MULTIPLE SWITCHING ELECTROMYOGRAPHY (EMG) ASSISTIVE COMMUNICATIONS DEVICE
A method, human interface device, and computer program product that provide improved multilevel switching from each bioelectrical sensor with inclusion of switch filtering based on extraneous events (e.g., spasms). A biosignal is received from a sensor device by an electronic processor of a first electrode switch device. In response to determining that the amplitude of the signal has changed from less than a first switch range to greater than the first switch range and less than the second switch range, the electrode switch device communicates a first switch signal to control the human interface system. In response to determining that the amplitude of the biosignal has changed from less than the second switch range to greater than the second switch range, the electronic switch device performs one of: (i) ignoring the instance and (ii) transmitting a second switch signal to control the human interface system.
1 . A system for controlling a host device using biosignals, the system comprising:
at least one biosignal sensor configured to be coupled to a user;
at least one electronic processor operatively coupled to the biosignal sensor;
a memory storing instructions that, when executed by the electronic processor, cause the electronic processor to:
(i) receive a biosignal sample stream from the biosignal sensor;
(ii) compute at least one feature of the biosignal sample stream comprising an amplitude and/or a rate-of-change value;
(iii) compare the at least one feature to at least one activation threshold; and
(iv) in response to the at least one feature satisfying a hysteresis condition relative to the at least one activation threshold, issue a switch signal; and
an interface configured to provide the switch signal to a human-interface device (HID) of the host device.
2 . The system of claim 1 , wherein the biosignal sensor comprises at least one of: a set of electromyography (EMG) electrodes; an EEG sensor; an EOG sensor; a galvanic skin-response sensor; a microphone configured to sense air flow or phonation; a pressure transducer; an accelerometer; a gyroscope; or a magnetometer.
3 . The system of claim 1 , wherein the at least one feature further comprises at least one of: root-mean-square (RMS) magnitude, a spectral-band power, a zero-crossing rate, or a wavelet-domain coefficient.
4 . The system of claim 1 , wherein the hysteresis condition requires that the at least one feature cross above the at least one activation threshold after descending below a lower threshold.
5 . The system of claim 1 , further comprising enforcing a signal-off gating requirement that requires the at least one feature to fall below a reset level before a subsequent switch signal is issued.
6 . The system of claim 1 , wherein comparing comprises monitoring the at least one feature over a rolling time window and determining whether the at least one feature exceeds the at least one activation threshold in at least one instance within the rolling time window.
7 . The system of claim 1 , wherein the electronic processor is further configured to apply at least one of: band-pass filtering, notch filtering, windowed smoothing, or outlier rejection to reduce artifacts and noise.
8 . The system of claim 1 , wherein the electronic processor maintains a baseline estimate and a signaling-level estimate and adjusts at least one activation threshold based on a trend of at least one of the baseline estimate or the signaling-level estimate to compensate for user fatigue and/or sensor-interface changes.
9 . The system of claim 8 , wherein adjusting comprises applying an exponential moving average or an adaptive scale factor to the at least one activation threshold.
10 . A method of multi-level switching using biosignals, the method comprising:
establishing, by an electronic processor, a plurality of activation levels including a first activation level V 1 and at least one additional activation level VN, and further establishing an upper ignore level VU greater than VN;
receiving a biosignal sample stream from a biosignal sensor;
comparing a feature of the biosignal sample stream to the plurality of activation levels; and
mapping excursions of the feature into different switch commands when between V 1 and VN, while treating excursions at or above VU as non-volitional and ignoring them.
11 . The method of claim 10 , further comprising applying a hysteresis rule requiring a transition from below a lower level to above V 1 within a time interval for a switch to be valid.
12 . The method of claim 10 , wherein the plurality of activation levels map to respective commands comprising at least two of: cursor select, cursor back, dwell-start, dwell-end, or mode change.
13 . The method of claim 10 , wherein VU is set as a function of at least one of: a baseline level of the biosignal, a recent maximum of the feature, or a fraction or offset relative to VN.
14 . The method of claim 10 , further comprising dynamically updating at least one of V 1 or VN based on a recent history of the feature's amplitude and/or frequency content.
15 . A method of motion-gated spasm suppression for biosignal switching, the method comprising:
a) receiving, by an electronic processor, a biosignal from a biosignal sensor coupled to the user;
b) receiving, by the electronic processor, three-axis accelerometer samples from an accelerometer coupled to the user;
c) computing a movement metric from the accelerometer samples; and
d) allowing a biosignal-based switch only when the movement metric is below a programmable threshold, and otherwise disallowing the switch as a spasm.
16 . The method of claim 15 , wherein the movement metric comprises a product of x, y, and z acceleration components compared to a threshold selected for the user.
17 . The method of claim 15 , wherein the movement metric comprises a magnitude of a three-axis acceleration vector with orientation compensation.
18 . The method of claim 15 , further comprising dynamically updating the programmable threshold according to at least one of: time-of-day, historical spasm frequency, or recent activity level.
19 . The system of claim 15 , wherein the inertial-measurement data comprises six-axis data samples from accelerometer and a gyroscopic samples.
20 . The system of claim 15 , wherein the inertial-measurement data comprises nine-axis data samples from accelerometer, gyroscopic, and magnetometer samples.