SYSTEM, METHOD AND STORAGE MEDIUM FOR MONITORING PHYSIOLOGICAL PARAMETERS - Patent application

The RF sensor system with adaptive signal processing addresses interference and false triggers in motion sensors, enabling reliable physiological monitoring and authentication by adjusting sensor parameters for optimal performance.

JP7577785B2Active Publication Date: 2024-11-05RESMED SENSOR TECH LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
JP2023074459
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2015-08-20
Filing Date
2023-04-28
Publication Date
2024-11-05
Estimated Expiration
2036-04-20

AI Technical Summary

Technical Problem

Existing motion sensors, such as continuous wave Doppler radar and pulsed Doppler sensors, face issues with false triggers, high sensitivity, and interference, particularly in intermediate range detection, and regulatory approval challenges, limiting their effectiveness in physiological monitoring and authentication.

Method used

A system utilizing radio frequency (RF) sensors with distance gating and adaptive signal processing to minimize interference, detect physiological characteristics like breathing and heart rate, and authenticate individuals based on unique biometric signals, incorporating processors to adjust sensor parameters for optimal performance and minimize interference.

Benefits of technology

The system effectively distinguishes between individuals by analyzing respiratory and cardiac features, reducing false triggers and interference, ensuring reliable authentication and monitoring, even in complex environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007577785000005
    Figure 0007577785000005
  • Figure 0007577785000006
    Figure 0007577785000006
  • Figure 0007577785000007
    Figure 0007577785000007
Patent Text Reader

Abstract

To provide a system for detection and identification of a human.SOLUTION: One or more sensors are configured for detection of characteristics of moving objects and living subjects for human identification or authentication. One or more processors, such as in a system of sensors or that control a sensor, may be configured to process signals from the one or more sensors to identify a person. The processing may include evaluating features from the signals such as a breathing rate, a respiration depth, a degree of movement and a heart rate, etc. The sensors may be radio frequency non-contact sensors with automated detection control to change detection control parameters on the basis of biometric identification, such as to avoid sensor interference.SELECTED DRAWING: Figure 7
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] (CROSS REFERENCE TO RELATED APPLICATIONS) This application is a continuation of U.S. Provisional Patent Application No. 62 / 149,839, filed April 20, 2015. No. 62 / 207,687, filed on August 20, 2015. No. 60 / 339,993, filed on May 13, 2003, the disclosure of which is hereby incorporated by reference in its entirety.

[0002] FIELD OF THEINVENTION The present technology relates to circuits and sensors for detecting characteristics of moving and living subjects. More specifically, the present technology focuses on signal characteristics specifically for human identification and authentication, e.g. For sensors such as distance-gated radio frequency motion sensing in health monitors, etc. Regarding. [Background technology]

[0003] Continuous wave (CW) Doppler radar motion sensors emit a continuous wave radio frequency (RF) carrier. It transmits a 100-MHz beam and mixes the transmitted RF with the return echoes to detect dots generated by moving targets. These sensors have a certain limit range. (i.e., they can receive signals for both near and far objects, and these receivers (The signal received is a function of the radar cross section.) This can result in false triggers, i.e. These can result in artifact interference. It may also have an undesirably high sensitivity.

[0004] A pulsed Doppler motion sensor is disclosed in U.S. Pat. No. 4,197,537 (Follen et al.) A short pulse is transmitted and its echo self-mixes with the transmitted pulse. The pulse width defines the range-gated region. When the transmit pulse ends, The mixed end and target return arriving after this time will exit unmixed.

[0005] U.S. Patent No. 5,966,090, entitled "Differential Pulse Radar Motion Sensor" (Mc A differential pulse Doppler motion sensor disclosed in Ewan et al. alternates between two different pulse widths. Then, by subtracting the Doppler response from each width, you get a fairly constant response vs. range. The system generates a range-gated "Doppler" sensing region that detects the

[0006] U.S. Patent No. 5,361,070 (McE Impulse radars, such as those described in

[1996] and

[2000] , have a very high frequency response related to the transmitted pulse width. It produces a narrow sensing area. It is called the "Pulsed Homodyne Field Disturbance Sensor" and is a US A two-pulse Doppler array such as that described in Patent No. 5,682,164 (McEwan) The dummy sensor transmits a first pulse and then a second pulse that mixes with the echo from the first pulse. Therefore, the range-gated sensing band has a defined minimum and maximum The UWB radar motion sensor is a worldwide RF They have the disadvantage of not having regulatory approval and have difficulty detecting objects at mid-range. and in some embodiments may be susceptible to RF interference.

[0007] A modulated pulse Doppler sensor is described in U.S. Patent No. 6,426,716 to McEwan. Range-gated microwave motion sensor with adjustable minimum and maximum detection These devices include associated power supplies to generate transmit and mixer pulses. RF oscillator with pulse generation and delay elements, signal transmission (TX) / reception (RX) antennas, or a pair of separate TX and RX antennas, and a mixer and range gated from the echo pulse. associated filtering, amplifying and demodulating elements to produce a Doppler signal. The RF receiver includes a detector / mixer having a

[0008] In U.S. Pat. No. 7,952,515, McEwan describes a particular holographic The company has disclosed a radar that adds a range gate to the holographic radar to detect specific down targets. McEwan specifically limits the response to the range region. When penetrating a material to image a defect, a cleaner crack on the imaged surface is obtained. This demonstrates that a taffle radar hologram can be obtained. This enables stacked hologram technology, where multiple imaged surfaces are The layers may be stacked in multiple directions.

[0009] In U.S. Pat. No. 7,994,968, McEwan describes a holographic radar. McEwan discloses an RF magnitude sampler for a narrow band Resolving interferometric patterns produced by wide-area holographic pulse radar. It describes what is possible.

[0010] In U.S. Patent Publication No. 2014 / 0024917, McMahon et al. configured to generate a vibration signal for emitting a gated sensing radio frequency pulse; The present invention describes physiological sensing sensors that can be configured to emit pulses. A radio frequency transmitter and a reflected pulse among the emitted radio frequency pulses are The received pulses include a receiver configured to receive a signal representing motion, sleep, breathing, and / or a time course of motion. Or it may be processed to detect physiological characteristics such as heartbeat.

[0011] Wearable wristbands like Nymi are a system of authentication - unique to each user. We introduce other variables that can be used in the ECG shape. nymi.com / )

[0012] US20100191076 A1(Aaron Lewicke,Yi Zhang , John D. Hatlestad) described daytime / nighttime respiratory rate monitoring. There are.

[0013] US8232866(William R. McGrath,Ashit Talu kder is a remote long standoff biometry system using microwave cardiac signals. This describes the block identification.

[0014] US832100 (Seth Snyder, Jasper Speicher) A biometric data display system and method are described.

[0015] WO2003000015(Mark D Wiederhold, Rodney P Meyer, Steven A Israel, John M Irvine Identification by analysis of physical and measurable changes is described.

[0016] US8762733(P.Alexander Derchak,Lance Mye rs) is an idea that uses physiological biometrics to determine physiological fingerprints. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A system and method for identity verification is described.

[0017] US20110040574(Ho Chung Nicholas Fung,Ch Yong Sang, Ph.D., a biometric health monitoring system This article describes the system.

[0018] US6993378(Mark D.Wiederhold,Steven A.I srael, Rodney P. Meyer, John M. Irvine The present study describes discrimination by analysis of measured changes.

[0019] Improve the sensor and / or its signal processing to sense characteristics for identification or authentication. There may be a need to improve.

[0020] For authentication and other purposes, it is preferable to be able to identify an individual.

[0021] The advent of wearable, contactless physiological and behavioral data collection has made it possible to identify specific individuals based on their personal biometrics. It detects and identifies your password from your fingerprint, rejects your data from other people (forged files), and to assess compliance with and use of the sensor (and potentially any delivered or associated services or treatments) Such characteristics can be derived from physiological and behavioral signals. can.

[0022] Detects sensor data being collected from specific users and micro- and macro-environments In addition to detecting the signal, the system can detect deviations from a normal (healthy) signal. for changes such as a worsening condition or an improvement due to treatment / therapy It is preferable that it is robust. Summary of the Invention

[0023] One aspect of some embodiments of the present technology is, for example, detecting physiological characteristics with radio frequency signals. This relates to sensors for detecting

[0024] Another aspect of some embodiments of the present technology is that the detection is performed using, for example, radio frequency signals. The present invention relates to a sensor for authenticating an individual from a signal characteristic obtained by the sensor.

[0025] Some versions of the technology may measure one or more human physiological parameters, such as health monitoring. The present invention may include a method or system for identifying a person for monitoring a parameter. The system may include one or more sensors for monitoring one or more physiological parameters of a person. The system may include one or more sensors for identifying a person. The one or more processors may be configured to process signals from the The process may include detecting respiratory, cardiac or motion features, such as detected respiratory rate, ... Evaluation of features including any one or more of depth, degree of detected motion, and detected heart rate. and the evaluation may include an evaluation comprising an analysis of one or more of the detected features. Cut.

[0026] In some versions, the one or more sensors include a radio frequency non-contact sensor. The detecting may include detecting a physiological characteristic during sleep of the person. The process may include detecting deep sleep. The process can include detecting REM sleep, said detection comprising detecting physiological characteristics of the person during wakefulness. The one or more sensors may include a detection of The one or more processors may include any one or more sensors described herein. The evaluation of the sensor signal may include detection and analysis of any one or more characteristics of the sensor signal described herein. It is possible.

[0027] In some versions, the one or more sensors include at least one of the one or more sensors. At least two sensors are constructed to minimize interference, such as radio frequency (RF) interference. The one or more sensors may be configured to perform range gating pulse timing, RF pulse the emission power level of the pulses, such as change the detection frequency of pulses, such as RF pulses, and / or the use of steerable antennas The adjusted position may be configured to minimize interference, such as RF interference. The one or more sensors may communicate via a wired or wireless link.

[0028] In some versions, the physiological recognition of the independent human source is optimized and other sources The biometric authentication technology can be adapted to dynamically adjust the performance of one or more sensors to reject unauthorized access. The system includes a control processor in communication with one or more sensors. The control processor may include a sensor that communicates with the one or more sensors to control the one or more Based on the identification of a person or an animal, by adjusting the detection control parameters of the sensor, The detection control parameters of one or more sensors can be adjusted. The detection control parameters are range gating, RF center frequency and RF power level. may include one or more of the following:

[0029] In some versions, the system includes a control processor in communication with one or more sensors. The control processor may include a human or motion sensor. The sensor may be configured to activate a further sensor system based on the identification of the object. The system may include a camera. Optionally, in the system, the processor: The detected biometric information is based on the identity of the person who was previously identified as a monitored person. ometric characteristics (e.g., as in a database associated with a person's identity) Some versions of the system may log or allow the user to log the In John, the processor identifies a person who has not been previously monitored. and suppressing logging of detected biometric characteristics based on the identified person. In some versions, the processor may be configured to enable or inhibit The selection may be configured to initialize biometric characteristic detection for a particular individual. Optionally, the processor may assign the newly detected biometric characteristic to the initialized biometric characteristic. The assessment can be performed by comparing the newly detected background characteristics with the background characteristics of the person. comparing the biometric characteristic with the initialized biometric characteristic. can.

[0030] Optionally, the evaluation may include a classification of features determined from the signal. The features may include one or more of the following: spectral peak ratio; Setting Optimizer Flag Vector; Peak-Trough Ratio; Filtered Respiration Rate; Breath Rate Inhalation fluctuation measure; In-band power of the sensor signal; Range of the sensor signal; Final respiratory rate; Maximum of the respiratory cycle Amplitude to minimum amplitude ratio; Highband power of the sensor signal; Average respiration rate; Periodic leg movement activity detection; Time turnover detection or turnover detection; and post-processing movements.

[0031] Optionally, said evaluation includes evaluating said characteristics, such as cardiac parameters, galvanic skin response parameters, data, exercise intensity parameters, breathing parameters, blood pressure parameters, coughing The signal may include one or more of a sleep parameter, a snoring parameter, and a sleep parameter. The classification may include a classification of a specified feature.

[0032] The evaluation may include comparing the determined characteristics with past characteristics. In some versions, the evaluation is based on the determined characteristics and / or average over a period of time. It may further include calculating a standard deviation value.

[0033] In some versions, one or more processors of the monitoring system may Received data detected by one or more sensors from the monitoring system is evaluated. Evaluation of the received data detects equivalence between multiple health monitoring devices, etc. Each monitoring device may include a setting classifier and a subject classifier. Each classifier may be configured to evaluate features from the signal to identify a person. The subject classifier is further configured to evaluate the past features to identify the person. The health monitoring device or system may include cameras that are sensitive to infrared light and infrared emitters. In some versions, the health monitoring device may further include a One or more processors in such a monitoring system detect events from the signal and To associate the event with a particular portion of a video captured by a camera that contains the event. The monitoring device or system may be configured as follows: The battery may further include a coil for wireless charging.

[0034] In some versions, the one or more processors detect biometric data indicative of an animal. When detecting the lock characteristics, the sensor detection power, sensor frequency, sensor distance gating or detection The device is configured to control one or more sensors to change other control parameters for the output. The one or more processors may generate a notification based on an identity of the person associated with the configuration parameters. access to setting parameters for any one or more of the light, sound and / or environmental devices; The one or more processors may be configured to access the detected The configuration parameters associated with the biometrics used to determine the lighting, sound and / or environmental equipment Any one or more of the settings may be configured to trigger.

[0035] In some versions, the system is in communication with the collection of sensors. The central controller may include a central controller that may be configured to recognize the positions of identified people within the structure. In some versions, one or more sensors may be configured to provide position confirmation. The one or more sensors may include a pass-through outlet. One or more sensors may include an AC plug and an AC power pass-through socket to power the sensor. One or more of the sensors may include a US power outlet for powering the sensor. One or more sensors may include wireless or or a network interface for wired network communication.

[0036] In some versions, the one or more processors determine the respiratory rate based on the identified person. The device may be configured to configure the operation of the therapy device. The configured operation may be The setting action can cause the respiratory treatment device to change the treatment. Can or can be changed.

[0037] Optionally, the one or more processors are operable to determine whether a biometric characteristic is evaluated in the identification. When the information is processed by a respiratory treatment device, the device is prompted to retrain for the identification of the person. The one or more processors may be configured to: The one or more processors may be configured to adjust the operation of the detected biometric data. Rely on different biometric characteristics to identify people depending on the quality assessment of their biometric characteristics The one or more processors may be configured to perform biometric processing to identify a person. The system may be configured to operate a registration process to initialize a baseline of network characteristics. The registration process may include guided breathing sessions or spontaneous breathing sessions. In some versions, one or more processors may configured to reject a biometric characteristic associated with identifying a person .

[0038] In some versions of the technology, one or more processors of the system are identified The one or more processors may be configured to set an alarm action based on the person who entered the room. The system may be configured to identify a primary user from one or more other users. One or more user parameters may be tracked over time to determine one or more respiratory, cardiac or or motion features. In some such cases, one or more of the respiratory, cardiac, or motion features may be detected. One of the following respiratory parameters: range, respiratory variability, shape, and inspiration / expiration ratio. Contains one or more of the following:

[0039] Optionally, the one or more processors may generate a user-defined classification profile based on the characteristics determined in the classification process. The classification process may be configured to classify the identity of the , hidden layer Markov model, logistic regression processing, linear kernel support vector machine The methods may include one or more of a radial kernel support vector machine and a radial kernel support vector machine. The classification process uses principal component analysis on a feature or set of features prior to classification. The classification process may include classifying real-time and offline characteristics. The classification process may include determining the identity of the user or primary user. Lazy integration of multiple classifiers and their outputs to generate an output posterior probability, such as the probability for or the output of a classifier.

[0040] In some cases, the system or method may be configured to detect the same or different persons. The system may include a plurality of sensors, the system being adapted to detect and control a sensed control parameter of the plurality of sensors. It can automatically adjust parameters such as: range, power, frequency, etc. The at least one of the number, the detection direction, and the radiation pattern.

[0041] Other aspects, features, and advantages of the present technology are part of this specification and are described by way of example only in connection with the principles of the technology. Further aspects of the present technology will become apparent from the following detailed description taken in conjunction with the accompanying drawings, in which: What is claimed is: [Brief description of the drawings]

[0042] Further exemplary embodiments of the present technology will now be described with reference to the accompanying drawings.

[0043] [Figure 1] FIG. 1 is a diagram of an exemplary detection apparatus suitable for implementation with a radio frequency sensor in some versions of the present technology. [Figure 2A] FIG. 1 illustrates a conceptual structure and process flow for evaluation of sensor signals suitable for some versions of the present technology. [Figure 2B] FIG. 13 illustrates further processing of the sensor signal for detection of an exemplary physiological indicator. [Diagram 3] 1A-1D illustrate a processing system and flow for calculating sleep and daytime human biometrics, respectively. [Figure 4] 1A-1D illustrate a processing system and flow for calculating sleep and daytime human biometrics, respectively. [Diagram 5] FIG. 2 illustrates a process for the calculation of characteristic features to generate a biometric fingerprint. [Figure 6] FIG. 1 illustrates the system process for enrolling (training) an identity and subsequent verification or rejection. [Figure 7] FIG. 1 illustrates the system process for enrolling (training) an identity and subsequent verification or rejection. [Figure 8]FIG. 13 shows heart rate and respiration rate signals for subject “User A” during a sleep period (approximately 8.3 hours) recorded using a non-contact pulsed continuous wave RF sensor at a distance of less than 1.8 m. [Figure 9] FIG. 9 shows heart rate and respiration rate histograms for the subject “User A” of FIG. 8 during a sleep period (approximately 8.3 hours) recorded using a non-contact pulsed continuous wave RF sensor at a distance of less than 1.8 m. [Figure 10] FIG. 13 shows signals with heart rate and respiration rate for subject “User B” during a sleep period (slightly over 6 hours) recorded using a non-contact pulsed continuous wave RF sensor at a distance of less than 1.8 m. [Figure 11] FIG. 11 shows heart rate and respiration rate histograms for subject “User B” of FIG. 10 during a sleep period (approximately 8.3 hours) recorded using a non-contact pulsed continuous wave RF sensor at a distance of less than 1.8 m. [Figure 12] 13 shows the cross-spectral density between HR and BR for a paced breathing frequency of 0.1 Hz. Registration can include paced breathing periods as well as spontaneous breathing. [Figure 13] FIG. 1 shows two sensors monitoring two people in a default configuration, where the biometric quality is degraded as the sensors receive overlaid respiration and heart rate signals from each individual, and movement of both individuals is detected. [Figure 14] Shown is two sensors monitoring two people. Biometric quality is excellent as the sensors are configured to minimize range and force to maximize clear and separated biometrics. [Figure 15] Figure 1 shows two sensors monitoring one person in a default configuration, where sensor_a "sees" person_b and in response a duplicate biometric is detected on at least one of the sensors the biometric quality is degraded. [Figure 16]Two sensors monitoring a person in a configuration designed to minimize range and force to maximize distinct and separated biometrics. Sensor_a goes into power save / search mode. Sensor_b detects distinct biometrics of person_b. [Figure 17] This diagram shows two sensors monitoring a bedroom where an animal (dog) has entered the room and is lying on the bed. Sensor_a goes into power saving / search mode while sensor_b monitors the non-human's heart rate and respiration rate signals. [Figure 18] A diagram showing three sensors monitoring three humans while having interference from other lifeforms (e.g., a dog). [Figure 19] FIG. 1 illustrates the processing of an exemplary classification system for user identification. [Figure 20] FIG. 1 illustrates the processing of an exemplary classification system for identification of two (or more) users in a sleep (or awake) environment in which sensor data is compared to confirm that the user detected at the bedside surface is the expected user. [Figure 21A] Fig. 1 shows a system according to the present technology: A patient 1000 wearing a patient interface 3000 receives a supply of pressurized air from an RPT device 4000. The air from the RPT device 4000 is humidified in a humidifier 5000 and travels along an air circuit 4170 to the patient 1000. A bed partner 1100 is also shown. [Figure 21B] FIG. 1 shows an RPT device 4000 being used with a patient 1000 having a nasal mask patient interface 3000. [Figure 21C] FIG. 1 shows an RPT device 4000 being used with a patient 1000 having a full-face mask style patient interface 3000. [Figure 22] FIG. 3 shows a non-invasive patient interface 3000 in the form of a nasal mask. [Figure 23A] FIG. 4000 illustrates an RPT device in accordance with one form of the present technology. [Figure 23B] FIG. 40 is a schematic diagram of a pneumatic circuit of an RPT device 4000 in accordance with one form of the present technology, with upstream and downstream directions indicated. [Figure 23C] FIG. 40 is a schematic diagram of electrical components of an RPT device 4000 according to one aspect of the present technology. [Figure 23D] 9D is a schematic diagram of an algorithm 4300 implemented in an RPT device 4000 according to one aspect of the present technology. In Fig. 9D, the solid arrows indicate the actual flow of information, for example by electronic signals. [Figure 23E] 9D in accordance with one aspect of the present technology. [Figure 24] FIG. 5 shows a humidifier 5000. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0044] 1. Background Looking at the current state of health care and the future of personal health, we see three interconnected categories: The challenges are clear: (i) collecting and monitoring data from our bodies; (ii) health budgets under pressure; and (iii) automation of healthcare.

[0045] (Early on the "Quantified Self" movement) As people receive more and more health education, they will be able to live longer. are alive, demanding more from their doctors, and are on a "power curve" There is some evidence that the “sales curve” is moving toward consumers. For a product to be successful, it must be attached to people's lives. , and should become a necessity. Prevention is much cheaper than cure, and health systems can You pay a doctor to take good care of you, but when you get sick, you don't pay. (i.e., in contrast to the way medicine has typically been paid for in the Western world) ) ancient Chinese system - we should move towards results-based measures.

[0046] The blurring of medical devices and services has been accompanied by the advent of automation and expert systems in medicine. These changes have led payers (e.g., insurance companies) to make funding decisions on a case-by-case basis. It is possible to seek to eliminate the gap and proactively manage hospital admissions in a planned manner. This facilitates the ability to monitor chronic diseases in the home.

[0047] 2. User Identity Therefore, to avoid impersonation, the biometric and behavioral data of the identified user To detect abnormal changes in user-specific patterns (i.e., to identify users), It is possible to distinguish between two people, known here as a generic biometric "fingerprint," A system that can

[0048] Unlike an actual human fingerprint or perhaps an iris scan, the "bio- A "metric fingerprint" necessarily implies a fuzzy (less precise) picture of a user's identity. It should be noted that the method is a 2D estimation and includes both biometric and behavioral modalities. Such physiological and behavioral parameters should be considered (e.g., whether the user feels more pain or not). Over time, users will improve their quality of life (by eating less, becoming healthier, and experiencing other life changes). It is expected to evolve with each generation.

[0049] The outlined system is intended to be able to detect user-specific patterns. Examples of uses include improving or managing a user's health and fitness level or therapeutic treatments. To encourage compliance (e.g., to reduce premiums or payment entitlements, or to provide additional to reduce out-of-pocket costs), authenticate users, and potentially monitor adherence to prescribed treatments This includes reviewing and / or authorizing payments or other incentives. In some cases, the detection of a particular user may be difficult (e.g., by using sensors to distinguish between different sessions). The current user whose biometric health parameters have been previously detected (e.g. during a session) ,To ensure that the health parameters are the same as the previous user ,detected previously. In such cases, previously detected health parameters from multiple sessions can be used. The analysis of data is of a single user, not mixed with the health parameters of different users. Thus, the sensor records of the detected parameters can be verified against the previously identified user. This can be based on biometric verification / authentication of the user.

