Questions about how individual parts of claim 1 might be changed. A question may describe something different from what the claim language recites, or a version of it; no view is offered on which.
1.1Peripheral sleep state sensing
What ifWhat if the wearable acquired wrist pulse, motion, and respiratory signals to estimate sleep state?
Claim language“acquire EEG”
An optical wrist sensor records pulse intervals, a three-axis accelerometer records movement, and a stretch belt records respiratory effort. A classifier uses pulse variability, movement counts, and breathing regularity to estimate sleep state over 30-second windows.
1.3Continuous slow-wave tracking
What ifWhat if the processor tracked slow-wave activity continuously and operated without assigning sleep stages?
Claim language“classify sleep stage from the EEG”
The processor filters the EEG and computes a rolling ratio of slow-wave power to broadband power. The controller uses the ratio, its rate of change, and a movement flag to decide when to enable stimulation.
1.4Stimulation during established N3
What ifWhat if the controller waited for established N3 before enabling stimulation?
Claim language“detect a N2-to-N3 stimulation opportunity”
A stage classifier requires two consecutive 30-second epochs scored as N3 before it enables the output. The controller then uses the filtered frontal EEG to time stimulation to individual slow waves and suspends output when the classifier exits N3.
1.5Phase-timed acoustic pulses
What ifWhat if the wearable used brief sounds to reinforce ongoing slow waves?
The controller estimates slow-wave phase from frontal EEG and triggers a pillow speaker near the predicted positive peak. The speaker emits short pink-noise bursts, and the controller pauses delivery when EEG or motion signals indicate arousal.
1.6Stopping by output count
What ifWhat if the controller ended each stimulation block after a preset number of delivered bursts?
Claim language“compute a N3-discharge index”
The output circuit acknowledges each completed burst, and a digital counter increments on each acknowledgment. When the counter reaches its preset limit, the controller disables output and starts a timed interval with stimulation suspended.
Section 2 · machine-generated
Variations
Further arrangements of the same subject matter, described here as of the publication date at the foot of this page.
Other field of use
Sleep modulation in shift work studies
A wearable EEG and stimulation headset operates in a controlled shift work study during scheduled daytime sleep. The controller uses elapsed time from EEG-detected sleep onset to track the session and enables slow-oscillatory stimulation when the EEG indicates a transition from N2 toward N3. Once the N3 index reaches its threshold or the dose limit is reached, the controller withholds stimulation and records latency to the next REM episode, its duration, and any interruptions. The study software uses those REM measurements and N3 duration to select the stimulation schedule for the next daytime session.
Component substitution
Hidden markov model for sleep staging
A hidden Markov model classifies EEG epochs using delta power, spindle density, and other spectral features as its observations. The hidden states represent wake, N1, N2, N3, and REM, and a transition matrix assigns probabilities to changes between stages. The controller enables stimulation when the current N2 probability and the predicted N3 probability both exceed configurable thresholds. After the N3 index or dose limit ends stimulation, the classifier continues tracking stages and supplies subsequent REM duration and continuity to the policy update.
Hydrogel contacts in a textile headband
A stretch textile headband holds separate snap-mounted silver–silver chloride EEG contacts and conductive hydrogel stimulation pads. The stimulation pads sit in anterior and suboccipital pockets, and insulated leads connect them to a current-controlled output while separate shielded leads connect the EEG contacts to the acquisition front end. The controller verifies contact impedance before enabling slow-oscillatory stimulation and suspends delivery when a pad exceeds its configured impedance limit. It withholds further stimulation after the N3-discharge criterion or dose limit is reached and uses EEG from the same headband to monitor subsequent REM.
Parameter or scale variation
Three window confirmation of N3 discharge
The controller evaluates the N3-discharge index in trailing 60-second EEG windows advanced every 15 seconds. It marks a window positive when both the fraction of epochs classified as N3 and median slow-wave power exceed their respective thresholds, and it resets the confirmation count on any negative window. Three consecutive positive windows or the configured session dose limit ends stimulation and starts the REM-permissive interval. The controller uses cumulative N3 time and the duration of the first subsequent REM episode to select a policy for the next session.
Combination with existing systems
Integration with overnight respiratory monitoring
The wearable exchanges timestamped data with a polysomnography recorder that acquires nasal airflow, thoracic and abdominal effort, and pulse oximetry. A respiratory event detector sends the stimulation controller an inhibit signal during scored apneas and hypopneas, and the controller requires a new N2-to-N3 opportunity before resuming. When the N3-discharge criterion or dose limit is reached, the controller withholds stimulation and continues monitoring REM. The policy update uses N3 duration and subsequent REM continuity alongside the recorded respiratory events to distinguish interruptions that coincide with respiratory disturbances.
