Gamma-Signal Human-Machine Interface for Sub-Second Actuator Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current non-invasive brain-machine interfaces (BMIs) are limited by low information transfer rates, primarily using slower rhythms, which restrict their ability to control actuators with sub-second reaction times, and they often ignore extracranial signals that could enhance their capabilities, posing ethical and health risks with invasive methods.
Innovation Solution
A method utilizing non-invasive devices to acquire and process gamma frequency brain and muscle signals, amplifying and filtering these signals to extract spectral features, which are then used to control actuators with real-time feedback, enabling faster and more versatile control of a wider range of tasks without invasive techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If non-invasive brain-machine interfaces use slower rhythms (alpha, beta bands) for signal acquisition, then the system maintains safety and avoids invasive procedures, but the information transfer rate remains limited and real-time control capability is reduced
Solution Approach 1:
The patent changes the frequency parameter from traditional slow rhythms (alpha, beta bands) to fast gamma rhythms (30-200 Hz). This parameter change enables non-invasive acquisition of high-frequency signals that carry more information, thereby increasing the information transfer rate while maintaining the safety of non-invasive procedures. The system acquires gamma rhythm signals from the brain and muscle signals from extracranial sources, processes these high-frequency signals in real-time, and uses them to control actuators with sub-second reaction times.
2Object-affected harmful factors
If non-invasive devices are used to acquire brain signals, then ethical and health risks are avoided, but the signal quality and information content are limited compared to invasive methods
Solution Approach 1:
The patent merges two signal sources: brain signals (electroencephalogram EEG) and muscle signals (electromyogram EMG). By combining these extracranial signals, the system compensates for the limited information content of individual non-invasive signals. The brain signals provide cognitive and motor intent information, while the muscle signals provide additional motor execution information, together enabling rich real-time control of actuators without invasive procedures.
3Device complexity
If traditional slow rhythm signals are used for actuator control, then the system remains simple to implement, but the reaction time exceeds sub-second requirements for real-time control
Solution Approach 1:
The patent utilizes the natural periodic oscillations of gamma rhythms in the brain and muscle signals. These high-frequency periodic signals occur naturally at 30-200 Hz, providing inherent temporal structure that enables fast processing. The system processes these periodic signals through spectral analysis and feature extraction to generate control commands with sub-second reaction times, maintaining relatively simple implementation while achieving real-time control performance.
Data Source
AI summary
This invention relates to a method to control at least one actuator to carry out at least one task comprising repetitive sequences of operations, shorter than 1000 milliseconds by acquiring of endogenously generated electrical potentials within gamma frequency range generated by the brain and/or by the muscles of the body, head, or eyes of a human user, by acquiring electrical signals with sampling rate of at least 250 samples/s, by extracting the features of the signals related to high-frequency gamma waves between 30-200 Hz, by classifying, mapping and converting the signals into actions in real time, and by sending feedback and neurofeedback to a human user to enable him to control the interface.