Hardware-Assisted Noise Detection for Biosignals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing biosignal detection systems face challenges in real-time noise mitigation due to reliance on software-based algorithms that are limited in detecting external interference and memory constraints, leading to incomplete noise identification and potential errors in signal quality.
Innovation Solution
A hardware-assisted noise detection system that identifies noise types in real-time, using separate channels and algorithms to mitigate noise, allowing for immediate discarding of bad signal epochs and adjusting gain/filter settings to improve signal quality, while prompting users to address noise sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If software-based artifact detection algorithms (ICA or SSP) are used to identify noise artifacts, then noise detection capability is provided, but real-time noise mitigation is limited and memory constraints are imposed
Solution Approach 1:
The patent replaces software-based artifact detection algorithms with a hardware-assisted noise detection system that uses dedicated noise detection circuits to identify and mitigate noise in real-time, eliminating the memory constraints and processing delays inherent in software-based approaches
Solution Approach 2:
The patent introduces a separate noise detection channel that acts as an intermediary to detect noise artifacts independently from the biosignal channels, allowing noise to be identified and mitigated without interfering with the primary biosignal acquisition process
2Reliability
If separate noise detection channels are added to identify external interference, then noise immunity is improved, but device complexity increases
Solution Approach 1:
The noise detection circuits are designed to detect multiple types of noise artifacts (electromagnetic interference, power line noise, motion artifacts) using a unified hardware architecture, allowing a single system to handle diverse noise sources without requiring separate specialized circuits for each noise type
Data Source
AI summary
A noise detection sensor structure provides hardware-assisted real-time artifact detection on a live bio signal of interest while that signal is being recorded. The structure is configured to detect and analyze the type of external noise, and to use information about the noise type (such as noise level, frequency range of noise, noise source) to improve signal quality and usability of the bio signal. Different algorithms can be used for each noise type detected. In some example cases, for instance, the structure is configured to at least one of extract noise from measurement signal, synchronize measurement with noise signal, change the software filter or analysis algorithm, change gain/filter settings of noise mitigation circuitry, and prompt user to assist as needed, based on the detected noise. Bad signal epochs can be discarded prior to recording, to efficiently use storage. Separate channel(s)/sensor(s) can be used, with different filtering/gain schemes, for noise detection.