Wireless Body Sensor Artifact Mitigation in Compressed Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing signal processing algorithms in wireless body area networks fail to effectively mitigate artifacts in biophysical signals, particularly motion and noise artifacts, which can lead to inaccurate interpretation of vital signs due to direct removal in signal space rather than compressed sensing space.
Innovation Solution
The method involves generating sampling instances based on a random seed, sensing motion information, and compressively sensing signal samples. Adjustments are made to sampling instances based on motion information or initial samples, allowing for improved artifact mitigation directly in the compressed sensing reconstruction domain, reducing aliasing and enhancing signal accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If artifact removal is performed directly in signal space using conventional algorithms, then the processing is simpler, but the accuracy of signal interpretation deteriorates due to aliasing from high-frequency components
Solution Approach 1:
The patent transforms the artifact removal operation from signal space to compressed sensing (CS) space by performing reconstruction with artifact mitigation in the CS domain. This dimensional transformation allows simultaneous achievement of low complexity (inherent to CS) and high accuracy (by avoiding aliasing in CS space)
Solution Approach 2:
The patent changes the fundamental parameter of where artifact removal is performed - from direct signal space operations to compressed sensing space operations. This parameter change enables the system to achieve both computational efficiency and superior signal interpretation accuracy by operating in the CS domain where aliasing is avoided
2Ease of manufacture
If conventional artifact detection algorithms are used, then the implementation is straightforward, but the robustness to acquisition errors deteriorates
Solution Approach 1:
The patent moves artifact removal from signal space to compressed sensing space, leveraging the inherent properties of CS reconstruction to provide robustness against acquisition errors while maintaining implementation feasibility through standard CS algorithms
3Measurement precision
If sampling is performed at high frequency to capture all components, then aliasing is reduced, but the energy consumption and data transmission load increase
Solution Approach 1:
The patent changes the sampling approach from traditional high-frequency Nyquist sampling to compressed sensing sampling. This parameter change enables accurate signal reconstruction at lower sampling rates by performing artifact mitigation in the CS domain, thereby reducing energy consumption and data transmission requirements while maintaining signal interpretation accuracy
Data Source
AI summary
Certain aspects of the present disclosure relate to a technique for mitigating artifacts of biophysical signals in a body area network. Information from multiple sensors (including motion information of the body) can be employed in mitigating the artifacts. The biophysical signals in the body area network can be compressively sensed.