Echocardiogram Analysis Using Convolutional Deep Learning

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The challenge in echocardiogram analysis lies in the difficulty of automated interpretation due to the vast search space of potential heart pathologies and the need for manual comparison of three-dimensional views from different planes, which is time-consuming and costly, especially with a high volume of follow-up examinations.

Innovation Solution

A computer-implemented method using convolutional deep learning networks to analyze pairs of consecutive echocardiograms, aligning them temporally, interpolating frames, and training a predictive model to automatically determine changes in cardiac function or structure, thereby simplifying the interpretation process.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If automated comparison of consecutive echocardiograms is performed, then diagnostic efficiency is improved and manual analysis burden is reduced, but the complexity of aligning three-dimensional views from different planes increases system complexity

Engineering Contradiction:
Improvediagnostic efficiencyVSAvoidsystem complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The system segments the complex task of echocardiogram comparison into distinct processing stages: image acquisition, temporal alignment, feature extraction, and change detection. Each stage handles a specific aspect of the analysis, making the overall complex system manageable and modular.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary alignment module that acts as a mediator between the raw echocardiogram images and the comparison algorithm. This intermediary component performs temporal alignment of three-dimensional views from different planes, transforming the complex alignment problem into a standardized input format that the change detection algorithm can process efficiently.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If manual interpretation of echocardiograms is performed, then diagnostic accuracy is maintained, but the time required for analysis increases and costs increase

Engineering Contradiction:
Improvediagnostic accuracyVSAvoidanalysis time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system enables self-service automated analysis by training the change detection algorithm on paired echocardiograms with expert annotations. Once trained, the system autonomously performs diagnostic analysis without requiring continuous manual intervention, maintaining diagnostic accuracy while significantly reducing analysis time and costs.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent incorporates feedback mechanisms where the automated system's predictions are compared against expert interpretations, and the system learns from these discrepancies through continuous training on annotated data. This feedback loop ensures diagnostic accuracy is maintained and improved over time while preserving the speed advantages of automation.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If extensive manual analysis of follow-up echocardiograms is performed, then changes in cardiac function are accurately detected, but the cost of examinations increases

Engineering Contradiction:
Improvechange detection accuracyVSAvoidcost
Core Design Contradiction:
Measurement precisionVSLoss of energy

Solution Approach 1:

The system performs preliminary training on a comprehensive dataset of paired echocardiograms with expert annotations before deployment. This preliminary action embeds the knowledge of change detection patterns into the algorithm, enabling accurate detection of cardiac function changes during actual use without requiring expensive manual analysis for each case.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent creates a computational model that copies the diagnostic reasoning process of expert cardiologists through training on annotated data. This virtual copy can then perform change detection analysis autonomously, maintaining the accuracy of expert review while eliminating the associated costs and resource requirements.

Inventive Principle:
Principle #26Copying

Data Source

PatentEP3704707B1A method and apparatus for analysing echocardiograms
Publication Date: 2024.01.10 KONINKLIJKE PHILIPS NV
  • EP3704707B1 patent drawingFigure 1
  • EP3704707B1 patent drawingFigure 2
  • EP3704707B1 patent drawingFigure 3

AI summary

There is provided a computer-implemented method for analysing echocardiograms, the method comprising: obtaining (302) a plurality of pairs of consecutive echocardiograms for a plurality of subjects from a database (200), each echocardiogram having an associated indication of the content of the echocardiogram; analysing (304) each pair of consecutive echocardiograms to determine an associated class, the class indicating whether there is a change or no change between the consecutive echocardiograms in the pair; for each pair of consecutive echocardiograms, determining (306) an abstract representation of each echocardiogram by performing one or more convolutions and/or reductions on the echocardiograms in the pair, the abstract representation comprising one or more features indicative of the class of the pair; and training (308) a predictive model to determine a class for a new pair of echocardiograms based on the abstract representations for the plurality of pairs of consecutive echocardiograms.