Positional obstructive sleep apnea detection system

Abstract

An obstructive sleep apnea detection device including an optical engagement surface adapted to engage a user's skin; a light source adapted to emit light from the optical engagement surface; a photodetector adapted to detect light at the optical engagement surface and to generate a detected light signal; a position sensor adapted to determine patient sleeping position; a controller adapted to determine and record in memory blood oxygen saturation values computed from the detected light signal and user position information from the position sensor; and a housing supporting the optical engagement surface, the photodetector, the light source, the position sensor, and the controller.

Description

INCORPORATION BY REFERENCE

All publications and patent applications mentioned in this specification are incorporated herein by reference in their entirety to the same extent as if each individual publication or patent application was specifically and individually indicated to be incorporated by reference.

BACKGROUND

Obstructive sleep apnea (OSA) occurs when tissue in the upper airway blocks the airway during sleep. The brain will sense the rise in CO₂, and will wake up the person so that breathing resumes. Such an event is called an apnea. A partial airway blockage causing an awakening is called a hypopnea. A person is unlikely to remember such awakenings, but sleep is disrupted. The severity of obstructive sleep apnea is measured by the frequency of awakenings, as shown in the table below.

Apneas + Hypopneas/Hour OSA Classification
0-5                     Normal
5-15                    Mild
15-30                   Moderate
30+                     Severe

Untreated, OSA not only leaves patients chronically fatigued, but it also carries significant health consequences. Unfortunately, despite a very high prevalence of OSA in the population, the vast majority of sufferers remain undiagnosed. Some estimates put the number of people with obstructive sleep apnea who are undiagnosed as high as 85%. In the US alone, this could mean 35 million undiagnosed sufferers. This large undiagnosed population is one of the biggest public health issues in sleep medicine.

There are many reasons for the low rate of diagnosis. The field of sleep medicine is relatively new. Awareness of clinical sleep problems and their causes remains relatively low. The understanding of sleep-disordered breathing is still advancing at a consistent rate, and will increase awareness further as more clinical consequences are characterized. Further, among some of the population there is a misconception that sleep-disordered breathing is not a serious condition, and is often dismissed as simple snoring. Awareness of the condition is presently significantly lower than the severity warrants.

A further impediment to diagnosis is the relatively onerous process through which most sufferers must pass to get diagnosed. Typically, an initial physician visit is followed by a second visit to a sleep specialist, and then an in-lab polysomnogram (PSG). Polysomnography is a very thorough observation while the user is sleeping. It is conducted in a laboratory setting, with a minimum of 22 electrodes and sensors placed on the patient, and with observation via video throughout the night. Many patients find this foreign sleep environment and observation inconvenient and intrusive. The very nature by which the data is gathered interferes with the typical, natural sleep habits of the test subject. Many labs also have long wait times for studies, especially for weekend studies, which are preferred by some patients to try to reduce the amount of disruption to their work routine. In-lab PSG studies also carry a high cost, often $2000-3,000 per night, a significant burden to the healthcare system. More recently, home sleep testing (HST) devices have been utilized to bring 4-8 channels of the traditional lab test into the home environment. Unfortunately, they still require a prescription, are not simple to use, and result in insufficient data about 20% of the time due to incorrectly placed or inadequately secured electrodes.

Some sufferers report not wanting to go through the whole process of the PSG only to find out the bad news of an apnea diagnosis. This form of wishful denial hinders the rate of diagnosis and postpones the initiation of treatment and its benefits. Many sufferers persist for years, suspecting they might have sleep apnea, but not acting to undertake the testing. This results in significantly increased healthcare spending, and reduced quality of life. Studies have calculated the cost of increased healthcare utilization by undiagnosed sleep apnea sufferers between $2000 and $12,000 additional USD per year. One study estimated the total economic cost of OSA (diagnosed and undiagnosed) to be between $65 and $165 billion annually. There are many impacts of sleep apnea that are difficult to fully estimate, such as the effect on interpersonal relationships of the sufferer.

In addition to these shortcomings, the PSG also has the challenge of trying to understand a complex patient given only one night of data. Due to the inconvenience and expense of PSG, most patients are only studied for one night. However, sleep quality can vary tremendously on a night-to-night basis. Many factors can influence sleep differently each night: exercise, diet, alcohol, stress, environment. It is a significant limitation of the PSG to only gather one night of data. Ideally, data would be gathered from each night of sleep throughout the typical fluctuations by these factors to more completely characterize a patient's sleep.

For those who do get diagnosed, the frontline therapy is Positive Airway Pressure (PAP). It is also called Continuous Positive Airway Pressure (CPAP), or Automatic Positive Airway Pressure (APAP). This disclosure refers to all such therapies as “PAP.” PAP is the most widely used and the most effective treatment for OSA. In PAP, a bedside compressor supplies pressurized air to the patient's airway through a hose and mask. The air pressure is set sufficiently high to maintain an open airway during sleep. The clinical data show significant benefit to the use of PAP therapy. Many patients report significant positive impact on their health and quality of life. However, many patients do not feel the benefit of PAP therapy all the time. This can lead to a patient perception that the treatment isn't effective, despite significant clinical evidence to the contrary. Some such patients then discontinue therapy, or only use it sporadically. The clinical data show that consistent use of PAP therapy yields the most benefit. Unfortunately, for these sporadic users, there is no convenient way to determine if the therapy is actually working, beyond user perception.

There is another category of patient who will use PAP therapy consistently for a while, then discontinue use once feeling good, assuming that they might be cured, or no longer need PAP treatment. Shortly thereafter they will again feel the fatigue and cognitive impairment reminiscent of their time before PAP therapy. Once they recognize this, they return to therapy. Some healthcare providers have termed this the yo-yo group, as they vacillate on and off therapy like a toy yo-yo going up and down.

Unfortunately for all PAP users, there is no convenient way to ascertain in a scientifically sound manner the effect of using and not using their PAP device on a nightly basis.

Sleeping position plays a very significant and greatly underappreciated role in Obstructive Sleep Apnea. When sleeping in a supine position, gravity's effect on the tongue, tonsils, soft palate and other structures of the upper airway is more pronounced. Clinical studies have shown that the majority of sleepers have at least twice as many blockages when back sleeping compared to side sleeping. The blockages that occur while supine sleeping are more severe: they last longer and result in more significant dips in blood oxygen levels. For some patients, the ratio of supine Apnea-Hypopnea Index (“AHI”) to lateral AHI exceeds 4 to 1. Sleeping position is one of the greatest determinants of the likelihood of an apnea or hypopnea occurring. Consequently, sleeping position is also one of the strongest determinants of PAP pressure needed to maintain a patent airway. During an in-lab titration study, PAP pressure settings are determined during REM stage, supine position sleeping. This is because during REM stage, supine sleeping the airway is most prone to collapse. Thus, PAP pressure settings are classically set to cover the worst-case scenario conditions. In fact, a titration study is not considered successful unless REM supine sleep is present. During some studies, technicians will enter the room and push the patient onto their back in order to capture REM supine sleep and the related pressure setting.

