ECG Noise-Filtering Device Using Savitzky-Golay and Wavelet Transform
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional ECG noise-filtering methods, such as low-pass, band-pass, and band-rejection filters, often fail to effectively filter noise without attenuating the real ECG signal, as they cannot immediately adjust their attenuation rates, leading to incomplete noise removal and potential loss of valuable signal data.
Innovation Solution
The combination of the Savitzky-Golay algorithm and the Stationary Wavelet Transform (SWT) algorithm is used in an ECG noise-filtering device to process ECG signals, where the Savitzky-Golay filter initializes noise cancellation, followed by SWT algorithms at multiple levels to isolate and remove noise, and inverse SWT to reconstruct the signal, with correlation calculations and PT-wave removal operations to refine the filtering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If traditional low-pass, band-pass, or band-rejection filters are used to filter noise from ECG signals, then noise filtering is achieved, but the real ECG signal is also attenuated and valuable signal data is lost
Solution Approach 1:
The filtering process is divided into multiple stages: first applying a low-pass filter to remove high-frequency noise, then performing wavelet transform to separate different frequency components, and finally selectively reconstructing the signal. This segmentation allows different filtering strategies to be applied to different frequency bands, effectively removing noise while preserving valuable ECG signal components.
Solution Approach 2:
The patent dynamically adjusts filtering parameters based on the characteristics of the ECG signal. By changing the cutoff frequencies and attenuation rates adaptively rather than using fixed filter parameters, the system can effectively filter noise while minimizing attenuation of the real ECG signal across different signal conditions.
2Object-affected harmful factors
If the attenuation rate of traditional filters is increased to improve noise filtering, then noise removal is enhanced, but the real ECG signal is also filtered more aggressively
Solution Approach 1:
Different attenuation rates are applied to different frequency components of the ECG signal. The wavelet transform decomposes the signal into multiple frequency bands, and each band receives appropriate filtering treatment. High-frequency noise components receive strong attenuation while low-frequency ECG components receive minimal or no attenuation, achieving local optimization of filtering quality.
Solution Approach 2:
The filtering system dynamically adjusts attenuation rates based on the instantaneous characteristics of the ECG signal. Rather than using a fixed attenuation rate, the system adapts the filtering strength to the signal conditions, applying stronger attenuation when noise is dominant and weaker attenuation when the ECG signal is strong, thereby maintaining measurement precision.
3Device complexity
If traditional filters are used with fixed frequency ranges, then simple filtering is achieved, but noise outside the specific frequency range cannot be filtered
Solution Approach 1:
The patent transitions from traditional single-dimensional frequency filtering to multi-dimensional signal processing by applying wavelet transform. This transforms the signal into a time-frequency representation, allowing noise at multiple frequency ranges to be identified and filtered simultaneously. The wavelet decomposition creates multiple resolution levels that capture both temporal and spectral characteristics of the ECG signal and its noise.
Data Source
AI summary
An Electrocardiography (ECG) noise-filtering device is provided in the invention. The ECG device includes a filter and a calculation circuit. The filter receives a first ECG signal and performs a Savitzky-Golay algorithm to generate a second ECG signal. The calculation circuit is coupled to the filter to receive the second ECG signal and processes the second ECG signal according to a Stationary Wavelet Transform (SWT) algorithm to generate a noise signal, and subtracts the noise signal from the second ECG signal to filter the noise signal in the first ECG signal.