HRV-Based VF Onset Prediction for Wearable Early Warning

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for predicting ventricular fibrillation (VF) are limited by the need for continuous ECG monitoring, unrealistic for widespread use, and lack clear time intervals for prediction, with existing algorithms not effectively addressing sudden cardiac death prevention in high-risk patients.

Innovation Solution

A machine learning model, specifically a temporal convolutional network (TCN), is trained on heart rate variability (HRV) parameters to predict VF onset by analyzing temporal beat activity, issuing alerts within a specified time frame using an Alarm Desirability Function (ADF) for supervised learning.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If continuous ECG monitoring is used for VF prediction, then prediction accuracy is improved, but device complexity and patient burden increase

Engineering Contradiction:
Improveprediction accuracyVSAvoidmonitoring system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts HRV parameters from ECG signals and uses them as simplified input features for the machine learning model, rather than using raw continuous ECG signals. This extraction approach reduces the complexity of the monitoring system while maintaining prediction accuracy, as HRV parameters capture essential cardiac variability information in a more manageable format.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent creates a simplified representation of cardiac activity through HRV parameter extraction, which serves as a copy or summary of the complex ECG signal. This copied information is then fed into the machine learning model, allowing prediction without requiring continuous complex ECG monitoring, thus reducing device complexity while preserving measurement precision.

Inventive Principle:
Principle #26Copying

2Reliability

If existing prediction algorithms are used, then VF detection capability is provided, but time interval for prediction is unclear and early warning is insufficient

Engineering Contradiction:
ImproveVF detection capabilityVSAvoidprediction time interval
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent implements preliminary action by training the machine learning model to predict VF onset based on historical HRV parameter patterns. The model analyzes temporal beat activity and identifies precursors to VF events, providing advance warning before the actual VF occurs. This allows intervention to be initiated earlier, reducing the loss of time compared to existing algorithms that detect VF only after onset or provide unclear time intervals.

Inventive Principle:
Principle #10Preliminary action

3Duration of action of stationary object

If wearable ECG monitors are deployed, then continuous monitoring is enabled, but patient compliance and realism decrease

Engineering Contradiction:
Improvecontinuous monitoring durationVSAvoidpatient compliance
Core Design Contradiction:
Duration of action of stationary objectVSEase of operation

Solution Approach 1:

The patent adopts a more practical approach by using simplified HRV parameter extraction rather than requiring complex continuous ECG monitoring. This allows the use of more affordable and patient-friendly wearable devices that can continuously monitor cardiac variability without requiring complex ECG lead configurations, thereby improving patient compliance while maintaining continuous monitoring capability.

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

4Reliability

If high-risk patients are identified for ICD implantation, then VF prevention is improved, but the fraction of eligible patients remains negligible

Engineering Contradiction:
ImproveVF prevention effectivenessVSAvoidapplicability to general population
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

The patent changes the parameters used for risk assessment by focusing on temporal beat activity patterns and HRV characteristics rather than relying on traditional ECG morphology analysis. This parameter transformation enables the detection of VF precursors in a broader population with higher sensitivity, potentially increasing the fraction of eligible patients for preventive intervention while maintaining high reliability in VF prediction.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3846677B1Predicting ventricular fibrillation
Publication Date: 2025.12.24 TECHNION RES & DEV FOUND LTD
  • EP3846677B1 patent drawingFigure 1
  • EP3846677B1 patent drawingFigure 2
  • EP3846677B1 patent drawingFigure 3

AI summary

A method comprising: at a training stage, training a machine learning algorithm on a training set comprising: (i) Heart Rate Variability (HRV) parameters extracted from temporal beat activity samples, wherein at least some of said samples include a representation of a Ventricular Fibrillation (VF) event, (ii) labels associated with one of: a first period of time immediately preceding a VF event in a temporal beat activity sample, a second period of time immediately preceding the first period of time in a temporal beat activity sample, and all other periods of time in a temporal beat activity sample; at an inference stage, receiving, as input, a target HRV parameters representing temporal beat activity in a subject; and applying said machine learning algorithm to said target HRV parameters, to predict an onset time of a VF event in said subject.