EMG Onset Detection via Dynamic Thresholding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electromyography processing methods fail to accurately remove noise caused by changes in skin conditions and electrode positions during repetitive movements, leading to inaccurate muscle movement analysis in athletes.
Innovation Solution
An electromyography processing apparatus that calculates root-mean-square values and uses sliding windows to detect onset sections and determine dynamic threshold values, effectively filtering out noise and improving muscle movement analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a fixed threshold value determined from static electromyography data is used for onset detection, then the detection method is simple and fast, but the accuracy deteriorates due to noise caused by changes in skin conditions and electrode positions during exercise
Solution Approach 1:
The patent applies dynamics by transitioning from a fixed threshold value to a time-varying threshold value that adapts to changing noise conditions during exercise. The threshold is continuously updated based on the current noise level estimated from the electromyography signal, allowing accurate onset detection despite changes in skin conditions and electrode positions over time
Solution Approach 2:
The patent changes the parameter of the threshold value from a static constant to a dynamic parameter that varies with time and noise conditions. By estimating the noise level at different time points and adjusting the threshold accordingly, the system maintains high detection accuracy throughout the exercise duration without requiring complex additional hardware
2Measurement precision
If the sliding window length for threshold value detection is increased, then the noise estimation becomes more accurate, but the response time to detect actual muscle onset increases
Solution Approach 1:
The patent makes the sliding window length a dynamic parameter that adapts based on the current exercise state. During periods of high noise variability, a longer window provides better noise estimation, while during stable periods or when rapid onset detection is needed, the window length is reduced to improve response speed. This dynamic adjustment resolves the contradiction between estimation accuracy and response time
Solution Approach 2:
The patent uses partial action by applying different sliding window lengths for different detection purposes. A longer window is used specifically for noise estimation when needed, while a shorter window is used for rapid onset detection, thus achieving both accurate noise characterization and fast response without requiring the full long window for both purposes simultaneously
Data Source
AI summary
An electromyography processing apparatus comprises a storage device that stores the electromyography data of the predetermined muscle; and an onset detection unit configured to determine that a portion is an onset portion based on the electromyography of a sliding window for onset detection and a threshold value; wherein the onset detection unit further comprises a threshold value determination unit configured to determine the threshold value based on the electromyography of a sliding window for threshold value detection.