AirMax is extremely miniaturized, and weighs only a few grams. It is applied over the nose easily.

Irregular or disrupted breathing (2-channel Airflow) AirMax captures airflow from each nostril independently, tracking breathing patterns including periods of reduced or paused airflow. These patterns may correlate with restless sleep, frequent waking, morning headaches, and daytime fatigue.

Heart rate and variability (PPG) A light sensor detects your heart rate and heart rate variability, showing how your autonomic nervous system responds during rest. Heart rate patterns - like slowing down during breathing pauses then spiking upon resumption - may correlate with breathing variations, while HRV patterns may correlate with recovery quality, stress levels, and sleep effectiveness.

Blood oxygen variability (Red/IR PPG channels) Unlike traditional pulse oximeters that heavily average readings, AirMax's proprietary algorithm calculates oxygen saturation beat-by-beat while flagging unreliable data from motion or noise. Measuring at the nasal ala (nostril) captures oxygen changes faster than finger-based sensors, revealing transient fluctuations that may otherwise be missed, and correlate with breathing variations, sleep quality, and morning grogginess.

Restlessness and head position (accelerometer) A precision accelerometer tracks movement, arousals, and head tilt during sleep. Frequent movement may indicate disturbed sleep that could improve with dietary or lifestyle changes. Head position tracking helps identify if breathing irregularities occur primarily in certain positions - whether on your back, side, or stomach - revealing which sleep postures to potentially avoid.

Snoring and environmental sounds (microphone) AirMax maximizes your phone's microphone sensitivity and records peak sound levels to capture snoring intensity and ambient noise like traffic, HVAC systems, or household disturbances. Loud or persistent snoring may correlate with breathing variations, while environmental noise may correlate with fragmented sleep and frequent awakenings.

Room illumination (light sensor) AirMax leverages your mobile phone's ambient light sensors to record the level of luminosity of the room. Too much light during the night or early morning can interfere with natural sleep rhythms.

EEG (via ZMax sync) AirMax respiration based data can be synchronized with EEG data coming from the ZMax headband. You may watch a video about ZMax and AirMax synchronized data or a second video validating the synchronization.

The Total version of AirMax outputs data in plain EDF file format. EDF data can be opened with any EDF viewer, or with widely available libraries in Python, C#, Matlab etc. Additionally the exported data is fully compatible with Hypnodyne's HDScorer software, for easy visualization of event markers, stimulation markers, and on-the-fly recalculation of the night for fine tuning detection settings.

SpO2 calculation is extremely sensitive to distortions in the shape of the PPG waveform. Most devices use a strong low-pass filter as a quick and simple solution. However, this approach has significant side effects. First, the desaturation signal can lag by up to 20 seconds. Second, filtering raises the minimum SpO2 value, resulting in a significant underestimation of the true desaturation.

With biosignals, analyzing artifact- free data is relatively easy. Identifying and rejecting artifacts, however, is much more challenging, as their morphology and characteristics can vary widely.

AirMax utilizes a number of custom algorithms to analyze the behavior of the PPG signal, reject malformed peaks, and identify unreliable SpO2 values. This means that AirMax SpO2 data is unfiltered, with a temporal resolution and lag of exactly one heartbeat.

During recovery from a desaturation, there may be only one or two heartbeats with a clean PPG waveform. These peaks can contain the true minimum SpO2 value for the event. AirMax can capture this value due to the precision of its analysis.

A confidence metric is provided for each event. Low-confidence events—with a trusted duration but an untrusted desaturation value—are shown in grey. The amount of 'grey' visible during a night also provides an indication of device- wearing quality: the better the sensor contact with the nostrils, the less PPG distortion occurs during each recovery, and the less grey appears on the chart.

• Miniaturized & Easy to Wear: Lightweight, non-intrusive design fits over the nose for effortless overnight use.
• Multi-Sensor Precision: Measures airflow, SpO₂, movement, head tilt, and sound for a complete picture of sleep physiology.
• Objective Data Collection: Captures breathing patterns, oxygen variability, snoring, restlessness, and environmental disturbances - data STOP-BANG and questionnaires can't provide.
• Scalable for Research: Ideal for perioperative, cardiovascular, metabolic, sleep, and population studies - works in labs, at home, or multi-site protocols.
• Easy Data Access & Sharing: EDF file format compatible with Python, Matlab, C#, and can be shared via phone apps instantly.
• ZMax Integration: Sync respiration with EEG data for complete multi-modal research. You may watch a video about ZMax and AirMax synchronized data or a second video validating the synchronization.

Breath holding – cardiogenic oscillations present

While holding your breath with the airway open, small pulses caused by each heartbeat can be seen.

Breath holding – cardiogenic oscillations absent

When the glottis is voluntarily closed, these heartbeat-related airflow pulses disappear.

Mouth breathing:

During mouth breathing, a regular airflow signal can still be detected at the nostrils.

Many clinical and research protocols use the STOP-BANG questionnaire to quickly flag participants at elevated risk for obstructive sleep apnea (OSA). STOP-BANG works because it is simple, validated, and easy to deploy at scale. Its limitation is equally well understood: it provides risk classification, not physiological evidence.

AirMax enables researchers to collect continuous, objective respiratory related signals that are not captured by questionnaire-based screening tools such as STOP-BANG, including event counts, oxygen saturation trends, temporal patterns, and night-to-night variability.

• Perioperative and anesthesia research: AirMax can be used to explore relationships between overnight respiratory physiology and postoperative observations in research settings, supporting hypothesis generation beyond self-reported risk scores.
• Cardiovascular and metabolic research: Quantitative physiological features (e.g., oxygen saturation dynamics, breathing pattern variability) can be analyzed alongside cardiometabolic variables to investigate associations in observational studies.
• Sleep and population-based studies: Wearable-derived data supports large-scale data collection and exploratory phenotyping in research cohorts, reducing reliance on resource-intensive laboratory studies.
• Large or at-home research cohorts: AirMax provides a scalable, low-burden method for collecting longitudinal physiological data to complement existing screening instruments.
• Studies incorporating STOP-BANG: AirMax does not replace questionnaires or redefine risk classification. Instead, it provides independently measured physiological data that may be used for validation analyses, subgroup exploration, and data-driven interpretation within research protocols.

In the morning, AirMax data can be easily shared using any app on your phone (e.g., email, WhatsApp, Telegram, etc.).

AirMax respiration data can be synced to real time EEG data collected with ZMax Total. For now, this requires ZMax Total to be running in wireless mode, so that packet timing can be shared. In the future some way of synchronizing AirMax and ZMax Lite data will be provided, albeit with some user input and not fully automated.

The first prototypes of AirMax are available. We will ship initial prototypes on a first come, first served basis. The order page is here.