Research Article
Yenkie KM, Diwekar UM and B
Abstract
Background: IVF is divided into four stages: superovulation, egg-retrieval, fertilization and embryo transfer. The superovulation stage has specific protocols which include daily injections of hormones, decided by regular monitoring and testing involving cost intensive methods like ultrasound, to enable multiple ovulations per menstrual cycle. However, there is a lack of systematic planning and hormonal dosage prediction for successful superovulation using apriori information. Methods: This work aims at developing a systematic outcome projection method for superovulation based on initial observations in an IVF cycle. The information about the size range and number of follicles was transformed to moment based information using the mathematical approach from crystallization literature. The follicle growth was modeled as a function of injected hormones and the properties of the follicles were represented in terms of moments. The model parameters were estimated using the data obtained from initial two days of observation. This information was used to project the follicle size distribution for the remaining cycle days. Results and Conclusion: The model assumptions and its correlation to batch crystallization prove promising on comparison of the simulated follicle size and number to the observed data for the patients. The model prediction accuracy is determined by the statistical analysis of clinical data available for 50 superovulation cycles. Thus, it can act as an indicator for the success or failure of the ongoing superovulation stage in the IVF cycle and the decision whether to continue with the procedure or abandon it and start from donor eggs can be made, thus saving treatment cost and time on unsuccessful attempts.