Studies have shown that a patient's required PAP pressure is significantly lower when not sleeping supine. One study found that the required pressure during side sleeping for all patients, not just those with positional OSA, was almost 3 cm H₂O lower. For those with positional OSA, the pressure difference between back and side can be much larger.

Studies are not required for most people to recognize the role of sleeping position on breathing. For centuries, it has been reported that people sleep louder when supine, and common advice suggests elbowing a loud snorer to get them off their back to quiet down.

Despite all this evidence, the role of sleeping position in OSA remains a mystery to the vast majority of OSA sufferers. Very few sufferers understand what role sleeping position plays in their apnea.

Accordingly, there is a compelling need for a way to determine the role of position, and the severity of oxygen desaturations in a convenient way. This has application for the undiagnosed as well as the diagnosed, those who are adherent to PAP therapy as well as those who find PAP therapy difficult to consistently use. All would benefit from a better understanding of their OSA, the role of sleeping position, and the nightly impact of their therapy. Beyond PAP therapy, the ability to understand the efficacy of alternative treatments for OSA and the role of sleeping position therein is also of great benefit.

SUMMARY OF THE DISCLOSURE

The present invention provides a positional obstructive sleep apnea detection system that obtains information about a subject's sleep position (e.g., sleeping on the back, sleeping on the side) and combines it with information indicative of a cessation or reduction in breathing (e.g., blood oxygen levels). In some embodiments, the detection system combines a pulse oximeter with a position sensor, optionally together in the same device or housing. Pulse oximetry is at the core of every sleep study to track blood oxygen saturation through the night. Significant apneas will usually drop blood oxygen saturation below standard levels. Some other embodiments may add optional other detectors to detect and possibly analyze, e.g., snoring sounds (microphone and/or accelerometer), respiratory rate (accelerometer and/or heart rate algorithm), respiratory effort (accelerometer, heart rate, and/or microphone), sleep/wake cycles (heart rate and/or accelerometer), airflow at the nasal opening (thermistor, flow sensor and/or pressure sensor). Some or all of this information may be used, e.g., to assess risk of, and possibly diagnose, sleep disordered breathing (especially sleep apnea). It may also be used by the subject to assess CPAP efficacy.

One aspect of the invention provides an obstructive sleep apnea detection device including an optical engagement surface adapted to engage a user's skin; a light source adapted to emit light from the optical engagement surface; a photodetector adapted to detect light at the optical engagement surface and to generate a detected light signal; a position sensor adapted to determine patient sleeping position; a controller adapted to determine and record in memory raw detected light signal values of, or blood oxygen saturation values computed from, the detected light signal and user position information from the position sensor; and a housing supporting the optical engagement surface, the photodetector, the light source, the position sensor, and the controller. In some embodiments, the device also includes an optional a communicator, such as a wireless transmitter or display, adapted to communicate the blood oxygen saturation values and user information position information. In some or all of these embodiments, the optical engagement surface may be shaped to engage skin on a forehead of the user.

In some or all of these embodiments, the housing also has a peripheral surface surrounding the optical engagement surface, the optical engagement surface extending downward from the peripheral surface. For example, the optical engagement surface may be disposed 0.1 mm to 2.5 mm from the peripheral surface, or 0.25 mm to 0.75 mm from the peripheral surface. The device may also have a flexible material disposed below the peripheral surface, with the optical engagement surface extending downward from a lower surface of the flexible material. The flexible material may optionally include an adhesive.

In some embodiments, the optical engagement surface may include a material preferentially absorbing light wavelengths corresponding to light wavelengths emitted by the light source. In some embodiments, the light source is configured to emit light through a light source opening in the optical engagement surface and the photodetector is configured to detect light through a photodetector opening in the optical engagement surface, the light source opening and photodetector opening being disposed 5 mm-10 mm apart. In some embodiments, the device has a tapered frustoconical surface extending from an opening in the optical engagement surface to an active surface of the photodetector, and the diameter of the frustoconical surface adjacent the photodetector may optionally be smaller than a diameter of the active surface of the photodetector. Some embodiments may have an opening in the housing adjacent the light source, the opening having a diameter smaller than a diameter of an active area of the light source.

In some or all of these embodiments, the position sensor includes an accelerometer. In some or all embodiments, the position sensor includes a set of instructions used by the controller to compute patient position from the detected light signal or from the blood oxygen saturation values.

Another aspect of the invention provides an obstructive sleep apnea detection device including: an optical engagement surface adapted to engage a user's skin; a light source adapted to emit light through a light source opening in the optical engagement surface; a photodetector adapted to detect light through a photodetector opening in the optical engagement surface and to generate a detected light signal; a position sensor adapted to determine patient sleeping position; a controller adapted to determine and record in memory blood oxygen saturation values computed from the detected light signal and user position information from the position sensor; and a housing supporting the optical engagement surface, the photodetector, the light source, the position sensor, and the controller, the housing comprising a peripheral surface surrounding the optical engagement surface, the optical engagement surface extending 0.1 mm to 2.5 mm downward from the peripheral surface.

In some embodiments, the optical engagement surface is disposed 0.25 mm to 0.75 mm from the peripheral surface. Some embodiments also include an optional a flexible material disposed below the peripheral surface, the optical engagement surface extending downward from a lower surface of the flexible material. The flexible material may include an adhesive.

In some or all of these embodiments, the light source opening and the photodetector opening are disposed 5 mm-10 mm apart.

Some or all of these embodiments may also include a communicator adapted to communicate the blood oxygen saturation values and user information position information, such as, e.g., a wireless transmitter disposed in the housing. In some embodiments, the optical engagement surface includes a material preferentially absorbing light wavelengths corresponding to light wavelengths emitted by the light source.

Some or all of these embodiments may also include a tapered frustoconical surface extending from the photodetector opening to an active surface of the photodetector. A diameter of the frustoconical surface adjacent the photodetector may be smaller than a diameter of the active surface of the photodetector. In some embodiments, the light source opening has a diameter smaller than a diameter of an active area of the light source.

In some embodiments, the position sensor includes an accelerometer, and in some embodiments, the position sensor includes a set of instructions used by the controller to compute patient position from the detected light signal.

Yet another aspect of the invention provides a method of obtaining obstructive sleep apnea information from a subject. In some embodiments, the method includes the following steps:

placing an optical engagement surface of a detection device against a skin surface of the subject (such as, e.g., the subject's forehead); transmitting light through the optical engagement surface into the skin surface of the subject; detecting light from the skin surface of the subject through the optical engagement surface; determining blood oxygen saturation from the detected light; obtaining subject position information with a position detector within the device; and storing the blood oxygen saturation and position information in memory.

Some embodiments include the step of time correlating blood oxygen saturation with position information. Some or all of these embodiments include the step of computing position information from blood oxygen saturation.

Some or all of these embodiments also include the step of communicating the blood oxygen saturation and position information, such as by, e.g., displaying blood oxygen saturation and position information and/or wirelessly transmitting the blood oxygen saturation and position information using an optional transmitter within the device. The method may also include the step of gathering position information from a second position sensor disposed on the subject's torso.

