Method and apparatus for monitoring chronic disease

Abstract

A method and apparatus for monitoring a patient's chronic disease state are disclosed. The method can include, with a processor, extracting, for each of a plurality of monitoring sessions, a respiration feature from a respiration signal indicative of a patient's respiration during a monitoring session, the respiration signal being derived from at least one sensor, and with the processor, computing a stability measure for the patient for the monitoring session, the stability measure representing an indication of a change point that has occurred in the monitoring session in a statistical distribution of the respiration feature.

Description

[0001] [Invention Specification]

[0002] Methods and equipment for monitoring chronic diseases

[0003] Divisional application statement

[0004] This application is a divisional application of Chinese invention patent application No. 201580040820.6, which was filed on May 25, 2015, with international application number PCT / AU2015 / 050273, and entered the Chinese national phase on January 22, 2017.

[0005] Cross-references to related applications

[0006] This application claims the benefit of Australian Provisional Patent Application No. 2014901975, filed on 26 May 2014, the entire contents of which are incorporated herein by reference.

[0007] Statements regarding federally sponsored research or development

[0008] not applicable

[0009] Name of the unit to which the joint research and development belongs.

[0010] not applicable

[0011] sequence list

[0012] not applicable Technical Field

[0013] This technology relates to the detection, diagnosis, treatment, prevention, and improvement of one or more chronic diseases. More specifically, this technology relates to medical devices or equipment and their use. Background Technology

[0014] Human respiratory system and its disorders

[0015] The body's respiratory system facilitates gas exchange. The nose and mouth form the entrance to the patient's airway.

[0016] The respiratory tract consists of a series of branching tubes, which become narrower, shorter, and more numerous as they penetrate deeper into the lungs. The primary function of the lungs is gas exchange, allowing oxygen to enter the venous blood from the air and expelling carbon dioxide. The trachea divides into the right and left main bronchi, which further divide into terminal bronchioles. The bronchi form the conduction airway and do not participate in gas exchange. Further branches of the respiratory tract lead to the respiratory bronchioles and ultimately to the alveoli. The alveolar region of the lungs is where gas exchange occurs and is called the respiratory zone. See *Respiratory Physiology*, 9th edition, 2011, published by Lippincott Williams & Wilkins, author John B. West.

[0017] A variety of breathing disorders exist in the human body. Specific disorders are characterized by particular event conditions, such as sleep apnea, shallow and slow breathing, and deep and rapid breathing.

[0018] Obstructive sleep apnea (OSA) is a form of sleep-disordered breathing (SDB) characterized by upper airway obstruction or impairment during sleep. This is due to a loss of normal muscle tone throughout the abnormal small upper airway and in the tongue, soft palate, and posterior pharyngeal wall regions during sleep. This symptom typically causes affected patients to stop breathing for 30 to 120 seconds each night, sometimes 200 to 300 seconds. This often causes excessive daytime sleepiness and can lead to cardiovascular disease and brain damage. Complications include general disturbances, particularly in middle-aged overweight men, but patients may not notice the problem. See U.S. Patent No. 4,944,310 (Mr. Sullivan).

[0019] Cheyne-Schönlein respiration (CSR) is another form of sleep-disordered breathing in patients. CSR is an impairment of the patient's respiratory controller, characterized by alternating periods of increased and decreased ventilation known as part of the CSR respiratory cycle. CSR is characterized by repeated deoxygenation and reoxygenation of arterial blood. CSR can be harmful due to the repeated hypoxia. In some patients, CSR is associated with recurrent awakenings during sleep, causing severe sleep disruption, increased sympathetic activity, and increased capillary volume reduction. See U.S. Patent No. 6,532,959 (Mr. Berthon-Jones).

[0020] Obesity-induced hypoventilation syndrome (OHS) is defined as a combination of severe obesity and chronic hypercapnia at wakefulness, without other known factors contributing to hypoventilation. Symptoms include shortness of breath, morning headache, and excessive daytime sleepiness.

[0021] Chronic obstructive pulmonary disease (COPD) can include any of several lower respiratory tract diseases that share common specific characteristics. These include increased resistance to airflow, prolonged expiratory phase, and loss of normal lung elasticity. Examples of COPD include emphysema and chronic bronchitis. COPD is caused by chronic tobacco smoking (a major risk factor), occupational exposure, air pollution, and genetic factors. Symptoms include: exertional dyspnea, chronic cough, and sputum production.

