Egg Cohort Distribution Using Blastocyst Potential Prediction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional methods for cryopreserving unfertilized eggs rely on visual examination and classification based on maturity, which is inefficient and does not account for the potential to develop into blastocysts, limiting the success of in-vitro-fertilization treatments.

Innovation Solution

A machine-learning model analyzes time-lapse images of unfertilized eggs to predict their potential to become blastocysts, enabling the distribution of eggs into cohorts based on these predictions to optimize the chances of obtaining viable embryos.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If visual examination and classification based on maturity is used for cryopreservation, then the process is simple and quick, but the accuracy of predicting blastocyst development potential is low

Engineering Contradiction:
Improveprediction accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces the manual visual examination system with an automated machine learning-based image analysis system. The machine learning model processes images of unfertilized eggs to predict blastocyst development potential, substituting human visual assessment with computational analysis that provides more accurate and consistent predictions.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces time-lapse imaging as an intermediary tool that captures developmental processes of unfertilized eggs over time. These images serve as intermediate data that feed into the machine learning model, enabling indirect observation and prediction of blastocyst potential without direct manipulation of the eggs.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If machine learning model with time-lapse images is used, then the prediction accuracy of blastocyst development is improved, but the time and computational resources required increase

Engineering Contradiction:
Improvesuccess rateVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent performs image capture and preliminary processing of unfertilized eggs before fertilization occurs. Time-lapse images are collected in advance, and the machine learning model analyzes these images to predict which eggs have high potential for blastocyst development. This preliminary assessment allows for optimized cohort distribution before the actual fertilization and development process begins.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent segments the egg population into distinct cohorts based on predicted blastocyst development potential. By dividing the unfertilized eggs into high-potential and lower-potential groups, the system can prioritize resources and attention on the most promising eggs, reducing the need to process and monitor all eggs equally throughout the development process.

Inventive Principle:
Principle #1Segmentation

3Productivity

If eggs are distributed into cohorts based on predictions, then the efficiency of obtaining viable embryos is improved, but the complexity of egg management increases

Engineering Contradiction:
ImproveefficiencyVSAvoidmanagement complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent applies different management strategies to different cohorts of eggs based on their predicted potential. High-potential eggs receive priority handling and are distributed to specific recipients or fertilization batches, while lower-potential eggs are managed differently. This localized quality-based management optimizes outcomes for each subgroup without requiring complex universal protocols.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent changes the management parameters (such as fertilization timing, recipient selection, and resource allocation) based on the predicted blastocyst potential of each egg cohort. Eggs with higher predicted potential are subjected to different handling protocols compared to those with lower potential, allowing optimized outcomes through parameter adjustment rather than uniform treatment.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4704105A1Systems and methods for distributing eggs for implantation according to their potential for successful fertilization
Publication Date: 2026.03.04 FAIRTILITY LTD
  • EP4704105A1 patent drawingFigure 1
  • EP4704105A1 patent drawingFigure 2
  • EP4704105A1 patent drawingFigure 3

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.