Still another aspect of the invention provides a method of performing a risk assessment of a subject for sleep disorders. In some embodiments the method includes the steps of obtaining sleep questionnaire information about the subject; obtaining blood oxygen saturation information from the subject during sleep; obtaining head position information from the subject during sleep; comparing the sleep questionnaire information, the blood oxygen saturation information and the head position information to a database of historical data; and determining the subject's risk of sleep apnea from the comparing step.

Yet another aspect of the invention provides a method of performing a risk assessment of a subject for sleep disorders. In some embodiments, the method includes the following steps: obtaining sleep questionnaire information about the subject; obtaining blood oxygen saturation information from the subject during sleep; obtaining head position information from the subject during sleep; comparing the sleep questionnaire information, the blood oxygen saturation information and the head position information to a database of data from a population of other subjects; and determining the subject's risk of sleep apnea from the comparing step.

In some embodiments, the invention provides a system for accurate assessment of the risk of obstructive sleep apnea and the analysis of the role of position in the subject's obstructive sleep apnea. In some embodiments, the invention obtains information about the subject's supine sleeping events and lateral sleeping events separately, as well as mixed AHI. The system may be used by the subject at home and does not require the use of a sleep lab.

In some embodiments, the device is powered by a battery, such as a coin cell, alkaline battery, LiPo battery or Li ion battery. The device may have charging circuitry, such as inductive charging circuitry, and it may be USB compatible using spring pins.

In some embodiments, the device may use an accelerometer, a gyroscope or a magnetoscope to obtain position information.

In some embodiments, the device may send and receive information via a wireless connection using, e.g., a Bluetooth, BLE, Zigbee, 802.11b, RF or other communication protocol, and the information can be sent to and received from a smartphone, which may send information to a network server. The device may also send information over a wired communication link, e.g., when not being worn by a sleeping subject.

In some embodiments, the device has a replaceable disposable adhesive backing that can be removably attached to a reusable electronics housing.

The system described herein allows for multiple nights of data gathering, without interfering with typical sleep habits, and in the comfort of the user's normal sleeping environment. This combination results in sleep data that is more relevant and representative of the user's sleep reality, allowing for more accurate and nuanced identification of sleep issues.

It is well documented that people do not sleep the same each night. Gathering multiple nights of sleep data enables a longitudinal view of sleep. Data gathered using this system can show changes in sleep due to environmental and health factors such as sleep hygiene, weight change, stress, alcohol intake, illness, allergies and exercise. Understanding the factors that influence the variability of sleep quality is critical to accurately addressing sleep disorders.

Further, the systems described herein allow for the collection of data without interfering with normal sleep patterns of the user.

Finally, the user's typical sleep environment is the only relevant test location. When sleeping in a new location, people have different reactions which inevitably influence their sleep quality.

Obstructive Sleep Apnea is position dependent in almost all cases. A person on his or her back will experience more frequent and more severe airway blockages as gravity pulls the tongue, tonsils, soft palate and other structures into the airway. Most OSA patients will experience at least twice as many events when sleeping on their backs as compared to their sides. Some will experience four to five times as many blockages when sleeping supine. In contrast, patients with central sleep apnea and other conditions that cause blood oxygen to drop during sleep are typically less position dependent. Since almost everyone spends time sleeping on both their back and their side, the combination of pulse oximetry data and position data can be used to help differentiate OSA from Central Sleep Apnea (CSA) and other non-position dependent conditions and situations.

The current invention combines sensors for optical measurement with sensors that can determine position. The device can be worn on the skin through the night. The position measuring component may or may not have to be calibrated on a given patient to be able to distinguish back sleeping form side sleeping. This device would record blood oxygen saturation and position for the duration of one or more nights. Then, importantly, this information can be analyzed to compare the frequency and/or severity of apnea events when the patient is back sleeping to the events occurring when the patient is side sleeping. This analysis can occur on the device itself, or on other devices using the data gathered by the device. The analysis can compare the frequency of low blood oxygen saturation events, the severity of those events, or a combination of frequency, severity and/or duration in the two sleeping positions. If the frequency, severity or duration of blood oxygen deficits is significantly greater during back sleeping than side sleeping, the data analysis would indicate OSA, while rough equality in the number, length and severity of blood oxygen deficits could indicate the possibility of CSA or another less common sleep disorder.

In some embodiments, a data graph is generated of blood oxygen saturation over time. The area above the curve, denoting the difference between the patient's recorded blood oxygen saturation and complete (100%) saturation over time, can be used to differentially diagnose OSA and CSA. In patients with OSA, the area above the curve will typically have a higher ratio of events per minute of back sleeping as compared to side sleeping. A patient with CSA and other conditions or situations will have a lower ratio of events per minute supine to lateral. This area above the curve can incorporate the frequency, duration and severity of blood oxygen desaturations.

The system described herein is able to distinguish between Central and Obstructive apnea. In one embodiment, the combination of position data and SpO₂ data can be used to determine the degree of Positional Obstructive Sleep Apnea (pOSA). Typically, pOSA is defined as having at least twice the number of events (apnea or hypopnea) when in the supine position. Further, desaturation events occurring in the supine position last longer and reach lower desaturation levels. The invention described herein offers the ability to determine the correlation between the subject's position and each desaturation event. This correlation is important for distinguishing CSA from OSA. OSA is position-driven the majority of the time. Studies have shown that up to 75% of OSA patients are pOSA. Nearly every patient has more events in the supine position. CSA, by contrast, is not nearly so position-driven. Although there are some case reports that CSA can be exacerbated by the supine sleeping position, the vast majority of CSA events are independent of position. Using this knowledge, we can estimate the likelihood that a user has CSA. If a user has significant events, but they are not correlated to the supine sleeping position, then CSA is strongly suspected. For these users, the application can suggest further studies using the device, as well as further studies using other standard of care testing equipment. In one embodiment, if CSA is suspected due to the initial results from the testing, then the user is directed to place the device on their chest for subsequent testing. This testing can be done using the same device and application, saving significant time, money, and convenience. On the subsequent test nights, the user places the device on the chest to gather motion data from that location. For patients with CSA, there is a cessation of breathing effort that is distinct and unique to CSA from OSA. A patient with CSA will have this cessation of breathing, which will show up in the motion data as a significant lapse in breathing effort motion. In this way, the system described herein can very reasonably distinguish OSA and CSA.

EMBODIMENTS

In one embodiment, the sensor can be located on the head or the torso, or both. With a sensor placed on both the head and the torso, more data can be gathered, and more subtle observations can be drawn from the comparison of the data from the two locations. The head sensor, for example on the forehead, can gather head position, pulse oximetry, and sound data. The sensor on the torso can be placed anywhere on the neck, clavicle, chest, belly, diaphragm. The torso sensor can detect motion and position data. From this data, the body position, respiratory rate, respiratory effort and heart rate can be determined.

One advantage of a second sensor on the chest area is that it can provide data on respiratory effort. Often during an apnea, there is increased respiratory effort as the body tries to get more oxygen but cannot due to the blockage. This effort is pronounced and can be captured by a motion sensor on the torso. Adding this additional channel capability can also satisfy testing channel requirements for certain health systems.

