ResNet-Based Echocardiogram View Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying standard views in echocardiograms are time-consuming and labor-intensive, particularly for large sample data, and primarily focus on 2D grayscale images with limited research on 3D imaging and color Doppler ultrasound images.
Innovation Solution
A multi-modal multi-view classification method for echocardiograms based on a deep learning algorithm, which involves collecting and preprocessing videos and images, annotating the data, dividing it into training, validation, and test sets, constructing a ResNet-based classification model, and evaluating its performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual selection of standard apical view images is used for post-processing analysis, then accurate identification of cardiac structures and functions can be achieved, but the process becomes time-consuming and labor-intensive for large sample data
Solution Approach 1:
The patent replaces the manual mechanical selection process with an automated deep learning-based image recognition system. The neural network model automatically identifies and classifies standard apical view images from echocardiogram sequences, substituting human operators while maintaining high accuracy in view identification and enabling rapid processing of large datasets
Solution Approach 2:
The system enables self-service automation where the echocardiogram analysis software automatically performs view classification without requiring manual intervention. The deep learning model independently processes images, identifies standard views, and prepares data for subsequent cardiac function analysis, making the system self-sufficient for the classification task
2Device complexity
If research focuses only on 2D grayscale images, then simpler processing can be achieved, but the system cannot utilize rich information from 3D imaging and color Doppler ultrasound images
Solution Approach 1:
The patent creates a universal image recognition system that can process multiple types of echocardiogram images including 2D grayscale, 3D imaging, and color Doppler ultrasound images. The deep learning model is designed to handle diverse image modalities and views (parasternal, apical, subxiphoid), making the system multi-functional and capable of utilizing rich information from all image types without requiring separate processing pipelines
Data Source
AI summary
The present disclosure provides a multi-modal multi-view classification method for echocardiograms based on a deep learning algorithm, including: collecting videos and images of multi-modal multi-view adult echocardiograms, and preprocessing the videos and images; annotating the preprocessed videos and images of adult echocardiograms to generate an adult echocardiogram dataset; dividing the adult echocardiogram dataset into a training set, a validation set, and a test set; constructing an adult echocardiogram view classification model based on a ResNet network, training the model using the training set, and selecting an optimal classification model using the validation set; evaluating performance of the optimal adult echocardiogram view classification model based on the test set; and inputting a to-be-tested image or video into the adult echocardiogram view classification model to obtain a classification result.