Related Experiment Video
Updated: Aug 28, 2026

Establishment of an Embryo Implantation Model In Vitro
Published on: June 21, 2024
Development and Internal Evaluation of an Integrated Patient-Embryo-Endometrial Scoring System for Predicting
Egehan Bilen1,2, Sultan Seda Doğan1, Müge Kovalı Sezer1
1Division of Reproductive Endocrinology and Infertility (IVF Center), Department of Obstetrics and Gynecology, Dokuz Eylül University School of Medicine, Balçova, 35340 Izmir, Turkey.
Abstract:
Background/Objectives: Implantation in in vitro fertilization (IVF) depends jointly on patient prognosis, embryo quality and endometrial receptivity, yet these are usually assessed in isolation. We developed and internally evaluated, as a proof of concept, an index combining all three. Methods: In this single-center prospective pilot study, 46 women undergoing embryo transfer were scored on the day of transfer for patient prognosis (SART model), embryo morphology and endometrial receptivity (Endo-R ultrasound score), each on a 0-10 scale and combined as (SART/10) × (Embryo + Endo-R)/2. The outcome was biochemical pregnancy (10 events). Discrimination (area under the curve, AUC), internal validation and incremental value beyond age were assessed by bootstrap resampling, cross-validation and likelihood-ratio tests. Results: The Integrated Score achieved the highest discrimination of the indices examined (AUC 0.86, 95% CI 0.71-0.98; cross-validated 0.80) but did not significantly outperform age alone (0.79; p = 0.33), and its embryo component contributed none (0.51; χ2 = 0.42, p = 0.52 when added to age). Only the difference from the Embryo Score survived Bonferroni correction. The Endo-R Score added information beyond both age and the SART Score (both p = 0.003). Conclusions: The index did not add measurable discrimination beyond patient age in this sample; the endometrial component carried independent information and is the domain worth pursuing. The prognosis model was applied outside its validated endpoint and population, so it ranks rather than calibrates risk. With 10 events these estimates are exploratory and the Youden-derived cut-off (≥4.2) is optimistic. The study contributes a specified rubric and open calculator for the multicenter validation that must follow.

