Seismocardiography Autoencoders for Reliable Heart Failure Indication

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing seismocardiography methods struggle to reliably identify heart failure due to varying amplitude measures caused by different underlying conditions, making it challenging to diagnose heart failure accurately.

Innovation Solution

A method using a trained autoencoder to analyze chest wall vibrations measured by an accelerometer, comparing reconstructed signal intervals with original intervals to determine heart failure indication based on correlation differences between healthy and unhealthy subjects.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If amplitude measures are used to identify heart failure in SCG signals, then diagnostic capability is improved, but reliability deteriorates due to varying amplitude measures caused by different underlying conditions

Engineering Contradiction:
Improvediagnostic capabilityVSAvoidreliability
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent transforms the SCG signal through multiple parameter changes: filtering to isolate specific frequency ranges, computing derivatives to extract rate-of-change information, and calculating temporal intervals between fiducial points. These transformations convert the unreliable amplitude measures into stable temporal interval measures that maintain diagnostic capability while improving reliability across different heart failure conditions

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent introduces temporal intervals between fiducial points as an intermediary parameter that mediates between the raw SCG signal and the final diagnostic indicator. These intervals serve as stable intermediate representations that are less sensitive to the varying amplitude measures caused by different underlying conditions, thereby improving reliability while preserving diagnostic information

Inventive Principle:
Principle #24Intermediary (Mediator)

2Loss of information

If both low frequency SCG components and audible components are sampled simultaneously using an accelerometer, then comprehensive cardiovascular information is obtained, but diagnostic complexity increases due to different approaches needed for different components

Engineering Contradiction:
Improvecomprehensive cardiovascular informationVSAvoiddiagnostic complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The patent segments the simultaneously sampled SCG signal into distinct components based on frequency characteristics. By applying frequency filtering, the signal is divided into low-frequency SCG components (below 5 Hz) and audible heart sound components (above 40-60 Hz), allowing each component to be analyzed with appropriate diagnostic methods while maintaining comprehensive cardiovascular information

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent dynamically processes the segmented signal components through derivative computations and temporal interval measurements. This dynamic transformation converts the complex multi-component signal into standardized temporal metrics that simplify diagnostic interpretation while preserving the comprehensive information from both SCG and heart sound components

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP4554470B1Autoencoders in quantitative seismocardiography
Publication Date: 2026.04.22 ACARIX AS
  • EP4554470B1 patent drawingFigure 1
  • EP4554470B1 patent drawingFigure 2
  • EP4554470B1 patent drawingFigure 3~4

AI summary

A technology for determining an indication of heart failure of a subject (18) is proposed. It comprises: obtaining (100a) a first signal interval (36) from a source signal recorded with an accelerometer (14) placed on the chest of a subject (18), wherein the first signal interval (36) corresponds to a first subinterval of a heart cycle; inputting (200a) the first signal interval (36) into a first autoencoder, wherein the first autoencoder is trained on the corresponding first signal intervals obtained from healthy subjects and outputs a reconstructed first signal interval (44), determining (300a) a first correlation between the first signal interval (36) and the reconstructed first signal interval (44), and determining (400) the indication of heart failure based on the first correlation.