[0050] The technology includes processes, such as software algorithms, and one or more sensors. read data from the servers and services, process the data, and use the newly collected data and A system that adapts system parameters based on a previously collected data set. Some or all of these sensors may be used to communicate physiological information to the end user of the system. Collect objective data readings.

[0051] One example of such a software algorithm and system is the gait counting accelerometer (multiple (multiple), compass (multiple), Global Positioning System (GPS) System: Built-in GPS (location setting / location awareness) and heart rate monitor Another exemplary system may include a smartphone with a sensor. accelerometers (for measuring body temperature), heart rate monitors (e.g., using optical or other methods), Galvanic skin response (GSR) measurement Devices, blood pressure monitors (e.g., derived from photoplethysmographic signals), and and / or (e.g., wireless RF biomotion (biomotor) devices as described above or discussed herein) The sensor may include a (motion) sensor-based (breathing) sensor.

[0052] Additional data is sent to the "lab on chip" via wireless link. From the user via an attachment module or, for example, via RFID (radio frequency identification) Directly from the drug dispenser via an identification chip such as a drug identification number (ID, radio frequency identification, etc.) , or from a measuring device such as a spirometer (peak flow meter).

[0053] 3. Detection of Physiological and Behavioral Information In one embodiment, the heart rate, breathing rate (e.g., shallow / slow) and other parameters are monitored while the user is in bed. Physiological parameters such as breathing depth (gasping or deep breathing), mobility (degree of movement) and its Equipment to collect other relevant parameters is provided at the bedside (or in or under the bed). The device may be positioned to the side of the device (on or against the user's body).

[0054] The system can operate continuously without requiring user input. The device can detect whether the sensor is active (e.g., every 2 4 hours, 7 days, or a week) to trigger its operation. Devices such as range gated RF sensors are used to A continuous space (e.g., a bedroom) to distinguish one user from other users The system can monitor movement patterns and breathing (e.g., range, To build a classification of features based on breath-to-breath changes, shape, and inhalation-to-exhalation ratios, The system can track the user's parameters over time. It may be configured to construct features.

[0055] The system will persistently track the main (dominant) or primary users of the system. The system can also detect when the user is asleep and build differential features. Track user habits such as the bedside and the time the user sleeps each week The system uses classification techniques such as neural networks (e.g., hidden layer Markov models). Use a classifier or logistic regression (L) to determine classification features for easier implementation. Offline and real-time regression can be used. The subject classification features are used to determine whether a detected person is a user or It allows us to determine the probability that another user is like the user's partner. Offline processing can also be calculated / determined during a sensing / monitoring session. This allows the system to be retrained based on real-time parameters In this regard, real-time classification is performed when a user is detected within the sensing range of the sensor. Offline, which may be calculated / determined after the sensing / monitoring session. Offline processing such as parameters can be done when the user is no longer within range of the sensor. This may be done using previously sensed data when the off-line feature (or features) The offline features or parameters are the product of offline processing. For example, instead of calculating and / or classifying based on very recent data ("real-time") Additionally, the offline processing step may be performed over the entire monitoring / sensing session (e.g., overnight) or or multiple sessions (e.g., overnight) for post hoc analysis. Such processing can be done, for example, when the bed partner wakes up and the bedroom / sensor This can be done after leaving the area. This way, the entire night / sleep session(s) A comprehensive review of all data can be provided. The data can be calculated and / or estimated, for example, for a particular sensing / monitoring set. In this case, the classification that has the most recent data during the analysis may be applied.

[0056] Linear kernel or radial kernel support vector machine Other techniques for classification, such as Support Vector Machine (SVM), may be implemented. Feature computation is a method to calculate feature sets prior to classification when a large number of very similar features are used. "whiten" cases (i.e., reduce such dimensional data before further processing) Principal Component Analysis (PCA) Multiple classifiers may be used. However, a "lazy integration" of the outputs can be used to form the output posterior probability.

[0057] When applied to multiple sensors, multiple sensors (e.g., placed on either side of a bed) By detecting the same or different people from sensors and sharing the data over a network, Thus, the sensor characteristics such as range, power, frequency, detection direction and / or emitted radiation pattern can be The sensor parameters are useful for areas where a large number of people can be present, such as patients in a hospital ward. In this regard, the emitted radiation pattern can be automatically adjusted to match the detection pattern. The area or detection region (i.e., the three-dimensional (3D) sensing volume - any area that may exist behind the sensor The detection pattern is related to range and direction. The sensor may be configured with a range (e.g., near and far gating), The emitted power level of the sensor (even in the far-field gates the SNR (signal-to-noise) is low) This also extends the scope, as the effective further detection is now much closer than before. The power can be adjusted to adjust the amount of power that affects the sensing environment when no one is present. Power can also relate to energy savings for lower power / battery usage. Radio frequency dielectric resonant oscillator A Dynamic Range Array (DRO) can consume significant power when used for sensors. Frequencies may vary due to coexistence (e.g. to avoid interference) or regulatory requirements of different countries. Automated detection direction can lead to poor device setup (e.g. If you do not point the user to the first optimal setting, the system will not give the user the actual appropriate setup. (It can automatically adapt to this setup without prompting you to move it.) For example, the sensor can be dynamically modified in a controlled and reversible manner in frequency and emission characteristics. The antenna may have a reconfigurable antenna.

[0058] Thus, the system can obtain information about, for example, movement, respiration, and / or is configured to detect and authenticate a user by detecting a heart rate parameter. It is possible.

[0059] For example, as shown in FIG. 1, some embodiments of the present technology may include detecting a user in the vicinity of a device. The present invention provides a sensing or detection device 100 useful for detecting a physiological characteristic of a patient. The sensor may be a stand-alone sensor or may be detected by a sensor of the device. Respiratory therapy is performed to provide automated treatment responses based on analysis of physiological characteristics. The device may be coupled to other devices, such as a medical device, a controller, and a flow generator. A respiratory treatment device equipped with such a sensor can be configured with such a sensor, A patient interface (e.g., a mask) responds to the detected physiological characteristic. An exemplary respiratory treatment device may be configured to adjust the pressure therapy applied to the patient. The present invention is disclosed in International Patent Application No. PCT / US2015 / 043204, filed on the 31st. No. 6,399,413, the entire contents of which are incorporated herein by reference.

[0060] A typical sensor for such a device is a radio frequency pulse for distance gating sensing. The transmitter can be used to emit radio frequency waves such as The receiver that can be optionally included in the combined device is a reflected version. The device may be configured to receive and process the reflected waves from the subject. Signal processing, such as a processor in the device that activates the sensor, is used to derive obtain.

[0061] For example, as shown in FIG. 2A, a transmitter may direct a radio frequency (RF) signal toward a subject, e.g., a human. Generally, the source of the RF signal is a local oscillator. The reflected signal is then received along with a portion of the original signal. The resulting signal is then amplified and mixed, and the output of the mixer may be filtered. The signal can contain information about a person's movement, breathing and cardiac activity, and can be processed in raw form (ra w: low) is called the motion sensor signal.

[0062] FIG. 2B illustrates the use of several of the raw sensor signals to generate an indicator of a physiological characteristic. 1 is a diagram illustrating some potential processing. The raw signal generally includes body movements, breathing, and The body movement includes components reflecting a combination of inhalation and cardiac activity. zero-crossing) or energy envelope detection algorithm (or more This can be identified by using a "crude algorithm" This may be used to create a "motion on" or "motion off" indicator. For example, Such a motion detection algorithm is described in U.S. Patent Application Publication No. 2009 / 020397. The method can be carried out according to the method disclosed in US Pat. No. 6,399,343, the entire contents of which are incorporated herein by reference. Respiratory activity is typically in the range of 0.1 to 0.8 Hz, and It is derived by filtering the original signal with a bandpass filter having an overband Cardiac activity is reflected in higher frequency signals, and such activity ranges from 1 to 10 Hz. by filtering with a bandpass filter having a passband in the range It can be done.

[0063] Such respiration and motion sensors are based on distance-gated RF motion detectors. The sensor receives a DC power input and detects the temperature within the detection range. Four analog motion channels with both in-phase and quadrature components of human breathing and movement signals In the case of a pulsed RF motion sensor, the sensor may be configured to provide a range Disaggregation gating can be useful for restricting motion detection to only a desired area or range. Therefore, detection using the sensor can be performed within a specified distance from the sensor. .

[0064] As another example, the type of sensor used may be that described in McEwan and McMahon, supra. A radio frequency (RF) based system using at least some of the techniques described in the patent document For example, ResMed's SleepMinder family of contactless sensors. International Publication WO2010 / 036700 and WO2010 / 091168, WO2008 / 057883, WO2007 / 143535 and WO2 The measurement and data processing techniques described in US Pat. No. 5,999,333 can also be used. In addition, accelerometers, piezoelectric or UWB (RF Ultra Wide Band) area) mattresses, passive infrared or other light sources such as cameras with skin color and motion detection capabilities Alternative techniques using chemical means may also be used. For example, the device is described in U.S. Pat. No. 009 / 0203972, International Patent Application No. PCT / US14 / 045814;201 No. 62 / 149,839, filed on April 20, 2005, and the same filed as this application. No. 62 / 207,670, filed on Implement motion detection algorithms or sensor techniques according to any method and sensor disclosed in and US Pat. No. 6,399,623, the entire disclosures of which are incorporated herein by reference. In addition, U.S. Provisional Patent Application No. 62 / 205,1 filed August 14, 2015 Any of the radio frequency sensors described in US Pat. No. 5,999,333 may be implemented in any version of the present technology. No. 6,399,413, the entire disclosure of which is incorporated herein by reference.

[0065] Monitor and classify sound patterns consistent with respiratory rate, chronic cough or snoring; To separate these from background noises such as fans, road noise and the like, A rophone can also be used. For example, it can be performed during a daytime nap. This is called "night-time sleep" monitoring and targets users in their sleep environment. Also, respiratory rate and heart rate, and other such physiological signals, may be measured using non-contact or contact-related techniques. The detection may also be performed using alternative sensors, such as technology-based sensors.

[0066] When the user gets out of bed, the user’s physiological parameters are monitored by the body wearable This can be selectively monitored by a sensor. For example, this can be done by a clip-based device. , stick-on patch (skin or lab-on-a-chip) or watch-style device the collected sensor parameters can be motion (via accelerometer) and status Top, location, galvanic skin response: GSR), (by optical, electrical or motion means) cardiac and respiratory rates and other related parameters This is called "daytime / wakefulness monitoring" and includes some or all of the following parameters: Such devices target users during waking activities. Earplugs like t, Basis, Apple, Withings or other products The device may be in a laptop, watch or wristband form factor.

[0067] Biometric data from one or more sensors may also be used to measure “liveness.” )” i.e. to further improve the reliability of live human detection, or face recognition, Using technologies such as skin color detection and / or microbrush detection to identify specific users Such data may be fused with video / camera image data to identify Fusion can be added to analyze further factors. For example, skin color detection can be used to improve face recognition. Used as a first step in computer vision, gesture recognition and other applications and is very robust to scaling, rotation and occlusion.

[0068] In some embodiments, the daytime and / or nighttime respiratory rate is monitored using a belt or bra. - Clipped to a chest band (e.g., spire.io device), or nasal cannula This can be achieved by wearable respiratory sensors such as accelerometers or photoplethysmography (PPG). The waveform can be extracted and captured from the thysmography (photoelectric volume pulse wave) signal. Such data may also be involved in the biometric identification described herein.

[0069] Pressure from implantable (i.e., implanted within the person being monitored) devices and It is possible to combine other data with the values ​​collected by the system. Also, baseline data collected outside the bedroom should be Thus, data from one or more sensors can be correlated with each other, e.g. A baseline of detected biometric data that can be attributed to a particular user establish a baseline and / or compare newly detected data with previously collected data By attributing newly discovered data to a specific user, The system can be evaluated upon biometric verification / identification of the user. The parameters can be compared to past parameters.

[0070] The system may also use contact sensor readings (e.g., readings from a body wearable sensor) data from sensors (e.g., Doppler, RF sensors) as well as video Aggregating and / or combining video / camera image data from the detection system The system can therefore provide a collective 24 / 7 service to suit the user's lifestyle. 7 Operation or part of it may be performed.

[0071] The algorithms and methods described herein access database storage. For example, computing devices (e.g., PCs, servers, cloud services, smartphones, etc.) The present invention may be implemented on one or more processors on a single host (such as a mobile device application or variant). The algorithm is based on a classification system that analyzes one or more data sources. It is possible to retrieve past data stored in the database. The plates and / or stored models may include user specific and population based demographic data. The probability weighting may be modified based on adaptive data.

[0072] 3.1 Sensor feedback The system has a feedback loop (e.g., a PC or a smartphone or a tablet) can be provided to the user or a third party (via a smart device such as a laptop) , and optionally to a Monitoring Centre for review, certification and optional arbitration. The system can also be used to help users with various breathing problems. Continuous positive airway pressure (CPAP) CPAP, adaptive servo ventilator such as ASV (abdominal pressure suction) and / or bilevel positive airway pressure (bilevel) machines. Such a device can provide feedback on the therapeutic effect of the treatment. is described in more detail herein.

[0073] In some embodiments, the sensor is connected to a CPAP device, e.g., a flow generator. or similar (e.g., Respiratory Pressure Therapy) It may be integrated with a respiratory therapy device such as a respiratory therapy (RPT) device or may be used as a The devices may be integrated into one another or may be configured to communicate together. Upon detecting the biometric of the person entering the system, the system It is therefore possible to cross-check whether the correct treatment was given. The program will determine the treatment of PAP devices, for example, CPAP, ASV or Vire based on the identification. Select the machine parameters (such as pressure or flow protocol) expected for the Bell machine. The system can be selectively reconfigured to respond to anticipated people (e.g. For example, if a previously detected and identified user cannot be identified, the monitor An alert can be flagged to the monitoring center and / or to the user. An example of a treatment device is described in more detail in previous and subsequent sections of this specification. do.

[0074] Similarly, when the expected biometric is detected, the subject's respiratory rate and heart rate , which can be used by the system to better tailor treatment to a specific user. For example, if an elevated heart rate is detected, the system will try to reduce the heart rate to within the expected range. To allow the patient to breathe more easily, a specific breathing curve can be input into a device that assists breathing. .

[0075] In some embodiments, the method includes administering a CPAP, bilevel, ASV, or other therapy to a patient. When heart rate feedback is received, therapy may or may not be in use. It tracks increased heart rate both when the heart is not in a circulatory system and when the heart is not in a circulatory system. Tracking increases in variability and / or tracking long-term trends in heart rate and respiratory rate dynamics is possible.

[0076] In this way, feedback can be provided to the monitoring server. do.

[0077] In a workplace health environment, such biometric data can be used in monitoring systems. For example, miners or truck drivers may , sleep disorder breathing screening, diagnosis, monitoring and sleep quality analysis equipment for these homes It can be installed in a room or bunk (e.g., in the driver's seat / berth of a vehicle). This type of sensor can be placed on the bed, not at the bedside, above or below. Specific subjects / users are actually monitored with the described biometrics By ensuring that monitoring information is shared across all users (i.e. A rostering system that ensures that players sleep in their assigned banks ng system), and can be used to In this regard, measures may be taken to ensure that persons identified / identified receive adequate sleep. The system will refer to the fatigue management system to detect people sleeping on the bunk and track their sleep quality. In one example, the user may have sleep apnea. , the user may not be compliant with the treatment (e.g., not wearing a CPAP mask) Thus, the use of treatment methods (especially for occupational health and safety reasons) is In some cases, the system may flag the user as being at high risk. In one embodiment, the system may analyze a user of the system before tracking and / or identifying the user. Permission may be requested from the user or operator.

[0078] Generally, in some versions, the sensed biometric data (e.g. Once identified as belonging to a particular person (by a classified "fingerprint"), the biometric information is The trick data is associated with the person's identity and stored, for example, in a database. It can be done.

[0079] Daytime, professional drivers and / or operators of heavy machinery or safety critical applications Others in the field (e.g., air traffic controllers) may use Plessey EPIC capacitive sensors. , monitored using physiological sensors such as radar or wearable devices. It is possible.

[0080] Exemplary collection of "awake" and "asleep" data is described below. The signal collection method from "wake" can be applied to "asleep" and vice versa. The same is true in the case of

[0081] 4. "Awake" Time Data Collection Examples of how information may be used by the system include:

[0082] 4.1 Heart rate a. Heart rate (HR) data is collected by the user continuously or semi-continuously. This is done by using a chest band (e.g., a strap such as those provided by Polar). sports monitoring band), or a watch that performs heart rate monitoring (e.g. asis watch, portable ECG, photoplethysmogram , ballistocardiogram or similar) Ideally, a low user impact device would be used, but this would be required for long periods of time (days, Suitable for daily monitoring over periods of time (months or years). b.HR data is based on product heart rate variability y:HRV) estimates. c. The recorded HRV parameters (e.g., short-term and long-term variations in heart rate) were The historical and demographic (expected) parameters of normal and chronically ill subjects are compared with those of normal and chronically ill subjects. The signatures are provided to a classifier. Also, a smoother variance is obtained based on low-pass filtering of the signal. The HRV of the participants can be used; this is mainly for the analysis of long-term fluctuations and trends. Used for comparison with a stripped version.

[0083] 4.2 Galvanic skin reaction a. Galvanic skin response (GSR) Also known as the electrodermal response The GSR measurement is performed using a wearable device (e.g., a Basis watch or other commercially available GSR measurement equipment). It can be recorded in the location. b. The GSR signal is used as a surrogate measure of sympathetic "fight or flight" activation. do. c. Combining the extracted GSR and HRV signals, sympathetic activation and parasympathetic nervous activation were The ratio between sympathetic and parasympathetic activation can be calculated. Balance is determined by the system's normal versus disease-progressing state (e.g., increased parasympathetic activity due to decreased It is used as an estimator of the sympathetic nervous activity (sympathetic nerve activity).

[0084] 4.3 Exercise intensity a. Variations in exercise intensity and duration are captured from "daytime" sensors and recorded in a database are analyzed for identified trends and fluctuations. b. The overlapped circadian / sleepiness model shows distinct patterns of activity and serenity throughout the day ( For example, after getting out of bed, going to work, then an early morning coffee break, then going to a meeting. Walking, lunchtime stroll, followed by afternoon sleepiness [the strongest sleep drive generally occurs around 1 p.m. between 12:00 and 13:00], reduced activity during rest, leaving work at home, and reading / watching television in the evening A decrease in sleep duration and a further decrease during sleep [the strongest desire for sleep generally occurs between 2 and 4 a.m. This could indicate an increase in employee activity during such times. c. Energy consumption is estimated by combining heart rate and GSR data along with a step counter. Alternatively, a surrogate for exercise intensity (e.g., Nike Fuel Band , FitBit, Jawbone Up or similar wearable device A separate step counter (pedometer) may be used.

[0085] 4.4 Respiratory parameters a. Respiration rate, depth and activity (as explained in the "Sleep" section below)

[0086] 4.5 Blood Pressure Parameters a. Derived from a wearable photoplethysmography-based sensor

[0087] 5. "Sleep" time data collection An example of how "sleep monitoring" information is used by the system is shown below. As follows:

[0088] 5.1 Respiration rate, depth and activity level The algorithm is implemented to detect patterns in the user's breathing rate and dynamics. The algorithm measures a person's baseline breathing patterns over days, weeks, months and / or years. Adaptively track respiratory rate, movement characteristics and respiratory waveform shape to build a profile of respiratory mechanics. This can be done.

[0089] The algorithm module generates an adaptive baseline for the user and Median, mean, interquartile range, skewness, kurtosis, minimum and maximum breaths over (e.g., 24 hours) By looking closely at respiratory parameters such as inspiration rate, the primary (but not exclusive) cause of sleepiness is In addition, the shape of the inspiratory / expiratory waveform and short-, mid- and long-term respiratory variations are Your baseline fitness will also affect these readings. can be done.

[0090] Respiratory chest movement, rocking detection and release, rolling in bed, and (e.g. physical Coarse and fine movement detection with multiple behaviors including scratching (due to irritation or discomfort) To determine physiological repetitive and variable movements, including those induced by As a group of algorithmic processing steps and associated digital signal processing, There is some overlap with the activity detection done in "being there" monitoring.

[0091] 5.2 Heart rate Heart rate can also be measured by contact (e.g., wearable devices such as the Basis watch that use optical methods) (Ballistocardiogram signals – from the skin surface) or Wavelet-based time-domain (derived from non-invasively detected cardiac mechanical motion) -From non-contact sensors like SleepMinder that use technologies like frequency conversion It can be estimated in a "night-time" bedroom environment.

[0092] 5.3 Coughing and snoring Using digital sampling of the audio signal recorded via a microphone The algorithm detects characteristic symptoms of snoring, snuffling, coughing or difficulty breathing. This can be achieved by using digital filter banks, frequency decomposition and Searching for "bursty" noises, i.e. "cough" noises using spectral analysis This can be achieved using a "signature" search or using morphological processing. The data are selectively cross-correlated with movement and breathing patterns.

[0093] 6. Signal Processing For one exemplary implementation of in-bed monitoring, the present invention , recorded by the radio frequency RADAR, digitized using an appropriate ADC module. The two recorded channels (in-phase I and quadrature Q) are analyzed. These RF signals are , continuous wave, or pulsed (e.g., ResMed's Sleep pMinder 5.8GHz and 10.525GHz sensors, equipment using FMCW method (The sensor may be applied to any device, device or other device.) In some cases, the sensor may be adapted as described in U.S. Pat. The sensor may be as disclosed in Published Application No. 2014 / 0024917, the entire contents of which are set forth herein. , which is incorporated herein by reference. The signal is fed to a filter bank, and the filters A series of digital filters, including bandpass filtering, are applied to the bank to remove low-frequency Detects and removes shaking information. The phase information in the two channels is compared to Generates a counterclockwise pattern. Suppresses signal foldover. Hysteresis and glitch detection are applied to ensure that the signal is within the reference frame of the sensor. It represents the relative direction of the motion source to the frame. Peak / trough detection and signal Tracking is additionally used to aid this process. Thus, the system detects the movement and Determine whether the object moves toward or away from the sensor and changes direction. It can be determined.

[0094] The spectral content of the signal is calculated using the Fast Fourier Transform and peak detection (frequency domain). This can be done using, for example, a discretized continuous wavelet transform. d continuous wavelet transform), appropriate basis selection This is done through time-frequency processing using selection and peak detection. The residual low frequency components are then It may also be processed to determine longer time scale trends.

[0095] Cheyne-Stokes Respiration Sleep disorders that include cyclical or periodic patterns such as In addition to causing aggravation, ed breathing patterns can also When such an SDB symptom manifestation is detected, Systems that know the subsequent application of treatments such as PAP are a valuable tool in understanding the changing biomechanics. You should know that you can automatically retrain your device to recognize tricks. (i.e., confirming pre- and post-CPAP therapy intervention). Once a disease condition affecting cardiac / motion activity has been treated, the system will begin the verification process. Can be reconditioned / retrained.

[0096] Cardiac patterns such as atrial fibrillation and tremors (including seizures), ventricular tachycardia, and others , and can provide an aspect of identity recognition. In such cases, treatment The input to the system related to the user is the idea that when searching for a treatment for his arrhythmia, This can reduce identity detection failures.

