Unfertilized Egg Cohorting Using Time-Lapse Blastocyst Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for cryopreservation and IVF rely on visual examination of unfertilized eggs for maturity, which is subjective and inefficient in predicting their potential for successful fertilization and development into blastocysts.
Innovation Solution
A machine-learning model is trained to analyze time-lapse images of unfertilized eggs, determining their potential to become blastocysts and distributing them into cohorts based on predicted outcomes, optimizing the chances of successful fertilization and development.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual examination methods are used to classify unfertilized eggs, then the process is simple and quick, but the prediction accuracy of fertilization success is low
Solution Approach 1:
The patent replaces the manual visual examination system with an automated machine learning system that processes images of unfertilized eggs. The ML model analyzes morphological features and time-lapse imaging data to predict fertilization success, substituting human subjective judgment with objective computational analysis that achieves superior accuracy.
Solution Approach 2:
The patent introduces time-lapse imaging technology as an intermediary between the egg sample and the prediction outcome. This intermediary captures detailed morphological changes over time, providing rich data for the ML model to analyze, thereby enhancing prediction accuracy without requiring direct manual intervention.
2Measurement precision
If machine learning analysis is implemented to predict blastocyst development, then prediction accuracy improves, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary analysis by capturing time-lapse images of unfertilized eggs during the incubation period before fertilization occurs. The ML model processes these pre-captured images to predict which eggs will successfully develop into blastocysts, enabling early selection and reducing the time needed for actual fertilization attempts.
Solution Approach 2:
The patent segments the egg selection process into distinct phases: image capture during incubation, ML model prediction of fertilization success, and subsequent prioritization of high-probability eggs for fertilization. This segmentation allows parallel processing and optimizes the timing of each step to minimize overall processing time.
Data Source
AI summary
Methods and systems are described for predicting probabilities of unfertilized eggs becoming blastocysts. For example, a method can include a trained machine-learning model receiving images of unfertilized eggs. The trained machine-learning model can determine predictions for each of the unfertilized eggs becoming blastocysts based at least on the appearance of the unfertilized eggs in the images. The predictions can be provided by the trained machine-learning model. Based on the predictions, a distribution of the unfertilized eggs into cohorts can be determined, including. Then, based on the cohorts, a second prediction of a cohort producing at least one blastocysts can be determined.