Arrhythmia Detection Using Wavelet Decomposition and CNNs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting arrhythmia in ECG signals using machine learning techniques are susceptible to noise, particularly in signals measured by wearable devices, leading to inaccurate classification.
Innovation Solution
A method involving the application of a decomposition algorithm, such as Discrete Wavelet Transform (DWT), to ECG signals, followed by processing with a two-stage convolutional neural network (CNN) architecture to classify the presence of arrhythmia.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If machine learning techniques are used to detect arrhythmia in ECG signals, then detection capability is improved, but accuracy deteriorates due to noise susceptibility
Solution Approach 1:
The ECG signal is divided into multiple segments or portions, and each segment is processed independently through the decomposition algorithm and CNN classification. This segmentation allows the system to handle noisy signals by analyzing local characteristics of each segment, improving overall classification accuracy while maintaining automated detection capability.
Solution Approach 2:
A decomposition algorithm (such as wavelet transform) is introduced as an intermediary between the noisy ECG signal and the CNN classifier. This intermediary process transforms the noisy signal into a decomposed representation that separates signal components from noise, enabling the CNN to achieve higher classification accuracy without sacrificing automation.
2Duration of action of moving object
If wearable devices are used for continuous ECG monitoring, then monitoring duration is improved, but signal quality deteriorates due to noise and interference
Solution Approach 1:
The decomposition algorithm is used to convert the harmful noise in wearable ECG signals into separable components. By transforming the noisy signal into frequency or scale domains, the system can identify and isolate noise from useful signal components, thereby maintaining measurement precision throughout long-term continuous monitoring periods.
Solution Approach 2:
The decomposition algorithm performs preliminary processing on the ECG signal before it is input to the CNN classifier. This preliminary action of decomposing and cleaning the signal in advance ensures that the subsequent classification operates on improved signal quality, enabling accurate detection during extended monitoring durations.
3Measurement precision
If standard ECG systems with multiple electrodes are used, then signal quality is improved, but practicality deteriorates for long-term and mobile monitoring
Solution Approach 1:
The patent replaces the complex mechanical system of standard 12-lead ECG equipment with a simplified wearable device that uses fewer electrodes. The decomposition algorithm and CNN classification compensate for the reduced electrode count by processing signals in the transformed domain, maintaining measurement precision while greatly improving ease of operation for continuous and mobile monitoring.
Data Source
AI summary
A system and method of detecting presence of arrythmia in an electrocardiogram (ECG) signal. The method comprises steps of applying a decomposition algorithm to one or more portions of the ECG signal, each portion corresponding to at least one heartbeat. The method also comprises, for each portion, selecting at least one output of the decomposition algorithm; providing the selected at least one output to a first trained convolutional neural network (CNN) arrangement, the first CNN arrangement generating coefficients of a predetermined size; and inputting the coefficients of the predetermined size to a second trained CNN arrangement, the second CNN arrangement trained to output a classification of whether arrythmia is present in the portion of the ECG.