[0097] 7. “Fingerprint” generation An important aspect of dealing with a large number of possible input signals is a robust estimate of the signal quality of each signal. It estimates the user's biometric "fingerprint" and checks whether this biometric marker is healthy. It is desirable to track and capture the evolution of a healthy to a diseased state, so that the signal is of high quality. Poor quality, adulterated or subject to objectionable processing (i.e. It is recognized that poor signal quality is not necessarily due to interference between sensors. For example, wearable heart monitors can be used to treat heart disease by inappropriate contact with the body. Wearing them at the same time may result in misinterpreted signals. Comparison of the data from other motion sensors in the same location can help to prevent misalignment of the sensor. In contrast to the long-term signal quality that is suspected due to the (e.g., potential signal impairment due to motion disturbing other sensors) It can be helpful to do so.

[0098] The user may not move the slider over any other artificial or natural motion source other than the user himself. By placing a step counter, you can trick the step counter into giving you an artificially high reading. It may be determined to provide a trade-off between the detected movement and / or other parameters. The time-varying statistics of the data distinguishes it from the user's predicted biometric "fingerprint" and invalidates it. Or it may be used to flag tampered values.

[0099] The system that manages the pattern recognition process must accept multiple input parameters. and a quality estimator is run to discard or flag any poor quality data. It is assumed that the input data may be overlapped in size and that this is To reduce any redundancy, we use randomized PCA (Main Component Analysis) It is recognized that various processes may be applied.

[0100] Figure 3 shows a human biometric (or multiple biometrics) in a sleep environment. Schematic of a system for calculating the In the case of digital sensors, the signal can be used directly. Filter out noise sources (e.g. main 50 / 60Hz) using filtering Heart rate, respiration rate and movement / activity are measured using a time-frequency method. by using short-time Fourier analysis or wavelet analysis Therefore, the presence of motion and breathing signals (physiological signals) in the sleep environment is , Distance-Gated Motion Sensor Module (or Piezo-Based Mattress Sensor or other types of sensors such as wrist-worn sensors) The sleep stages are then detected using the decomposed motion signal, which includes the physiological signal. The biometric features are extracted and the "fingerprint" / biometric is estimated. It distinguishes between humans and non-humans, whether or not the human has been previously known to the system. Behavioral data includes long-term trends in sleep duration (e.g., typical weekday / weekend variations). Long-term trends), bedtime (weekdays / weekends), sleep efficiency, number of awakenings, time it takes to fall asleep Sleep latency, REM and deep sleep ratios, etc. - These are the results of weekend social activities behaviors, such as those affected by voluntary sleep restriction nothing.

[0101] FIG. 4 shows a method for calculating human biometric characteristics from physiological parameters and, optionally, Dynamic characteristics (e.g. diurnal activity, location, heart rate / breathing rate spikes or The diagram shows a system that combines behavioral data (e.g., patterns of change in the road surface, etc.) for daytime signals. The available sensors and desired user interface can include physiological signals as well as Depending on the use case, sleep time and daytime detection may be combined. It is understood that each of the above may be considered on an individual basis.

[0102] Figure 8 shows a non-contact pulsed continuous wave RF sensor at a distance of less than 1.8 m. The heart rate (upper part) of the subject “User A” during the sleep session (approximately 8.3 hours) was recorded. Figure 9 shows the associated heart rate (lower panel) and respiratory rate (lower panel) signals for "User A". Figure 10 shows a histogram of the sleep and respiratory rates of a (slightly more than 6 hours) sleep session. The signals with heart rate and respiratory rate of a different subject “User B” during the session and the associated The heart rate and respiratory rate histograms are shown. Such signals can be evaluated by the processing of the present technique. It can be valued.

[0103] For example, Figure 5 shows a schematic diagram for the computation of characteristic features to generate a biometric fingerprint. The process is shown in Fig. 1. By distinguishing the biometric parameters input to the user classifier, Thus, user recognition can be performed (e.g., cardiac and respiratory signals can be used for such purposes). These features can then be calculated and combined (see Figure 5). If HR is not available or detected, the system will The system can then fall back to BR (respiratory rate) during part or all of the period; thus, the system detects Depending on the quality of the biometric characteristics acquired, different biometric methods can be used to identify a person. Any of the characteristics shown in FIG. 5 or identified below may be used. The (e.g., two or more) features are evaluated as part of a biometric fingerprint. Possible:

[0104] [Respiration / respiratory signal related parameters] - Daytime and / or nighttime respiratory rate variability (variability that characterizes the user) - This can be between breaths or over longer time scales (e.g. 30, 60, 90 seconds or more). (for a longer period of time) - Safety over time (related to variability) - Standard deviation of respiratory rate - Breathing depth (shallow, deep, etc.) and the relative amplitude of adjacent breaths -Average or mean respiratory rate - Mean trimmed to reject outliers (e.g., 10%) -Awake or asleep (i.e., the detected sleep state of the user) - Sudden changes in breathing rate (sudden acceleration or deceleration) during quiet periods and REM sleep -Median (50th percentile) -Interquartile range (25th to 75th percentile) -5th to 95th percentile -10th to 90th percentile -Histogram shape -skewness -kurtosis - Peak frequency over time - The ratio of the second and third harmonics of the peak frequency -Percentage of valid data (valid physiologically relevant data) -Autocorrelation of individual signals -Characteristic patterns in the spectrogram -Awake or asleep -Relative percentage of REM and deep sleep

[0105] [Heart / Heart Signal] -Heart rate variability (e.g., derived from a ballistocardiogram) between heart beats) and during longer defined moving windows - e.g., 30, 60, 90 seconds) -Temporal variability (beat-to-beat / respiratory variability) -average - Cropped average (10%) - Standard deviation -Median (50th percentile) -Interquartile range (25th to 75th percentile) -5th to 95th percentile -10th to 90th percentile -Histogram shape -skewness -kurtosis -Safety over time -Peak frequency over time - Ratio of the second and third harmonics of the peak frequency -Percentage of valid data (valid physiologically relevant data) -Awake or asleep -Autocorrelation of individual signals -Characteristic patterns in the spectrogram

[0106] [Cardiac and respiratory signals] - Magnitude Squared Cross Spectral Density (with a moving window) re Cross Spectral Density) -Cross Coherence - Respiratory sinus arrhythmia peak -LF / HF ratio, which indicates the parasympathetic / sympathetic balance of the autonomic nervous system - Cross-correlation, cross-coherence (or cross-spectrum) of cardiac and respiratory signal estimates density) - Characteristic patterns of movement over longer time scales, i.e., observed in the signal Statistical behavior - Detect and compare movement patterns (e.g., during sleep) between these cardiac and respiratory signals Some users are able to sleep more restfully and some are unable to sleep well. It may disappear.)

[0107] For example, a reference distribution of motion and / or other characteristics can be compared to the calculated distribution. It is possible to use the nonparametric Kolmogorov-Smirnov method. nov) goodness-of-fit tests of registered histograms defined in terms of variability The extraction of parameters can be done using a number of techniques (such as STFT or wavelets). This can be achieved using time-frequency analysis techniques.

[0108] Any single parameter (e.g., average heart rate or a particular feature such as resting or deep sleep) Although the average respiration rate in a given sleep stage) cannot be used to distinguish between users, More advanced systems can combine multiple features and group the features accordingly. The early integration approach is preferred, where the loop feeds the classifier with training data (labels If data (labelled data) is available, a supervisory classification system can be used, which This allows a large training data set to be run on the system to generate model parameters. The data resulting from the first use (daytime or nighttime signals) will be provided to the Biomet Feedback to allow the user's lick "fingerprint" to be augmented with user-specific accuracy This allows the user to update the user-specific classifier. This can be achieved through the sample acquisition template being stored in a database. A matching step is then performed to verify the identity. The lack of such detailed training data leads to sparse coding. coding) (e.g., LCC local coordinates derived from the image processing field Semi-supervised or supervised, using techniques such as encoding or other encoding techniques A learning feature hierarchy (e.g., deep learning) approach is used.

[0109] Decision-based neural networks are used after the template matching step has been performed. can be used to verify the decision.

[0110] The physiological fingerprint of the user is determined based on the "correct" (i.e., expected) This can be seen, for example, in the bedroom table (ni The non-contact sensor placed on the stand monitors user A. User A can get out of bed (e.g., take a bath) User B (e.g., can go to the room or work), and another user B (e.g., can go to the bedside The child(ren), baby or pet can move within the sensor range. The sensor data processor should sense the parameters of user B, so This can be confused with user A. More advanced processors can confuse user B. It senses the "fingerprint" / pattern and distinguishes it from A. The static sensor knows the user set. If so, the system will access supervised training data as part of the processed learning. If the user is unknown (i.e. unknown to the system) semi-supervised (i.e. user A knows but B, C, etc. don't) ) Pattern recognition can be done without distinguishing labels, or without supervision. Unsupervised or semi-supervised systems are available to the system. The system creates a “fingerprint” of a distinctive combination of a user’s physiological and behavioral parameters. " Multiple days / nights of data may be required to generate a

[0111] While enrolling the system using contactless or minimal contact sensors in the bedroom, The subject is asked to perform a specific breathing movement in front of the sensor - e.g., a guided deep breathing signal. When prompted, the system will perform deep or can register users with certain shallow breathing patterns. The level of consistency between the signal and other HR features can be used as a baseline input. A more complex registration may be required to train the system and / or monitor sensor data during the day. To better understand the mechanism by which spontaneous breathing can be monitored, it is recommended to record spontaneous breathing over a longer period (e.g., overnight signals). request.

[0112] The algorithmic processing steps of the registration process and subsequent identity verification are shown in FIG. 6 and 7. In FIG. 6, the system for registration is first configured. This can actually be automatic or manual; for example, to register a sleep session This can be initiated manually, but the system automatically collects data generated during the overnight period. Dynamically scans the device to verify origin and generate a biometric signature ("fingerprint") for the user To generate it, we calculate features (e.g., model weights for a classifier). This will be discussed later. The biometrics are stored in a database for search and comparison. Daytime registration is The data from the above monitoring devices is collected and a user log is generated to generate a signature. This behavioral aspect is what malicious users can do to prevent respiratory infections. Behavioral data has more merit than physiological data because it is easier to mimic cardiac signals and can be assigned the lowest weighting value.

[0113] In FIG. 7, the first step is to obtain a valid signal (e.g., the sensor baseline of an accelerometer). The presence of humans in the area of ​​the RF sensors (more than just human activity) The signals are detected (meaning breathing and cardiac signals are detected), features are calculated, and machine learning The algorithm outputs a biometric estimate. The system then Access the database to generate the appropriate sleep and / or wake model (for the detected signal) Based on the probability that a particular user is detected, the system checks for possible Identity and access management (IaA) to check candidates Access to the IAM (Integrated Access Management) system. A comparison is then made to authenticate the human biometric as valid or (if applicable) to determine whether the biometric is correct. Approve or check the limit level - or use another device to ensure it is the same database) or reject the biometric (do not authenticate the user). A decision is made.

[0114] 8 and 10 show user A and user B (as detected by the RF sensor). Sample graphs of heart rate and respiration rate for User B are shown. The associated histograms of these signals are The tograms are shown in FIGS.

[0115] FIG. 12 shows the heart rate of a subject breathing at a sustained rate of 0.1 Hz (6 deep breaths / min). The cross-spectral density between the parts of the respiratory rate and the simultaneous respiratory rate is shown. The method can be used to characterize the relationship between systems), where the peak is 0. The joint effect on heart rate and respiratory rate at 1 Hz is shown.

[0116] The aim of this system is to detect and reject data from animals sleeping in the bed - for example, dogs and cats. For example, some nighttime recordings can be made possible by the use of long-range sensors. If the bed detects a second person, it will detect and reject the BR and HR components of the second person in the bed. It can be for one day or for the entire night if the first person is away.

[0117] More generally, to use a particular monitoring technique (e.g., over a period of time) Employers or insurance companies that offer incentives to users (to demonstrate activity / exercise) In the case of the company, the estimated biometric parameters are used to verify that the authorized user technology and in fact other parties (either unintentionally or intentionally disguised) This is used to check that the

[0118] The relationship of specific sleep stages to respiratory and / or heart rate parameters is discussed in detail below. Since the beat rate statistical parameters are not specifically under the subject's voluntary control, The system can provide additional insight into the trick identifier. For example, the system can The device may be configured to detect whether the user is in REM (dreaming) or deep sleep. For example, the system is based on International Patent Application No. PCT / US2013, filed on September 19, 2013. No. 13 / 060652 and International Patent Application No. PCT / US filed on July 8, 2014 As disclosed in the method of 2014 / 045814, the sleep state or the determination to prepare Methods for determining the amount of oxidized cellulose can be carried out using the methods disclosed in US Pat. No. 6,399,323, the entire contents of which are incorporated herein by reference. Preferentially, deep sleep is selected to better counteract the effects of daytime stressors. REM sleep can sometimes cause larger breathing changes. However, it can also cause symptoms of SDB, PTSD (post-traumatic stress disorder), etc. or other problems may affect you to a greater extent. Also analyzed may be first stage light sleep parameters.

[0119] Critically, when a person sleeps, he or she is consciously directing their breathing into a particular pattern. Therefore, separate wake and sleep models are used to reduce the possibility of signal manipulation. In fact, separate monitors can be generated for deep, REM or other sleep stages. Thus, in the identity authentication process, A person's fingerprint may be used at any or each of these stages. These sleep stage specific fingerprints are then used to identify specific sleep stages in the identification process. Identification of a person in a particular sleep stage is evaluated based on appropriate characteristics determined at each sleep stage. This can be done in relation to:

[0120] In this regard, the system may provide analysis of the data being analyzed, including multiple nights and days. The data can be trended over time.

[0121] In addition, the breathing pattern detected by the PAP device or RPT can The conduit and / or mask, or as referred to herein, may form part of a "fingerprint." The patient interface or patient circuit, like the other devices described in detail above, It may contain additional sensing features such as a speedometer or a conduit (e.g., CPA The movement of the respiratory tract (P) can be detected and characterizable movement information can be extracted. In this way, PAP devices can be made compliant with regulations. To do this, the user's biometric "fingerprint" can be checked.

[0122] A simple increase in respiratory rate can be associated with non-chronic conditions such as the common cold. Please note that.

[0123] A multi-parameter holistic analysis of individual health helps the system to understand the situation and This means that relevant data can be associated with the correct user. Insurance companies offer discounts to users of their systems based on users meeting certain health improvement goals. Such goals may include decreasing average heart rate, decreasing respiratory rate, and increasing exercise intensity. etc. can be achieved.

[0124] Galvanic responses recorded by simple actigraphic sensors or wearable sensors Skin response and / or skin temperature changes are used to augment this biometric classification. If specific heart rate and GSR are changed, they can be cross-referenced to exercise intensity. can indicate whether a stressful event or activity affects a particular change. For each individual, people can measure their heart rate, breathing rate, and other parameters in response to exercise or stress events. and GSR changes. Therefore, the user's inherent stress level and The ability to deal with the problem is a key component of biometrics when combined with other available parameters. It can serve as a check marker.

[0125] Longer term patterns in exercise intensity and steps taken can be used to gauge a user's specific behavior. can be correlated with

[0126] The system detects theft of monitoring equipment (e.g., devices with RF sensors). i.e. to prevent unauthorized or unknown user biometrics from being detected. Such a determination may be made based on the If available, for example, location data could be included as part of the "fingerprint" It is possible to reference data (e.g., GPS coordinates).

[0127] The system may be configured in a hospital environment where different patients are monitored in different beds in the hospital. Readings from different patients are taken as they are moved or discharged and the bed is reused. The system can be used to automatically separate hospital data collected from monitoring devices. / Clinic data sets are provided to patients to take home for long-term monitoring It can be automatically associated with the second device.

[0128] This means that an attacker / harmful act, for example, can detect the presence of a second RF sensor under attacker / harmful act control. Replay attacks attempt to imitate or replay RADAR signals. Note that the first RF sensor can be very robust against the second It can detect radiation and interference from RF sensors and flag an alarm signal to the system. Physiological signals can provide new biometrics every few seconds. In effect, the system can provide continuous authentication.

[0129] The system for identifying the biometrics of a first person and a second person in a bed comprises a sensor adjust the range and / or power level of the sensor to detect the desired person (e.g., the first person) To achieve this, a control signal can be fed back to the RF sensor.

[0130] Thus, in some cases, the system may aggregate multiple information sources and smartly process these data. and / or to provide user and data access to specific information, e.g., during nocturnal / sleep recording. It is possible to identify users by receiving feedback on the device. Other versions of the system can target only certain parameters or The parameters may not be adapted to the specific user parameters.

[0131] Potential advantages of some technology versions may include: 1. The system detects physiological patterns based on cardiac and / or respiratory signals and can be identified. 2. The signal is transmitted via a pulsed radio frequency (RF) sensor or group of sensors. It can be captured. 3. The system is able to identify the biometric parameters of the user. That is, the heart rate and / or breathing rate and / or amplitude recorded by the non-contact sensor. Based on this, verify that the recorded signal came from a given user. can be done. 4. The system updates biometrics during sleep and wakefulness stages. It is possible. 5. Sleep states reduce / remove conscious variability imposed on the signal (e.g., OH& As mentioned in the S use case (i.e., respiratory parameters that can also affect heart rate) (This reduces the risk of misclassifying turns or of users "faking" them). 6. Prioritize biometrics during deep (optimal) or REM (suboptimal) sleep You can date. 7. The system communicates with the server to retrieve existing identity templates / models. can be checked and identity determined. 8. The system can register users using awake or sleep stages. Cut. 9. The system will respond to specific deep or shallow breathing patterns of defined inhalation / exhalation duration and depth. Users can register when the call is presented. 10. Systems using pulsed radio frequency (RF) sensors should be To mitigate “replay” attacks, a second unauthorized RF sensor is initiated. It can detect and notify you if 11. The system can detect the health condition of a person and provide guidance to the flow generator. lock and key, and cardiac adaptive servo ventilation Includes smart adaptation of the ASV (adaptive ventilator) to new health conditions Input can be received when a treatment / therapy etc. is created so that it can be processed. The system ensures that appropriate people are using the therapy device in the environment for which it was designed. otherwise, it may be better suited to new / unknown users. can be reset to the boundary. 12. The system performs day and night monitoring of web pages, e.g., in databases and libraries. from the available sensor (as well as other sensors such as accelerometers and video for daytime respiration rate) Using "survivability" data) breathing rate, heart rate, movement patterns, galvanic skin response, Biometric data, including blood pressure, can be updated. 13. The system is based on a system in which the data of a second or third person is centralized in the first person. To avoid mistakes caused by the second biometric data, a second biometric data may be used for some or all of the recording period. It is possible to detect when a trick is detected. 14. The system detects animals, such as dogs or cats, sleeping or awake in the sensor area. can be detected. 15. A system that detects two people (two biometrics) in a bed is the first To better detect people, a control signal is sent to the sensor / processor. The sensor behavior (power / distance gating / directivity) can be adjusted (biometric With additional details on tuning the sensor based on the load and selecting the lowest sensor power for a given situation (Examples and examples). 16. The system can be cloud-based, so that the system can Multiple fixed or mobile objects in different locations so that multiple objects (multiple objects) do not need to act together The system can track a person's biometrics through sensors. For example, Tracking a person from a first sensor in one room to a second sensor in a second room and automatically comparing physiological data from the first and second sensors.

[0132] 8.0 Multi-Sensor Collaboration with Biometric Feedback When multiple sensors cooperate, e.g., sensors installed at remote locations or co-located From a system with a control processor (either continuously or as part of the pairing process) ) When communicating over a wired or wireless link, the characteristic biometric parameters are , to optimize physiological recognition of the human as an independent source and reject other sources. It can be used to dynamically tune the performance of one or more sensors.

[0133] Two or more sensors located adjacent to each other to minimize RF or other types of interference. When exchanging information, the biometric information of the human or human beings detected by each sensor is The identity and / or associated biometric quality of a person may be determined by one or more sensors. This can be used as a basis for adjusting the control parameters of the The available parameters are range, gating pulse timing and (within acceptable regulatory limits) Including the emitted power level or the RF detection frequency (e.g., center frequency) of the sensor Such modifications can be made by analog sensors provided that the control signals are configured to allow this. The signal may be provided to the sensor device, but is most easily accomplished with a digitized sensor design.

[0134] The sensor should consume the lowest possible power to achieve good signal quality biometrics. This may be done for environmental reasons and if the sensor is battery powered. This may be advantageous because of the additional ability to be powered by other sensors or devices. This can minimize the possibility of interference with other devices.

[0135] 8.1 Biometric Control of Multi-Sensor Arrays The sensors may be configured to operate in close proximity to one another while maintaining minimal RF interference. In this regard, two sensors can be placed on opposite sides of the bed, eliminating the need for a conventional arrangement. The device may be programmed to monitor the sleeping individual with a specific The biometric characteristics of the user can be detected by the respective sensors. Each sensor can be initialized with such user data, so that the sensor However, the data in the figure may be used to include a baseline of data that can be attributed to the user. During further operation, the sensor receives overlapping signals from each person, as shown in FIG. For example, the quality of the sensor may be degraded because the signal from human_a may be detected. The sensor _a is programmed to measure the respiration, heart rate and / or or a motion signal, whereby the signal received with respect to human_a Similarly, a robot programmed to detect signals from human_b can The sensor _b receives overlapped breathing, heart rate and / or movement signals from the person _a. , thereby reducing the received signal for person_b.

[0136] To avoid such degradation, sensors can be adjusted to minimize interference. For example, as shown in FIG. 14, the separate individuals are placed on opposite sides of the bed and are asleep in a common arrangement. The two sensors are programmed to monitor the body, with each sensor having its own range and It can be programmed to minimize interference by adjusting the power. The biometrics are programmed to receive biometrics from human_a and human_b, respectively. The paired sensors, sensor_a and sensor_b, maximize distinct and separated biometrics. For example, the range and power of the When detecting a biometric that is not specifically identified by the user's "fingerprint," security In the example of FIG. 14, sensor_b can reduce its detection range. Detect a biometric that is recognized as being associated with the initialized user and If it detects a biometric that it does not recognize as being associated with an authenticated user, Sensor_b is a biometric that is recognized as being associated with the initialized user. only until it detects (e.g., via power or distance gating adjustments) its detection range The sensor_a can be programmed in the same way. Since the sensor only detects biometric signals from a single person, two sensors The biometric quality received by both sensor_b and sensor_b may be excellent. do.

[0137] In some embodiments, the devices are positioned on opposite sides of the bed to monitor the sleeping individual. The two sensors are programmed to monitor only a single individual. For example, as shown in FIG. 15, a signal from a person_b can be detected. The programmed sensor _b can collect breathing, heart rate and / or That is, for example, the sensor can receive a motion signal from its initialized The user can be recognized from the detected biometric data. A sensor _ programmed to detect signals from a (currently non-existent) human _b a can receive breathing, heart rate and / or movement signals from human _b. For example, the sensor may determine that the detected biometric data is a valid user ID for the sensor. Therefore, it is possible to recognize that the sensor_a is not coming from sensor_b. b) can be replicated, or otherwise If there is interference between the signals, the biometric quality will be degraded.

[0138] To avoid duplicate biometrics or other interference, A sensor that is programmed to receive a signal from a When detecting only metric, it may enter power saving mode and / or search mode. In this mode the sensor can, for example, reduce or terminate sensing or sensing range. Therefore, as shown in FIG. 16, while one sensor is in a slip state, The sensor can operate with minimized interference. In this regard, sensor_b is The device is programmed to detect breathing, heart rate and / or movement signals from the Sensor_a can then be programmed to detect signals from (now non-existent) person_a. The sensor_a may determine that the human_a is not present and enter power saving / search mode. Switching to mode avoids duplicating the biometric detected by sensor_b. Power saving and / or search modes can be used to disable the sensor while waiting for a particular individual to return. The range and power can be limited. Thus duplicate biometrics are no longer received. I can't.

