ECG-Based Diastolic Function Assessment Using Time-Frequency Models

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current heart testing methods, such as electrocardiograms (ECGs), fail to detect subtle abnormalities in diastolic function, leading to misdiagnoses, while echocardiograms, the gold standard, are expensive and typically reserved for confirmed heart issues.

Innovation Solution

A system and method that uses ECG measurements, combined with machine-learned computational models, to estimate echocardiogram parameters indicative of diastolic function, utilizing time-frequency transforms and patient demographics to provide quantitative and qualitative indicators of diastolic function.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If echocardiography is used to measure diastolic function parameters, then measurement precision is improved, but device complexity and cost increase

Engineering Contradiction:
Improvediastolic function parameter measurementVSAvoidechocardiography equipment
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces the mechanical/electronic imaging system of echocardiography with an electrocardiogram-based computational model. The system uses electrical signal processing and machine learning algorithms to estimate diastolic function parameters from ECG data, eliminating the need for complex ultrasound hardware while achieving comparable measurement precision.

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

Solution Approach 2:

The patent creates a computational copy or digital twin of the echocardiography measurement process. By training machine learning models on data from patients with known diastolic function (obtained through echocardiography), the system learns to predict these parameters from ECG signals alone, effectively copying the diagnostic capability without requiring the original complex equipment.

Inventive Principle:
Principle #26Copying

2Measurement precision

If echocardiography is used to assess diastolic function, then measurement precision is improved, but ease of operation deteriorates

Engineering Contradiction:
Improvediastolic function parameter measurementVSAvoiddiastolic function assessment
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent extracts the essential diagnostic information from the complex echocardiography process and encapsulates it within a simplified ECG-based computational model. By separating the complex image acquisition and analysis steps from the final diagnostic output, the system provides precise measurements through a much simpler operational interface that requires only standard ECG recording.

Inventive Principle:
Principle #2Taking out (Extraction)

3Ease of operation

If conventional ECG analysis methods are used, then ease of operation is maintained, but measurement precision deteriorates

Engineering Contradiction:
ImproveECG analysisVSAvoiddiastolic function parameter measurement
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent fundamentally changes the parameters used for ECG analysis by introducing time-frequency transform domain features and machine learning models rather than traditional time-domain measurements. This transformation enables the extraction of subtle patterns and relationships in the ECG signal that are invisible to conventional analysis methods, significantly improving measurement precision while maintaining operational simplicity.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4252651B1Electrocardiogram-based assessment of diastolic function
Publication Date: 2026.04.29 HEARTSCIENCES INC
  • EP4252651B1 patent drawingFigure 1
  • EP4252651B1 patent drawingFigure 2
  • EP4252651B1 patent drawingFigure 3

AI summary

Diastolic function may be assessed by operating one or more machine-learned computational models on electrocardiograms or electrocardiogram-derived features to compute quantitative diastolic indicators, including estimates of echocardiography parameters conventionally measured by echocardiography. Parameters such as those derived from time-frequency transforms of the electrocardiograms are used as input to the model(s) and/or computed within the model(s).