Smartphone-Assisted CSCTAS-Derived Sperm Kinematics and Clinical Pregnancy after Intrauterine Insemination in a Turkish Cohort of Couples with Unexplained Infertility.

Abstract

Conventional semen analysis provides limited information on sperm kinematic behavior in couples with unexplained infertility undergoing intrauterine insemination (IUI). We used a laboratory based, microscope mounted smartphone-assisted Computerized Sperm Counting and Tracking Analysis System (CSCTAS) to characterize sperm kinematics in prepared samples and compare them by ultrasound-confirmed clinical pregnancy after IUI. In this retrospective single-center Turkish cohort, 184 first IUI cycles between January 2018 and December 2019 were analyzed. Prepared samples were assessed under standardized Makler chamber conditions. Manual-CSCTAS agreement for total motility was assessed by Bland-Altman analysis, and single-variable receiver operating characteristic analyses were supplementary. Clinical pregnancy occurred in 29 cycles (15.8%). Pre- and post-preparation semen parameters were similar between groups. Mean CSCTAS-derived wobble (WOB) was higher in pregnancy-positive cycles (0.83 ± 0.03 vs. 0.81 ± 0.09; p = 0.004; Cohen's d = 0.24), whereas beat-cross frequency (BCF) was lower at the nominal level (7.28 ± 0.70 vs. 7.61 ± 1.05 Hz; p = 0.039; d = - 0.33). Although statistically significant, the differences had small effect sizes. Bland-Altman analysis showed a mean bias of - 3.44% points, 95% limits of agreement of - 24.69 to 17.81, and proportional bias. Single-variable area under the curve (AUC) values were 0.526 for WOB and 0.555 for BCF, and both 95% confidence intervals included 0.50. These findings suggest exploratory population-level digital motility signals; however, the small effect sizes and near-chance single-variable AUC values indicate limited predictive performance and do not support their use as standalone predictors of individual IUI outcome.