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.
1.1Sleep timing from brain activity
What ifWhat if scalp electrical signals determine the sleep period?
Claim language“based on first motion data”
Scalp electrodes record brain activity, and a processor labels 30-second epochs as awake or asleep from their waveforms. The processor sets the period boundaries at the first and last sleep epochs in the overnight recording. A separate motion sensor supplies data for the restlessness calculation within those boundaries.
1.2Oxygen drops as sleep data
What ifWhat if the sleep fitness data track blood oxygen drops during sleep?
Claim language“based on a restlessness associated with the user”
A finger sensor estimates blood oxygen saturation once per second using red and infrared light. For each analysis window, the processor records the largest drop from the preceding 60-second median and the number of seconds spent at least three percentage points below that median.
1.3Fixed thresholds for breathing classification
What ifWhat if the processor applies the same fixed thresholds to each sleeper's current sleep fitness data?
Claim language“historical information associated with the user”
The processor calculates seconds of restlessness and counts movement bursts in each 30-second window. It labels a window as disturbed breathing when it contains at least three bursts and at least ten seconds of restlessness, using the same preset rule for every recording.
1.3Continuous scoring from airflow reduction
What ifWhat if the sleep fitness score comes from the accumulated magnitude of airflow reduction across the sleep period?
Claim language“a classification of a plurality of temporal windows”
A nasal pressure sensor supplies a waveform from which the processor estimates the breathing amplitude envelope. The processor integrates the proportional deficit below the sleeper's reference amplitude and divides the result by the sleep duration. It maps that average deficit to a 0–100 score, with larger deficits producing lower scores.
1.4Spoken sleep score
What ifWhat if a speaker delivers the sleep fitness score as spoken audio?
Claim language“for presentation on a display device”
After the sleep period ends, a processor converts the numeric score to a spoken phrase using a speech synthesizer. A digital-to-analog converter and amplifier drive a bedside speaker when the user presses a playback button.
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 monitoring in veterinary recovery kennels
A pressure mat beneath the kennel bedding records movement and pressure changes from a recumbent animal. A processor identifies sleep intervals from sustained low movement and counts posture shifts during each interval to quantify restlessness. It classifies breathing windows using pressure oscillations, the restlessness measure, and baseline recordings from the same animal. A kennel display shows the percentage of classified windows with disturbed breathing.
Component substitution
Bed motion sensing with piezoelectric film
A piezoelectric film strip beneath the mattress replaces a body-worn accelerometer and feeds a charge amplifier connected to a bedside processor. The processor uses sustained low movement to identify sleep intervals and counts large transients during those intervals as a measure of restlessness. It extracts breathing cycle intervals from smaller oscillations and classifies each window using their variability, the restlessness measure, and the sleeper's historical baseline. The display shows the proportion of classified windows with regular breathing.
Parameter or scale variation
Overlapping windows with coverage weighting
A wrist accelerometer samples motion at 25 hertz, and a processor groups the samples into 30-second epochs to identify the sleep period. Within that period, the processor measures restlessness as the fraction of seconds containing movement bursts and evaluates breathing motion in 60-second windows spaced 15 seconds apart. It classifies each window using breathing rhythm variability, the restlessness measure, and the preceding 14 nights of usable recordings. For each 15-second interval, it averages the binary disturbance classifications of the windows that cover it and displays the mean of those interval values as a percentage.
Combination with existing systems
Sleep scoring with positive airway pressure
A bedside processor receives wrist accelerometer data and timestamped airflow and mask leak measurements from a positive airway pressure device. It aligns the streams on a common timeline and uses wrist motion to identify sleep intervals and quantify restlessness. For each window, it compares airflow variability and reductions in breath amplitude with the user's historical values while accounting for mask leak and restlessness. The display places the nightly score beside the therapy usage timeline and marks windows with severe mask leak as unclassified.
Motion scoring with overnight pulse oximetry
An under-mattress motion sensor and a finger pulse oximeter send timestamped measurements to a bedside processor. The processor identifies sleep intervals from motion, measures restlessness from movement bursts, and aligns each breathing window with oxygen saturation measurements from that window and the following 30 seconds. It classifies windows using breathing motion, oxygen saturation drops relative to the user's overnight baseline, and restlessness, while rejecting oxygen measurements flagged for poor signal quality. The display shows the nightly score and the timestamps of disturbed windows.
Automated or data-driven variant
Personalized scoring with a trained classifier
A chest accelerometer records gross body movement and smaller oscillations associated with breathing. Software identifies the sleep interval and builds a feature vector for each window from breathing period variability, changes in oscillation amplitude, movement burst count, and the fraction of the sleep interval spent moving. A gradient boosted tree classifier trained on windows with reference breathing labels receives these features along with their deviations from the user's historical medians. Software thresholds the predicted disturbance probabilities and displays the percentage of classified windows below the threshold.