In another embodiment, the same sensor device is worn over multiple nights. Multiple nights of data helps accumulate additional information about the user's sleep quality. As sleep quality varies nightly, multiple nights of data provide a much stronger indication of the user's sleep issues.

In one embodiment, the sensor is worn on the head for multiple nights to gather baseline information on the user's sleep, such as oxygenation, movement, and respiratory disturbances. Once this pattern has been established, the same sensor can then be worn on the torso, chest, or belly. In this location, the sensor will detect the movement of the diaphragm and chest corresponding to respiratory effort. If the sensor detects long pauses of movement, this is an indication of possible central sleep apnea. In this way, the same device can be used over multiple nights in two different locations to differentiate between OSA and CSA.

The system consists of wearable sensors, a data analysis platform and algorithms, and a user interface for gathering user input and communicating results.

Wearable Sensors

The device includes optical sensors for obtaining a photoplethysmography (PPG) signal. Light is emitting into the tissue in at least two different wavelengths. The resulting light reflected back into the optical sensor is measured at each wavelength. The results are compared to determine the oxyhemoglobin concentration. The optical sensors enable measuring blood oximetry, pulse rate and heart rate variability (HRV).

The device also includes sensors for determining position and acceleration. In a preferred embodiment, this is done by an accelerometer. The accelerometer data is used to determine head position of the sleeper. As discussed herein, sleeping position is of critical importance for respiratory sleep disorders. Motion data can also be used to identify restless leg syndrome (RLS), a sleeping disorder that involves urges to move the legs. Accelerometer data can also indicate wake and sleep intervals of the user, and be used to calculate total sleep time. RLS is characterized by the urge and sensation to move the legs during sleep and sleep transitions. The unique signature of RLS movement is detectable in the accelerometer data. RLS movements are jerky and repetitive, which can be discerned from a typical sleeping motion signal. In one embodiment, RLS risk can be assessed from user-inputted data on a companion app or questionnaire, or from to measurements and signatures in the forehead accelerometer data. Then, if RLS is suspected, additional night(s) of observation and measurement with the device can be done, with the device placed on the legs to measure movement. This approach can confirm suspected RLS with great certainty, due to the exceptionally pronounced motion signal on the legs during RLS events. Actigraphy is the study of motion data to determine physiologic events. In one embodiment, motion data from the accelerometer is used to determine sleep stage. During REM sleep, the user is typically nearly motionless, due to the body's protective mechanism to prohibit acting out one's dreams. Thus, motion data can discern suspected REM stage sleep from times of little or no motion. During Stage 1 and 2 sleep, the respiratory signal, found in the motion data, slows as the users transitions to sleep.

The user's sleeping position can also be determined from the PPG waveform gathered by the device. The waveform characteristic changes depending on sleeping position from supine sleep to lateral sleep. Even the right and left lateral sleeping positions are distinguishable. As typical sleepers change position 20 or more times each night, the device will typically have many opportunities to assess what each position looks like on the PPG waveform.

Since the sleeping position affects the PPG waveform, it is also advantageous to use the position data gleaned from the accelerometer to inform the processing of the PPG data. For instance, since we know the amplitude of the AC component of the PPG is reduced in the side-lying position, we can use different filters to process the PPG in this circumstance to yield more accurate SpO₂ results. Correspondingly, when the accelerometer data indicates that the user is supine sleeping, we can process the PPG signal accordingly with the knowledge that the amplitude will be larger.

The device can determine when a user sits up, stands up, and leaves the bed. This is especially useful for tracking visits to the bathroom in the night. Nocturia, the nighttime urge to urinate, is a symptom of sleep apnea. Once using treatment such as CPAP, many users report a reduction in nocturia. The device is able to quantify this reduction, offering positive reinforcement to the user.

The motion data gathered by the accelerometer can also be used to strengthen the measurements from the oximeter. In pulse oximetry, significant error in the signal can be created during motion events. Using the motion sensor, these motion events can be identified. In one embodiment, the device algorithm will flag SpO₂ values that are calculated from data gathered during a motion event as determined by the motion sensor. In one embodiment, the algorithm will discard SpO₂ values that are associated with motion events above a threshold that renders the SpO₂ significantly less accurate. Under typical use, the sensor will be able to achieve accuracy within about +/−3% for SpO₂. Data gathered during motion events that reduce the accuracy beyond a threshold, such as +/−5%, may be discarded. In another embodiment, when there is a motion a different set of coefficients is used to calculate the SpO₂ which take into account motion circumstances.

In one embodiment, the PPG signal itself is used to determine when there are motion events and to consequently filter out the calculations for the SpO₂. Motion events show up in a pronounced way in the PPG waveform as an increased signal to noise ratio. In another embodiment, this motion signature in the PPG can be coupled with the data from the motion sensor to determine when to discard or adjust the SpO₂ data calculation.

Motion data can also identify coughing, enabling the measurement of sickness and allergies. Coughing typically results in significant, sudden movement in predominantly one of the 3-axes of accelerometer data. This signature is recognized by the software and recorded as a coughing event. Measuring and tracking coughing empowers users to understand the triggers in their sleeping environment that can lead to disrupted sleep.

In another embodiment, motion data is used to identify vibration caused by snoring. The motion signal signature from snoring is characterized by a steadily increasing vibration lasting a typical 1-2 seconds, and timed to coincide with the respiratory cycle. Snoring will occur up to 15 times per minute, but more typically 6-12 times per minute. The motion data exhibits values that fluctuate as they increase, followed by a more rapid drop off in motion. In some snorers, the data reveal long periods without significant motion, punctuated by much more significant motion events as the pauses are broken with sudden snoring.

The user's respiratory rate can be determined from the device sensor data. The PPG waveform contains adequate signal for the derivation of the respiratory rate using signal processing techniques described in the art. Actigraphy data from the accelerometer is capable of revealing the respiratory rate as well. Respiratory rate shows up as a sinusoidal pattern in the motion data. Respiration rate can be determined from the processing of this signal. Combining the motion data derived from the motion sensor with the PPG signal data can strengthen the accuracy of the respiratory rate calculation. Additionally, respiratory effort, especially when there is respiratory distress, can be determined from the processing of the data. Significantly increased motion in the respiratory signal indicates respiratory distress. This can occur during an apnea or hypopnea, and provides another signal to identify and confirm respiratory events.

In some embodiments, a microphone captures the sound signal, and allows for analysis of snoring sounds and pauses in breath. The microphone can be incorporated in the device. Alternatively, the system can use an existing microphone in a smartphone, phone, tablet, or other connected device.

In some embodiments, the device includes conductive sensors for the forehead for gathering an electroencephalogram (EEG) signal. EEG can be used to indicate brain activity, sleep staging, and identify arousals during sleep.

In some embodiments, pressure and flow sensors are included to assess airflow. This can be realized through the placement of an open tube just beneath the nares, which is in communication with a pressure sensor placed in the device. Alternatively, an additional patch with wireless connectivity can be placed just beneath the nares and utilize pressure, temperature, sound, or flow sensors to assess respiratory airflow. Alternatively, an additional patch placed under the nares can be connected to the device with a thin wire cable for power and/or data communication.

