Scenario

An IVF laboratory is developing a decision-support tool to help embryologists choose which embryo to transfer for a current patient.

For each patient, the embryologist typically reviews a small cohort of embryos from the same IVF cycle, often around 5–9 fertilized eggs or embryos. The final decision is made within this patient-specific cohort: which embryo, or shortlist of embryos, should be prioritized for transfer?

The clinic also has a larger historical dataset of embryos from previous patients. This dataset includes embryo images, morphology attributes, patient and cycle context, hormone measurements, and whether each embryo was associated with pregnancy. The embryologist wants to use this historical data to make a more data-informed decision, while still relying on expert judgment for the final selection.

This creates two layers of information:

Layer	Role in the Scenario
Current patient embryo cohort	The actual options being considered for transfer.
Historical clinic dataset	Contextual evidence used to understand patterns among similar embryos from previous cases.

The embryo selection process must consider multiple forms of evidence, including morphology, developmental timing, PGT results, patient and hormone context, lab context, historical outcome patterns, similarity to prior cases, and uncertainty or conflicting evidence.