[0139] In certain embodiments, more than one sensor may be used. For example, the sensor shown in FIG. As mentioned above, three sensors in close proximity can be used. In this regard, the first sensor, “Alph "Sensor_a" is located on the bedroom table on the first side of the bed The first person (human_a) may be an initialized user of sensor_a. Person_b, who shares a bed with person_a, can sit on the second opposite side of the bed. A human being, who may be an initialized user of sensor_b, is placed on the room table. The second sensor, “beta” (sensor_b), is directed toward b. In one example, human_b has a larger body mass index than human_a or human_c. Furthermore, human_c may be a large subject having a body mass index (BMI). The person may be sleeping in a bed on the far side of the thin partition of _a. A human_c, who may be a selected user, may be monitored by a sensor_c.

[0140] The system uses sensor_a at odd frequencies, sensor_b at even frequencies, and sensor_c at odd frequencies. In this way, the R between the sensor groups can be configured to operate at several frequencies. F interference can be minimized.

[0141] A biometric signal may be detected from each sensor device. Based on the metric signals, humans a, b, and c are identified based on their "fingerprints". In this regard, the sensor_a may be When the person leaves the room, the biometrics identified as person_b (on the other side of the bed) In other words, the biometric signal of person_b can be detected. was detected by both sensors a and b. Therefore, the system control process The sensor sends or generates a control signal to sensor_a to adjust the range gating. This can shorten the time required for authentication and allows for continuous monitoring of detected biometrics. Additionally, the system control processor may selectively reduce the power level of sensor_a. It can be done.

[0142] When sensor_a detects human_a, the sensor is calibrated or otherwise activated. The power state can be restored to normal. (c) The biometric of person_a is correctly detected, and the biometric of person_b is not correctly detected. When an unconfigured detector is configured, the control parameters for that detector configuration are saved as the optimal settings. It is possible.

[0143] Additionally, if an individual or animal not intended to be monitored comes within range of the sensor, The sensor may be programmed to change further controls. For example, FIG. As shown in the figure, a dog named Dog_a climbs over Human_a and onto the center of the bed. The sensor can detect non-human biometrics. Upon such detection, the sensor detects the dog_a’s biometric by minimizing the sensor_a’s In order to allow human_a and human_b to be preferentially detected by sensor_b, Distance gating and / or power parameters of the sensors (e.g., two sensors_a and You can adjust both the power and capacitance (b).

[0144] In this manner, the behavior of the sensor may be, for example, by a sensor control processor or other system control processor, including any It may be adjusted one or more times during a monitoring session.

[0145] In any case where optimizing the sensor parameters does not result in an improvement in the biometric quality The user or users may be prompted to adjust the orientation of one or more sensors. do.

[0146] A multi-sensor (including sensor_a, sensor_b, and sensor_c) as shown in FIG. In one example of a support setup, the sensor can selectively perform a location check (e.g., GPS or or other input data) and / or to set the detection frequency to minimize interference In this regard, the processor may include a processor programmed to: In this way, an RF detection frequency suitable for the location of the sensor can be determined and used. Once the sensor determines its location is within the United States, it transmits data (e.g., frequency and geographic Access the target position table and detect the target from the table in relation to the US You can set the center frequency, FCC spectrum mask and other settings. Appropriate ETSI (European Telecommunications Sta According to the European Telecommunications Standards Institute (EESTA) spectrum mask Similar parameters may be used in parts of the EU and may vary depending on the region (or local regulations). In this way, the center frequency of the sensor can be licensed in other countries. The center frequency of the Sa_a is automatically set to 10, 527, 125kHz, and the power level is set to 43.8%. The distance gating (time of flight) can be adjusted to 92.5cm, and the detection roll-over The center frequency of sensor_b is automatically adjusted to 10, 525, 00 0 kHz, the power level can be adjusted to 93.2% and the distance gating is 16 The center frequency of sensor c is automatically adjusted to provide a roll-op at 5.2 cm. The output level can be adjusted to 80.0% and the frequency can be adjusted to 10, 522, 375kHz. The separation gating is adjusted to provide a roll-op at 150.0 cm.

[0147] In one example, person_a has a heart rate of 51.4 bpm with an interquartile range of 6.3 bpm during deep sleep. 55.7 bpm during REM sleep with a median heart rate of 1.0 bpm and an interquartile range of 9.1 bpm m. The personal biometrics of person_a are the median, The shape of the resulting histogram and the medium to high consistency between the respiratory and cardiac signals are best. The respiratory rate of person_a during sleep is 13.1 br / min (+ / - 3.2 br / min) ) during REM sleep and 13.9 br / min (+ / - 1.1 br / min) during REM sleep.

[0148] Person_b had an average heart rate of 77.5 bpm with a noticeable acceleration of almost 30 bpm throughout the night. and increased beat-to-beat variability due to underlying intermittent arrhythmias. A mean respiratory rate of 2 br / min was observed during periods of significant heart rate acceleration to be greater than 20 br / min. The change in br / rate during REM sleep was 4.5br / min. Respiratory arrest (apnea and hypopnea) was detected during follow-up. Tachycardia is associated with these apneic episodes. HR / BR consistency is generally consistent with HR and B R is low due to the significant variability in both. The indices of skewness and kurtosis are unevenly distributed. This indicates that.

[0149] The average heart rate of human_c is 59.8 bpm, which is particularly stable during deep sleep. During EM sleep, variability increased dramatically to oscillations between 2 and 23 bpm. was 17 br / min and the minimum was 15.2 br / min.

[0150] When dog_a enters the space between human_a and human_b, sensor_b detects the dog_a as described above. For example, the unrecognized / uninitialized Upon detecting the biometric, the processor controlling the sensor generates a detection voltage for sensing. Changes in sensors such as force, frequency or other control parameters can be controlled. For example, the center frequency is adjusted to 10,525,000kHz and the power level is adjusted to 84.9%. distance gating can be adjusted to provide roll-over at 145.6cm This can be done (for example, by adjusting the range, power level or detection direction). ) Only human_a is detected, and dog_a is rejected (i.e., not detected). 2) By limiting the sensor_a parameters, it helps to distinguish between dogs and humans. Similarly, the sensor_b parameters are such that only human_b is detected and dog_a is rejected. The system also automatically adjusts to ensure that unknown biometrics are detected. It can also flag breathing rate, breathing depth and heart rate parameters. A flag can be used to indicate that the data matches an animal. A normal dog taking a rest can take 2 4 breaths per minute, but such a respiratory rate in humans is limited by congestive heart failure or C Many chronic diseases, such as chronic obstructive pulmonary disease (OPD), are characterized by Shallow breathing (relative to the average or individual baseline) is accompanied by differences in heart rate variability. When dogs or cats are subjected to vigorous exercise or high temperatures, they may develop eczema more rapidly with their open mouths over a period of time. A dog's normal breathing range is about 20-30 breaths / minute. The average heart rate is generally 60-100 bpm for large breeds and 60-100 bpm for small breeds. A typical cat's respiratory rate range is 100-140 bpm, while a puppy's is around 220 bpm. The blood pressure is 20-40 bpm and the heart rate is 140 and 220 bpm. Adding food to a cat or dog tends to increase the total lung capacity of the mammal; Therefore, the sensor control parameters are The data may also include one or more images that are recognized by the processor as depicting a non-human animal. Such biometric characteristics may be adjusted upon detection.

[0151] Similarly, as another example, when person_a leaves the bed, sensor_a detects a center frequency of 10 , 527, 125kHz and goes into power save / search mode with power levels between 0-100% Distance gating (time of flight) can be rolled out from 50 to 300 cm. If the existence of sensor_b is known in advance, the search for sensor_a The mode is to adjust the power percentage and roll to minimize possible interference pulses entering sensor_b. The op distance can be limited.

[0152] In some embodiments (e.g., using classical phased array or digital beamforming) Steerable antennas adjust the sensor's coverage range and minimize interference. For example, the sensing "cones" of sensor_a and sensor_b can be used is adjusted to completely eliminate the biometric evidence of dog_a.

[0153] Optionally, in some cases, for example if the sensor does not identify the initialized user, Other systems may be triggered based on the biometric identification of the sensor. For example, As shown in

[17] , the system’s processor may detect unexpected biometrics (e.g. , dog biometrics) to alert (e.g., human_a and human_b's smart device), can trigger or send an alert. The processor upon detection of such a situation may activate the webcam or (for example) turn on the speaker. An audio alert can be triggered via the

[0154] By detecting certain biometrics, light and / or sound and / or environment A custom configuration of parameters can be set (i.e., the controller that performs the identification / processor). For example, two partners sharing a bedroom Consider a system that uses smart tagging to detect a first partner. TV, sports channels (or Netflix or a similar online service) Switch to a smart TV (stream via the Internet) and turn off the main lighting in the bedroom. The user is then detected as asleep and soft jazz music is played over the Hi-Fi system. The Rakuten is turned on and the TV is turned off. After the user enters a deep sleep state, the Hi-Fi The volume gradually decreases and then disappears when a second partner is detected in the bedroom. , Hi-Fi / Radio can play music from your stored music collection or from streaming music services. Popular music via services (e.g. Spotify, Deezer, Apple) / Configured to play the latest top 20 hits; Additionally, the room lights are dimmed and bedside lighting has an orange / red tint with low white / blue content. The TV is configured to display a sunset video sequence (with low backlight intensity). The second partner can then either display a fireplace or simulate a fireplace. When the second partner falls asleep, the music and TV amplitude is reduced until the user enters a deep sleep. , gradually fade / diminish until both are off. If you are in a location where you are, the sports TV / video stream will be sent to the first partner's wireless headset. The pop music is then transmitted to the wireless headphones of the second partner (also If available, the TV polarizer will direct the sports feed to the first and transmitting the fireplace video to a second user. , a customized bedroom by detecting the biometrics of one or more partners (or your living room, for that matter). To do this, alarms are customized based on the first and second partners. For example, the first partner may want to wake up later than the second partner. Thus, to avoid waking the first partner, a low amplitude Centralized lighting together with an alarm can be used for the second partner. is based on an identification programmed into the system and made by the system's processor. can be accessed.

[0155] For example, the system may detect additional animals and / or (unrecognized) humans in the sensor's detection void. For example, as shown in Figure 17, In this way, dog_a (or actually any other animal or human) enters the sensor detection space and the system Optionally, an image or video of the event can be triggered. The information may be scanned and stored and / or transmitted to authorized persons for review. Triggered cameras (e.g. networked, low-light, illuminator-linked Integrated infrared, thermal or other types of cameras or webcams can be used to identify unrecognized individuals. To emphasize this point, one or more When the biometrics of an unknown person such as a thief / intruder are found in the security area, Video and / or audio sensors are accessible and can generate warnings. If the information is available or under the control of a system, for example, a monitoring centre may contact the police before A recording or live transmission of video and / or audio is made for review by the Such biometric detection may be performed in conjunction with other detection devices of attributes. When used in a building, PIR (passive infrared), other microwave intruder sensors, glass break detectors (vibration motion or acoustic), magnetic contact switches, shock sensors, video motion triggers or (e.g. together with pressure-activated devices (e.g., mats) to reduce false alarms in intruder alarm systems. It can be used.

[0156] When carried out in an apartment block or office building, such biometric testing The fire brigade checks the human (biometric) number provided by the fire alarm system. Check (and selectively correlate with access card systems, if in operation) Such RF sensors can be used to detect and detect the presence of a standard video feed. In contrast to , it can continue to function in smoky or hot conditions (and RF can, for example, For example, full video processing and the same private life for possible use in the bathroom / bedroom etc. (Not closely related). User biometrics are stored in a central controller or a set of sensors. Once registered in a system such as a computer that receives an identification from The computer can output a list of names and locations within the building / structure. The central computer identifies the location of each sensor and determines which sensor in the sensor set has a detection range. maintain data to provide such a list in case of emergency in relation to the identification of people within the It can be held.

[0157] Such RF sensors can be powered in a number of ways. For example, RF The sensor may also be placed in an AC pass-through outlet (i.e., the mains socket / Pikiba with a very slim form factor that can be plugged into an AC outlet piggy-back AC outlet; USB socket is , may also be provided on the socket faceplate for secondary use of the device. A night lighting feature may also be provided on the device when placed in a hallway, for example. , can be activated by motion detection using an RF sensor. It should detect within the area of ​​the socket / AC outlet (covering part or the whole of the room). via Wi-Fi or other wireless communication means, or via the mains wiring ( For example, via Homeplug or a similar standard. An example is provided where an over-the-air USB adapter or cable is provided. -Wireless or via the USB link itself) communication is dependent on the adapter or cable. It is integrated and draws / sources power from the 5V provided by the USB specification. This allows: The RF sensor can be placed anywhere a USB cable is provided, for example a wall phone It can be plugged into a machine / tablet charger, can be plugged into a laptop, etc. By plugging such a device into a laptop or tablet, the device It is understood that the user's biometrics can be checked in front of or near the AC Either the central device or the USB inline device option A very low-cost and simple device for monitoring spatially for biometric properties. It provides the functionality of simply plugging in the device.

[0158] 9.0 Further methodologies for identification / differentiation As previously mentioned, the identification methodology involves the use of a sensor to distinguish between entities adjacent to the sensor. In further examples below, multiple presences may be used to identify specific users. It can be distinguished so as to obtain

[0159] 9.1 Introduction The majority of adults sleep with partners who potentially have different sleep patterns. A nurse can get into bed before the primary (or first) user of the sleep sensor. or can remain in bed after the primary user has left. To avoid misidentifying a user as the main user, the process distinguishes between them. In some cases this can be used by the system to can be evaluated individually (i.e., separated from the perceived data of the partner) .

[0160] 9.2 Methodology 9.2.1 Experimental design A pilot study was conducted in which five healthy subjects were recruited. Each subject slept with a partner. The 10.5GHz non-contact sensor is aimed at the patient's chest and connected to a bedside unit. The device was placed at the subject's (or user's) bedside. See Figure 1. Tactile sensors were installed to record data 24 hours per day for the duration of the study. The primary user (hereinafter referred to as the user) is the person who sleeps when he or she goes to bed and when his or her partner goes to bed. They also kept a diary detailing when they got up in the morning and when their partner got up. Yes, it was provided.

[0161] The sensor data has a start and end time equal to the time the device was first turned on. The 24-hour recording was split into two parts: (a) user data, (b) part For each record, (c) data missing, and (d) 'insufficient information' were identified. Created manually (based on user-provided diary entries in and out of bed) .

[0162] 9.2.2 Sensor signals and pre-processing An exemplary sensor system outputs two analog voltage signals representing the I and Q signals of the sensor. The sensor generates two short pulses of radio frequency energy at 10.5 GHz. The first pulse acts as the main transmit pulse and the second pulse The first pulse is reflected from a nearby object and received again by the sensor. The distance traveled by the echo pulse is determined by the time the echo signal is This time delay induces a time delay between the echo signal and the reference signal when the echo signal reaches the sensor. This introduces a phase shift between the mixer pulse and the echo pulse inside the receiver. By multiplying (mixing) the phases of the echo pulses, a signal proportional to any phase shift in the echo pulse is generated. A moving object (e.g. a person breathing) is detected by electronics in the circuit. (de Chazal, P., Fox, N., O'Ha Re, E. et al. Sleep / wake measurement using non-contact bio-motion sensors. J. Sleep R es.,2011,20:356~366).

[0163] The sensor data is filtered (low pass and anti-aliasing) prior to analysis. The signal is anti-alias filtered and in parallel for motion analysis. The data is then filtered by anti-aliasing. For example, a 30 second period or other suitable period is analyzed.

[0164] An initial presence detection was performed followed by sleep / wake detection. Movement and activity levels were monitored every The respiratory rate was calculated at 1 Hz. Overall sleep state analysis (identified A combination of light sleep, deep sleep and REM sleep may also be performed.

[0165] 9.2.3 Feature Extraction and Examination Afterwards, features (e.g., 25 features (Table 1)) were extracted (e.g., for all records). (Every 30 second epoch). Thus, the processor uses the following features in the identification process: The method may be configured to calculate or determine any one or more of the following: To do this, we use, for example, one-way analysis of variance (ANOVA) to verify that the induced features are consistent with the user and and during a known period of partner use, and each individual characteristic is compared to the Each feature can be evaluated to see whether it can meaningfully distinguish between users. Also, the average value across the entire record and the average value from past data for a specific subject This significantly increases the number of available features.

[0166] Table 1 (Example feature names and descriptions) [Table 1A] [Table 1B]

[0167] When implementing a classifier system to identify / recognize a specific person, the features are For example, every night, overnight, one or more epochs, etc.) to users and partners Any one of many such calculated features, etc. can serve to provide a statistically significant distinction between users and partners. The characteristics in Table 1 have a mean score of 0.05 or less when evaluated individually using one-way ANOVA. Have a p-value (where we consider a result to be statistically significant when p<0.05) This was found to significantly distinguish user data from partner data. These features are used to generate a classifier to distinguish between user data and partner data. indicates usefulness for inclusion in the model.

[0168] 9.2.4 Exemplary Classifier Training Such a classifier can be adopted using data from sensor recordings. For registration, all partner data is used as part of the training set. An equal number of user epochs are randomly selected to complete the training set. It is possible.

[0169] A feature selection routine (e.g., sequential forward feature selection) is used to idation) (e.g., 10-fold cross-validation) and classification methods, e.g., log We select a subset of these features that can be combined using stick regression to improve the user experience. Or you can best identify whether your partner was present or not. The selected features are used as a classifier for implementation in a processing device to identify the person. It is possible.

[0170] 9.2.5 Evaluating the performance of the classifier The trained classifier can be tested using data from the sensor recordings. The Sting set contains all partner data and all user data. This can be done.

[0171] 9.3.0 Further Example Classifiers 9.3.1 Separate classifiers for motion and non-motion Many potential features are observed during either the motion section of the data or the non-motion (breathing) section of the data. During the operation, non-sense characters (i.e. not-a-number / NaN For example, during the movement section, the settings optimizer and breathing rate The feature may be equal to NaN. Similarly, the motion feature may be equal to 0. Thus, in some versions, detection features such as those in Table 1 , a group with meaningful features during movement, and a group of clear breathing where no movement was detected. These may be organized into distinct groups, such as groups that have meaningful characteristics between them. The groups may have some overlapping characteristics that are meaningful in both scenarios. A classifier model can then be used for each group of features, such as the sensor signals The motion flags set by the analysis are used to classify breathing or motion in each epoch. This can be used as a trigger to select the model of the motion frame (see FIG. 19). Depending on the lag, a particular epoch may be associated with a particular user (e.g., user or partner) In classifying whether a particular condition is related to a particular disease, different sets of features, etc. may be considered.

[0172] 9.3.2 Developing classifiers using only subject-based metrics In some versions, the features may also be subject-based or signal quality-based. The images can be classified or grouped and used to generate two classifiers. Signal quality based features are based on real-life data, i.e., sensors capture data 24 hours a day. Both the user and the partner may be in a position where they are constantly on one side of the bed for a period of time. The user who is within a certain distance range from the sensor and the partner who is within another distance range from the sensor Two annotated beds of data (annotated tw This is useful when training a model using out-of-bed data. Using such a dataset, signal quality-based features are combined with subject-based features. The 'Set up' of Figure 19 can be used to develop a classifier. Eliminate quality-based features and include only subject-based features, e.g., breathing rate and movement. By this, the unannotated data from two distinct subjects are merged to generate a 'subject classifier'. It is possible to develop a 'Subject classifier' (see Figure 19). Recording begins when the user gets into bed and gets ready to sleep, and when the user wakes up in the morning When a record is terminated, the record typically contains a significant period of partner data. No. Partner data is only available when the primary user is not in bed but their partner is present. - Signals only exist when the user is present and the partner is not visible at all e) 'Reflected from the main user while not in use. In this scenario, both user data and All signal quality based features are captured because partner data is captured using identical settings are eliminated during classifier development for this 'subject classifier'. may be better suited to different scenarios due to the attributes of the training data used. This can be achieved by using a set of classifiers to detect when a sleep session begins or ends. While subject classifiers may be more useful for identifying who is in the middle of the day, It may be more useful to identify if someone was taking a nap.

[0173] 9.3.3 Use of Historical Data The general classification model may be used for an initial period of use, overnight or possibly longer. This allows for subject-specific over-the-counter (OTC) regressions that can be used to improve model performance over time. This allows the collection of data that is periodically updated and used. As the time period is extended, it gradually becomes more tailored and provides a more accurate classifier. An exemplary process for such a classification model is shown with reference to FIG. For example, during initial use or setup, the setup classification process in 1920 For identification of The set classification is a breathing classification and / or a movement classification in 1924 and 1926, respectively. A subject specific classification process 1928 may also be included. A more specific subject classification process 1928 can be implemented in 1930 and 1932. Processes for respiration classification and / or motion classification may also be used. The feature extraction is better than the set of features evaluated by the classification process in 1920. Other sets of features, such as those taken from unit 1922, can be evaluated, but such There may be overlaps in such sets of features.

[0174] The classifier combiner process 1940 may, for example, in 1940, A classification can be selected from the subject classification and set classification processes. The output is a binary decision. The probability (percentage) that the user is present when the If the probability is greater than the threshold, the output is that a partner is present. The threshold includes the duration or hours of sleep per day (but not The classifier set may vary over time based on a number of variables, including but not limited to: The matcher may vary as a function of time (e.g., number of sessions or collected usage data). The method may be operable to selectively select the output of a classifier process or the like. For example, You can choose only the settings category at first, and then later (e.g. for a period or session) Alternatively, you can select only subject classification instead. The classification process is then applied to both the classification output and the classification output. The weighting values ​​initially performed by the set classification process can be selectively modified. A higher weighting can be given to the identifications made over time, The discrimination weighting of the subject classification process can be gradually increased. data stream (see section 8.3.1 Separate classifiers for motion and non-motion) There is a separate breathing and movement classification process to account for possible NaN or 0 features in However, a separate cardiac feature classifier may also be included.

[0175] 9.4.0 Alternative Machine Learning Models Optionally, other supervised, semi-supervised, or unsupervised machine learning models For example, anomaly detection, support vector machine or clustering classification methods can be implemented. Methods may be provided in further system implementations.

[0176] 10.0 Further Methodologies for Distinguishing Multiple Users / Subjects FIG. 20 illustrates, for example, a case in which multiple devices communicate with each other and / or a central processing unit, e.g. For example, when a server processes data from sensors, the number of devices monitored by N devices is FIG. 20 shows a process that may be implemented in a system for discrimination of N subjects. Only two sensor devices (100-A and 100-B) are shown.

[0177] For example, each sensor device may access data received from other sensor devices in a common neighborhood. It is possible to access or process the data and apply a hierarchical classification process. For example, each sensor device (1 and 2) in FIG. 20 is initially, for example, By comparing the sensor data or features extracted from the sensor data from the , evaluate whether other devices are detected or not, 2010a, 2010b. The term "feature extraction" refers to the process in which data from device (2) is extracted from both devices (1) (within the dotted area called 100a) as well as device (2) (within the dotted area called 100b). Similarly, data from device (1) is also processed on both devices. This allows for comparison between the data collected by both devices and ensures that the same person, or to be able to determine whether two individuals were being monitored. If both devices are monitoring the same person, the data with the better signal quality will be used. The classification can be used to determine which user it is. If you are monitoring two individuals, this diagram would be: Main User We will proceed with data analysis assuming this to be the case.

[0178] If not, analysis of the devices that detected the pattern will be carried out in 2012a, 2012b. That is, each device can then proceed normally in the sense of each user's By collecting and evaluating the patient's own data (e.g., respiratory characteristics, cardiac characteristics, sleep characteristics, etc.), The monitoring sessions of different users can be conducted independently.