Additional sensors in some embodiments include an ambient light sensor to assess the amount of ambient light in the sleeping environment, gas sensors to assess the amount of key respiratory gases such as CO₂ which plays a key role in regulating respiratory drive, and other environmental sensors to assess air quality of the sleeping environment.

The sensor device consists of an electronics housing, electronics, and a flexible skin attachment element. In a preferred embodiment, the electronics housing is reusable and can be recharged. In another embodiment the electronics and housing are disposable. A new flexible skin attachment element can be attached for each use. The flexible skin attachment element can be an adhesive layer beneath the housing. Alternatively it can be a thicker layer with adhesive on both sides and a flexible layer in between made of a spacer material such as foam. In another embodiment the flexible skin attachment element can be an overpatch with an adhesive underside and a means to be applied or all or a portion of the housing to secure it to the user. A further alternative embodiment has the electronics housing located within a headband secured to the forehead. Another embodiment integrates the electronics into an eyeshade. A further embodiment secures the electronics in an ear plug which is fitted partway into the ear canal. Another embodiment incorporates the electronics into a sleep hat, placing the sensor over the forehead.

Sensor Locations

In one embodiment, the device is placed on the forehead. The forehead provides an ideal placement for gathering the signals of interest. In this embodiment, the device may be placed on the upper half of the forehead, closer to the hairline than the eyebrows. The upper forehead has excellent perfusion, resulting in a good signal. The device is placed anywhere aligned between the two eyes, above the eyebrows, and below the hairline. Placement in the upper portion of this described box often results in the best signals. The vasculature of the upper half of the forehead is characterized by multiple smaller vessels resulting in better perfusion. This area has a better AC/DC signal ratio on the PPG, which allows for a better signal to noise ratio.

Other locations where the device can be placed and achieve adequate signal include the ear lobe, the ear canal, behind the ear on the skull, the nostril, the nasal septum, the clavicle, the neck, the laryngeal prominence, the back, the sternum, the chest, or the belly among others.

User Interface

In one embodiment, a smartphone, tablet or similar communication device (hereafter smart device) is integral to the system. An application on the smart device collects information from the user. The application leads the user through a streamlined collection of user data including age, gender, height, weight and self-reported sleep quality. This collection may include a survey based on clinically validated sleep questionnaires such as STOP-BANG, Epworth Sleepiness Scale, Berlin questionnaire. These questions have been shown to identify apnea with good accuracy. Incorporating this data into our system enhances the accuracy of the assessment. Additionally, the application includes certain questions designed to identify individuals who may have more complex sleep disorders and may benefit from a polysomnography test. Questions about leg movements during sleep, as reported by the user or bed partners, can help identify Restless Leg Syndrome (RLS). The application survey may also include questions to identify users with more complicated health issues such as congestive heart failure or pulmonary insufficiency, which necessitate consultation with a physician.

Data Analysis

The data analysis platform takes data from the sensors over one or, preferably, multiple nights and combines it with the results from the questionnaires to create a very accurate, real-world, and personalized assessment of the user's sleep. The ability to combine measured, objective sleep data with user-reported data from validated questions is unique to this device. In one embodiment, the data from the objective measurements will be given priority over the user-reported data. However, in cases where the clinical conclusion from the measured data is not very clear, the subjective user-reported data is used to determine the next step for the user. For example, if the measured overnight SpO₂ data does not show significant desaturations, but the user reported data shows significant fatigue and sleep complaint, the application would suggest further study to examine the cause of the fatigue since OSA would not be suspected. In this scenario, the application would suggest investigating further testing for Respiratory Effort Related Arousals (RERA), insomnia, or other sleep disorders. This is just one example to illustrate the power of combining the measured sleep data with user reported data to help determine the cause of sleep issues. This data can optionally be compared to historical controls to give the user an idea of where they stand compared to their peers of similar age, gender, body mass index and sleep habits. Research shows that providing reference to a peer group enhances the impact of the sleep study results and can serve as a powerful motivator for behavior change.

Influencing Behavior

An important function of the system we describe herein is the ability to positively influence user behavior. One key way to achieve this is through a feedback loop about therapy. In addition to identifying and diagnosing sleep apnea, the system can provide feedback on the use of various treatment therapies. A PAP user can use the device to verify the efficacy of the PAP device. If desired, different levels of PAP therapy can be tested and assessed using the device. For newly diagnosed patients, feedback is especially important. The sensor can provide useful feedback for these users. Most new users of PAP therapy fall into one of two categories: those who immediately feel benefit from therapy and quickly become consistent users, and those who feel less acute immediate benefit. For the latter group, quantitative feedback on exactly how much PAP therapy is improving their breathing at night is extremely useful in helping convince them how the benefits outweigh the inconveniences of therapy. With the sensor, they can track their progress on PAP. Research has shown that usage during the first few weeks of PAP therapy is very indicative of long-term adherence.

Users who choose alternative therapies can benefit from the apnea detection system as well. There are several alternatives to PAP therapy. PAP therapy has repeatedly been shown to be the most effective, but only if used. For patients who cannot or choose not to use PAP, some options exist including: oral appliances, surgeries, positional therapy, lifestyle changes, oral pressure therapy. However, the efficacy of these therapies is difficult to assess. Many users are not followed closely or given sleep testing assessment once on these therapies, or following surgery. Consequently, they may be risking their health if their apnea is insufficiently treated. The apnea detection system can be used in conjunction with all alternative therapies to assess their efficacy. This improved access to thorough assessment empowers users to utilize alternative therapies with confidence, or discontinue spending time and money on approaches that do not benefit them.

BRIEF DESCRIPTION OF THE DRAWINGS

The novel features of the invention are set forth with particularity in the claims that follow. A better understanding of the features and advantages of the present invention will be obtained by reference to the following detailed description that sets forth illustrative embodiments, in which the principles of the invention are utilized, and the accompanying drawings of which:

FIG. 1 is a perspective view of a wearable sensor according to an embodiment of the invention.

FIG. 2 shows the system in use. The wearable sensor is on the forehead of a user. A smartphone is shown in the foreground, presenting the findings of the sensor.

FIGS. 3A and 3B are top views of wearable sensors showing embodiments of the invention.

FIGS. 4A, 4B, 4C, and 4D show one embodiment of the wearable sensor in multiple views. FIG. 4A is a perspective view; FIG. 4B is a top view; FIG. 4C is a side view; and FIG. 4D is another side view.

FIGS. 5A, 5B, 5C, and 5D show an embodiment of the wearable two-part sensor having a housing and a skin-contacting element. FIG. 5A is an exploded perspective view; FIG. 5B is a top view; FIG. 5C is an exploded side view; and FIG. 5D is another exploded side view.

FIG. 6 shows an exploded view of the wearable sensor with the housing, skin-contacting element, and some key internal components.

FIG. 7 shows a block diagram of an embodiment of the system.

FIG. 8 shows wearable sensor assembly adapted to be placed on the ear lobe.

FIG. 9 shows a wearable sensor assembly augmented by a thermistor and/or pressure sensors placed at or near the nostrils.

FIG. 10A shows a top view of a wearable sensor assembly according to another embodiment.

FIG. 10B shows a side view of the wearable sensor assembly of FIG. 10A.

