Embryo Image Analysis Using Deep Neural Networks for Pregnancy Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for evaluating mammalian embryos are subjective and lack consistency, making it difficult to identify those with a higher likelihood of resulting in pregnancy while screening out less viable embryos.
Innovation Solution
A system and method utilizing a microscope, video camera, and processor to convert digital images of embryos into greyscale, detect and expand boundaries, segment, and isolate embryos for analysis with deep neural networks and machine learning models to predict pregnancy outcomes and determine embryo grade and stage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual inspection and subjective grading by embryologists is used, then the process is simple and quick, but the evaluation lacks consistency and objectivity
Solution Approach 1:
The patent replaces the mechanical/subjective visual inspection system with an automated image processing and machine learning system. Digital images of embryos are processed through algorithms that objectively measure morphological features, replacing the embryologist's subjective grading with quantifiable data analysis, thereby improving evaluation consistency while accepting increased system complexity
Solution Approach 2:
The patent introduces an intermediary layer between the embryo and the evaluation outcome: a computer-based image analysis system that captures digital images, processes them through standardized algorithms, and generates objective measurements. This intermediary eliminates the variability of human judgment while maintaining the essential function of embryo assessment
2Reliability
If automated image processing and machine learning are implemented, then objectivity and standardization improve, but the system complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-training machine learning models with large datasets of embryo images and outcomes before actual use. The system performs preliminary image processing steps (segmentation, feature extraction) automatically, and the pre-trained models provide reliable predictions without requiring complex real-time adjustments during embryo assessment
Solution Approach 2:
The patent uses digital copies of embryos in the form of high-resolution images captured by microscopy systems. These digital replicas allow for repeated analysis, storage, and sharing without affecting the actual embryos, enabling complex computational analysis while keeping the physical system simple
Data Source
AI summary
A method and system for predicting embryo grade, stage and/or pregnancy outcome in a mammalian embryo, which includes observing a plurality of mammalian embryos with a microscope, observing a plurality of digital images of mammalian embryos with a camera, converting the plurality of digital images of mammalian embryo from RGB to greyscale, detecting, diluting, and expanding the boundaries of the mammalian embryo followed by segmenting, cropping and isolating the digital images or utilizing the plurality of digital images of mammalian embryos that are both original images and mask images with a convolutional neural network that minimizes pixel classification errors that provide semantic representations to provide information about embryo qualities with a processor electrically connected to the camera to predict embryo grade, stage and/or pregnancy status of the plurality of mammalian embryos utilizing either a deep neural network segmenter, an autoencoder for extracting features, or a deep neural network.