[0179] However, if multiple devices detect the same pattern (i.e., the same person is If the object is detected by multiple sensors, parallel classifiers (e.g., two or more) are installed on each device. Each device then processes the detected subjects as described in, for example, FIG. By implementing a process comparable to that of A classifier can be implemented to identify / identify when the sensed data is The data is used to set up a classification process (e.g., with historical data) that targets subject recognition. a, 2020b and / or subject-specific classification processes 2028a, 2028b The sorter combiners 2040a, 2040b may then sort the data based on the previously monitored It is possible to determine the case regarding recognized subjects. For example, sensor monitoring can be performed by using the accumulated information identified for previous sessions. To continue monitoring users, it can be carried out in 2012a and 2012b. If there is no recognized user, the sensor data can be ignored or the sensor The user can then log in with any sensors / monitors from previous sessions that can be attributed to other users. Monitor new users in 2012a and 2012b in a non-cumulative manner against the tracking data. You can proceed to tally.

[0180] The potential advantage of this approach is that it is possible to have multiple people (e.g., a room with two sensors) 2 people) and requires biometric analysis for a larger population. It is customizable to a specific person (for example, two people) in a way that does not require a Therefore, the distinction is made regarding finite groups (e.g., household / bed partners or room patients). Also, when the detection ranges overlap, both of the two sensors may be single-ended. It also enables the output of multiple sensors to be shared and communicated when detecting a bed user. , which allows system-level separation of sleep / breathing / movement / heart rate. Therefore, the bed shape And size, bedside locker, in the context of widely varying people, The robustness of the deployed system can be significantly improved.

[0181] FIG. 20 shows two parallel or coexisting sensor processes, but with a further flow processing block. By duplicating the data on multiple devices and sharing it, more than two people can be monitored. It is understood that this may be the case.

[0182] In this regard, various technologies are used to analyze sensor data. For example, to enable sensor-to-sensor (point-to-point) communication, For this purpose, wireless communication (e.g., Wi-Fi Direct (ad-hoc)) is used to selectively communicate with sensors, etc. Other forms of communication (e.g., wired or other wireless communication) may be implemented. It is also possible.

[0183] 11.0 Additional Optional Features (i) In some cases, the sensor device may be wirelessly charged (e.g., using a QI wireless charger). Such a rechargeable sensor product can be integrated with a rechargeable product. The device may have a battery (e.g., lithium) and coil that allows it to be used. Such a sensor product would then be charged during the day when placed in close proximity to a wireless charger. After charging, during a monitoring session (e.g., overnight near the user) The sensor product may be inserted into a holding structure for directing the sensor product for use. For example, the sensor Mates with a wall mount, stand or bed mount that will receive your rechargeable sensor product The wireless charger can then be used to charge other devices (e.g., smartphones) overnight. Can be used to charge the device.

[0184] (ii) In some cases, the sensor device or the like is microphone and / or infrared sensitive. Integrated cameras (e.g., those that do not filter out IR or have other infrared detection capabilities) Each of the cameras may include an infrared emitter (e.g., a camera device optimized to emit infrared light). For example, one or more IR LEDs with a diffuser / lens may also be included. So, the sensor device may have a function to detect (for example, stored in the device memory and / or transmit wireless communication) The sensor and camera are both in a dark bedroom. It can record user movements in real time or at night by event ( For example, SDB episodes detected via non-contact or minimal contact sensors, PLM / R specific types of movement such as LS, abnormal movements, respiratory rate) or (e.g. non-contact Snoring, wheezing, coughing, etc. (detected via microphone / transducer) If other events that match the common time frame are detected, the processor Link / associate with segments of video captured by the IR camera during a video shoot Then, the related / concatenated segments (and optionally the periodic time frames) can be (e.g., for display on a display device such as a telephone or computer) Indexing videos to be accessed together to view important events of the night So, for example, you can record SDB events in a video For example, the user can see that they have stopped breathing. You can see that he regains breathing after a while. The device is a smart device (e.g., a sensor device with an integrated monitor / display). or a smartphone, laptop or computer that communicates with the sensor device This allows for a very fast video review of a night's sleep by events, for example. It is possible.

[0185] (iii) Such camera-based sensors may also provide real-time or near-real-time For example, the sensor may be implemented for remote viewing by a patient or other user. for sleep and other states (e.g., the wakefulness of monitored subjects during deep sleep) Thus, for example, Serve as an infant monitor / telemedicine specialist monitor in a clinic / hospital environment. It is possible.

[0186] 12.0 Treatment For example, some versions of current technology include therapeutic devices to treat respiratory disorders. Various treatments can be used to prevent or improve breathing. One can take advantage of such treatment to prevent respiratory disease.

[0187] Nasal continuous positive airway pressure (CPAP) therapy is used to treat obstructive sleep apnea (OSA). The mechanism of action is that the continuous positive airway pressure acts as a pneumatic splint, blocking the soft palate and tongue. By pushing the catheter toward the posterior oropharyngeal wall, it is possible to prevent obstruction of the upper airway. .

[0188] Non-invasive ventilation (NIV) therapy is used to ensure that patients are breathing fully and / or receiving adequate oxygen levels. It does some or all of the work of breathing to help keep the upper airway open and help maintain a good Ventilatory support is provided to the patient through a non-invasive patient interface. NIV has been used to treat CSR, OHS, COPD, NMD and chest wall disease.

[0189] 12.1 Diagnostic and Therapeutic Systems In some versions of the technology, the treatment is performed in conjunction with the aforementioned identification device / system. Such a system may be provided by a treatment system or device that can be operated. The devices can also be used to diagnose a condition without treating it.

[0190] The treatment system includes a respiratory pressure therapy (PRT) device, an air circuit, a humidifier, and a patient interface. It consists of interface and data management.

[0191] The patient interface may, for example, provide a flow of breathable gas to the respiratory device. The flow of breathable gas is delivered through the nasal cavity. and / or a mask in the mouth, a tube in the mouth or a tracheostomy tube in the user's trachea Depending on the therapy being applied, the patient interface may form a seal. For example, in the patient's face, the pressure is sufficiently different from the surrounding pressure for the treatment to be effective. The gas is delivered at a positive pressure of, for example, about 10 cmH2O relative to the ambient pressure. In other forms, such as high-flow air, a positive pressure of approximately 10 cmH2O is required to deliver air to the airways. The patient interface need not include a seal that can be provided.

[0192] 12.2 Respiratory Pressure Therapy (RPT) Devices Air pressure generators are known for various applications, such as industrial scale ventilation systems. However, medical air pressure generators have more general requirements such as reliability, size and weight. There are certain requirements that cannot be met by conventional air pressure generators. In addition, there are devices designed for medical use. Regardless of installation, comfort, noise, ease of use, efficiency, size, weight, manufacturability, cost and reliability are all considered. It is implied by one or more defects in sexuality.

[0193] One known RPT device used to treat sleep-disordered breathing is CPAP therapy. The S9Sleep Theapy System is manufactured by ResMed. RPT Another example of a device is a ventilator. ResMe adult and pediatric ventilators d Ventilators such as the Stellar™ series are designed to reduce NMD, OHS and CO Invasive and noninvasive treatments for a variety of patients treating a number of conditions, including but not limited to PD. and can provide support for non-invasive non-dependent ventilator therapy.

[0194] RPT devices typically consist of a pressure generator such as a motor-driven blower or compressed gas reservoir. The RPT device comprises a resuscitator and is configured to supply pressurized air to the patient's airway. The outlet is connected via a pneumatic circuit to the patient interface.

[0195] 12.1 Exemplary Treatment Devices of Option As previously discussed, in one form, the present technology provides a method for treating and / or monitoring a respiratory disorder. The instrument or device may include an instrument or device for ringing the air circuit 41. 70 to the patient interface 3000. In the following description, the RPT device may be an RPT device 4000 for detecting the presence of a target object. 24.

[0196] 12.2 Patient Interface A non-invasive patient interface 3000 according to one aspect of the present technology has the following functional aspects: Seal forming structure 3100, plenum chamber 3200, positioning and stabilizing structure 3300 and a connection port 3600 for connecting to an air circuit 4170, a vent 3400, an air One form of connection port 3600 for connecting to the airway circuit 4170 and a forehead support 3700. In some configurations, the functionality is provided by one or more physical components. In some embodiments, one physical component may serve one or more functions. During use, the device can be used to facilitate the delivery of pressurized air to the airways. The seal-forming structure 3100 is positioned to surround the entrance to the patient's airways so as to

[0197] 12.3 RPT device The RPT device 4000 according to one aspect of the present technology comprises mechanical and pneumatic components 41. 00, an electrical component 4200, and one or more algorithms 43 The RPT device 4000 is programmed to run the The housing 4010 includes an upper portion 4012 and a lower portion 4014, which are connected to each other. The outer housing 4010 may include one or more panels 4015. The RPT device 4000 may include one or more internal components of the RPT device 4000. The RPT device 4000 may include a chassis 4016 that supports the components. may include 18.

[0198] The air passage of the RPT device 4000 includes an inlet filter 4112, an inlet muffler 4122, A pressure generator 4140 (e.g., a blower 4142) capable of supplying pressurized air and an outlet One or two sensors such as the Fuller 4124, the pressure sensor 4272 and the flow sensor 4274 It may include one or more airway items such as more transducers 4270 .

[0199] One or more airway items may be attached to a removable air block 4020. The air block 4020 can be located within the outer housing 4010. In one embodiment, the air block 4020 is supported by the chassis 4016 or forms part of it.

[0200] The RPT device 4000 includes a power supply 4210, one or more input devices 4220, a central A controller 4230, a treatment device controller 4240, a pressure generator 4140, or Further, a protection circuit 4250, a memory 4260, a converter 4270, a data communication interface An input / output device 4290 may comprise a display 4280 and one or more output devices 4290. The electrical components 4200 are mounted on a single printed circuit board unit (PCBA) 4202. Alternatively, the RPT device 4000 may include one or more PCBAs 4202. is fine.

[0201] 12.3.1 Mechanical and Pneumatic Components of the RPT Device The RPT device 4000 may include one or more of the following components in an integrated unit: In an alternative embodiment, one or more of the following components may be included: Each of these can be positioned as a separate unit.

[0202] 12.3.1.1 Air filters An RPT device 4000 in accordance with one form of the present technology includes an air filter 4110 or a plurality of air An air filter 4110 may be included.

[0203] In one form, the air inlet filter 4112 is located upstream of the pressure generator 4140. It is located at the beginning of the airway.

[0204] In one form, an air outlet filter 4114, such as an antibacterial filter, It is located between the air block outlet 4020 and the patient interface 3000.

[0205] 12.3.1.2 Mufflers An RPT device 4000 in accordance with one form of the present technology includes a muffler 4120 or a plurality of mufflers. It may contain Lar 4120.

[0206] In one form of the present technology, the inlet muffler 4122 is an air passage upstream of the pressure generator 4140. It is located in.

[0207] In one form of the present technology, the outlet muffler 4124 is connected to a pressure generator 4140 and a patient interface. It is located in the air passage between the Face 3000.

[0208] 12.3.1.13 Pressure generator In one form of the present technology, the pressure device 4140 that provides the pressurized air is a controllable blower. For example, the blower 4142 may be one or more fans housed in a volute. The pressure generator 4140 includes a brushless DC motor 4144 having an impeller of, for example, For example, a positive pressure of about 120 liters / min can be achieved in the range of about 4 cmH2O to about 20 cmH2O. or other form of air supply or flow in the range of approximately 30 cmH2O. .

[0209] The pressure generator 4140 is under the control of the therapy delivery controller 4240 .

[0210] In other embodiments, the pressure generator 4140 may be a piston-driven pump, a high pressure source (e.g. The pressure regulator may be a bellows or pressure regulator connected to a pressure source (eg, compressed air reservoir).

[0211] 12.3.1.4 Transducers The transducer may be internal or external to the RPT device. An external transducer is, for example, placed in the air circuit. or forms part of the air circuit, such as a patient interface. The external transducer transmits data to the RP A form of non-contact sensor, such as a Doppler radar motion sensor, that transmits or transfers information to the device The state may be as follows.

[0212] In one form of the present technology, one or more transducers 4270 are connected to a pressure generator 4140. One or more transducers 4270 are located upstream and / or downstream of the flow. The characteristics of each air flow such as volume, pressure or temperature at that point in the air path The device is constructed and arranged to generate data indicative of a characteristic of the device.

[0213] In one form of the present technology, one or more transducers 4270 are connected to a patient interface. Located proximal to 3000.

[0214] In one embodiment, the signal from the converter 4270 is low pass, high pass or band pass filtered. - Filtered by filtering etc.

[0215] 12.3.1.5 Anti-spillback valves In one form of the present technology, an anti-spillback valve 4160 is connected to the humidifier 5000 and the air block. The anti-spillback valve is located between the humidifier 5000 and the humidifier 4020. It is constructed and arranged to reduce the risk of leakage, for example to the motor 4144.

[0216] 12.3.1.6 Air circuits In accordance with one aspect of the present technology, an air circuit 4170 is configured such that the air flow is 0 and Patient Interface 3000 while in use. The present invention is a conduit or tube constructed and arranged to provide a

[0217] 12.3.1.7 Supplemental oxygen In one form of the present technology, supplemental oxygen 4180 is provided, for example, upstream of air block 4020, in the air One or more of the air passageways, such as to the circuit 4170 and / or to the patient interface 3000 or more points.

[0218] 12.3.2 RPT Equipment Electrical Components 12.3.2.1 Power supply In one form of the present technology, the power supply 4210 is integrated into the external housing of the RPT device 4000. In another form of the present technology, the power supply 4210 is internal to the RPT device 400. 0 external housing 4010.

[0219] In one form of the present technology, the power supply 4210 provides electricity only to the RPT device 4000. In another form of the present technology, a power supply 4210 may be provided to power the RPT device 4000 and the humidifier 5000. Provides electricity to both.

[0220] 12.3.2.2 Input Devices In one form of the present technology, the RPT device 4000 includes a button to allow an operator to interact with the device. It includes one or more input devices 4220 in the form of a button, switch, or dial. A button, switch or dial can be accessed through a physical device or a touch screen. In one embodiment, the control may be a software device that controls the operation of the control. The modules may be physically connected to the housing 4010 or, in another embodiment, may be connected to a central control The communication may be wireless with a receiver in electrical communication with the controller 4230 .

[0221] In one form, the input device 4220 allows an operator to select values ​​and / or menu options. The system is constructed and arranged so that the options can be selected.

[0222] 12.3.2.3 Central Controller In one form of the present technology, the central controller 4230 is an x86 INTENL processor. The RPT device 4000 is a processor suitable for controlling the RPT device 4000, such as a processor.

[0223] A central controller suitable for controlling an RPT device 4000 according to another aspect of the present technology. The 4230 is based on the ARM Cortex-M processor from ARM Holdings. For example, the STM from ST MICROELECTRONICS A 32 series microcontroller may be used.

[0224] Other central controllers suitable for controlling the RPT device 4000 according to further alternative forms of the present technology. The Controller 4230 is a choice of ARM9-based 32-bit RISC CPUs. For example, STR from ST MICROELECTRONICS A 9 series microcontroller may be used.

[0225] In certain alternative forms of the present technology, a 16-beat RISC CPU controls the RPT device 400. For example, the TEXAS INSTRU Programming from the MSP430 family of microcontrollers manufactured by MENTS A processor may be used.

[0226] In another form of the present technology, the central controller 4230 is a dedicated electronic circuit. In one form, the central controller 4230 is an application specific integrated circuit (ASIC). In another form, the central controller 4230 is implemented as discrete electronic components. Includes.

[0227] The central controller 4230 controls one or more converters 4270 and one or Further, the input device 4220 and the input signal(s) from the humidifier 5000 are received. It is structured as follows.

[0228] The central controller 4230 may transmit the output signal(s) to one or more output devices 4 290, a treatment device controller 4240, a data communication interface 4280 and a humidification The device 5000 is configured to provide

[0229] In some forms of the present technology, a central controller 4230 may 4260 and stored in a persistent computer readable storage medium, such as memory 4260. 4300, such as one or more algorithms described herein. In some embodiments of the present technology, the method may include any of the above-mentioned methods. Thus, the central controller 4230 can be integrated with the RPT device 4000. However, in some forms of the present technology, some methodologies may be implemented by remotely located devices. For example, a remotely located device may determine the control settings of a ventilator and Respiratory-related events can be detected by analyzing stored data from any of the sensors described. can be detected.

[0230] The central controller 4230 controls various sensors (e.g., transducers 4270), data communications The interface 4280 is a single computer that interacts with the memory 4260 as well as other devices. The functionality of the controller 4230 may include more than one controller. Thus, as used herein, the term "center" refers to the Limiting the architecture to a single controller or processor that controls other devices For example, alternative architectures may include more than one controller or processor. This may include a distributed controller architecture that includes, for example, A separate local (i.e., within the RPT device 4000) that performs part of the algorithm 4300 or a remotely located controller, or more than one that stores a portion of the algorithm. The computer program may further include a local or remote memory for An algorithm, when expressed as a set of high-level human-readable code (e.g., C++, Visual Basic, other object-oriented languages, etc.) or low / machine level instruction language ( The algorithm(s) may include Such code or instructions may be implemented in a controller, such as an ASIC or DSP. Can be burned or performs the tasks required by the algorithm(s) Execution ported to a specially programmed DSP or general-purpose processor for It may be a possible runtime.

[0231] 12.3.2.4 Clock The RPT device 4000 includes a clock 4232 connected to the central controller 4230. This can be done.

[0232] 12.3.2.5 Therapy Device Controller In one form of the present technology, the therapy device controller 4240 includes a central controller 423 0 is a control module 4330 which forms part of the algorithm 4300 that executes.

[0233] In one form of the present technology, a therapy device controller 4240 includes an integrated circuit for dedicated motor control. For example, the MC33035 brushless DC motor controller manufactured by ONSEMI It is one form of controller.

[0234] 12.3.2.6 Protection circuit The RPT device 4000 according to the present technology may include one or more protection circuits 4250. .

[0235] One form of the protection circuit 4250 according to the present technology is an electrical protection circuit. One form of protection circuit 4250 is a temperature or pressure safety circuit.

[0236] 12.3.2.7 Memory In accordance with one form of the present technology, the RPT device 4000 includes a memory 4260, e.g., a non-volatile memory. In some configurations, the memory 4260 is a battery-powered static In some forms, the memory 4260 may include volatile RAM. may include.

[0237] The memory 4260 may be located on the PCBA 4202. The memory 4260 may include an EEPROM. It can be in the form of OM or NAND flash.

[0238] Additionally or alternatively, the RPT device 4000 may be, for example, a Secure Dingi A removable memory card like the standard SD card 4260 Includes.

[0239] In one form of the present technology, the memory 4260 stores one or more algorithms. A computer, such as the system 4300, that implements one or more of the methodologies described herein. A non-transitory computer-readable storage medium having computer program instructions thereon Work as.

[0240] 12.3.2.8 Converters The transducer may be internal to the device 4000 or external to the RPT device 4000. The converter may be located on the air delivery circuit 4170, for example at the patient interface, or can form part of the external transducer. Or it may be in the form of a non-contact sensor such as a transmitting Doppler radar motion sensor.

[0241] 12.3.2.8.1 Flow rate The flow transducer 4274 according to the present technology may be a differential pressure transducer, such as a The differential pressure transducer can be based on the SDP600 series. and one of the pressure transducers is in fluid communication with a respective first and second pressure transducers in the flow restricting element. It is connected to a second point.

[0242] In one example, a signal representing the total flow (Qt) from the flow transducer 4274 is The signal is received by the controller 4230.

[0243] 12.3.2.8.2 Pressure A pressure transducer 4272 according to the present technology is placed in fluid communication with the pneumatic path. An example of a pressure transducer 4272 is a sensor from the HONEYWELL ASDX series. Other suitable pressure transducers are the NPA series from GENERAL ELECTRIC. These sensors are

[0244] In use, the signal from the pressure transducer 4272 is received by the central controller 4230. In one form, the signal from the pressure transducer 4272 is transmitted to the central controller 42 30 before being received.

[0245] 12.3.2.8.3 Motor Speed In one form of the present technology, a motor speed converter 4276 is and / or used to determine the rotational speed of the blower 4142. Motor Speed ​​Converter 4 The motor speed signal from 276 may be provided to the therapy device controller 4240. The rotor speed transducer 4276 may be, for example, a speed sensor such as a Hall effect sensor.

[0246] 12.3.2.9 Data communication systems In one form of the present technology, a data communication interface 4280 can be provided, The data communication interface 4280 can be connected to a remote external communication network. 4282 and / or a local external communication network 42 84. The remote external communications network 4282 is connected to a remote external device 4286. The local external communication network 4284 can be connected to a local external device 4288. It is possible to connect.

[0247] In one embodiment, the data communication interface 4280 is connected to the central controller 4230. In another embodiment, the data communication interface 4280 is part of a central The controller 4230 may include an integrated circuit or processor separate from the controller 4230 .

[0248] In one embodiment, the remote external communications network 4282 is the Internet. The data communication interface 4280 is a wired connection (e.g., via Ethernet or optical fiber). bar) or wireless protocols that connect to the Internet (e.g., CDMA, GSM, LAN TE).

[0249] In one embodiment, the local external communications network 4284 is Bluetooth or The wireless communication system may utilize one or more communication standards such as the Wireless Receiver Protocol or the Consumer Infrared Protocol.

[0250] In one embodiment, the remote external device 4286 is connected to one or more computers, e.g., a network In one embodiment, the remote external device 4 The 286 may be a virtual computer rather than a physical computer. Even in this case, such remote external device 4286 must be operated by an appropriately authorized person, such as a clinician. can access it.

[0251] The local external device 4288 may be a personal computer, a mobile phone, a tablet or It may be a remote control.

[0252] In one embodiment, the interface includes, for example, an RF motion sensor as described herein. 1. The sensor may communicate with a sensor such as any one of the sensors described in .

[0253] 12.3.2.10 Optional displays and output devices, including alarms The output device 4290 according to the present technology may be one or more of a visual, audio and tactile unit. The visual display may be a Liquid Crystal Display (LCD) or It may be a light emitting diode (LED) display.

[0254] 12.3.2.10.1 Display Drivers The display driver 4292 receives input and displays it on the display 4294. , symbols or images, converts them into commands and displays them on the display 4294. The character, symbol or image is displayed.

[0255] 12.3.2.10.2 Display The display 4294 responds to commands received from the display driver 4292. The device is configured to visually display letters, symbols or images in response to the display. The ray 4294 may be an 8-segment display, in which case the display driver 42 92 converts each character or symbol, such as the number "0", into eight logic signals, Each of the eight segments can be activated to display a specific letter or symbol. Indicate whether

[0256] 12.3.3 RPT Device Algorithm 12.3.3.1 Pre-processing module The pre-treatment module 4310 according to the present technology includes a transducer 4270, e.g., a flow sensor 42 74 or pressure sensor 4272 as input and 4320. Calculate one or more output values ​​to be used as forces.

[0257] In one form of the present technology, the output values ​​are the interface or mask pressure Pm, the respiratory rate Includes the flow rate Qr and the leakage flow rate Ql.

[0258] In various aspects of the present technology, the pre-processing module 4310 performs one of the following algorithms: Includes one or more of: Pressure Compensation 4312, Vent Flow Estimation 4314, Leak Flow Estimation 431 6. Estimating respiratory flow 4317, determining ventilation 4311, determining target ventilation 4313, estimating respiratory rate 43 18, and backup rate determination 4319.

