Oocyte Quality Analysis Combining Morphology and Pressure Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current embryo selection methods in IVF treatments are subjective, labor-intensive, and lack a standardized grading system, leading to inaccurate assessments of oocyte viability due to reliance on manual annotation and single-type information analysis.
Innovation Solution
A system utilizing machine learning models to analyze both morphological and mechanical features of unfertilized oocytes through image sequences and pressure measurements, providing objective and quantitative evaluations of oocyte quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual annotation and grading systems are used to assess oocyte quality, then embryologists can perform subjective evaluation, but the process is labor-intensive and error-prone with low accuracy
Solution Approach 1:
The patent replaces the manual mechanical grading process with an automated machine learning system that processes oocyte images and sensor data. The system uses trained models to automatically assess oocyte quality, eliminating the need for manual annotation by embryologists while improving accuracy through objective, data-driven evaluations.
Solution Approach 2:
The system enables self-service assessment where the machine learning model autonomously evaluates oocyte quality without requiring human intervention for each assessment. The model has been pre-trained on datasets and can independently perform grading, selection, and quality assessment tasks.
2Measurement precision
If single-type information (e.g., images only) is used for oocyte analysis, then the analysis process is simpler, but the assessment accuracy is insufficient
Solution Approach 1:
The patent merges multiple types of sensor information including images, pressure sensor data, and other biophysical measurements into a unified analysis framework. The machine learning system processes this multi-modal data together to comprehensively assess oocyte quality, leveraging the complementary information from each sensor type to improve prediction accuracy.
Solution Approach 2:
The system creates a composite information structure by combining data from multiple sensor types. The machine learning model processes this composite dataset that integrates visual, mechanical, and other biophysical properties, enabling more robust and accurate oocyte viability predictions than any single sensor type could provide alone.
3Stability of the object's composition
If standardized grading systems are implemented, then oocyte selection becomes more consistent, but no universally adopted standard currently exists leading to variability
Solution Approach 1:
The patent develops a universal machine learning-based grading system that can consistently assess oocytes across different laboratories and conditions. The system is designed to be adaptable to various oocyte types and assessment scenarios, providing a standardized yet flexible framework that can be universally applied without requiring laboratory-specific customization.
Data Source
AI summary
The present disclosure generally relates to systems and methods for evaluating viability of oocytes. In some implementation examples, an image sequence associated with an oocyte being aspirated into a pressure tool is obtained. Using a segmentation model, objects associated with the oocyte can be identified. Based on the objects identified, features such as morphological features and an aspiration depth associated with the oocyte can be determined. At least some of the features can then be fed into a machine learning model to generate an oocyte grade that indicates a likelihood of the oocyte developing into a usable blastocyst. Optionally, the oocyte grade can be presented to a user via an interactive user interface.