EEG Noise Estimation via Acoustic Envelope Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electroencephalogram measurements are susceptible to low-frequency electrical noises from electro-acoustic transducers, which are not effectively filtered out due to the belief that the frequency bands do not overlap, leading to interference and discomfort in wearable devices.
Innovation Solution
An electroencephalogram measurement apparatus that includes an electro-acoustic transducer, an amplitude envelope extraction section, a frequency analysis section, and a noise estimation section to identify and subtract electrical noise from the electroencephalogram signals using transform rules and frequency analysis, allowing for accurate noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If electrodes are placed in proximity to electro-acoustic transducers in wearable devices, then ease of operation and device integration are improved, but electrical noise from the transducers interferes with electroencephalogram measurement precision
Solution Approach 1:
The patent segments the signal processing into multiple stages: raw signal acquisition, noise component identification through frequency analysis, noise subtraction, and clean electroencephalogram extraction. This segmentation allows the system to handle noisy signals from proximity electrodes by separating the electroencephalogram components from transducer-induced noise components in the frequency domain
Solution Approach 2:
The patent introduces an intermediary signal processing system that includes frequency analysis means and noise subtraction means. This intermediary processing chain acts as a mediator between the noisy raw signals and the final clean electroencephalogram output, enabling the system to maintain both close electrode-transducer proximity and high measurement precision
2Object-affected harmful factors
If conventional frequency filtering is applied to remove electrical noise, then noise reduction is attempted, but low-frequency noise from electro-acoustic transducers cannot be effectively filtered due to frequency band overlap assumptions
Solution Approach 1:
The patent employs feedback mechanisms where the output of frequency analysis feeds into noise component identification, which then informs the noise subtraction process. The system continuously monitors the frequency characteristics and adjusts noise removal accordingly, creating a closed-loop feedback system that adapts to varying noise conditions while preserving electroencephalogram signal integrity
Solution Approach 2:
The patent changes the approach from simple frequency band filtering to a more sophisticated parameter-based noise removal method. By analyzing frequency characteristics, amplitude envelopes, and temporal patterns, the system dynamically identifies and removes noise components based on multiple parameters rather than relying on fixed frequency band assumptions, enabling effective low-frequency noise removal
Data Source
AI summary
An electroencephalogram measurement apparatus includes: an electroencephalogram measurement section for measuring an electroencephalogram of a user by using a plurality of electrodes; an electro-acoustic transducer for presenting an acoustic signal to the user, the electro-acoustic transducer being in a vicinity of at least one electrode among the plurality of electrodes while the electroencephalogram measurement section is worn by the user; an amplitude envelope extraction section for extracting an amplitude envelope of the acoustic signal presented by the electro-acoustic transducer; a frequency analysis section for applying a frequency analysis to the amplitude envelope extracted by the amplitude envelope extraction section; and a noise estimation section for estimating an electrical noise which is mixed at the at least one electrode by using a previously provided set of transform rules and the extracted amplitude envelope.