Overview

In vitro fertilization (IVF) outcomes depend on many factors, and clinicians currently lack reliable tools to predict which embryo transfers will succeed. This project builds machine learning models to estimate pregnancy success from preimplantation genetic testing (PGT) data.

Data

We cleaned and analyzed approximately 8,600 embryo records, linking PGT results with transfer outcomes to build a comprehensive training dataset.

Models and Delivery

The project produced predictive models and a Flask web application that allows clinicians to input embryo data and receive estimated transfer success probabilities. The tool is designed to support — not replace — clinical decision-making.

Mentorship

As part of this project, I mentored a student through a Python crash course, designing a coding curriculum around the research so he could write the initial analysis before we scaled to larger models.

Interactive Tools

Try the embedded IVF calculators below.

Single Embryo Predictor

Cumulative Calculator