Wireless soft scalp electronics and virtual reality system for brain-machine interfaces
An exemplary wireless soft scalp electronic system and method are disclosed that can actuate commands for a brain-machine interface (BMI) or brain-computer-interface (BCI) by performing real-time, continuous classification. e.g., via a trained neural network, of motor imagery (MI) brain signals or of steady-state visually evoked potential (SSVEP) signals. In some embodiments, the exemplary system is configured as a low-profile, portable system that includes microneedle electrodes that can acquire EEG signals for a brain-machine interface controller. The microneedle electrodes may be configured as soft imperceptible gel-less epidermis-penetrating microneedle electrodes that can provide improved contact surface area and reduced electrode impedance density, e.g., to enhance EEG signals and the signal classification accuracy. The microneedle electrodes can be further integrated with soft electronics that can be mounted locally in proximity to the electrodes to reduce obtrusive wiring and improve signal acquisition quality.
1. A system comprising:
a set of low-profile EEG sensors, each comprising an array of flexible epidermis-penetrating microneedle electrodes fabricated on a flexible-circuit substrate, the flexible-circuit substrate operatively connected to an analog-to-digital converter circuitry operatively connected to a wireless interface circuitry; and
a brain-machine interface operatively connected to the set of low-profile EEG sensors, the brain-machine interface comprising:
a processor; and
a memory operatively connected to the processor, the memory having instructions stored thereon, wherein execution of the instructions by the processor causes the processor to:
receive EEG signals acquired from the low-profile EEG sensor;
continuously classify brain signals as control signals via a trained neural network from the acquired EEG signals; and
output the control signals to a virtual reality environment controller to actuate a command in a VR scene generated by the virtual reality environment controller to be viewed by a subject.
2. The system of claim 1 , wherein the command causes a set of movements of an extremity in the VR scene, and wherein the trained neural network is configured to classify the brain signals for the set of movements.
3. The system of claim 1 , wherein the set of low-profile EEG sensors are connected to the brain-machine interface over a set of stretchable flexible connectors.
4. The system of claim 1 , wherein the microneedle electrodes have expanded contact surface area and reduced electrode impedance density.
5. The system of claim 1 , further comprising:
a wearable soft headset comprising a low-modulus elastomeric band.
6. The system of claim 1 , wherein the trained neural network comprises a spatial convolutional neural network.
7. The system of claim 1 , wherein the set of low-profile EEG sensors is placed along a scalp of the subject for motor imagery.
8. The system of claim 1 , wherein the set of low-profile EEG sensors is placed along a scalp of the subject for steady-state visually evoked potentials (SSVEP) measurements.
9. The system of claim 8 , wherein the virtual reality environment controller to configured to generate split-eye asynchronous stimulus (SEAS) in the virtual scene for a real-time text speller interface.
10. The system of claim 1 , wherein the execution of the instructions by the processor further causes the processor to:
transmit the acquired EEG signals to a remote or cloud computing device executing a retraining operation of the trained neural network; and
receive during run-time operation of the virtual reality environment controller an updated trained neural network from the remote or cloud computing device.
11. The system of claim 1 , wherein a plurality of the flexible epidermis-penetrating microneedle electrodes of the array each is at least 500 μm in height (e.g., 800 μm) to mount on a hairy scalp with a base width of about 350 μm and has an area of about 36 mm 2 .
12. A method comprising:
providing a set of low-profile EEG sensors placed at a scalp of a user, wherein the set of low-profile EEG sensors each comprises an array of flexible epidermis-penetrating microneedle electrodes fabricated on a flexible-circuit substrate, the flexible-circuit substrate operatively connected to an analog-to-digital converter circuitry operatively connected to a wireless interface circuitry;
receiving, by a processor or a brain-machine interface operatively connected to the set of low-profile EEG sensors, EEG signals acquired from the low-profile EEG sensor;
continuously classifying, by the processor, brain signals as control signals via a trained neural network from the acquired EEG signals; and
outputting, by the processor, the control signals to a virtual reality environment controller to actuate a command in a VR scene generated by the virtual reality environment controller to be viewed by a subject.
13. The method of claim 12 , wherein the set of low-profile EEG sensors are placed directly on the scalp without conductive gel or paste.
14. The method of claim 12 , wherein the set of low-profile EEG sensors includes i) a reference array of flexible epidermis-penetrating microneedle electrodes placed at an apex position on the scalp and ii) six arrays of flexible epidermis-penetrating microneedle electrodes releasably attached to a low-modulus elastomeric band at a first frontal position, a second back position, and at four side positions for motor imagery measurements.
15. The method of claim 12 , wherein the set of low-profile EEG sensors includes i) a reference array of flexible epidermis-penetrating microneedle electrodes placed at a back position on the scalp and ii) four arrays of flexible epidermis-penetrating microneedle electrodes releasably attached to a low-modulus elastomeric band at a back region of the scalp for steady-state visually evoked potentials (SSVEP) measurements.
16. The method of claim 12 , further comprising:
transmitting, by the processor, the acquired EEG signals to a remote or cloud computing device executing a retraining operation of the trained neural network; and
receiving, by the processor, during run-time operation of the virtual reality environment controller an updated trained neural network from the remote or cloud computing device.
17. A non-transitory computer-readable medium having instructions stored thereon, wherein execution of the instructions by a processor of a brain-machine interface controller causes the processor to:
receive EEG signals acquired from a set of low-profile EEG sensors placed at a scalp of a user, wherein the set of low-profile EEG sensors each comprises an array of flexible epidermis-penetrating microneedle electrodes fabricated on a flexible-circuit substrate, the flexible-circuit substrate operatively connected to an analog-to-digital converter circuitry operatively connected to a wireless interface circuitry, wherein the set of low-profile EEG sensors are placed directly on the scalp without conductive gel or paste;
continuously classify brain signals as control signals via a trained neural network from the acquired EEG signals; and
output the control signals to a virtual reality environment controller to actuate a command in a VR scene generated by the virtual reality environment controller to be viewed by a subject.
18. The computer-readable medium of claim 17 , wherein the set of low-profile EEG sensors includes i) a reference array of flexible epidermis-penetrating microneedle electrodes placed at an apex position on the scalp and ii) six arrays of flexible epidermis-penetrating microneedle electrodes releasably attached to a low-modulus elastomeric band at a first frontal position, a second back position, and at four side positions for motor imagery measurements.
19. The computer-readable medium of claim 17 , wherein the set of low-profile EEG sensors includes i) a reference array of flexible epidermis-penetrating microneedle electrodes placed at a back position on the scalp and ii) four arrays of flexible epidermis-penetrating microneedle electrodes releasably attached to a low-modulus elastomeric band at a back region of the scalp for steady-state visually evoked potentials (SSVEP) measurements.
20. The computer-readable medium of claim 17 , wherein the execution of the instructions further causes the processor to:
transmit the acquired EEG signals to a remote or cloud computing device executing a retraining operation of the trained neural network; and
receive during run-time operation of the virtual reality environment controller an updated trained neural network from the remote or cloud computing device.