[0259] 12.3.3.1.1 Pressure compensation In one form of the present technology, the pressure compensation algorithm 4312 is A signal indicative of the pressure in the pneumatic path near the outlet of 20 is received as an input. The rhythm 4312 estimates the pressure drop in the air circuit 4170 and controls the patient interface 300 The estimated pressure Pm at 0 is provided as an output.

[0260] 12.3.3.1.2 Vent flow estimation In one form of the present technology, the vent flow assessment algorithm 4314 is The patient interface 3000 receives as input the estimated pressure Pm at the patient interface 300 Estimate the air vent flow rate Qv from vent 3400 at 0.

[0261] 12.3.3.1.3 Leakage flow rate estimation In one form of the present technology, the leak flow estimation algorithm 4316 calculates the total flow Qt and The vent flow rate Qv is received as an input and the leakage flow rate Ql is estimated. The flow estimation algorithm 4316 may be implemented to estimate the flow rate over a period of time long enough to include several respiratory cycles, e.g. For example, by calculating the average of the difference between the total flow rate and the vent flow rate Qv over a period of about 10 seconds. The leakage flow rate Ql is estimated using the above formula.

[0262] In one form, the leak flow estimation algorithm 4316 is The total flow rate Qt, vent flow rate Qv, and estimated pressure Pm at 00 are received as inputs. The leak conductance is calculated to obtain the leak flow rate Ql. Determine the leakage flow rate Ql as a function of conductance and pressure Pm. The leak conductance is a low value equal to the difference between the total flow rate Qt and the vent flow rate Qv. The low-pass filtered average of the non-vented flow rate and the pressure Pm. The time constant can be calculated as the square root of the square root of the sigma-squared coefficient, where the low pass filter time constant is a function of the time constant that covers several respiratory cycles. The leakage flow rate Ql is a value that is sufficiently long to cover the leakage conductance, for example, about 10 seconds. It can be estimated as the product of the pressure function Pm and the

[0263] 12.3.3.1.4 Respiratory flow estimation In one form of the present technology, the respiratory flow estimation algorithm 4317 calculates the total flow Qt, The vent flow rate Qv and the leak flow rate Ql are received as inputs, and the vent flow rate Qv and the leak flow rate Ql are calculated from the total flow rate Qt. The respiratory flow of air to the patient, Qr, is estimated by subtracting the flow Ql.

[0264] In another aspect of the present technology, the respiratory flow estimation algorithm 4317 is Possible proxies for respiratory flow include: nothing: - Patient 1000 chest respiratory movements - the current drawn by the pressure generator 4140 -Motor speed of pressure generator 4140 -Trans-thoracic impedance of 1000 patients The respiratory flow proxy value is measured using a transducer 4270 in the RPT device 4000, e.g., a motor speed A sensor 4276, or a sensor external to the RPT device 4000, such as a respiratory motion sensor or a transthoracic impedance sensor.

[0265] 12.3.3.1.5 Ventilation determination In one form of the present technology, a ventilation decision algorithm 4311 receives a respiratory flow Qr. and determines a measurement, Vent, that indicates the current patient ventilation.

[0266] In some embodiments, the ventilation decision algorithm 4311 estimates the actual patient ventilation. Determine the ventilation measure Vent, which is

[0267] In one such embodiment, the ventilation measure Vent is measured at a corner of 0.11 Hz. The frequency is selectively filtered by a low-pass filter, such as a second-order Bessel low-pass filter. It is half the absolute value of the respiratory flow Qr, which is filtered.

[0268] In one such embodiment, the ventilation measure Vent is calculated as total alveolar ventilation (gross alveolar ventilation (i.e., non-anatomical dead space ventilation) This requires an estimate of the anatomic dead space. The patient's height (or arm-span in the case of severe skeletal deformity) can be used as a good predictor. Then, the total alveolar ventilation, for example, is the same as the measured ventilation of the actual patient determined as described above, and is less than the product of the estimated anatomical dead space and the estimated spontaneous respiration rate Rs. -span). and is less than the product of the estimated anatomical dead space and the estimated spontaneous respiration rate Rs.

[0269] In other embodiments, the ventilation determination algorithm 4311 determines a measured value of ventilation Vent that is approximately proportional to the ventilation of the actual patient. One such embodiment estimates the peak respiratory flow Qpeak throughout the inspiratory portion of the cycle. This procedure, which involves sampling the respiratory flow Qr, and many other procedures, generate measured values that are approximately proportional to ventilation as long as the shape of the flow waveform does not change very much (where two breaths are taken as similar when the flow waveforms of the breaths standardized in time and amplitude are similar). Some simple examples include the central positive respiratory flow, the median of the absolute values of the respiratory flow, and the standard deviation of the flow. The absolute value of the respiratory flow using a positive coefficient, and even some arbitrary linear combinations of any of the several order statistics using both positive and negative coefficients, are approximately proportional to ventilation. Another example is the average of the respiratory flow at the mid-K ratio (time-based) in the inspiratory portion, where 0 < K < 1. There are many measured values that are directly proportional to ventilation when the shape of the flow waveform is constant. In other forms, the ventilation determination algorithm 4311 does not rely on the respiratory flow Qr, but instead determines a measured value of ventilation Vent that is a proxy for the current patient's ventilation, such as the oxygen saturation (SaO2) or the partial pressure of carbon dioxide (PCO2), obtained from a suitable sensor attached to the patient 1000.

[0270] 12.3.3.1.6 Determining target ventilation In one form of the present technology, a central controller 4230 receives a current ventilation measurement Ven The measured value of t is taken as input and the target value Vtgt for the measured ventilation is determined. , performs one or more target ventilation determination algorithms 4313.

[0271] In some forms of the present technology, there is no target ventilation determination algorithm 4313, e.g., RP By hard-coding during configuration of the T device 4000 or via input device 4220 The target ventilation, Vtgt, is predetermined by manual input.

[0272] Adaptive servo-ventilation (AS V) In other forms of the present technology, such as therapy (described below), target ventilation determination algorithm 4 313 calculates the target ventilation Vtgt from the value Vtyp that indicates the typical recent ventilation of patient 1000. Perform calculations.

[0273] In some forms of adaptive servo-ventilation therapy, the target ventilation Vtgt is set to a typical The ventilation Vtyp is calculated as a smaller but higher percentage. The percentages are (80%, 100%), or (85%, 95%), or (87%, 92%). may be in the range.

[0274] In other forms of adaptive servo-ventilation therapy, the target ventilation Vtgt is set according to typical modern ventilation It is calculated as slightly greater than a single multiple of Vtyp.

[0275] A typical modern ventilation Vtyp is measured over multiple time periods over some pre-determined time scale. The distribution of recent ventilation measurements Vent over a period of time tends to be clustered. That is, it is a measure of the central tendency of the current ventilation measurements over recent history. Target Ventilation Determination In one embodiment of the algorithm 4313, the recent history is on the order of a few minutes, but In any case, the deviation should be longer than the time scale of the Cheyne-Stokes waxing and waning periods. The target ventilation decision algorithm 4313 determines the typical maximum ventilation from the current ventilation measurement Vent. To determine the approximate ventilation Vtyp, any of a variety of well-known measures of central tendency can be used. One such measurement can be made using the current This is the output of a low pass filter for the Vent measurements.

[0276] 12.3.3.1.7 Respiratory rate estimation In one form of the present technology, the respiration rate estimation algorithm 4318 estimates the respiration rate for the patient 1000. The method receives as input the respiratory flow rate Qr, which is the rate at which the patient is responsive to the respiratory rate, and generates an estimate of the patient's spontaneous breathing rate Rs.

[0277] The respiration rate estimation algorithm 4318 estimates the respiration rate when the patient 1000 is breathing spontaneously, i.e. That is, when the RPT device 4000 is not delivering a "back-up breath" (described below). It is possible to estimate the spontaneous breathing rate Rs over a period of time. Therefore, the respiration rate estimation algorithm 4318 is servo-assisted (pressure-assisted) When the pressure assist is low, one actual In an embodiment, the respiration rate is estimated over the period when the respiration rate is below 4 cmH2O, which is Such periods are more likely to reflect spontaneous respiratory effort.

[0278] In some forms of the present technology, the respiration rate estimation algorithm 4318 may Since the breathing rate during sleep can be substantially different from the breathing rate during wakefulness, Anxiety typically produces a higher respiratory rate than prevails during sleep. When patients focus on their breathing process, their breathing rate typically remains constant, consistent with normal arousal. This is lower than that during sleep or during sleep. and International Patent Application PCT / AU2010, the entire disclosure of which is incorporated herein by reference. / 000894 can be used to identify waking breathing periods from respiratory flow Qr .

[0279] In some forms of the present technology, the respiration rate estimation algorithm 4318 estimates the respiration rate during the period of interest. The reciprocal of one of various well-known statistical measures of central tendency of the absorption duration Ttot (r eciprocal) to estimate the spontaneous respiratory rate Rs. It is preferable to reject outliers or at least be robust to them. One such measurement is where the lower and upper K ratios of aligned breath durations are discarded. The constant, i.e. the trimmed mean, and the remaining The average calculated for breath duration is robust to outliers. For example, if K is 0.25, This corresponds to discarding the upper and lower quartiles of the respiratory duration Ttot. The median is another robust measure of central tendency, but it is sometimes used when the distribution is strongly bimodal (b imodal) can give unsatisfactory results. A simple average is It can also be used as a measure of respiratory rate, but is sensitive to outliers. For example, consecutive time intervals exceeding 45 breaths / min or less than 6 breaths / min are An initial interval filtering step may be used to exclude outliers from the calculation. Other filtering mechanisms that may be used in combination with interval filtering include: Exclude breaths that are not part of a sequence of N consecutive spontaneous breaths (where N is , some small integer (e.g., 3) and consecutive spontaneous early and late exhalations. Eliminating inspiration, e.g., eliminating the first and last breaths of a sequence of four breaths The rationale for the latter mechanism is that during spontaneous breathing sequences, First and last breaths, and generally early and late breaths, can be atypical; e.g. Initially, spontaneous breathing can occur as a result of arousal, and eventually spontaneous breathing can occur as a result of arousal. A spontaneous breath is a dry breath that results in a backup breath that ends the spontaneous breath sequence. can be longer due to reduced respiratory drive .

[0280] In some forms of the present technology, the respiration rate estimation algorithm 4318 may be configured to estimate an initial estimation period. The initial estimate of spontaneous respiratory rate Rs is then used by the therapy engine module 4320 to After starting the process, estimate the spontaneous respiratory rate Rs using a longer estimation period than the initial estimation period. To improve statistical robustness, we continuously update the initial estimation period. It could be 20 minutes of normal spontaneous breathing, but then the estimated duration would be some maximum duration, e.g. For this estimation, the duration of such Rather than using a rolling window, a low pass filter on the breath duration is As the session progresses, the participants are given progressively longer reaction times (or, more precisely, progressively lower covariates). The frequency bands may be used having different frequencies.

[0281] In some forms, a short-term estimate with appropriate treatment of central tendency, such as a trimmed mean, The period (e.g., 10 minutes) measurements are input to an appropriate low-pass filter to convert them into time units or This provides an estimate of Rs that varies over time scales of The averages sampled are for moving windows of breath duration data lasting several hours or days. This can occur when the breath duration data needs to be calculated in real time, potentially resulting in a large amount of This has the advantage that no data needs to be stored and processed.

[0282] In some forms of the technology, the measurements are taken over a short period of time, particularly over a single breath. The respiratory rate measured may also be used instead of respiratory duration in the central tendency measures described above. , which generally provide similar, but not identical, results.

[0283] 12.3.3.1.8 Backup Rate Determination In one form of the present technology, the backup rate determination algorithm 4319 is Receives as input the spontaneous respiration rate estimate Rs provided by the constant algorithm 4318. The backup rate Rb is returned. In the event of a significant lack of spontaneous respiratory effort, the patient 1000 is provided with a backup breath, i.e. and the proportion of those who continue to receive ventilatory support.

[0284] In one embodiment of the pre-processing module 4310, a backup rate determination algorithm 431 9, and instead the backup rate Rb is input to the RPT device via, for example, the input device 4220. The RPT device 4000 may be manually provided thereto or may be hard-coded during construction of the RPT device 4000.

[0285] In one form, known as adaptive backup rate, the backup rate decision algorithm Rhythm 4319 determines the backup rate Rb as a function of the spontaneous breathing rate Rs. In this embodiment, the function is to convert the backup rate Rb from the spontaneous breathing rate Rs to 2 breaths / min. In another embodiment, the function is determined as the backup The respiratory rate, Rb, is determined as the spontaneous respiratory rate, Rs, multiplied by a constant slightly less than one.

[0286] In one form known as the variable backup rate, the backup rate decision algorithm The rhythm 4319 determines the backup rate Rb as a function of time. b is the sustained timed backup rate The spontaneous backup rate (STBR) is some fraction of the final target backup rate. As the spontaneous backup rate (SBR) The fraction is initialized to a known value. The fraction may be 2 / 3, or 3 / 4, or any other positive value less than 1. SBR is the time when the most recent inspiration was a spontaneous (i.e., patient-triggered) breath. STBR is the inverse of the timeout period for a backup breath when How can the information be determined (e.g., by manual entry or hard coding as described above)? The breathing rate may be set to some typical rate, such as 15 bpm or 20 bpm. The backup rate Rb changes from SBR to STBR as time elapses since the spontaneous breathing. The increase is in accordance with a predetermined profile, e.g., a series of The profile may follow a step or continuous linear profile, said profile being: The backup rate Rb is selected to reach the STBR after a predetermined interval. Intervals are measured in units of time, such as 30 seconds, or in terms of patient breaths, such as 5 breaths. It can be determined.

[0287] In some forms of variable backup rate, the backup rate Rb is changed from SBR to ST The predetermined interval of increase towards the BR may be a function of the adequacy of the current ventilation. In one embodiment suitable for servo-ventilation, a target value Vtgt is present for measuring ventilation. The backup rate is increased when the current measurement of ventilation Vent is less than the target ventilation Vtgt. Approach STBR quickly.

[0288] In one form of variable backup rate known as adaptive variable backup rate, The backup rate determination algorithm 4319 is determined by the respiration rate estimation algorithm 4318. The current estimated spontaneous breathing rate Rs is provided as a function of the backup as a function of time. The adaptive backup rate Rb is determined in the same way as the adaptive backup rate determination. The determination is based on the SBR to ST over a predetermined interval, which may be a function of the adequacy of current ventilation. Increase the backup rate Rb towards the standard respiratory rate (STBR), e.g., 1 5 bpm. A reliable estimate of the spontaneous respiratory rate Rs is obtained by the respiratory rate estimation algorithm. If available from the 4318 standard, STBR is the current estimate multiplied by some constants. The SBR can be set to the spontaneous breathing rate Rs, which is the rate at which the STB It may be set to some fraction of R. In one embodiment, a fraction, e.g., 2 / 3, is used to Occasional long respiratory periods in patients with a relatively low respiratory rate, e.g., 12 breaths / min To accommodate this, the spontaneous breathing rate Rs was set to a low value, such as 0.55, during the initial estimation period. It can be determined.

[0289] In some configurations, the currently estimated spontaneous breathing rate Rs is multiplied to obtain STBR. The constant used is slightly higher than 1 to provide more aggressive ventilation during apnea. For example, it can be 1.1, which is preferred in short apneas. The constants are particularly important in cases where the difficulty of resynchronizing with the patient is a particular issue. If it turns out to be a problem for the patient, it can be slightly lower than 1, e.g. A low backup ratio may result in a longer post-expiratory pause during which resynchronization typically occurs. This makes resynchronization easier.

[0290] 12.3.3.2 Treatment Engine Module In one form of the present technology, the therapy engine module 4320 is Estimates of the pressure Pm, the respiratory flow rate Qr, and spontaneous breathing rate for the patient at the 3000 Receives one or more of Rs as input and provides one or more treatment parameters as output In various embodiments, the treatment engine module 4320 may implement the following algorithm: Includes one or more of: Phase Determination Algorithm 4321, Waveform Determination Algorithm 4322 , Inspiratory Flow Limitation Decision Algorithm 4324, Apnea / Hypopnea Decision Algorithm 4325 , a snore detection algorithm 4326, an airway patency determination algorithm 4327, and a treatment parameter Parameter determination algorithm 4329.

[0291] 12.3.3.2.1 Phase determination In one form of the present technology, the phase determination algorithm 4321 is indicative of respiratory flow Qr. It receives as input a signal and provides as output the phase Φ of the patient's 1000 current respiratory cycle.

[0292] As discrete phase determination In some known implementations, the phase output Φ is a discrete variable. The embodiment may include, for example, detecting the onset of spontaneous inspiration and expiration, respectively, by setting the The phase output Φ is a binary phase output having an inhalation or exhalation value expressed as a half revolution value. The trigger and cycle points are determined by the phase change from expiration to inspiration and from inspiration to expiration, respectively. The RPT device 4000, which "triggers" and "cycles", is a discrete position In one embodiment of binary phase determination, the phase output Φ is When the quantity Qr has a value exceeding a positive threshold, a discrete value of 0 (thereby causing the RPT device 4 000 ("triggering"), and when respiratory flow Qr has a value more negative than the negative threshold , with discrete values ​​of 0.5 revolutions (thereby "cycling" the RPT device 4000) It is decided that

[0293] Another embodiment of the discrete phase determination is to determine a value for one of inspiration, mid-inspiratory pause, and expiration. The 3-phase output Φ is provided.

[0294] In another form known as continuous phase determination, the phase output Φ varies from, for example, 0 to 1 revolution. A continuous value that varies from 0 to 2π radians. The device 4000 triggers and In one embodiment of continuous phase determination, successive values ​​of phase Φ are determined based on the respiratory flow. The phase determined in this embodiment is determined using a fuzzy logic analysis of the quantity Qr. The continuous values ​​of are often called "fuzzy topology". Fuzzy Topology Determination Algorithm43 In one embodiment of 21, the following rules apply to respiratory flow Qr: 1. If respiratory flow is 0 and increasing rapidly, the phase is 0 RPM. 2. If the respiratory flow rate is constant and a large positive number, the phase is 0.25 revolutions. 3. If respiratory flow is 0 and falls rapidly, the phase is 0.5 revolutions. 4. If the respiratory flow rate is constant at a large negative number, the phase is 0.75 revolutions. 5. Respiratory flow is constant at 0, and the 5-second low-pass filtered absolute value of respiratory flow is If it is large, the phase is 0.9 revolutions. 6. If the respiratory flow is positive and the phase is expiration, then the phase is 0 RPM. 7. If respiratory flow is negative and phase is inspiration, then phase is 0.5 revolutions. 8. If the absolute value of the 5-second low-pass filtered respiratory flow is large, the phase is at 20 seconds. It is low pass filtered by a constant and increases at a constant rate identical to the patient's respiratory rate.

[0295] The output of each rule is a function whose phase is the result value of the rule and whose magnitude is the fuzzy degree to which the rule is true. The respiratory flow rate can be expressed as a vector. The degree is determined by the appropriate membership function. The results are then processed using several functions, such as the centroid function. In such combinations, the rules are weighted equally or , may be weighted differently.

[0296] In another embodiment of the continuous phase determination, the inhalation time Ti and the exhalation time Te are determined first. The phase Φ is then calculated based on the amount of inspiration time that has elapsed since the previous trigger instant. Half the fraction of time Ti, or the amount of expiration time that has elapsed since the previous cycle instant plus 0.5 revolutions The period is determined as the 1st year of the 1st lunar month plus half the percentage of Te between the 1st lunar month and the 2nd lunar month (whichever is more recent).

[0297] In some forms of the present technology suitable for pressure-assisted ventilation therapy (described below), the phase-determining algorithm The algorithm 4321 maintains the trigger even when the respiratory flow Qr is significantly smaller, for example during apnea. Thus, the RPT device 4000 is configured to receive spontaneous pulses from the patient 1000. Deliver "back-up breaths" when there is a lack of spontaneous / timed effort. In this configuration, known as the stationary / timed (S / T) mode, phase determination The algorithm 4321 is provided by the backup rate determination algorithm 4319. A backup rate Rb can be used.

[0298] The phase determination algorithm 4321, which uses the "fuzzy phase", follows the "momentum" rule. By including in the fuzzy phase rule, the S / T mode is calculated using the backup rate Rb. The effect of the momentum rule is to impose a continuous phase on other rules. When there is no characteristic of the respiratory flow rate Qr to be carried forward, the backup rate Rb is used to In one embodiment, the ventilation measurement Vent (described below) is carried forward to the inhalation. ) is well below the ventilation target value Vtgt (which will be described later), Momentum rules are weighted more heavily in the combination. (vs.) Rapid increase in pressure support in response to mild to moderate hypoventilation results in increased ventilation The ventilation target can be achieved very close to the target ventilation. other times (when the patient is not in central apnea) without being pushed in unnecessarily. The ventilation is adjusted close to the target to allow the patient to breathe at a rate significantly lower than the normal rate. It is preferable that momentum rules be given a low weighting when ventilation is in progress. A momentum rule has a low weighting when it is close to the target value but above it by some amount. When applied, adequate ventilation requires relatively high pressure backup at a rate well below the backup rate. This can be easily achieved with the aid of a back-up breather. This allows targeted ventilation to be delivered with lower pressure support. This is desirable for a variety of reasons, the main one being reducing mask leakage. The purpose is to

[0299] In summary, in the fuzzy phase determination algorithm 4321 implementing the S / T mode: However, there is a dilemma in choosing the weighting value of the momentum rule, including the backup rate Rb. If the weighting is too high, patients may be "pushed" by the backup rate. If the weighting value is too low, Pressure support can become excessive. Therefore, the S / It is preferable to provide a method for performing T-mode.

[0300] The phase determination algorithm 4321 (discrete or continuous without momentum rule) is The S / T mode is implemented using a backup rate Rb in a manner known as limited backup. A timed backup can be implemented as follows: The algorithm 4321 may, for example, monitor the respiratory flow rate Qr as described above. The system attempts to detect the onset of spontaneous inspiration by detecting the spontaneous effort of inspiration. Starts at a certain period after the previous trigger instant whose duration is equal to the reciprocal of the backup rate Rb If it is not detected within an interval known as the backup timing threshold, the phase decision algorithm Algorithm 4321 sets the phase output Φ to the value of inspiration (which causes the RPT device 40 00). The RPT Device 4000 is triggered and a backup breath is delivered. When the phase begins to change, the phase determination algorithm 4321 may, for example, monitor the respiratory flow Qr. The phase output Φ is then adjusted to detect the start of spontaneous expiration. (thereby cycling the RPT device 4000).

[0301] As in the variable backup rate system described above, the backup rate Rb changes over time. When the SBR is increased to STBR in response to the load, the backup timing threshold is opened for a longer period. That is, the RPT device 4000 uses more backup breathing. gradually become less alert to the absence of spontaneous breathing effort as Such an RPT device 4000 will be able to provide backups when needed. If the patient prefers to breathe at a lower than standard rate while still providing breathing, Patients are less likely to feel "pushed"

[0302] STBR in a variable backup rate system is the same as the adaptive variable backup rate system mentioned above. As in the previous system, the backup breathing rate is adjusted to the patient's estimated spontaneous breathing rate Rs. is delivered at a rate appropriate to the patient's own recent spontaneous breathing efforts.

[0303] 12.3.3.2.2 Waveform determination In one form of the present technology, a therapy control module 4330 controls a pressure generator 4140 The therapeutic effect of the controlled waveform template Ο(Φ) varies as a function of the phase Φ of the patient's respiratory cycle. The treatment pressure Pt is supplied.

