Wearable ECG Peak Detection Using Wavelet Complexity Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wearable patch-type electrocardiogram measurement devices face challenges in detecting R-peaks due to unstable and noisy signals caused by user movement, resulting in small signal magnitude and difficulty in peak detection.
Innovation Solution
A bio-signal measurement apparatus that wavelet transforms sensed electrocardiogram signals into multiple levels, uses complexity values to detect effective peaks by calculating average peak time values and determining peak occurrence times based on pre-set critical complexity levels, employing algorithms like Shannon entropy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a patch-type wearable device is used to enable portable ECG measurement, then ease of operation and portability are improved, but measurement precision deteriorates due to small signal magnitude and unstable baseline
Solution Approach 1:
The patent applies wavelet transformation to convert the ECG signal from time domain to time-frequency domain, creating multiple decomposition levels. This dimensional transformation allows the system to analyze signal characteristics at different scales, effectively extracting meaningful R-peak information even when the original signal magnitude is small and contaminated by noise.
Solution Approach 2:
The patent changes the parameter space by computing complexity values (Shannon entropy) of wavelet-transformed signals at multiple decomposition levels. By transforming the detection criterion from direct amplitude thresholding to complexity value analysis, the system can reliably detect R-peaks in low-magnitude signals that would be indistinguishable from noise in the time domain.
2Measurement precision
If wavelet transformation is applied to improve peak detection accuracy, then measurement precision is improved, but device complexity increases due to multiple transformation levels and complexity calculations
Solution Approach 1:
The patent segments the ECG signal processing into distinct stages: pre-processing, wavelet transformation at multiple levels, complexity calculation for each level, and R-peak detection. This segmentation allows the system to apply computationally intensive wavelet transformation only to relevant signal portions and process different decomposition levels independently, managing complexity through structured modular processing.
Solution Approach 2:
The patent performs wavelet transformation at multiple decomposition levels (excessive action) to ensure robust peak detection, but practically implements a limited number of levels (partial action) based on signal characteristics. The system calculates complexity values for each level but uses thresholding and merging strategies to avoid processing all possible levels, balancing computational effort with detection accuracy.
3Measurement precision
If multiple wavelet transformation levels are processed to detect candidate peaks, then measurement precision is improved, but loss of time increases due to multiple complexity calculations
Solution Approach 1:
The patent performs pre-processing of the ECG signal before wavelet transformation, including baseline removal and normalization. This preliminary action prepares the signal for more efficient wavelet transformation and reduces the computational burden of subsequent complexity calculations, thereby reducing overall processing time while maintaining detection accuracy.
Solution Approach 2:
The patent implements a multi-level wavelet transformation strategy where not all decomposition levels are fully processed. The system skips levels that contribute minimally to R-peak detection based on signal energy distribution, rushing through the processing of less informative levels while maintaining thorough analysis at critical levels where R-peak information is concentrated.
Data Source
AI summary
One or more embodiments relate to a bio-signal measurement apparatus and a bio-signal measurement method for detecting peaks and a computer program for executing the method. Sensed ECG signal is wavelet transformed into one or more levels, and an effective peak of a bio-signal is detected using a complexity value of a signal surrounding the peak in the converted signals.