Abstract
The sperm quality of some males is in a critical state, making it hard for clinicians to choose the suitable fertilisation methods. This study aimed to develop an intelligent nomogram for predicting fertilisation rate of infertile males with borderline semen. 160 males underwent in vitro fertilisation (IVF), 58 of whom received rescue ICSI (R-ICSI) due to fertilisation failure (fertilisation rate of IVF ≤30%). A least absolute shrinkage and selection operator (LASSO) regression analysis identified sperm concentration, progressively motile spermatozoa (PMS), seminal plasma anti-Müllerian hormone (spAMH), seminal plasma inhibin (spINHB), serum AMH (serAMH) and serum INHB (serINHB) as significant predictors. The nomogram was plotted by multivariable logistic regression. This nomogram-illustrated model showed good discrimination, calibration and clinical value. The area under the receiver operating characteristic curve (AUC) of the nomogram was 0.762 (p < .001). Calibration curve and Hosmer-Lemeshow test (p = .5261) showed good consistency between the predictions of the nomogram and the actual observations, and decision curve analysis showed that the nomogram was clinically useful. This nomogram may be useful in predicting fertilisation rate, mainly focused on new biomarkers, INHB and AMH. It could assist clinicians and laboratory technicians select appropriate fertilisation methods (IVF or ICSI) for male patients with borderline semen.