomeres and micropipettes during micromanipulation, reducing reliance on manual skill.

The system combined image segmentation and robust tracking algorithms to maintain accuracy under real-world challenges like variable lighting and partial occlusions. It achieved reliable detection and uninterrupted tracking across multiple frames, validating feasibility for real-time operation. These results indicate strong potential to cut preparation time, reduce micropipette misalignment risk, and improve consistency in delicate biopsy steps.

With further developments, this technology could streamline IVF workflows, lower error rates, and reduce costs for fertility clinics, while improving patient outcomes. It represents an early step toward next generation assisted reproductive systems, where automation enhances precision and accessibility in reproductive healthcare.