Despite four decades of refinement, in vitro fertilisation live birth rates plateau at 25-30% globally, largely due to limitations in subjective embryo selection. We evaluated whether artificial intelligence-assisted embryo selection improves clinical pregnancy rates compared to conventional morphological assessment. This multi-centre prospective cohort study was conducted across six university-affiliated IVF clinics in the United Kingdom between January 2023 and March 2024. We enrolled 1,187 patients undergoing single blastocyst transfer, with 1,172 completing the analysis (585 in the control group and 587 in the intervention group). The study compared a 6-month control phase using conventional morphological grading against a 6-month intervention phase employing the iDAScore v1.0 deep learning algorithm. The primary outcome analysis employed logistic regression, adjusted for female age, body mass index, anti-Müllerian hormone level, and infertility diagnosis. Clinical pregnancy rates were significantly higher in the AI-assisted group (45.7%, 268/587) compared to the conventional group (38.1%, 223/585), with an adjusted odds ratio of 1.40 (95% confidence interval, 1.11-1.77; p = 0.004). This represents a 40% increase in pregnancy odds and an absolute improvement of 7.6 percentage points. AI-assisted embryo selection significantly improves clinical pregnancy rates whilst maintaining the safety standard of single embryo transfer.
Agboola et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: