BrainWAVE: A Flexible Method for Noninvasive Stimulation of Brain Rhythms across Species

BrainWAVE circuit design. A, Flicker presentation for mice in a clear enclosure with a strip of light-emitting diodes (LEDs) and a speaker. B, Flicker presentation for humans with LED-lined goggles and headphones. C, Flicker presentation for humans with an LED-frame surrounding a computer monitor and earbuds. D, A circuit diagram of an audio flicker circuit with an audio amplifier to allow for volume adjustments. E, A circuit diagram of an LED visual flicker circuit with a metal–oxide–semiconductor field-effect transistor (MOSFET) to allow a 12-V voltage source to power the LEDs. F, Left, An Arduino Uno microcontroller (top) and Arduino shield (bottom). A MOSFET and wires are soldered on the Arduino shield. Middle, The Arduino shield fits on top of the Arduino Uno and both sit in a custom 3D-printed case. Right, The Arduino and BrainWAVE stimulator circuit shield are enclosed in a 3D-printed case. Ports in the case allow a USB Type-B cable, an LED cable, and a power cable to plug into the Arduino and shield. See Extended Data Table 1-1 for a list of BrainWAVE stimulator parts. See Extended Data Figure 1-1 for a physical diagram of an Arduio-Visual BrainWAVE circuit. See Extended Data 2 for detailed instructions on how to assemble the device.

A user-friendly BrainWAVE graphical user interface (GUI) to perform many types of experiments. We developed a user-friendly application to run a variety of experiments involving visual and/or auditory stimulation. Preprogramed tasks are found under the task dropdown menu and include four different tasks. First, a classical flicker task, with exposure to 5.5 Hz (θ-like), 40 Hz (γ-like), 80 Hz, and random nonperiodic flicker at visual, audiovisual and auditory modalities. Second, a flicker duration task, exposing subjects to a given modality and frequency of flicker for minutes at a time. Third, a flicker frequency task, which allows exposing subjects to up to 26 different frequencies of flicker of a given modality. Fourth, a single pulse evoked potential task, where subjects are exposed to single visual, audiovisual and auditory 12.5-ms pulses. The stimuli parameters are set in entry boxes for stimulus duty cycle and tone (sound frequency). The comments box is used to write and save time-stamped experiment notes during the experiment. Developed for testing in human participants, each task includes tests for comfort to determine the optimal brightness and volume of the stimuli that are comfortable to the subject (adjusted on the device), tests for safety to determine whether the intended flicker stimuli induce adverse events, experimental tasks, control occluded condition (where subjects wear a sleep mask and earplugs), and measures of brightness and volume used. See Extended Data 2, BrainWAVE stimulator guide, for instructions on how to set-up and run an experiment using the BrainWAVE GUI.

Measuring BrainWAVE stimulator output signals. A, Visual and auditory stimuli generated by signal generators are measured with a light meter and decibel meter, respectively. Connecting these devices to an oscilloscope allows visualization and quantification of frequency and duty cycle. B, Ideal 40-Hz visual signal with a 50% duty cycle. C, Ideal 40-Hz audio signal (8-kHz tone) with a 50% duty cycle. Inset, a zoomed-in view of the signal at the stimulus onset showing an 8-kHz sinusoid. D, Ideal 40-Hz visual (blue) and audio (orange) signals in phase, both with a 50% duty cycle. E, Example of an ideal 40-Hz visual signal (blue) with a 50% duty cycle in phase with a 40-Hz audio signal (orange) with a 4% duty cycle. F, Example of a real 40-Hz visual signal (dark blue) and its light output (light blue) recorded with an oscilloscope. G, Example of a real 40-Hz audio signal (dark orange) and its audio output (light orange) recorded with an oscilloscope. H, Same as F, but of a BrainWAVE stimulator with an inverted MOSFET, which turns the lights off with a high-voltage signal and turns LEDs on with a low-voltage signal. I, Same as F, but with a BrainWAVE stimulator that uses a light source with a slower onset time. A delay is observed in the onset timing between the visual signal (dark blue) and measured output (light blue). See Extended Data Table 3-1 for recommended light and sound levels.

Sensory flicker entrains brain regions in humans and rodents. A, Top, Schematic of a typical electroencephalogram (EEG) recording setup. Participants with EEG caps sat in front of a computer monitor (∼2 feet away) framed by LEDs and received auditory stimulation through earbuds. Bottom, Example EEG trace recorded from the center electrode (Cz) during flicker stimulation. B, Left, Power spectral density averaged across all channels in an example EEG recording during baseline (top) and 40-Hz audiovisual flicker stimulation (bottom). Right, Heatmaps of mean power at 40 Hz averaged across subjects before (top) and after (bottom) 40-Hz stimulation. C, Mean power at 40 Hz during baseline (“Baseline”) and during 40-Hz audiovisual flicker stimulation (“40 Hz”), and mean power at neighboring frequencies (31–39 and 41–49 Hz) during 40-Hz audiovisual stimulation (“Around 40 Hz”). Mean power at 40 Hz is significantly higher during 40-Hz stimulation than power at 40 Hz during baseline (n = 10 participants, p = 0.002, two-sided Wilcoxon signed-rank test) or power at neighboring frequencies (n = 10 participants, p = 0.014, two-sided Wilcoxon signed-rank test). Each colored line represents a single participant. * p ≤ 0.05, ** p ≤ 0.01. D, Example setup of a human intracranial recording during flicker stimulation. Local field potential recordings were obtained from treatment-resistant epileptic patients implanted with intracranial electrodes for seizure monitoring. A computer controlled the delivery of sensory stimuli via a custom-made BrainWAVE stimulator circuit, which sent its output to a set of LED-lined goggles and earbuds. E, Example of modulation to visual flicker recorded in lingual gyrus. Left, Axial slice of preimplant T1 MRI overlaid with postimplant computerized tomography (CT) scan, showing the location of the recording depth electrode. Highlighted in red is the electrode for data on right. Example recording trace before and during stimulation where the start of stimulation is indicated with a dashed line (below). Right, Power spectral density averaged across 15 trials of either 40-Hz visual (blue) or random visual stimulation (green). Shaded areas represent SEM (n = 15 trials in 1 participant, p = 0.00005, paired t test). F, Top, Schematic of in vivo electrophysiology in head-fixed mice. Mice running on a spherical treadmill received sensory flicker stimulation through a strip of LEDs placed above the mouse and a speaker to the right. Bottom, Example trace of local field potentials in mouse auditory cortex during 40-Hz audiovisual flicker. G, Power spectral density comparison between 40-Hz flicker stimulation (blue) and baseline (gray), and between 40-Hz flicker and random (green) condition. H, Firing rate modulation during 40-Hz audiovisual stimulation (left), baseline (middle), and random (right) conditions in mouse hippocampus. Colors above indicate if light was on (yellow), off (black), or varied trial-to-trial (gray).