Specific version of a part
Wrist acceleration for sleep timing
A wrist-mounted three-axis accelerometer samples at 25 Hz, and a processor calculates activity counts in 30-second epochs. A sleep and wake classifier uses the counts from each epoch and its neighbors to locate sleep onset and final waking.
Mattress motion for restlessness
A piezoelectric strip beneath the mattress records motion during the detected sleep period. The processor counts movement bursts with at least two seconds between bursts and stores the count and total burst duration for each 30-second window.
Personal thresholds from seven nights
The processor stores restlessness measurements from the sleeper's seven most recent recorded nights and calculates their 95th percentile. It labels a current 30-second window as disturbed breathing when the restlessness measurement exceeds that threshold. It subtracts the percentage of windows with that label from 100 to produce the sleep fitness score.
Bedside score and restlessness display
A bedside liquid-crystal display presents the sleep fitness score as a number from 0 to 100. A timeline beneath the score shows the duration of restlessness in each 30-second window of the sleep period.
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.
1.1determining a sleep period of a user based on first motion data of the user captured by one or more motion sensors;
Sleep cycle analysis component 256 analyzes sensor information, environmental information and subjective information from the person and constructs a more accurate sleep architecture for the individual. This can involve adjusting the times associated with each cycle of the sleep architecture as well as drawing a correlation between the data from one or more sensors and the particular person's entry or exit from the different cycles.
A · US 7,524,279 B2 · ¶ 38
Sleep cycle analysis component 256 analyzes sensor information, environmental information and subjective information from the person and constructs a more accurate sleep architecture for the individual. This can involve adjusting the times associated with each cycle of the sleep architecture as well as drawing a correlation between the data from one or more sensors and the particular person's entry or exit from the different cycles.
B · US 2009/0264715 A1 · ¶ 38
1.2determining, for the sleep period of the user, sleep fitness data based on a restlessness associated with the user as determined according to second motion data of the user captured by the one or more motion sensors during the sleep period of the user;
For example, the number of wakeups and corresponding duration of wakeups can be determined based upon movement data collected from movement sensor 222 in addition to other vital signs taken from the various other vital signs measurements.
A · US 7,524,279 B2 · ¶ 34
For example, the number of wakeups and corresponding duration of wakeups can be determined based upon movement data collected from movement sensor 222 in addition to other vital signs taken from the various other vital signs measurements.
B · US 2009/0264715 A1 · ¶ 34
1.3calculating a sleep fitness score associated with the sleep period, based on a classification of a plurality of temporal windows within the sleep period as representing disturbed breathing for the user or not breathing disturbed for the user, the classification of a temporal window among the plurality of temporal windows being determined based on historical information associated with the user and the sleep fitness data; and
Upon execution, the instructions receive a subjective measure of sleep quality from the person after the sleep, create a sleep quality index from the objective measure of sleep quality and subjective measure of sleep quality, correlate the sleep quality index and a current sleep system settings with a historical sleep quality index and corresponding historical sleep system settings.
A · US 7,524,279 B2 · abstract
Typically, the instructions are loaded in a memory where they execute to generate an objective measure of sleep quality from the sleep data from the person and gather environmental data during the sleep by the person.
B · US 2009/0264715 A1 · abstract
1.4providing, for presentation on a display device, at least one of the sleep fitness score or the sleep fitness data.
An empty cell means only that no passage was reproduced there.
Statements in these documents
Further, data collected during these one or two night stays are often too short and unpredictable to make accurate assessment of problem.
A · US 7,524,279 B2 · ¶ 6
After each sleep, the method obtains a subjective measure of sleep quality from the person useful in creating a sleep quality index from the objective measure of sleep quality and subjective measure of sleep quality.
A · US 7,524,279 B2 · ¶ 8
Further, data collected during these one or two night stays are often too short and unpredictable to make accurate assessment of problem.
B · US 2009/0264715 A1 · ¶ 6
After each sleep, the method obtains a subjective measure of sleep quality from the person useful in creating a sleep quality index from the objective measure of sleep quality and subjective measure of sleep quality.
B · US 2009/0264715 A1 · ¶ 8
Cite aspatent.fyi, US 2026/0354753 A1, claim 1, first published 2026-11-19T03:31:31-05:00, https://patent.fyi/US20260354753/, SHA-256 5cf5dd63f9afdbb2a4e63acf9c8becea6282aa65b64b17d1df1ffddf27a7e732