From Theory to Practice: Developing an AI-Driven Framework for Predictive Ovulation Trigger Timing in ICSI | ||
| Journal of Reproductive Medicine and Embryology | ||
| Articles in Press, Corrected Proof, Available Online from 11 April 2025 | ||
| Document Type: Narrative (literature review) | ||
| DOI: 10.21608/jrme.2025.372151.1040 | ||
| Authors | ||
| Ashraf Abo Ali* 1, 2; Karim mekhimer1, 3; Omar Abo Ali1, 2; Aya elsayed1, 3; Omar zakaria1, 3; Hassan Maghraby*2, 3 | ||
| 1Madina Fertility Center, Madina Women Hospital, Alexandria, Egypt. | ||
| 2Egyptian Foundation of Reproductive Medicine and Embryology (EFRE), Egypt. | ||
| 3Obstetrics and Gynecology Department, Faculty of Medicine, Alexandria University, Egypt. | ||
| Abstract | ||
| The timing of ovulation trigger administration is a critical challenge in assisted reproductive technologies (ART), where improper timing can lead to suboptimal oocyte retrieval and fertilization outcomes. Despite its significance, there is no standardized approach to determine the optimal timing, leading to clinical variability. This study aims to develop a predictive model using Meta AI to determine the optimal timing for ovulation trigger administration, with the goal of maximizing oocyte yield and the number of mature metaphase II (MII) oocytes retrieved on the day of oocyte pick-up (OPU). By incorporating a comprehensive set of clinical variables, this model seeks to guide clinicians and patients in making evidence-based decisions regarding ovulation induction, even in the absence of real-world data, ultimately improving the efficiency and outcomes of in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI) procedures. A literature review identified key factors influencing ovulation trigger timing, including patient demographics, ovarian reserve markers (AMH, AFC), stimulation parameters, and hormonal levels. Logistic regression was selected as the model due to its simplicity and interpretability. The model was evaluated using performance metrics such as accuracy, precision, recall, F1 score, and area under the curve (AUC). Three predictive approaches were proposed: a Follicle-Based Trigger Model (FBTM), a refined FBTM integrating AMH and AFC, and a Trigger Day Predictive Score (TDPS) model. Hypothetical results suggest these models could improve ovulation trigger timing and ART outcomes. Further empirical validation is required for clinical application. | ||
| Keywords | ||
| Ovulation trigger; predictive model; Meta AI; ICSI; machine learning | ||
|
Statistics Article View: 26 |
||
