Neural Network ECG Interference Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electrocardiogram (ECG) signals acquired during medical procedures are often distorted by external interference, such as power grid spectral lines and their harmonics, which existing techniques struggle to suppress in real-time, necessitating prolonged processing and delaying the presentation of undistorted signals to physicians.
Innovation Solution
A neural network, specifically an autoencoder with multiple layers, is trained using undistorted ECG signals and interference signals with spectral lines and harmonics, allowing for the suppression of interference in real-time ECG signals by applying the trained model to both the ECG and external interference signals, producing a clean ECG signal within seconds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional signal processing techniques are used to suppress interference in ECG signals, then interference suppression is achieved, but processing time is prolonged and real-time presentation is delayed
Solution Approach 1:
The neural network is trained in advance using clean ECG signals and synthetic interference signals containing spectral lines and harmonics. This preliminary training enables the network to rapidly process real ECG signals in real-time without requiring prolonged processing during actual medical procedures
Solution Approach 2:
Traditional mechanical signal processing techniques (filtering, averaging) are replaced with a neural network-based computational system. The neural network processes ECG signals through learned patterns rather than conventional algorithmic operations, achieving both high interference suppression accuracy and real-time processing speed
2Measurement precision
If multiple processing phases are used to extract clean ECG signals, then signal quality is improved, but processing complexity and time are increased
Solution Approach 1:
The patent extracts and removes interference components (spectral lines and harmonics) from ECG signals using a neural network. The network is trained to identify and separate interference patterns from genuine cardiac signals, extracting only the relevant medical information while discarding noise
Solution Approach 2:
The neural network is designed to handle multiple types of interference simultaneously (power line noise, muscle artifacts, electrode contact noise) through a single unified model, eliminating the need for multiple specialized processing phases and reducing overall system complexity
Data Source
AI summary
A method includes, receiving a first electrocardiogram (ECG) signal, which is acquired in a heart of a patient and is distorted by interference. One or more external signals that sense the interference concurrently with acquisition of the first ECG signal are received from one or more sources external to the heart. A second ECG signal, in which the interference is suppressed relative to the first ECG signal, is produced by applying a trained Neural Network (NN) to the first ECG signal and to the one or more external signals.