Oral Presentation ESA-SRB-NZSE-CaSR 2026 in conjunction with ENSA

Higher sperm curvilinear velocity is linked to good quality blastocyst formation (143971)

Sahar Shahali 1 2 , Jennifer Lorimer 1 3 , Rob McLachlan 3 4 , Deirdre Zander-Fox 3 4 5 , Klaus Ackermann 6 , Moira O'Bryan 2 , Reza Nosrati 1
  1. Mechanical and Aerospace Engineering , Monash University, Melbourne, Victoria, Australia
  2. School of Biosciences, Faculty of Science, University of Melbourne, Melbourne, Victoria, Australia
  3. Monash IVF Group, Melbourne, Richmond Victoria, Australia
  4. Department of Obstetrics and Gynaecology, Monash University, Melbourne, Clayton, Victoria, Australia
  5. Department of Obstetrics and Gynaecology, Adelaide University, Adelaide, South Australia, Australia
  6. SoDa Labs and Department of Econometrics and Business Statistics, Monash Business School, Monash University, Melbourne, Victoria, Australia

Background: Sperm motility is a key part of routine semen assessment, but outputs reflect the sperm population rather than the individual sperm used for intracytoplasmic sperm injection (ICSI). Therefore, it remains unclear whether the kinematics of the sperm used for ICSI provide information about the likelihood of the embryo forming a good-quality blastocyst. To address this knowledge gap, we compared the kinematic parameters of individual sperm used for ICSI between embryos that did and did not develop into a good-quality blastocyst by day 6.

                                    

Methods: We analysed 594 linked sperm-embryo ICSI cycles records, including 243 embryos that formed a good-quality blastocyst by day 6 and 351 that did not. Nine sperm kinematic parameters were compared across the full dataset and within groups. Mixed-effects logistic regression included cohort as a random effect. A multivariable model tested the independent contribution of each parameter. XGBoost–SHAP, a machine-learning approach, was used to explore non-linear patterns across combined motility measures.

 

Results: Motility parameters showed similar distributions in sperm linked to embryos that did and did not form a good-quality blastocyst. No individual parameter showed a strong direct link with good-quality blastocyst formation, with Spearman correlations close to zero (−0.05 to 0.06). The univariable mixed-effects models showed the same pattern, with all 95% confidence intervals crossing one. However, in the multivariable model, each standard deviation increase in curvilinear velocity (VCL) was linked to 31% higher odds of good-quality blastocyst formation. In other words, higher curvilinear velocity was linked to a greater likelihood that the resulting embryo would form a good-quality blastocyst by day 6, after accounting for the other selected motility parameters. SHAP analysis further showed that VCL contributed nonlinearly to the model output.

 

Conclusion: VCL was the most informative sperm motility parameter for good-quality blastocyst formation when considered alongside other sperm kinematic.