FIG. 10C shows a perspective view of the wearable sensor assembly of FIG. 10A.

FIG. 11A shows a top view of a wearable sensor assembly.

FIG. 11B shows a side view of the wearable sensor assembly of FIG. 11A.

FIG. 11C shows a top perspective view of the wearable sensor assembly of FIG. 11A.

FIG. 11D shows a bottom perspective view of the wearable sensor assembly of FIG. 11A.

FIG. 12 shows a wearable sensor assembly positioned on the neck of a user.

FIG. 13 shows a wearable sensor assembly positioned behind the ear of a user.

FIG. 14 shows a wearable sensor assembly positioned inside the ear canal of a user.

FIG. 15 shows a wearable sensor assembly in two different locations: the forehead and either the chest, belly, or torso.

FIG. 16A shows a side view of a sensor according to another embodiment in place on a user's skin.

FIG. 16B shows a side view of the sensor of FIG. 16A.

FIG. 17 shows a bottom view of the sensor of FIGS. 16A-B.

FIG. 18A is another bottom view of the sensor with additional dimensions.

FIG. 18B shows a midline cut side view of the sensor.

FIG. 18C shows a detail cutaway side view of the photodetector.

FIG. 18D shows a detail cutaway side view of the emitter element.

FIG. 19 shows placement of the sensor on the user's forehead.

FIG. 20A shows a raw PPG signal from a supine positioned subject.

FIG. 20B shows a raw PPG signal from a left lateral positioned subject.

FIG. 20C shows a raw PPG signal from a right lateral positioned subject.

DETAILED DESCRIPTION

Referring to FIG. 1, a perspective view of one embodiment of a wearable sensor assembly 100 portion of the system is shown. The housing for the electronics 120 and the flexible skin attachment element 110 are shown assembled together. The electronics housing 120 can be made of a hard plastic, or a softer elastomer, or a combination thereof. The flexible skin attachment element 110 can be made of a foam, fabric, polymer, elastomer, or similar flexible material that is compatible with a skin-contacting application.

FIG. 2 shows an embodiment of the system in use. The wearable sensor assembly 100 is shown attached to the forehead of a sleeping user. A smartphone 130 is shown in the foreground displaying the results of the positional sleep apnea assessment by the system. A key feature of the system is shown in FIG. 2: the ability to indicate the degree to which sleep apnea is affected by sleeping position. This information empowers users to better understand and treat their apnea. For example, as shown in FIG. 2, the display of the smartphone 130 provides information about AHI when the user is on his back and on his side as well as the likelihood that the user will experience an apnea event.

In FIG. 3A an embodiment of the wearable sensor assembly 100 is shown with a flexible skin attachment element 110 made of elastomer or foam. FIG. 3B shows an embodiment of the wearable sensor assembly 100 with a flexible skin attachment element 110 made of fabric, paper, or woven material. In each case, the flexible skin attachment element is constructed to be able to conform to the anatomy of the wearer, particularly to the curvature of the forehead.

FIG. 4A shows a perspective view of the wearable sensor assembly 100 with the flexible skin attachment 110 and the electronics enclosure 120.

FIG. 4B shows a top view of the wearable sensor assembly 100.

FIG. 4C shows an end view of the wearable sensor assembly 100 illustrating a possible curvature of the flexible skin attachment 110.

FIG. 4D shows a side view of the wearable sensor assembly 100 illustrating a possible curvature of the flexible skin attachment 110.

FIG. 5A shows a perspective view of the wearable sensor assembly 100, with the two main elements separated for illustration. The flexible skin attachment element 110 can be placed over the electronics enclosure 120 to secure it to the user.

FIG. 5B shows a top view of the wearable sensor assembly 100.

FIG. 5C shows an end view of the wearable sensor assembly 100.

FIG. 5D shows a side view of the wearable sensor assembly 100.

FIG. 6 shows an exploded view of one possible construction of the wearable sensor assembly 100, illustrating several of the key components. The power supply 146 can be a lithium polymer or lithium ion battery. The power supply 146 could also be a primary cell, coin cell, fuel cell battery, or other rechargeable or single use chemistry cell or cells. The position sensor 140 can be an accelerometer, a 6-axis accelerometer, a magnetometer, gravitometer, gyroscopic sensor or other similar device for determining position, including a mechanical element such as a mercury switch. The wireless connection transmitter and receiver 142 can be a Bluetooth radio, Bluetooth Low Energy (BLE), radio, Zigbee, 802.11b, nRF, GSM, a module containing any of these elements, or other similar component to enable 2-way communication between the wearable sensor and a network or device such as a smartphone. The pulse oximetry measurement can be provided by a pulse oximeter module or individual components selected to perform pulse oximetry per the desired requirements. The pulse oximetry components are shown by 144. Signals from the sensor are sent to a microprocessor 148 for processing. The microprocessor 148 may be selected from a group of microprocessors standardly available of suitable specifications, or it may be custom fabricated specifically for the purpose herein. Memory (not shown) can be integrated as part of the microprocessor 148 or added as a separate component.

FIG. 7 is a block diagram showing the key components of the wearable sensing module 202 and their relationships to each other and to a remote connected device 216. Light signals are emitted by the pulse oximetry module 206, and received back by the module 206. These signals are sent to the microprocessor 210 where they are further processed. Additional sensing signals from the accelerometer 211 and microphone 212 are sent to the microprocessor 210 as well. Optionally, a pressure sensor 220 and flow sensor 222 can be added to the system. Flow can be derived by a pressure differential calculation as is known in the art, or by use of a thermistor 218. Pressure measurements and thermistor measurements would be taken just at/near the nostrils. Commonly, such sensors are placed on the upper lip just below the nose. Information from the various sensors and any calculations performed by the microprocessor may be sent wirelessly to a remote connected device 216 (such as, e.g., a smartphone or computer) by a transmitter/receiver 214. Further processing of the data from the sensors and information from the microprocessor may be performed by the connected device 216.

FIG. 8 shows wearable sensor assembly 300 adapted to be placed on the ear lobe. In this case, the pulse oximetry assembly would use transmittance oximetry instead of reflectance oximetry as is used on the forehead. This arrangement would also allow for the capture of head position during use. The user interface can specify which ear is to be used, or this can be discerned from the accelerometer data and an algorithm.

FIG. 9 shows a wearable sensor assembly 400 augmented by a thermistor and/or pressure sensors in a housing 404 placed at or near the nostrils. The sensors in housing 404 communicate with the forehead mounted sensor housing 402 via conductors 406. The device is shown with the nostril sensors and the forehead wearable sensor assembly. Alternatively, and not shown, the pressure and flow data could be coupled to an ear-worn sensor assembly. Further, the thermistor and pressure and flow sensors could be housed in a separate assembly with its own power supply, memory, processor, and wireless communication capability.

FIG. 10A shows a top view of the wearable sensor assembly 100.

FIG. 10B shows a side view of the wearable sensor assembly 100.

FIG. 10C shows a perspective view of the wearable sensor assembly 100.

FIG. 11A shows a top view of the wearable sensor assembly 100.

FIG. 11B shows a side view of the wearable sensor assembly 100.

