Beyond the Inner Cell Mass: Evaluating the Impact of Inner Cell Mass Visibility on Implantation Prediction in Static...

Embryo selection in IVF often relies on inner cell mass (ICM) morphology as a predictor of implantation. However, in static two-dimensional images, ICM visibility may be limited by technical factors like orientation or focus, rather than actual quality. Since most IVF labs still rely on static images for embryo evaluation, it is important to reconsider the emphasis placed on ICM and investigate other morphological features that may support clinical decision-making.

To assess whether the visibility of the inner cell mass (ICM) in static blastocyst images affects the accuracy of embryologists in predicting clinical pregnancy, defined by the presence of fetal cardiac activity.

A retrospective cohort study conducted at a single IVF center, analyzing 300 single vitrified warmed blastocyst transfer cycles between 2021 and 2024. Two senior embryologists independently evaluated static blastocyst images, blinded to clinical metadata. Clinical pregnancy, confirmed by fetal cardiac activity at 7 to 8 weeks, was used as the ground truth for measuring prediction accuracy.

This retrospective study analysed 300 single vitrified warmed blastocyst transfers performed at a single IVF centre between 2021 and 2024. Only Day 5 blastocysts that re-expanded within 2-3 hours post-warming were included. Static images captured just before transfer were evaluated independently by two senior embryologists, who were blinded to clinical outcomes and to each other's assessments. ICM visibility (Good/Poor) and binary implantation predictions were recorded. Clinical pregnancy was determined by the presence of fetal cardiac activity at 7-8 weeks of gestation.

Chi square tests were performed to compare prediction accuracy between good and poor ICM visibility groups for overall, positive and negative predictions. A P value less than 0.05 was considered statistically significant.

A total of 300 blastocysts were analysed to assess the impact of ICM visibility on the accuracy of implantation prediction by embryologists. While embryos with good ICM visibility comprised 63% of cases (n = 189), prediction accuracy was slightly higher for embryos with poor ICM visibility (44.14%) than for those with good visibility (39.68%), although the difference was not statistically significant. Chi square analysis showed no significant association between ICM visibility and accuracy for either positive (P = 0.0652) or negative (P = 0.9220) predictions.

Our findings suggest that the predictive accuracy of embryo outcomes does not significantly differ whether the ICM is clearly visible or not in the static 2D image. This supports the need to focus on other morphological and morphometric features that can be consistently evaluated. AI tools analysing these features may offer more reliable and objective clinical decision support.