[0022] Neuromuscular diseases (NMD) are a broad term encompassing a wide range of conditions and disorders that impair muscle function directly through intrinsic muscle pathology or indirectly through neuropathology. Some NMD patients are characterized by progressive muscle damage leading to decreased mobility (requiring a wheelchair), dysphagia, respiratory muscle failure, and ultimately, respiratory failure resulting in death. Neuromuscular diseases can be classified as rapidly progressive or chronically progressive: (i) rapidly progressive disorders: characterized by worsening muscle damage over several months and resulting in death within a few years (e.g., juvenile amyotrophic lateral sclerosis (ALS) and Duchenne muscular dystrophy (DMD)); (ii) variable or slowly progressive disorders: characterized by worsening muscle damage over several years and only slightly reducing the likelihood of life expectancy (e.g., limb-girdle type, face-shoulder-arm type, and myotonic muscular dystrophy). Symptoms of respiratory failure in neuromuscular diseases (NMD) include: progressive general weakness, dysphagia, shortness of breath at exercise and rest, fatigue, drowsiness, morning headache, difficulty concentrating, and mood changes.

[0023] Chest wall disorders are deformities of the thoracic cavity that cause inefficient coupling between the respiratory muscles and the thoracic cavity. Chest wall disorders are typically restrictive and contribute to the potential for chronic hypercapnia-related respiratory failure. Scoliosis and / or kyphosis can cause severe respiratory failure. Symptoms of respiratory failure include: dyspnea on exertion, peripheral edema, orthopnea, periodic pleural infections, morning headache, fatigue, poor sleep quality, and loss of appetite.

[0024] Heart failure (HF) is a fairly common and serious clinical condition characterized by the heart's inability to meet the body's oxygen demands. Due to its prevalence and severity, managing HF presents a significant challenge to modern healthcare systems. HF is a chronic condition that is inherently deteriorating. The progression of HF is often characterized by prolonged periods of relative stability (even with reduced cardiovascular function), emphasizing acute episodes. During these acute episodes, patients experience progressively worsening symptoms such as shortness of breath (difficulty breathing), gallop rhythm, increased pressure in the pharyngeal veins, and orthopnea. This is often accompanied by significant obstruction (which increases fluid in the lung cavities). This excess fluid often results in a weight gain of several kilograms. However, in many cases, physicians have limited options to help patients regain stability before significant obstruction occurs, and in many cases, hospitalization is required. In extreme cases, without appropriate treatment, patients may experience acute decompensated heart failure (ADHF) events, sometimes referred to as decompensation.

[0025] treat

[0026] Continuous positive airway pressure (CPAP) has been used to treat obstructive sleep apnea (OSA). The assumption is that continuous positive airway pressure acts as a blowing splint and can prevent upper airway obstruction by pushing the soft palate and tongue forward and away from the posterior pharyngeal wall.

[0027] Non-invasive ventilation (NIV) provides ventilator support to patients through the upper airway to help them achieve full breathing and / or maintain adequate oxygen levels by performing some or all of their breathing. Ventilator support is delivered through a non-invasive patient interface. NIV has been used to treat Cheyne-Stokes respiration (CSR), obesity-related hypoventilation syndrome (OHS), chronic obstructive pulmonary disease (COPD), muscle disease (MD), and chest wall disorders.

[0028] Treatment System

[0029] The aforementioned treatments can be provided using treatment systems or devices. These systems and devices can also be used to diagnose conditions without providing treatment.

[0030] The treatment system may include a respiratory pressure therapy device (RPT device), an air loop, a humidifier, a patient interface, and data management.

[0031] Patient Interface

[0032] A patient interface can be used to interface a respiratory device with a user, for example, by providing an airflow. The airflow can be provided to the nose and / or mouth through a nasal mask, to the mouth through a tube, or to the user's trachea through a tracheostomy tube. Depending on the treatment to be administered, the patient interface may, for example, form a seal with the patient's facial area to facilitate gas delivery at a pressure sufficiently different from ambient pressure to achieve therapeutic effect, such as a positive pressure of approximately 10 cm H2O. For other forms of treatment, such as oxygen delivery, the patient interface may not include a sufficient seal to facilitate gas delivery to the airway at a positive pressure of approximately 10 cm H2O.

