Chest Video Respiration Sensing With Vibration Noise Subtraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to accurately determine respiratory information in automotive applications due to noise interference from non-respiratory chest displacements caused by vibrations, making it difficult to measure vertical chest displacements effectively.
Innovation Solution
A method that utilizes a chest video signal to compute both chest displacement and deformation signals, estimates noise contribution from non-respiratory sources, and subtracts this noise to obtain a noise-reduced chest displacement signal, using techniques like spectral subtraction and adaptive filtering to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If vertical chest displacements are used directly for capturing respiration, then the measurement approach is simple, but the measurement precision deteriorates due to noise from vibrations
Solution Approach 1:
The patent introduces an intermediary processing system that includes: (1) obtaining both vertical displacement signals and deformation signals from the chest area, (2) computing a noise estimate by combining these signals through cross-spectral analysis, and (3) removing the estimated noise from the displacement signal. This intermediary processing chain acts as a mediator between the simple displacement measurement and the accurate respiration detection, filtering out vibration noise while preserving respiratory information.
Solution Approach 2:
The patent combines multiple signal types (vertical displacement signal and deformation signal) to create a composite measurement approach. By fusing these different signal modalities and processing them together through spectral analysis and noise estimation, the system achieves more accurate respiration detection than would be possible with displacement signals alone, analogous to using composite materials to achieve superior properties.
2Measurement precision
If noise removal processing is applied to the chest displacement signal, then the measurement precision improves, but the device complexity increases
Solution Approach 1:
The system performs self-service by using the chest's own deformation signal to generate the noise estimate for removing noise from the displacement signal. The deformation signal, which contains information about chest mechanics, is used to characterize and remove noise from the displacement measurement, allowing the system to clean its own data without requiring external reference measurements or complex additional sensors.
Solution Approach 2:
The patent implements a feedback mechanism where the computed noise estimate is fed back into the signal processing chain to remove noise from the displacement signal. The cross-spectral analysis and noise estimation create a closed-loop system that continuously refines the respiration signal by subtracting the estimated noise component, improving measurement precision through iterative signal refinement.
Data Source
Figure 1~3
Figure 4~5
Figure 6
AI summary
The present invention relates to a method, device and system for determining respiratory information of a subject. To enable the determination of respiratory information in a more robust and precise way, the method comprises obtaining a chest video signal of the subject, the chest video signal including a time series of images of the subject acquired over time and covering at least part of the subject's chest area; computing a chest displacement signal from the obtained chest video signal by determining vertical displacements of the chest in a direction substantially along or parallel to the subject's longitudinal body axis; computing a chest deformation signal from the obtained chest video signal by determining deformations of the chest; computing a noise-estimate related to the non-respiratory noise contribution in the chest displacement signal using the chest displacement signal and the chest deformation signal; removing at least part of the non-respiratory noise contribution in the chest displacement signal by subtraction of the noise-estimate from the chest displacement signal to obtain a noise-reduced chest displacement signal; and determining the respiratory information from the noise-reduced chest displacement signal.