FIG. 11C shows a perspective view of the top of wearable sensor assembly 100.

FIG. 11D shows a perspective view of the bottom of wearable sensor assembly 100.

FIG. 12 shows the wearable sensor assembly 100 positioned on the neck of a user. In this position on the front of the user's neck, the sensor can obtain signals for pulse oximetry, pulse, position, movement, snoring (auditory and vibrational signals), respiratory effort and respiratory flow movement.

FIG. 13 shows the wearable sensor assembly 100 positioned behind the ear of a user. In this position over the bone behind the user's ear, the sensor can obtain signals for pulse oximetry, pulse, position, movement, snoring—both auditory and vibrational signals, respiratory effort and respiratory flow movement. This position is out of the way for the user and should not interfere with normal sleep habits. Additionally, this location is less visible.

FIG. 14 shows the wearable sensor assembly 100 positioned inside the ear canal of a user. In this position the sensor can obtain signals for pulse oximetry, pulse, position, movement, snoring—both auditory and vibrational signals, and respiratory effort. This position is out of the way for the user and should not interfere with normal sleep habits. Additionally, this location is less visible.

FIG. 15 shows the wearable sensor assembly 100 in two different locations. One location is on the forehead. Another location is on the chest, belly, or torso. This can be achieved with two sensors worn simultaneously, or with one sensor worn in each location on different nights. The advantages of the forehead placement are discussed elsewhere. The chest placement offers several additional advantages. The motion of the chest during sleep yields a distinct motion signature from the sensor. This signal can be used to determine respiratory effort. Specifically, this is useful in differentiating Obstructive Sleep Apnea from Central or Mixed Sleep Apnea. Central Sleep Apnea is characterized by the absence of respiratory effort. Therefore, if a user has shown evidence of apnea with previous testing, using the sensor placed on the chest can determine if there are corresponding pauses in respiratory effort. By comparing the results from a sensor placed on the forehead to those from a sensor placed on the chest, it can be determined if Central Sleep Apnea is responsible. In some cases, this signal differentiation will be strong enough to indicate a diagnosis. In some cases, the signal differentiation may only be strong enough to indicate that further testing is justified to definitively diagnose the user. Central sleep apnea (CSA) is a neurologic disease in which breathing stops during sleep even though the airway is open. Both OSA and CSA disrupt sleep and carry significant long-term health consequences. These disorders require different therapies, and thus it is important for a diagnostic test to distinguish them.

FIG. 16A shows an embodiment of the wearable sensor assembly 100 in side view placed onto the skin tissue 150. The sensor interface element 160 extends downward from a peripheral portion of the housing 105 to form an optical engagement surface to directly couple the optical elements 170 with the skin tissue 150. This direct coupling helps ensure a quality raw optical signal is captured by the photodetector. FIG. 16B shows a side view of the wearable sensor assembly 100 of this embodiment. The manner in which the optical coupling elements 170 protrude into the surface of the skin, beyond the surface of the flexible skin attachment element 110, is shown. A typical distance for this approximately 0.5 mm, with a typical range of 0.25 mm to 0.75 mm. However, this distance may vary to be as small at 0.1 mm or as large as 2.5 mm or larger. The elements must contact the skin sufficiently to produce the desired optical coupling and without an air gap for the optimal optical signal. However, if the elements protrude too far into the skin, they can significantly decrease blood perfusion and result in a poorer photoplethysmography signal. The optimal distance, as described herein, is needed to result in a favorable signal. The optical coupling elements 170 are comprised of optically-clear materials. To successfully optically couple the optical components with the skin, it is important that any air gaps are eliminated. Thus, a material that can eliminate air gaps and is optically clear fits two of the key parameters. One such material used in the instant invention is a 2 part epoxy. This material can flow during manufacturing to eliminate air gaps and is optically clear upon hardening.

FIG. 17 shows an underside view of another embodiment of a wearable sensor assembly 100, the side that is in contact with the skin tissue. Of note is the optical engagement surface of the sensor interface element 160, which engages the tissue. Within this element is the light source 180 and the photodetector 190. As is well known in pulse oximetry, the light source 180 may emit light at two different wavelengths (e.g., red light and infrared light), and the photodetector 190 may detect the time-based changes in absorption of the two light wavelengths to be used in a calculation of the user's blood oxygen saturation. A key dimension is the noted by A, the distance from the center of the energy emitting element 180 to the center of the energy detecting element 190. In one embodiment, this dimension A is 7 mm. In other embodiments, it can range from 5 mm-10 mm and yield a quality signal. It is important to be far apart enough so that the light energy emitted from the light source 180 must travel through enough blood-perfused tissue, including pulsatile blood, to result in a signal with enough time-variable, or AC, component to allow for an accurate blood oxygen saturation measurement. The emitting and detecting elements should be close enough together so that the detected time-variable signal is primarily composed of light energy from the emitter and not any other sources.

FIGS. 18A and 18B show an underside view and a side view, respectively, of one embodiment of the wearable sensor assembly 100. The optical engagement surface of the raised sensor interface element 160 performs at least two key functions. One, it helps optically couple the device's optical components with the skin, as discussed herein. Second, it is also constructed of a light-blocking material specifically chosen to not allow passage of the wavelengths of interest emitted by the device. Some possible materials include thermoset and elastomeric materials which absorb the wavelengths of interest, including but certainly by no means limited to: ABS, polycarbonate, Santoprene™ thermoplastic, butyl styrene, and others. The wavelengths of interest are typically nominal 670 nm, 850 nm, and 900 nm.

FIG. 18B shows a side view of one embodiment of the wearable sensor assembly taken along the line B-B in FIG. 18A. Key components within the assembly include the photodetector element, the energy emitting element, the accelerometer, the wireless communication component, the microprocessor, and the battery. Also shown is the charging port, herein pictured as a USB connector in FIG. 18A. The adhesive flexible skin attachment element has been omitted from the figure for clarity. The microprocessor controls the function of the device. Energy is emitted from the energy emitting element into the tissue. In a preferred embodiment, the emitted energy is light in two distinct wavelengths of approximately 650 nm and 930 nm. Other wavelengths are possible, with one being in the red spectrum and the other in the infrared spectrum. This light passes through the tissue and then a portion of it is received by the photodetector. The photodetector receives the light energy and outputs a corresponding voltage to the strength of the intensity corresponding to each wavelength. From this data, the PPG waveform is determined, and the SpO2 is calculated. The PPG is the graph of the AC component of the intensity of these received signals over time. The SpO₂ is calculated by: for a given time t, calculating ΔA_red=(Red_t−Red_t-1)/((Red_t+Red_t-1)/2) and ΔA_IR=(IR_t−IR_t-1)/((IR_t+IR_t-1)/2); then graph ΔA_red vs ΔA_IR, the slope of the best fit line is then R. The R value is then adjusted by coefficients to yield the SpO₂, for instance using: (x)R²+(y)R+z=SpO₂. In one embodiment, the SpO₂ is calculated by the microprocessor on the device. In another embodiment, the raw signal data from the photodetector and accelerometer is sent by the wireless communication component to another device or service for processing. In yet another embodiment, the raw signal data undergoes preprocessing prior to being sent by the wireless communication component to another device or service for further processing.

