<![CDATA[Introduction: Artificial intelligence (AI) based clinical decision support systems (CDSS), including tools such as Opt-IVF, are increasingly applied in assisted reproductive technology to support individualized controlled ovarian stimulation. These systems are designed to tailor gonadotropin dosing, enhance oocyte yield, and improve in vitro fertilization (IVF) outcomes while simultaneously reducing adverse effects and treatment costs. However, the overall impact of AI-assisted CDSS on hormone dosing optimization and IVF outcomes has not yet been comprehensively synthesized. Methods: A systematic search of major scientific databases was conducted to identify studies evaluating AI- or machine learning–based decision support tools applied during ovarian stimulation in IVF cycles. Eligible studies included observational studies, retrospective cohorts, and clinical validation studies reporting hormone dosing parameters and IVF-related outcomes. Primary outcomes comprised initial and cumulative gonadotropin doses and the number of retrieved metaphase II (MII) oocytes. Secondary outcomes included indicators of ovarian response and treatment safety. Data were extracted and synthesized, and meta-analytical techniques were applied where appropriate. Result: Six studies fulfilled the inclusion criteria. Across these studies, AI-driven CDSS consistently reduced both starting and cumulative gonadotropin doses compared with conventional clinician-guided dosing. Meta-analysis demonstrated that AI-assisted dosing achieved comparable numbers of retrieved MII oocytes, indicating non-inferior reproductive outcomes despite lower hormone exposure. Several studies also reported enhanced personalization of stimulation protocols based on individual patient characteristics, including age, anti-Müllerian hormone levels, antral follicle count, and body mass index. No increase in adverse events, including ovarian hyperstimulation syndrome, was reported. Conclusion: AI-driven clinical decision support tools effectively optimize hormone dosing during IVF cycles while maintaining comparable reproductive outcomes. These technologies represent a promising strategy for personalized ovarian stimulation, potentially reducing medication burden and treatment costs without compromising efficacy. Further large-scale randomized studies are required to confirm long-term clinical benefits and to standardize AI implementation in IVF practice.]]>