Automated or data-driven variant
Constrained policy selection across sleep sessions
An on-device contextual bandit selects an N2-to-N3 trigger threshold and a session dose limit from a table of preset combinations. Its inputs include baseline N3 duration, recent sleep-onset latency, and the previous session's REM continuity. The stimulation controller keeps the selected settings for the session, withholds stimulation when the N3-discharge index or dose limit reaches its threshold, and scores the result using N3 duration and subsequent REM duration with a penalty for EEG-detected arousals. The bandit updates the score estimate for the selected combination after each session with usable N3 and REM recordings.
Specific version of a part
Frontal dry electrode acquisition
Two spring-loaded conductive pads on a fabric headband form a differential frontal EEG channel. A high-input-impedance amplifier feeds an analog anti-alias filter and a converter that samples at 250 samples per second.
Frontal and mastoid current path
Conductive hydrogel pads contact the forehead and the skin over a mastoid process. An isolated constant-current output drives the electrode pair and checks the voltage across it during delivery.
Timed suspension after threshold
The controller ramps the output amplitude to zero when the N3-discharge index reaches its threshold or the dose counter reaches its limit. It then holds the output at zero for a configurable interval while EEG acquisition and sleep staging continue.
Weighted N3 and REM scoring
At the end of each night, the processor calculates a weighted sum of standardized N3 duration and subsequent REM bout length. It ranks tested timing and dose settings by their mean score across nights and selects the highest-scoring setting for the next session.
Section 3 · quoted from USPTO records
Other work
Passages reproduced word for word from two earlier-published U.S. patent documents, placed beside the language of claim 1. No characterization is offered. Each passage is marked A or B for the document it comes from; ¶ numbers are the paragraph numbers in the USPTO’s published full text of that document.
1.1an EEG front end configured to acquire EEG from a sleeping subject
In a preferred embodiment, a circuit contains a microcontroller, optionally with additional on-board RAM or flash memory, which digitizes EEG data acquired through one or more ADCs (and optional upstream amplifier chains), and then spectrally processes the EEG data to determine which sleep cycle the subject is in.
A · US 2008/0319505 A1 · ¶ 20
1.2a transcranial electrical stimulation circuit coupled to at least one anterior electrode and at least one posterior, nuchal, suboccipital, occipital, mastoid, or neck electrode
The prototype implementation is adapted to record from 1-4 EEG electrodes, to do spectral analysis of EEG signals between 0 and 200 Hz, and to deliver 1 Hz oscillatory current amplitude from two frontal electrodes (with two mastoid electrodes providing the current return path) at the onset of the first epoch of slow wave sleep during a sleep episode.
A · US 2008/0319505 A1 · ¶ 26
1.3one or more processors configured to: classify sleep stage from the EEG
1.4detect a N2-to-N3 stimulation opportunity
The device then induces a programmable, oscillating current through one or more output DACs (and optional downstream amplifiers) across the scalp, triggered by the occurrence of a particular spectral signature, the onset of the first epoch of slow-wave sleep in a sleep period.
A · US 2008/0319505 A1 · ¶ 20
1.5deliver slow-oscillatory transcranial electrical stimulation through a limbic-targeted electrode montage during the N2-to-N3 stimulation opportunity
From oscillatory transcranial current stimulation to scalp eeg changes: a biophysical and physiological modeling study.
B · US 2017/0340882 A1 · ¶
1.6compute a N3-discharge index from one or more N3 or slow-oscillation features
1.7after the N3-discharge index reaches a criterion or after a maximum stimulation dose is reached, terminate, reduce, or withhold further N3-promoting stimulation and enter a REM-permissive interval
1.8monitor at least one subsequent REM feature during the same night
1.9update or select a stimulation policy according to a dual-stage sleep-architecture objective comprising at least one N3 feature and the at least one subsequent REM feature.
An empty cell means only that no passage was reproduced there.
Statements in these documents
In addition, such a device may be used for consumer-targeted memory augmentation applications, as well as enabling a wide variety of treatments for disorders ranging from epilepsy to stroke to Parkinson's disease.
A · US 2008/0319505 A1 · ¶ 4
There is no device that has combined these two powerful technologies, nor have they been combined in a small, portable, inexpensive format.
A · US 2008/0319505 A1 · ¶ 5
For this reason, optimization algorithms based on globally defined, cortical targeting data are needed.
B · US 2017/0340882 A1 · ¶ 4
While conventional TMS and tDCS technologies allow for only one or two stimulation sites, the multi-electrode approach perfectly complements this scientific and therapeutic need.
B · US 2017/0340882 A1 · ¶ 6
Cite aspatent.fyi, US 2026/0355514 A1, claim 1, first published 2026-11-19T12:21:55-05:00, https://patent.fyi/US20260355514/, SHA-256 40ebf9b8b0bd70bf82a3e484096ba9a9d817846e1e1e9c3d752077cdd91d2ef7