FIG. 18C shows a detailed side view (marked “C” in FIG. 18B) showing the photodetector 190 in the housing 105. Of note is a tapered frustoconical surface 194 extending from the photodetector 190 and tapering down to an opening 196 in the housing. This downward taper angle is important to allow the maximum amount of light signal to reach the sensor. In one embodiment, the angle of the taper is 45 degrees, although it can range from 5-85 degrees. The dimension G denotes the diameter of the opening of the aperture. In one embodiment, this dimension G is slightly smaller than the active surface 192 of the photodetector 190. FIG. 18C also shows a dimension H, the distance from the surface 192 of the photodetector 190 to the opening 196 in the housing. In one embodiment, this distance is minimized to allow for the optimal signal to pass through. Optically clear material 198 is disposed in the volume extending between the opening 196 and the active surface 192 of the photodetector 190. Other device components supported within the housing include a position sensor (e.g., accelerometer), a controller and a communicator (e.g., wireless transmitter).

FIG. 18D shows a detailed cutaway side view of the light source 180. In one embodiment, the W dimension, which describes the light source opening 184 in the housing, is sized to be smaller than the active area 182 of light source 180. Dimension K shows the distance from the surface of the energy emitting element to the exterior surface of the housing. In one embodiment, this distance is minimized to allow for an optimal amount of light to pass through. Optically clear material 186 is disposed in the volume extending between the opening 184 and the active area 182 of the light source 180. In one embodiment, the light source is a pair of light emitting diodes (LED). In one embodiment, these LEDs operate in the wavelengths of approximately 650 nm and approximately 930 nm. Many other wavelengths are possible. Some specific wavelengths found to work well are 656-657 nm and 930-932 nm.

FIG. 19 shows a placement zone of the device 100 for on the forehead 200 of the user according to one embodiment of the invention. While it is possible to get adequate readings anywhere on the forehead, placing the device 100 in the upper half of the forehead results in significantly better signal quality. The zone illustrated in FIG. 19 is above and between both eyes, and the upper half of the forehead as defined by the area between the eyebrows and the hairline. This area is characterized by more microvasculature and smaller vessels which typically result in a better perfusion signal. This yields a larger AC component of the PPG waveform, allowing for more detail to be derived from the signal. The readings taken from this zone can often be doubly strong, with twice the perfusion index, as readings taken from the lower half of the forehead. This upper half zone is preferred for the most accurate signal results.

FIG. 20A shows a portion of the PPG signal taken from a subject in the supine position. The amplitude of the signal indicates the perfusion index, the ratio of the pulsatile bloodflow to the non-pulsatile blood. The signal also shows a characteristic shape with relatively narrow peaks. In contrast, the waveforms shown in FIG. 20B and FIG. 20C show a lower amplitude and perfusion index that is about one-half that of FIG. 20A. Additionally, the waveform in FIG. 20B is flattened on top and shows a distinctly different shape than that of FIG. 20A Likewise, the waveform in FIG. 20C is also more rounded on top and shows a distinctly different shape than that of FIG. 20A. These changes in the waveform and the perfusion index can be used to determine the position of the sleeper. First, the PPG waveforms are characterized in each of the supine and non-supine cases, which allows threshold levels to be determined for each case. These threshold levels are used to compare to a given PPG waveform to output a position determination of the sleeper. In another embodiment, the waveforms for supine and non-supine sleep are compared using digital signal processing techniques to characterize the shape of each waveform. This is used to determine thresholds for waveform shape, which are then applied to a given PPG waveform to determine the user's position.

Another advantage of the system and methods described herein is the ability to identify other oxygen-related illnesses and conditions. Such conditions include altitude illness, lung diseases, COPD, emphysema, bronchitis, pulmonary edema, fibrosis, pneumonia, congenital defects, congestive heart failure, anemia, narcotics use, anesthetics use, and certain poisonings like cyanide.

Claims

1. An obstructive sleep apnea detection device comprising: an optical engagement surface adapted to engage a user's skin;a light source adapted to emit light from the optical engagement surface;a photodetector adapted to detect light at the optical engagement surface and to generate a detected light signal;a position sensor adapted to determine patient sleeping position;a controller adapted to determine and record in memory blood oxygen saturation values computed from the detected light signal and user position information from the position sensor;a housing supporting the optical engagement surface, the photodetector, the light source, the position sensor, and the controller; anda tapered frustoconical surface extending from a photodetector opening in the optical engagement surface to an active surface of the photodetector, a diameter of the frustoconical surface adjacent the photodetector being smaller than a diameter of the active surface of the photodetector.

2. The device of claim 1 further comprising a communicator adapted to communicate the blood oxygen saturation values and user information position information.

3. The device of claim 2 wherein the communicator comprises a wireless transmitter disposed in the housing.

4. The device of claim 1 wherein the optical engagement surface is shaped to engage skin on a forehead of the user.

5. The device of claim 1 wherein the optical engagement surface is disposed 0.1 mm to 2.5 mm from the lower surface of the flexible material.

6. The device of claim 1 wherein the optical engagement surface is disposed 0.25 mm to 0.75 mm from the lower surface of the flexible material.

7. The device of claim 1 wherein the flexible material comprises adhesive.

8. The device of claim 1 wherein the optical engagement surface comprises material preferentially absorbing light wavelengths corresponding to light wavelengths emitted by the light source.

9. The device of claim 1 wherein the light source is configured to emit light through a light source opening in the optical engagement surface and the photodetector is configured to detect light through the photodetector opening in the optical engagement surface, the light source opening and photodetector opening being disposed 5 mm-10 mm apart.

10. The device of claim 1 further comprising an opening in the housing adjacent the light source, the opening having a diameter smaller than a diameter of an active area of the light source.

11. The device of claim 1 wherein the position sensor comprises an accelerometer.

12. The device of claim 1 wherein the position sensor is adapted to determine patient position from the detected light signal or from the blood oxygen saturation values.

13. A method of obtaining obstructive sleep apnea information from a subject, the method comprising: adhesively attaching a flexible material of a detection device to a skin surface of the subject;protruding an optical engagement surface of the detection device below a lower surface of the flexible material into a skin surface of the subject;transmitting light through the optical engagement surface into the skin surface of the subject;detecting light from the skin surface of the subject through the optical engagement surface;determining blood oxygen saturation from the detected light;obtaining subject position information from blood oxygen saturation; andstoring the blood oxygen saturation and position information in memory.

14. The method of claim 13 wherein the skin surface comprises a skin surface on the subject's forehead.

15. The method of claim 13 further comprising time correlating blood oxygen saturation with position information.

16. The method of claim 13 further comprising communicating the blood oxygen saturation and position information.

17. The method of claim 16 wherein communicating comprises displaying blood oxygen saturation and position information.

18. The method of claim 16 wherein communicating comprises wirelessly transmitting the blood oxygen saturation and position information.

19. The method of claim 18 wherein communicating comprises wirelessly transmitting the blood oxygen saturation and position information from a transmitter within the device.

20. The method of claim 13 further comprising gathering position information from a second position sensor disposed on the subject's torso.