[0304] In one form of the present technology, the waveform determination algorithm 4322 performs treatment parameter determination. The phase determination algorithm 4321 used by the algorithm 4329 For the domain of phase values ​​Φ, we define a waveform template Π(Φ) with values ​​in the range [0, 1]. provide.

[0305] In one form suitable for either discrete or continuous valued phases, the waveform template Π (Φ) has a value of 1 for phase values ​​below 0.5 revolutions and a value of 0.5 for phase values ​​above 0.5 revolutions. is a square wave template with a value of 0 for In this case, the waveform template Π(Φ) consists of two smoothly curved parts, i.e., 0.5 Rotate the phase value to a smoothly curved scale from 0 to 1 (e.g., a raised cosine (raised cosine) rise, and for phase values ​​exceeding 0.5 revolutions, It includes a smoothly curved (e.g., exponential) decay from An example of such a "smooth and pleasant" waveform template is a raised cosine There is a smooth decay (as Φ approaches one revolution, the limit of Π becomes exactly 0). "Shark fin" waveform balance with quasi-exponential The rate is.

[0306] In some forms of the present technology, the waveform determination algorithm 4322 may include a RPT device 400 Selects a waveform template Π(Φ) from the waveform template library according to the setting of 0. Each waveform template Π(Φ) in the library is a lookup table of values ​​Π for phase values ​​Φ. In another embodiment, the waveform determination algorithm 4322 may be provided as a Parameterized by one or more parameters (e.g., the time constant of the exponentially curvilinear part) The waveform template Π(Φ) is simply generated using a predetermined function form. The parameters of the function form are determined in advance. or may depend on the current state of the patient 1000.

[0307] The waveform determination algorithm 4322 determines whether the inhalation (Φ=0 rotations) or exhalation (Φ=0.5 rotations) In some embodiments of the present technology suitable for discrete binary phase of is the discrete phase Φ and time t measured since the most recent trigger moment (the expiration-to-inspiration transition). We compute the waveform template Π “on the fly” as a function of both In one such embodiment, the waveform determination algorithm 4322 determines the waveform template. Φ,t is calculated into two parts (inhalation and exhalation) as follows:

number

[0308] 12.3.3.2.3 Inspiratory flow limitation detection In one form of the present technology, a processor is configured to: 4324.

[0309] In one form, the algorithm 4324 receives as input a respiratory flow signal Qr. , providing as an output a metric of the extent to which the inspiratory portion of the breath exhibits inspiratory flow limitation.

[0310] In one form of the present technology, the inspiration portion of each breath is calculated based on the phase Φ estimated at each instant in time. For example, the inhalation portion of a breath is identified by values ​​of respiratory flow where the phase Φ is less than or equal to 0.5. A number of equally spaced points (e.g., 65) representing time points are -Time curves are interpolated by the interpolator. Then the curves described by the points are , scaled by the scaler to have a unity length (duration / period) and unity area The scaled respiration is then scaled to remove the effects of varying respiration rate and depth. The signal is compared in a comparator to a pre-stored template representing normal, unobstructed breathing. At any time during inspiration from this template, the specified threshold (typically 1 scaling Any missed breaths (more than 100 breaths per minute) such as coughs, sighs, swallows, and hiccups are rejected. For unrejected data, the first such scaled point shift is The moving average is calculated by the central controller 4230 over the preceding few inspiration events. This is repeated over similar inhalation events for a second such point, etc. So, for example, 65 scaled data points are returned to the central controller. Generated by the 4230, it is a transition of several preceding inspiration events, e.g., three events. represents a moving average. A moving average of continuously updated values ​​of (e.g., 65) points is , the "scaled flow" designated below as Qs(t). Alternatively, a single inspiration event may be utilized in place of the moving average.

[0311] From the scaled flows, two shape factors for determining partial blockage are calculated: It is possible.

[0312] A shape factor of 1 is the average ratio of the central (e.g., 32) scaled flow points. This is the average ratio of the total (e.g. 65) scaled flow points. This ratio is If it exceeds 1, breathing is considered normal. If the ratio is below 1, breathing is considered obstructed. The ratio of about 1.17 is taken to mean that partially obstructed breathing is is taken as the threshold between the maximum oxygen intake and the maximum oxygen consumption required to maintain adequate oxygen delivery for a typical user. This is equivalent to the degree of obstruction that occurs when

[0313] A form factor of 2 is the unit scale taken across the intermediate (e.g., 32) points. The RMS deviation from the normalized flow is calculated. An RMS deviation of about 0.2 units is considered positive. An RMS deviation of 0 is considered to be a completely flow-limited breath. The closer the MS deviation is to 0, the more the breath is considered to be flow-limited. do.

[0314] Form factors 1 and 2 may be used alternatively or in combination. In another embodiment, the number of sampled points, breaths, and center points may be any of the above. Also, the thresholds may be different than those mentioned above.

[0315] 12.3.3.2.4 Determination of Apnea and Hypopnea In one form of the present technology, a central controller 4230 detects apnea and / or respiratory One or more algorithms 4325 are performed to detect the degradation.

[0316] In one form, the one or more apnea / hypopnea decision algorithms 4325 include: It receives the flow rate Qr as an input and outputs a flag to indicate that an apnea or hypopnea has been detected. Provide as a force.

[0317] In one form, apnea is a period of time that is determined by a function of respiratory flow Qr. It is said to be detected when the flow threshold is exceeded. The function is the peak flow, the average flow over a relatively short period of time, Average flow, or a flow between the average flow and the peak flow for a relatively short period of time, e.g., RMS flow can be determined. The flow threshold can be a relatively long term flow measurement.

[0318] In one form, the hypopnea is a function of respiratory flow Qr for a predetermined period of time. A function is said to be detected when it falls below the second flow threshold for a relatively short period of time. average flow during a period, or the intermediate flow between the average flow and the peak flow for a relatively short period, e.g., RM The second flow threshold may be a relatively long-term flow measurement. The second flow threshold is greater than the flow threshold used to detect apnea.

[0319] 12.3.3.2.5 Snoring Detection In one form of the present technology, a central controller 4230 may include one or more One or more snore detection algorithms 4326 are performed.

[0320] In one form, the snore detection algorithm 4326 receives a respiratory flow signal Qr as an input. The system receives a signal from the input and provides as output a metric of the extent to which snoring is present.

[0321] The snore detection algorithm 4326 determines the strength of the flow signal in the 30-300 Hz range. The snore detection algorithm 4326 may include detecting background noise, e.g. For example, the respiratory flow signal Qr may be swept to reduce the sound of the airflow in the system from the blower 4142. The method may further comprise the step of filtering.

[0322] 12.3.3.2.6 Determination of airway patency In one form of the present technology, a central controller 4230 may be configured to To do this, one or more algorithms 4327 are performed.

[0323] In one form, the airway patency algorithm 4327 receives a respiratory flow signal Qr as an input. and determining the power of signals in the frequency range of about 0.75 Hz and about 3 Hz. The presence of a peak in the frequency range is considered to indicate an open airway. The absence of a peak is considered to indicate a closed airway. It is seen as showing the way.

[0324] In one embodiment, the frequency range in which the peak is sought is the frequency range in which the small forced oscillations of the treatment pressure Pt occur. In one embodiment, the forced vibration has an amplitude of about 1 cmH2O. The frequency is 2 Hz.

[0325] In one form, the airway patency algorithm 4327 receives a respiratory flow signal Qr as an input. The system receives and transmits a cardiogenic signal to determine the presence or absence of a cardiogenic signal. The absence of a cardiogenic signal indicates a closed airway. It is considered to be an indication.

[0326] 12.3.3.2.7 Determination of Treatment Parameters In some forms of the present technology, a central controller 4230 controls the therapy engine module. 4320, using the values ​​returned by one or more other algorithms, one or more treatment parameter determination algorithms for determining one or more treatment parameters; Do 329.

[0327] In one form of the present technology, the treatment parameter is the instantaneous treatment pressure Pt. In one embodiment, the treatment parameter determination algorithm 4329 uses the following formula: The therapeutic pressure Pt is determined using the

number

[0328] The waveform determination algorithm 4322 indexes the waveform template Π(Φ) by the phase Φ. If the parameters are provided as a look-up table of selected values, the treatment parameter determination algorithm 29 is the closest to the current phase value Φ returned by the phase determination algorithm 4321. by searching a new lookup table entry or by straddling the current phase value Φ. Equation (1) is applied by interpolating between the two entries.

[0329] The values ​​of the amplitude A and the base pressure P0 depend on the pressure treatment mode selected in the manner described below. It can be set by the treatment parameter determination algorithm 4329.

[0330] 12.3.3.3 Treatment Control Module A therapy control module 4330 according to one form of the present technology includes a therapy engine module 43 20. Receives treatment parameters as input from a treatment parameter determination algorithm 4329. , controls the pressure generator 4140 to deliver air flow in accordance with the treatment parameters.

[0331] In one form of the present technology, the treatment parameter is a treatment pressure Pt, and the treatment control module The controller 4330 controls the pressure generator 4140 to generate mass at the patient interface 3000. The pump delivers a gas flow whose pressure Pm is equal to the treatment pressure Pt.

[0332] 12.3.3.4 Detection of fault conditions In one form of the present technology, the processor may include one or more methods for detecting a fault condition. A defect condition detected by one or more of the methods may be determined to be at least one of the following: Can also contain one: -Power outage (no power or insufficient power) -Transducer defect detection -Failure to detect the presence of a component -Operating parameters outside recommended ranges (e.g. pressure, flow, temperature, PaO2) -Failure of the test alarm to produce a detectable alarm signal.

[0333] Upon detection of a fault condition, the corresponding algorithm determines the presence of the fault by one or more of the following: Signal the presence of: -Initiation of audio, visual and / or kinetic (e.g. vibration) alarms - Send messages to external devices -Incident logging.

[0334] 12.4 Humidifier In one form of the technology, an air or gas insulator is provided for delivery to the patient relative to the surrounding air. A humidifier 5000 (for example, as shown in FIG. 24) is provided for varying the humidity. Typically, the humidifier 5000 increases the absolute humidity prior to delivery to the patient's airways ( Used to increase the temperature of an air stream (relative to the surrounding air).

[0335] 12.5 Glossary For purposes of disclosure of this technology, in some forms of this technology, one or more of the following definitions may be used: may apply. In other aspects of the technology, alternative definitions may apply.

[0336] 12.5.1 General Air: In some forms of this technology, air means atmospheric air, and in other forms of this technology In this state, air is composed of some or all of the breathable gases, such as oxygen-rich atmospheric air. It means a combination of

[0337] Respiratory Pressure Therapy (RPT) air into the airways at a treatment pressure that is typically positive relative to the atmosphere Transmission of supplies.

[0338] Continuous Positive Airway Pressure (CPAP) Therapy: Therapeutic pressure is applied almost throughout the patient's breathing cycle. In some forms, the pressure at the entrance to the airways is slightly reduced during exhalation. High during breathing and slightly lower during inhalation. In some forms, the pressure varies between different breathing cycles of the patient. For example, it increases when it detects signs of partial upper airway obstruction, and If there are no signs of

[0339] Patient: A human being, whether or not suffering from a respiratory disease.

[0340] Automated Positive Airway Pressure (APAP) Therapy: Minimal and maximal, depending on the presence or absence of SDB events CPAP therapy, where the treatment pressure can be automatically adjusted between inhalations, for example, from inhalation to exhalation.

[0341] 12.5.2 Respiratory cycle patterns Apnea: By some definitions, an apnea is a period of respiratory flow, e.g., 10 seconds, or Obstructive apnea occurs when the patient is unable to breathe despite their efforts. It is said to occur when air cannot flow due to a blockage in the airway. When apnea is detected by decreased respiratory effort despite the presence of Mixed apnea occurs when a reduced or absent respiratory effort causes airway obstruction. It is said to have occurred when it occurs simultaneously with obstruction of the ventricle.

[0342] Breathing rate or respiratory rate (Rs): the rate at which a patient is spontaneously breathing, usually in breaths per minute It is measured in.

[0343] Duty cycle: The ratio of inhalation time Ti to the total respiratory duration Ttot.

[0344] Effort (Breathing): Respiratory effort is the effort made by the patient's spontaneous breathing while trying to breathe. case.

[0345] Expiratory portion of the breathing cycle: the period from the start of the expiratory flow to the start of the inspiratory flow.

[0346] Flow limitation: Flow limitation is a condition in which the patient's efforts do not result in a corresponding increase in flow. This is interpreted as a situation during inspiration. When flow restriction occurs during the inspiratory portion of the respiratory cycle, When flow limitation occurs during the expiratory portion of the respiratory cycle, the expiratory flow is is described as a restriction.

[0347] Hypopnea: According to some definitions, hypopnea is preferably a decrease in flow, In one form, hypopnea is not an interruption of the flow. When a decrease in respiratory effort causes hypopnea, central apnea occurs. In one form in adults, any of the following can be considered hypopnea: (i) The patient's breathing slows by 30% for at least 10 seconds with an associated desaturation of 4% Is; or (ii) The patient's breathing is reduced (50% or less) for at least 10 seconds and % have associated desaturation or arousal.

[0348] Inspiratory portion of the breathing cycle: The period from the start of the inspiratory flow to the start of the expiratory flow. This is called the intake portion of the intake cycle.

[0349] Patency (airway): The degree to which the airway is open or the degree to which the airway is open. A patent airway Airway patency is quantified with a value of one (1) as open and zero (0) as closed. do.

[0350] Positive end-expiratory pressure (PEEP): Pressure above atmospheric pressure in the lungs that exists at the end of expiration.

[0351] Peak flow (Qpeak): The maximum value of flow during the inspiratory portion of the respiratory airflow waveform.

[0352] Respiratory flow / airflow, patient flow, respiratory airflow (Qr): these synonyms are RPT This may be understood to refer to the device's estimate of respiratory flow, as opposed to the actual flow experienced by the patient. Respiratory flow is the "true respiratory flow" or "true respiratory airflow" usually measured in liters per minute represent.

[0353] Tidal Volume (Vt): The amount of air inhaled or exhaled during normal breathing without extra effort. The amount of air.

[0354] (Inspiratory Time) (Ti): The duration of the inspiratory portion of the respiratory flow waveform.

[0355] (Expiratory) Time (Te): The duration of the respiratory portion of the respiratory flow waveform.

[0356] (Total) time or breathing period (Ttot) - the inspiratory portion of one respiratory flow waveform The total period between the start of the inspiratory portion of the respiratory flow waveform and the start of the subsequent inspiratory portion of the respiratory flow waveform.

[0357] Upper airway obstruction (UAO): includes both partial and total upper airway obstruction. This is flow-limited. Increasing the pressure difference across the upper airway increases flow slightly or may even decline. (Sterling Register Action)

[0358] Vent: A measure of the total amount of gas exchanged through a patient's respiratory system. It includes one or both of the inspiratory and expiratory flows per unit time. Expressed as volume per minute, This amount is often called "minute ventilation". Minute ventilation is sometimes simply given as volume. This is understood as the capacity per minute.

[0359] 12.5.3 RPT Device Parameters Flow: The instantaneous volume (or mass) of air delivered per unit time. Flow and ventilation are simply Although it has the same dimensions as the volume or mass per unit time, the flow rate is measured over a much shorter period of time. Flow is nominally positive during the inspiratory portion of the patient's respiratory cycle and is therefore The results are negative for the expiratory portion of the patient's respiratory cycle. In some cases, the reference to flow rate is , a reference to a scalar quantity, i.e. a quantity that has only a magnitude. A reference to is a reference to a vector quantity, i.e. a quantity that has both magnitude and direction. Yes. The flow rate is given by the symbol Q. 'Flow rate' is sometimes shortened to simply 'flow'. Total flow The ventilation flow rate Qt is the air flow exiting the RPT device. The ventilation flow rate Qv is the flow rate that flushes the exhaled gas. The leakage flow rate Ql is the air flow that exits the vent so that the patient interface can be The respiratory flow, Qr, is the rate at which the patient's respiratory system The airflow that is admitted into the system.

[0360] Leak: The word leak is considered to be an unintended air flow. In one example, a leak is a mass This can occur as a result of a faulty seal between the mask and the patient's face. Thus, leakage can occur in the swivel elbow to the periphery.

[0361] Pressure: force per unit area. Pressure is measured in cmH2O, gf / cm 2 , hectopascals It can be measured in a range of units, including: 1 cmH2O is 1 g-f / cm 2 is the same as, approximately 0.98 hectopascals. In this specification, unless otherwise stated, pressure is expressed in cm It is given in units of H2O. The pressure at the patient interface (mask pressure) is given by the symbol Pm. while representing the target value to be achieved by the mask pressure Pm at the current instant in time. Pressure is given by the symbol Pt.

[0362] 12.5.4 Ventilation Terminology Adaptive servo ventilator (ASV): A device that has variable ventilation rather than a single fixed target ventilation. Variable target ventilation is based on several aspects of the patient, e.g. the patient's breathing characteristics. We can learn from some of these characteristics.

[0363] Backup rate: The rate at which the ventilator would deliver to the patient if not triggered by spontaneous breathing efforts. The ventilator parameter that determines the rate at which breathing can be performed (typically in breaths per minute).

[0364] Cycled: The end of the inspiratory phase of the ventilator. The moment the ventilator delivers a breath to a spontaneously breathing patient. When this occurs, at the end of the inspiratory portion of the breathing cycle, the ventilator stops delivering a breath. It's been a good cycle.

[0365] Expiratory positive pressure (EPAP) is the base pressure that is applied during breathing and is fluctuating between breaths. The device generates a desired mask pressure that it attempts to achieve at a given time.

[0366] End-Expiratory Pressure (EEP): The pressure at which the ventilator reaches the end of the expiratory portion of the breath at the desired mask pressure. If the pressure waveform template II(Φ) is zero at the end of expiration, That is, when Φ=1, if II(Φ)=0, then EEP is equal to EPAP.

[0367] IPAP: The desired mask pressure that the ventilator attempts to achieve during the inspiratory portion of the breath.

[0368] Pressure Support: Pressure during the inhalation of the ventilator that exceeds the pressure during the exhalation of the ventilator It is a number that indicates the increase in the pressure of the inhaler, and is generally expressed as the difference between the maximum and base pressure during inspiration (e.g., PS = IPA In some contexts, pressure support refers to the pressure that the ventilator actually achieves. It refers to the difference that one intends to achieve rather than that which one has made.

[0369] Servo ventilator: A ventilator that measures the ventilation of a patient and has a target ventilation. adjusts the level of pressure support to direct the patient's ventilation towards the target ventilation. Force support minus minimum pressure support.

[0370] Spontaneous / Timed (S / T): a ventilator that attempts to detect the onset of breathing in spontaneously breathing patients or some other mode of the device, but if the device does not detect breathing within the required time. If the patient is not responsive, the device will automatically begin delivering breaths.

[0371] Swing: Equivalent term for pressure support.

[0372] Triggered: When the ventilator delivers a breath of air to a spontaneously breathing patient, patient effort It is said to be triggered at the beginning of the inhalation portion of the breathing cycle.

[0373] Typical Recent Ventilation: Typical recent ventilation Vtyp is the ventilation rate over a given time scale. The most recent ventilation measurements tend to cluster around them, i.e. A measure of the central tendency of historical ventilation measurements is a good representative of recent ventilation. do.

[0374] Ventilator: A device that provides pressure support to the patient to facilitate breathing. A mechanical device that does some or all of the work.

[0375] 12.6 Other Notes A portion of the disclosure of this patent document contains material which is subject to copyright protection. The copyright owner is , so long as it appears in the patent file or records of the Patent and Trademark Office, There is no objection to the facsimile reproduction of a patent disclosure. Copyright of whatever is used.

[0376] Unless the context clearly indicates otherwise and unless a range of values ​​is given, Intervening values ​​are between the upper and lower limits of the range, up to 1 / 10 of the unit of the lower limit, and Any other stated or intervening values ​​in the stated ranges are encompassed by the art. The upper and lower limits of these intervening ranges are not to be construed as limiting any specifically excluded limits in the stated range. Subject to limitations, the intervening ranges may be independently included and are encompassed within the scope of the present technology. If the range specified includes one or both of the limits, either or both of those included limits The range excluding the above is also included in the technology.

[0377] Furthermore, one or more values ​​may be disclosed that are incorporated herein by reference. Unless otherwise specified, such values ​​may be approximate and may vary based on practical engineering practice. To the extent that an implementation allows or requires, such values ​​may be used to any convenient significant digit. I want you to understand that it is possible.

[0378] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this technology pertains. Any methods or materials similar or equivalent to those described herein may also be used in the practice or manufacture of the present technology. can be used in testing, but a limited number of exemplary methods and materials are described herein.

[0379] If a particular material is identified as preferably used to construct the component Obviously alternative materials with similar properties can be used as substitutes. Unless otherwise specified to the contrary, any and all components described herein may be manufactured It is understood that they may be manufactured together or separately.

[0380] As used in this specification and the appended claims, the singular forms "a," "an," and "the" It should be noted that the plural is included unless the context clearly dictates otherwise. be.

[0381] All publications mentioned herein are incorporated by reference in their entirety for purposes of illustration only and are not intended to be limiting unless otherwise specified. The publications discussed herein are incorporated by reference to disclose and describe the The disclosures contained herein are provided for purposes of disclosure prior to the filing date of this application. The disclosure herein cannot be construed as giving the applicant a right to antedate such publication by reason of prior invention. The publication dates listed may be different from the actual publication dates and should be confirmed independently. .

[0382] Moreover, in interpreting this disclosure, all terms should be used in the broadest reasonable manner consistent with the context. In particular, the words "comprises" and "comprises" should be interpreted in a "sing" refers to an element, component or step in a non-proprietary manner The element, component, or step referenced is interpreted as existing or Other elements, components not used or explicitly referenced herein are subject to change without notice. It should be combined with steps 1 or 2.

[0383] The subject headings used in the Detailed Description are included solely for the reader's ease of reference and are not intended to be limiting. shall not be used to limit the subject matter found throughout the disclosure or claims. The subject headings should not be used to construe the scope of the claims or the claim limitations.

[0384] Although the technology described herein has been described with reference to specific embodiments, these embodiments It should be understood that the above is merely illustrative of the principles and applications of the technology. The names and symbols may imply specific details that are not required to practice the technology. For example, the terms "first" and "second" are used unless otherwise specified, but are optional. It is not intended to indicate any arbitrary order, but can be used to distinguish between distinct elements. Furthermore, although the process steps in a methodology may be described or illustrated sequentially, Such an order is not required. Those skilled in the art will appreciate that such an order may be modified and It will be appreciated that the features may be implemented simultaneously or synchronously. cormorant.

[0385] Thus, numerous modifications may be made to the exemplary embodiments without departing from the spirit and scope of the technology. It will be appreciated that other arrangements may be devised without departing from the scope of the invention.

