Method for detecting motion artifacts in respiratory signals

Abstract

According to the method for detecting the motion artifacts in the respiratory signals, the difference value between the upper envelope and the lower envelope of the respiratory signals is calculated, the threshold value is obtained in a self-adaptive mode in combination with the sliding window technology, and therefore accurate detection of the motion artifacts is achieved. The method is simple, convenient, rapid and efficient, the motion artifacts can be accurately detected, and a solid foundation is provided for subsequent artifact filtering.

Description

Technical Field

[0001] The present application relates to a respiratory signal detection technology, and in particular to a method for detecting motion artifacts in a respiratory signal. Background Art

[0002] Motion artifacts in respiratory signals refer to errors or noise introduced by movement or physical activity when collecting or recording respiratory signals. Such artifacts can interfere with the analysis and processing of signals, affecting the accurate identification and analysis of respiratory patterns, which is particularly important in health monitoring, disease diagnosis, and biomedical engineering.

[0003] Motion artifacts in respiratory signals mainly come from human body movement and changes in the position of the device. Specifically, they include changes in body position (such as standing, sitting, and lying), strenuous exercise (such as running, jumping, etc.), improper wearing of the device (such as loose sensors or changes in position), and external environmental interference (such as changes in temperature and humidity, air flow, etc.). These factors will cause chest or abdominal movement, which will cause distortion or noise interference to the respiratory signal, affecting the accuracy and stability of the signal.

[0004] At present, the detection methods of motion artifacts in respiratory signals have made certain progress, and the research mainly focuses on technologies such as signal processing, machine learning, and multi-signal fusion. Common methods include using filtering techniques (such as low-pass and band-pass filtering) and signal decomposition (such as wavelet transform, principal component analysis, and empirical mode decomposition) to remove artifacts; at the same time, combining motion sensors (such as accelerometers) with synchronous acquisition of respiratory signals, and performing artifact compensation through real-time motion monitoring. In addition, machine learning and deep learning methods (such as support vector machines, convolutional neural networks, and long short-term memory networks) have also been applied to the automatic identification and removal of motion artifacts. Although certain results have been achieved, due to the complexity and diversity of motion artifacts, how to improve the real-time, accuracy, and adaptability of the method remains a major challenge in this field. Summary of the invention

[0005] In view of the above problems, the present application aims to propose a method for detecting motion artifacts in a respiratory signal.

[0006] The method for detecting motion artifacts in a respiratory signal of the present application comprises: Extract the upper envelope and lower envelope of the breathing signal obtained during the subject's sleep; Calculating the difference between the extracted upper envelope and the lower envelope to obtain an envelope difference signal; A sliding window is added to the envelope difference signal; for the envelope difference signal in each sliding window, a threshold is used for judgment, and the part of the envelope difference signal that is greater than the threshold is the position of the motion artifact in the sliding window.

[0007] Preferably, the threshold is Perform calculations; in, is the threshold value, is the envelope difference signal within the sliding window, is the window width of the sliding window.

[0008] Preferably, the window width of the sliding window is ,in is the sampling rate of the breathing signal.

[0009] The method for detecting motion artifacts in respiratory signals of the present application calculates the difference between the upper envelope and the lower envelope of the respiratory signal and adaptively obtains the threshold value in combination with the sliding window technology, thereby realizing accurate detection and positioning of motion artifacts. The patented method is simple, fast and efficient, and can accurately detect motion artifacts, providing a solid foundation for subsequent artifact filtering. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 Schematic diagram of respiratory signals and motion artifact detection recorded using a smart chest strap during sleep; Figure 2 Schematic diagram of the respiratory signal envelope difference and threshold in the sliding window. DETAILED DESCRIPTION

[0011] The motion artifacts during sleep are mainly the turning over motion artifacts. ( Figure 1 black line), and the upper envelope ( Figure 1 red line) and the lower envelope ( Figure 1 blue line) and then calculate the difference ( Figure 1 green line).

[0012] For the envelope difference , it is necessary to find a suitable threshold to detect the turning motion artifact. The present invention proposes a sliding window adaptive threshold method to detect the endpoint of the motion artifact. Specifically, Figure 2 As shown, the envelope difference Signal plus sliding window, the window width is set to (Fs is the sampling rate of the respiratory signal), there is no overlap between the sliding windows. Figure 2 (a) is the envelope difference signal within the window, and then for each envelope difference within the window Signal , remove the mean and take the absolute value (Formula 1) ( Figure 2 (b) ), use formula 2 to calculate the threshold of each window . Figure 2(c) The motion signal measured by the accelerometer during sleep can be used as the gold standard for motion signal detection. , we can determine the location of the turning motion artifact, and the area greater than the threshold is the area where the artifact is located. Figure 2 It can be clearly found that the motion artifact detection method proposed in this patent has good coincidence between the motion artifact segments detected by the motion artifact detection method and the motion artifact measured by the accelerometer.

[0013] (Formula 1); (Formula 2).

[0014] The motion artifact detection method proposed in this application has significant advantages. On the one hand, it is simple to implement and has low computational complexity; on the other hand, the method is not only suitable for rapid detection of motion artifacts in long-term respiratory signals, but also has good applicability in real-time motion artifact detection.

[0015] Unless otherwise defined, all technical and / or scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which the invention relates. The materials, methods and examples mentioned in this application are illustrative only and not limiting.

[0016] Although the present invention has been described in conjunction with specific embodiments, within the spirit of the invention of this application, those skilled in the art may make appropriate substitutions, modifications and changes, and such substitutions, modifications and changes still fall within the scope of protection of this application.

Claims

1. A method for detecting motion artifacts in a respiratory signal, comprising: Extract the upper envelope and lower envelope of the breathing signal obtained during the subject's sleep; Calculating the difference between the extracted upper envelope and the lower envelope to obtain an envelope difference signal; A sliding window is added to the envelope difference signal; for the envelope difference signal in each sliding window, a threshold is used for judgment, and the part of the envelope difference signal that is greater than the threshold is the position of the motion artifact in the sliding window.

2. The method for detecting motion artifacts in a respiratory signal according to claim 1, characterized in that: The threshold is passed Perform calculations; in, is the threshold value, is the envelope difference signal within the sliding window, is the window width of the sliding window.

3. The method for detecting motion artifacts in a respiratory signal according to claim 1, characterized in that: The window width of the sliding window is ,in is the sampling rate of the breathing signal.

Images