Search papers, labs, and topics across Lattice.
This study introduces a context-aware hierarchical Bayesian model that leverages high-resolution environmental data from IVF laboratories to improve pregnancy rate predictions. By engineering 55 temporal features that capture incubator dynamics, the model significantly reduces prediction error to 1.27%, outperforming traditional raw averages. The hierarchical approach also enables effective sharing of environmental effects across clinics, achieving a notable 64% error reduction in predictions for a specific age group in a Northern European clinic.
Context-aware features can slash IVF prediction errors by over 60%, revealing critical insights hidden in environmental data.
IVF pregnancy rates are routinely modeled using patient-level variables, while high-resolution laboratory environmental data remain underutilized. We show that this is a missed opportunity. Rather than relying on raw sensor averages, we engineer 55 context-aware temporal features, including rolling thermal stability, simultaneous temperature-humidity adherence, peak stress duration, and post-stress recovery speed, that capture the dynamics of incubator microenvironments. On 61 weeks of data from an Asian IVF clinic, these features reduce cross-validated prediction error to 1.27%, compared to 3-5% for raw averages. We then train a hierarchical Bayesian Beta regression model that shares environmental effects across an Asian and a Northern European clinic via partial pooling, while preserving site-specific baselines. On held-out data from the Northern European clinic, the model achieves R2 = 0.86 and a 64% error reduction for the 35-39 age group over a naive baseline, demonstrating that structured environmental monitoring contains clinically meaningful, transferable signal.
