Optimizing trigger timing in minimal ovarian stimulation for In Vitro fertilization using machine learning models...

Area of Science:

• Reproductive Medicine
• Artificial Intelligence in Healthcare
• Biostatistics

Background:

• Optimal trigger timing (TT) is vital for oocyte maturation and release in In Vitro Fertilization (IVF).
• Determining precise TT is challenging due to multiple variables, impacting IVF success rates, especially in minimal ovarian stimulation protocols.

Purpose of the Study:

• To develop a machine learning multi-output model for predicting IVF outcomes (retrieved oocytes, mature oocytes, fertilized oocytes, usable blastocysts).
• To enhance trigger shot timing (TT) by identifying patients with potentially suboptimal timing in minimal stimulation cycles.

Main Methods:

• Development of a machine learning multi-output predictive model.
• Utilizing the model to analyze trigger shot timing in minimal ovarian stimulation cycles.
• Statistical validation of the model's accuracy and performance.

Main Results:

• Approximately 27% of treatments utilized a suboptimal trigger shot day.
• Optimizing TT with the AI model showed a potential 46% increase in usable blastocyst production.
• The model accurately predicted IVF outcomes within a 48-hour window post-trigger shot.

Conclusions:

• AI-driven predictive models can serve as valuable tools for optimizing trigger shot timing in IVF.
• Enhancing TT through AI can improve IVF success rates, particularly in minimal ovarian stimulation.
• The developed AI model contributes to making the IVF process safer and more efficient.