[0386] It will be further understood that any reference herein to subject matter known in the art is not an admission that such subject matter is generally known by those of ordinary skill in the art, unless a contrary indication is indicated. In order to maintain the disclosure matters at the time of filing of the present application, the contents of claims 1 to 92 at the time of filing of the present application are added as follows. (Claim 1) 1. A system for identifying persons for monitoring physiological parameters of one or more persons, comprising: one or more sensors for monitoring physiological parameters of the one or more persons; one or more processors configured to process signals from the one or more sensors to identify the person, the processing including evaluation of one or more features including respiratory features, cardiac features, or movement features; Including, the system. (Claim 2) The system of claim 1 , wherein the evaluation of the features includes an analysis of at least one of a detected breathing rate, a detected breathing depth, a detected degree of movement, and a detected heart rate. (Claim 3) The system of claim 1 or 2, wherein the one or more sensors include a radio frequency non-contact sensor. (Claim 4) The system of claim 2 , wherein the monitoring includes detecting a physiological characteristic of the person while sleeping. (Claim 5) The system of claim 1 , wherein the processing comprises detecting a sleep stage of the person. (Claim 6) The system of claim 1 , wherein the processing comprises detecting deep sleep of the person. (Claim 7) The system of claim 1 , wherein the processing comprises detection of REM sleep in a person. (Claim 8) The system of claim 1 , wherein the monitoring comprises detecting physiological characteristics of the person during their waking hours. (Claim 9) The one or more sensors include A radio frequency non-contact sensor; A biomotion sensor; A wearable sensor; The system according to any one of claims 1 to 8, comprising any one or more of the following: (Claim 10) The system of claim 1 , wherein the evaluation of the one or more processors comprises detection and analysis of any one or more features, including biometric and behavioral features. (Claim 11) 11. The system of claim 1, wherein the one or more sensors are configured to minimize interference between at least two of the one or more sensors. (Claim 12) The system of any one of claims 1 to 11, wherein the one or more sensors vary control parameters for range gating pulse timing, emitted power level of the pulses, detection frequency of the pulses and / or positioning of the steerable antenna to minimize interference. (Claim 13) 13. The system of claim 1, wherein the one or more sensors communicate via a wired or wireless link. (Claim 14) 14. The system of claim 1, wherein characteristic biometric parameters are applied to dynamically adjust performance in at least one of the one or more sensors to optimize physiological recognition of stand-alone human sources and reject other sources. (Claim 15) 15. The system of claim 1, further comprising a control processor in communication with the one or more sensors, the control processor in communication with the one or more sensors and adapted to adjust detection control parameters of the one or more sensors based on identification of the human, other person, or animal by the one or more sensors. (Claim 16) The system of claim 15 , wherein detection control parameters of the one or more sensors include one or more of range gating, center frequency, and power level. (Claim 17) 18. The system of claim 1, further comprising a control processor in communication with the one or more sensors, the control processor configured to activate a further sensor system based on identification of the human, other person, or animal. (Claim 18) The system of claim 17 , wherein the further sensor system comprises a camera. (Claim 19) 19. The system of claim 1, wherein the processor logs the detected biometric characteristic based on identifying the person as a previously monitored person. (Claim 20) 20. The system of claim 1, wherein the processor ceases logging the detected biometric characteristic upon identifying the person as a previously unmonitored person. (Claim 21) 21. The system of claim 1, wherein the processor is configured to initialize biometric characteristic detection for a particular person. (Claim 22) 22. The system of claim 21, wherein the processor compares the newly detected biometric characteristic to an initialized biometric characteristic to identify the particular person. (Claim 23) 22. The system of claim 21, wherein the evaluation further comprises comparing the newly detected biometric characteristic to an initialized biometric characteristic. (Claim 24) The evaluation includes a classification of features determined from the signal, the features being: Spectral peak ratio, Set optimizer flag vector, Peak-to-trough ratio, Filtered respiratory rate, A measure of respiratory variability, the in-band power of the sensor signal, Sensor signal range, final respiratory rate, The ratio of maximum to minimum amplitude of the respiratory cycle, High bandwidth power for the sensor signal, average respiratory rate, Periodic foot movement activity detection, Detection of turnover, and Post-processed movement, The system according to any one of claims 1 to 23, comprising one or more of the following: (Claim 25) 24. The system of claim 1, wherein the evaluation comprises classification of features determined from the signals, the features comprising one or more of: cardiac parameters, galvanic skin response parameters, exercise intensity parameters, respiratory parameters, blood pressure parameters, cough parameters, snoring parameters, and sleep parameters. (Claim 26) 26. The system of claim 24 or 25, wherein the evaluation comprises a comparison of the determined characteristics with past characteristics. (Claim 27) 27. The system of claim 24, wherein the evaluation further comprises calculating an average and / or a standard deviation value over a period of time from the determined characteristics. (Claim 28) 28. The system of claim 1, wherein the one or more processors of the system evaluate received data detected by one or more sensors from another monitoring system. (Claim 29) 30. The system of claim 28, wherein evaluating the received data includes determining comparable sensing between a plurality of health monitoring devices. (Claim 30) 30. The system of any one of claims 1 to 29, wherein the monitoring device comprises a set classifier and a subject classifier, each classifier configured to evaluate features from the signals to identify a person. (Claim 31) 31. The system of claim 30, wherein the subject classifier is further configured to evaluate historical characteristics to identify a person. (Claim 32) 32. The system of any one of claims 1 to 31, wherein the health monitoring device further comprises a camera sensitive to infrared light and an infrared emitter. (Claim 33) 33. The system of claim 32, wherein the one or more processors of the health monitoring device are configured to detect an event from the signal and associate the event with a particular portion of video captured by the camera that contains the event. (Claim 34) 34. The system of claim 32 or 33, wherein the health monitoring device further comprises a battery and a coil for wireless charging of the battery. (Claim 35) 35. The system of any one of claims 1 to 34, wherein the one or more processors are configured to control the one or more sensors to alter sensor detection power, sensor frequency, sensor distance gating, or other control parameters for sensing based on detection of a biometric characteristic indicative of an animal. (Claim 36) 35. The system of any one of claims 1 to 34, wherein the one or more processors are configured to access setting parameters for any one or more of light, sound, and / or environmental devices based on an identity of a particular person associated with the setting parameters. (Claim 37) 37. The system of claim 36, wherein the one or more processors are configured to trigger settings in any one or more of the lights, sounds, and / or environmental devices according to setting parameters associated with the detected biometric of the identified particular person. (Claim 38) 38. The system of any one of claims 1 to 37, further comprising a central controller in communication with the collection of sensors, the central controller configured to provide confirmation of the location of an identified person within the structure. (Claim 39) 39. The system of claim 1, wherein the one or more sensors include a night light. (Claim 40) 40. The system of claim 1, wherein the one or more sensors include a pass-through outlet. (Claim 41) 41. The system of claim 1, wherein the one or more sensors include an AC plug for supplying power to the sensor, and an AC power pass-through outlet. (Claim 42) 42. The system of claim 1, wherein the one or more sensors include a USB plug for providing power to the sensor, and a USB pass-through adapter. (Claim 43) 43. The system of claim 1, wherein the one or more sensors include a network interface for wireless or wired network communication. (Claim 44) 44. The system of any one of claims 1 to 43, wherein the one or more processors are configured to configure operation of a respiratory treatment device based on the identified person. (Claim 45) 45. The system of claim 44, wherein the configuration action allows or denies treatment by the respiratory treatment device. (Claim 46) 45. The system of claim 44, wherein the configuration action changes therapy on the respiratory treatment device. (Claim 47) 47. The system of any one of claims 44 to 46, wherein the one or more processors are configured to retrain for identification of the person when biometric characteristics evaluated in the identification are processed by the respiratory treatment device. (Claim 48) 48. The system of any one of claims 1 to 47, wherein the one or more processors are configured to adjust operation of the one or more sensors upon determination of sensor signal quality. (Claim 49) 49. The system of any one of claims 1 to 48, wherein the one or more processors are configured to rely on different biometric characteristics to identify the person depending on a quality assessment of the detected biometric characteristics. (Claim 50) 50. The system of claim 1, wherein the one or more processors are configured to operate an enrollment process to initialize a baseline of biometric characteristics for identifying a person, the enrollment process including a guided breathing session or a spontaneous breathing session. (Claim 51) 51. The system of claim 1, wherein the one or more processors are configured to reject a biometric characteristic associated with identifying the person if frequency interference is detected. (Claim 52) 52. The system of any one of claims 1 to 51, wherein the one or more processors are configured to set an alarm action based on the identified person. (Claim 53) 53. The system of any one of claims 1 to 52, wherein the one or more processors are configured to identify a primary user from one or more other users. (Claim 54) 54. The system of claim 1, wherein the system is configured to track one or more user parameters over time and construct classification features based on at least one of the one or more respiratory, cardiac or motion features. (Claim 55) 55. The system of claim 54, wherein at least one of the one or more respiratory, cardiac, or motion features includes one or more of the following respiratory parameters: range, respiratory variability, shape, and inspiration to expiration ratio. (Claim 56) 56. The system of any one of claims 1 to 55, wherein the one or more processors are configured to classify a user's identity from characteristics determined in a classification process. (Claim 57) 57. The system of claim 56, wherein the classification process includes any one or more of a neural network, a hidden layer Markov model, a logistic regression process, a linear kernel support vector machine, and a radial kernel support vector machine. (Claim 58) 58. The system of claim 56 or 57, wherein the classification process includes using main component analysis on the features prior to classification. (Claim 59) 59. The system of any one of claims 56 to 58, wherein the classification process includes real-time and offline characteristics. (Claim 60) 60. The system of any one of claims 56 to 59, wherein the classification process includes multiple classifiers and a delayed integration of their outputs to generate output posterior probabilities. (Claim 61) 61. The system of any one of claims 1 to 60, comprising a plurality of sensors arranged to detect the same or different persons, the system being configured to automatically adjust parameters of the plurality of sensors, the parameters including at least one of range, power, frequency, detection direction, and radiation pattern. (Claim 62) 1. A method for identifying a person for monitoring physiological parameters of one or more persons, comprising: monitoring a physiological parameter of the one or more persons with one or more sensing sensors; processing, by one or more processors, signals from the one or more sensors to identify the person, the processing including evaluating features including one or more respiratory features, cardiac features, or movement features; The method includes: (Claim 63) 63. The method of claim 62, wherein evaluating the characteristic includes analysis of at least one of a detected respiration rate, a detected respiration depth, a detected degree of movement, and a detected heart rate. (Claim 64) 64. The method of claim 62 or 63, wherein the one or more sensors include a radio frequency non-contact sensor. (Claim 65) 65. The method of any one of claims 62 to 64, wherein said evaluating comprises detection and analysis of any one or more characteristics, including biometric and behavioral characteristics. (Claim 66) 66. A method according to any one of claims 62 to 65, further comprising activating a further sensor system based on identification of the human, other person, or animal by the one or more sensors, using a control processor in communication with the one or more sensors. (Claim 67) 67. The method of any one of claims 62 to 66, further comprising using a processor to log the detected biometric characteristic based on identifying the person as a previously monitored person. (Claim 68) 68. The method of any one of claims 62 to 67, further comprising using a processor to cease logging the detected biometric characteristic based on identifying the person as a previously unmonitored person. (Claim 69) 69. The method of any one of claims 62 to 68, further comprising the step of initializing biometric characteristic detection for a particular person. (Claim 70) 70. The method of claim 69, further comprising the step of: in a processor, comparing the newly detected biometric characteristic with initialized biometric characteristics to identify the particular person. (Claim 71) 70. The method of claim 69, further comprising the step of: in the processor, comparing the newly detected biometric characteristic with the initialized biometric characteristic. (Claim 72) The evaluating step includes classifying features determined from the signal, the features comprising: Spectral peak ratio, Set optimizer flag vector, Peak-to-trough ratio, Filtered respiratory rate, A measure of respiratory variability, the in-band power of the sensor signal, Sensor signal range, final respiratory rate, The ratio of maximum to minimum amplitude of the respiratory cycle, High bandwidth power for the sensor signal, average respiratory rate, Periodic foot movement activity detection, Detection of turnover, and Post-processed movement, The method according to any one of claims 62 to 71, comprising one or more of the following: (Claim 73) 73. The method of any one of claims 62 to 72, wherein the evaluating comprises classifying features determined from the signals, the features comprising one or more of: a cardiac parameter, a galvanic skin response parameter, an exercise intensity parameter, a respiratory parameter, a blood pressure parameter, a cough parameter, a snoring parameter, and a sleep parameter. (Claim 74) 74. The method of claim 72 or 73, wherein said evaluating comprises comparing said determined characteristic to a previous characteristic. (Claim 75) 75. The method of any one of claims 72 to 74, wherein the evaluating further comprises calculating in a processor an average and / or standard deviation value over a period of time from the determined characteristics. (Claim 76) 76. The method of any one of claims 62 to 75, further comprising using the one or more processors to control the one or more sensors to modify sensor detection power, sensor frequency, sensor distance gating, or other control parameters for sensing based on detection of a biometric characteristic indicative of an animal. (Claim 77) 76. The method of any one of claims 62 to 75, further comprising using the one or more processors to access setting parameters for any one or more of light, sound, and / or environmental devices based on an identity of a particular person associated with the setting parameters. (Claim 78) 78. The method of claim 77, further comprising using the one or more processors to trigger settings in any one or more of light, sound, and / or environmental devices according to setting parameters associated with the detected biometric of the identified particular person. (Claim 79) 79. The method of any one of claims 62-78, further comprising using the one or more processors to configure operation of a respiratory treatment device based on the identified person. (Claim 80) 80. The method of claim 79, wherein the step of configuring operation allows or denies treatment by the respiratory treatment device. (Claim 81) 80. The method of claim 79, wherein the step of configuring the operation modifies therapy on the respiratory treatment device. (Claim 82) 82. The method of any one of claims 62-81, further comprising operating, using the one or more processors, an enrollment process to initialize a baseline of biometric characteristics for identifying the person, the enrollment process including a guided breathing session or a spontaneous breathing session. (Claim 83) 83. The method of any one of claims 62 to 82, further comprising using the one or more processors to set an alarm action based on the identified person. (Claim 84) 84. A method according to any one of claims 62 to 83, configured in the one or more processors to identify a primary user from one or more other users. (Claim 85) 85. The method of any one of claims 62 to 84, further comprising tracking one or more user parameters over time to construct classification features based on at least one of one or more respiratory, cardiac or motion features. (Claim 86) 86. The method of claim 85, wherein at least one of the one or more respiratory, cardiac, or motion features includes one or more of the following respiratory parameters: range, respiratory variability, shape, and inspiration to expiration ratio. (Claim 87) 87. The method of any one of claims 62 to 86, further comprising, in the one or more processors, classifying the identity of the user from the characteristics determined in the classification process. (Claim 88) 88. The method of claim 87, wherein the classification process includes one or more of a neural network, a hidden layer Markov model, a logistic regression process, a linear kernel support vector machine, and a radial kernel support vector machine. (Claim 89) 89. The method of claim 87 or 88, wherein the classification process includes using main component analysis on the features prior to classification. (Claim 90) 90. The method of claim 88 or 89, wherein the classification process includes real-time and offline characteristics. (Claim 91) 91. The method of any one of claims 88 to 90, wherein the classification process includes multiple classifiers and a delayed integration of their outputs to generate output posterior probabilities. (Claim 92) 92. The method of any one of claims 62 to 91, further comprising using multiple sensors positioned to detect the same or different persons and automatically adjusting parameters of the multiple sensors, the parameters including at least one of range, power, frequency, detection direction, and radiation pattern. [Explanation of symbols]

[0387] 100 Detection device 1000 patients 1100 Bed Partner 1922 Feature Extraction Unit 1928 Classification process 1940 Classifier Combinator Process 3000 Patient Interface 3100 Structure 3200 Plenum Chamber 3300 Structure 3400 Vent 3600 Connection Port 3700 Forehead support 4000 RPT equipment 4010 Outer Housing 4012 Upper 4014 Lower 4015 Panel 4016 Chassis 4018 Handle 4020 Air Block 4100 Air Components 4110 Air Filter 4112 Inlet Air Filter 4114 Outlet Air Filter 4120 Muffler 4122 Inlet muffler 4124 Exit muffler 4140 Pressure Generator 4142 Blower 4144 Motor 4160 Back valve 4170 Air Transmission Circuit 4180 Supplemental Oxygen 4200 Electrical Components 4202 PCBA 4210 Power supply 4220 input device 4230 Central Controller 4232 Watch 4240 Therapy Device Controller 4250 protection circuit 4260 Memory 4270 Converter 4272 Pressure Sensor 4274 Flow Sensor 4276 Motor Speed ​​Sensor 4280 Data Communication Interface 4282 Remote external communication network 4284 local external communications network 4286 remote external device 4288 local external device 4290 output device 4292 Display Driver 4294 Display 4300 Algorithm 4310 Processing Module 4311 Ventilation Decision Algorithm 4312 Pressure Compensation Algorithm 4313 Target ventilation determination algorithm 4313 Target ventilation determination 4314 Vent flow rate estimation algorithm 4316 Leakage flow rate estimation algorithm 4317 Respiratory flow estimation algorithm 4318 Respiration rate estimation algorithm 4319 Backup Rate Decision Algorithm 4320 Treatment Engine Module 4321 Phase Determination Algorithm 4322 Waveform Determination Algorithm 4324 Inspiratory Flow Limitation Decision Algorithm 4325 Apnea / hypopnea decision algorithm 4326 Snoring Detection Algorithm 4327 Airway Patency Determination Algorithm 4329 Treatment Parameter Decision Algorithm 4330 Treatment Control Module 4340 method 4500 methods 5000 humidifier 2020a Configuration Classification Process 2020b Configuration Classification Process 2028a Specific Classification Process 2028b Specific Classification Process 2040a Sorting and Combining Machine 2040b Sorting and Combining Machine

Claims

1. 1. A physiological parameter monitoring system for identifying a person and monitoring a physiological parameter of the identified person, comprising: one or more sensors, including a bio-motion sensor, for monitoring one or more physiological parameters of the person; one or more processors configured to process signals from the one or more sensors to identify a person, the processed signals including bio-motion signals from the bio-motion sensors, the processing including evaluation of features including one or more respiratory, cardiac or movement features, whereby the one or more processors determine that a current user whose biometric parameters are detected is the same as a previous user whose biometric parameters were previously detected, the one or more processors configured to process the bio-motion signals to detect one or more sleep stages, and associate the detected one or more sleep stages with a person identified by evaluation of the features; A physiological parameter monitoring system comprising:

2. The physiological parameter monitoring system of claim 1 , wherein the one or more processors are configured to process signals from the one or more sensors to identify a person by evaluating one or more biometric signatures.

3. The physiological parameter monitoring system of claim 2 , wherein the one or more biometric signatures include a sleep stage specific fingerprint.

4. 3. The physiological parameter monitoring system of claim 2, wherein the one or more processors comprise a user identification process configured to evaluate characteristics determined from the user's sleep stages using the one or more biometric signatures.

5. 5. The physiological parameter monitoring system of claim 1, wherein the one or more processors are configured to adjust one or more sensing control parameters of the one or more sensors based on processing including evaluation of the characteristics if they detect that a current user is not a previous user.

6. 6. The physiological parameter monitoring system of claim 5, wherein the one or more sensing control parameters include one or more of a detection range of the one or more sensors, a power level of the one or more sensors, a detection frequency of the one or more sensors, a detection direction of the one or more sensors, and a radiation pattern of the one or more sensors.

7. 7. The physiological parameter monitoring system of claim 1, wherein: (a) the one or more sensors are configured as a bedside unit; or (b) the bio-motion sensor includes a radio frequency non-contact sensor and the physiological parameter is detected by a radio frequency signal.

8. A physiological parameter monitoring system according to any one of claims 1 to 7, configured to determine, by means of a microphone, one or more sounds made by a person.

9. The physiological parameter monitoring system of claim 8 , configured to use the microphone to monitor and classify a user's sound patterns separated from background noise.

10. A physiological parameter monitoring system according to any one of claims 1 to 9, wherein the one or more processors are configured to refrain from recording a detected biometric characteristic based on an identification of the person as a person who has not been previously monitored.

11. 11. The physiological parameter monitoring system of claim 1, wherein: (a) the evaluation of the features includes a classification of features determined from signals from the one or more sensors, the classified features including one or more of a galvanic skin response parameter, an exercise intensity parameter, a blood pressure parameter, a cough parameter, a snoring parameter, and a sleep parameter; or (b) the one or more processors are further configured to operate an enrollment process to initialize a baseline of biometric features for identifying a person based on signals from the one or more sensors.

12. 1. A method of monitoring physiological parameters of an identified person by one or more processors of a physiological parameter monitoring system adapted to identify the person, comprising: receiving one or more monitored physiological parameters from one or more sensors on the person, including a bio-motion sensor; processing signals from the one or more sensors to identify a person, the processed signals including bio-motion signals from the bio-motion sensors, the processing including evaluation by the one or more processors of features including one or more respiratory features, cardiac features or motion features, whereby the one or more processors determine that a current user whose biometric parameters are detected is the same as a previous user whose biometric parameters were previously detected, the one or more processors processing the bio-motion signals to detect one or more sleep stages, and associating the detected one or more sleep stages with a person identified by evaluation of the features; The method comprising:

13. The method of claim 12 , wherein the one or more processors process signals from the one or more sensors to identify a person by evaluating one or more biometric signatures.

14. The method of claim 13 , wherein the one or more biometric signatures include a sleep stage specific fingerprint.

15. The method of claim 13 , wherein the one or more processors operate a user identification process using the one or more biometric signatures to evaluate characteristics determined from the user's sleep stages.

16. The method of any one of claims 12 to 15, wherein the one or more processors adjust one or more sensing control parameters of the one or more sensors based on processing including evaluating the characteristics if the one or more processors detect that the current user is not the previous user.

17. 17. The method of claim 16, wherein the one or more sensing control parameters include one or more of a detection range of the one or more sensors, a power level of the one or more sensors, a detection frequency of the one or more sensors, a detection direction of the one or more sensors, and a radiation pattern of the one or more sensors.

18. 18. The method of any one of claims 12 to 17, wherein (a) the one or more sensors are configured as a bedside unit, or (b) the bio-motion sensor includes a radio frequency non-contact sensor and the physiological parameter is detected with a radio frequency signal.

19. The method according to any one of claims 12 to 18, wherein the monitoring system uses a microphone to determine one or more sounds made by a person.

20. 20. The method of claim 19, wherein the monitoring system uses the microphone to monitor and classify a user's sound patterns separated from background noise.

21. The method of any one of claims 12 to 20, wherein the one or more processors refrain from recording the detected biometric characteristic based on identifying the person as a previously unmonitored person.

22. 22. The method of claim 12, further comprising: (a) evaluating the features includes a classification of features determined from signals from the one or more sensors, the features including one or more of a galvanic skin response parameter, an exercise intensity parameter, a blood pressure parameter, a cough parameter, a snoring parameter, and a sleep parameter; or (b) the one or more processors operating an enrollment process to initialize a baseline of biometric characteristics for identifying a person.

23. A processor-readable storage medium having processor-executable instructions stored thereon that, when executed by one or more processors, cause the one or more processors to identify persons for monitoring physiological parameters of the one or more persons, The processor executable instructions include: instructions for accessing one or more monitored physiological parameters of the person from one or more sensors, including a bio-motion sensor; instructions for processing signals from one or more sensors to identify a person, the processed signals including bio-motion signals from the bio-motion sensors, the processing including evaluating, by the one or more processors, features including one or more respiratory, cardiac, or movement features, whereby the one or more processors determine that a current user whose biometric parameters are detected is the same as a previous user whose biometric parameters were previously detected, the one or more processors processing the bio-motion signals to detect one or more sleep stages, and associating the detected one or more sleep stages with a person identified by evaluation of the features; A processor-readable storage medium comprising:

24. A processor-readable storage medium storing processor-executable instructions which, when executed by one or more processors, cause the one or more processors to perform a method according to any one of claims 12 to 22.

Citation Information

Patent Citations

  • System and method for snoring detection and confirmation

    JP2009539499A

  • Body information measuring device

    JP2014210137A

  • Monitoring, predicting and treating clinical episodes

    US20140350351A1

  • System and method for determining sleep stage

    WO2014047310A1

  • Systems and methods for high resolution distance sensing and applications

    WO2014055755A1