[0033] Respiratory Pressure Therapy (RPT) device

[0034] Air pressure generators are known for their use in a variety of applications, such as industrial-scale ventilation systems. However, air pressure generators for medical applications have more specific requirements that cannot be met by ordinary air pressure generators, such as the reliability, size, and weight requirements of medical devices. In addition, stable devices designed for medical treatment may have disadvantages, including one or more of the following: comfort, noise, ease of use, efficiency, size, weight, manufacturability, cost, and reliability.

[0035] Humidifier

[0036] Delivering an unhumidified airflow can cause respiratory dryness. Using a humidifier at the patient interface with an RPT device produces humidified air, reducing nasal dryness and increasing patient respiratory comfort. Except in colder climates, warm air applied to or near the patient interface is generally more comfortable than cold air.

[0037] Monitoring System

[0038] Patients with heart failure or chronic obstructive pulmonary disease (COPD) who wish to monitor their condition at home may benefit from this technology to prevent or mitigate potential clinical events such as heart failure decompensation or COPD flare-ups. Features proposed or used for predicting clinical events include body weight, B-type natriuretic peptide (BNP) levels, nighttime heart rate, and changes in sleep posture. Sleep multidimensional mapping (PSG) is a standard system for monitoring cardiopulmonary disorders. A typical PSG setup involves placing 15 to 20 contact sensors in the body to record various bodily signals, such as electroencephalograms (EEG), electrocardiograms (ECG), and electrooculograms (EOG).

[0039] Heart failure has been shown to be highly associated with sleep-disordered breathing (SDB). More specifically, Cheyne-Stokes respiration (CSR) is often caused by instability in the body's respiratory control system, one cause of which is heart failure. The severity of Cheyne-Stokes respiration can be represented by a set of characteristics indicating the degree to which breathing during sleep resembles typical Cheyne-Stokes respiration, i.e., "Cyyne-like respiration" characteristics. Furthermore, characteristics indicating the severity of obstructive sleep apnea, such as the Apnea / Shallow Breathing Index (AHI), have shown to be independent predictors of death from ADHF events and hospitalization related to ADHF events. Values ​​and variations in these sleep-disordered breathing characteristics can provide useful information about the likelihood of ADHF events. Contact sensor modalities, such as nasal masks or oronasal tubes capable of monitoring and analyzing respiratory parameters during sleep to extract SDB characteristics, have been discussed in the context of monitoring chronic cardiopulmonary diseases. Implantable sensors have been used to monitor chest impedance and arrhythmias to predict ADHF events.

[0040] Heart failure monitoring systems based on the aforementioned sensor models tend to be unsatisfactory because they require good patient compliance. For example, weight-based monitoring systems that rely on patients recording their daily weight are wearable but impractical for long-term monitoring; or, these systems are invasive or intrusive. The use of implantable devices is only feasible for a subset of heart failure patients who are suitable for such devices.

[0041] SleepMinder (ResMed Sensors Technologies, Dublin, Ireland) is a contactless bedside monitor suitable for long-term monitoring of chronic diseases. SleepMinder comprises a bio-motion transceiver sensor operating on the Doppler radar principle with ultra-low power (less than 1mW), operating in the unlicensed 5.8GHz frequency band. SleepMinder can measure human movement, particularly respiratory movement, within a range of 0.3 to 1.5 meters; in situations with two people in bed, the combination of a high-performance sensor design and intelligent signal processing allows SleepMinder to measure only the respiratory movements of the person closest to the sensor. SleepMinder is suitable for long-term monitoring of chronic diseases because it is non-intrusive and does not exhibit significant compliance issues. Summary of the Invention

[0042] This technology aims to provide medical devices for monitoring cardiopulmonary disorders or other chronic diseases that offer advantages in one or more of the following: improved comfort, cost, efficiency, ease of use, and manufacturability.

[0043] The first aspect of this technology relates to devices for monitoring cardiopulmonary disorders or other chronic diseases.

[0044] Another aspect of this technology relates to methods for monitoring cardiopulmonary disorders or other chronic diseases.

[0045] One form of this technology includes a chronic disease monitoring device that extracts respiratory features from a patient's respiratory signals during each monitoring phase, and calculates a stability measurement from a statistical analysis of the time series formed by continuous values ​​of the respiratory features across multiple monitoring phases. The stability measurement represents an indication of the points of change that have occurred in the statistical distribution of the respiratory features during that monitoring phase. An alert can be generated if the stability measurement meets criteria.

[0046] Another form of this technology includes a method for monitoring chronic diseases, which includes calculating a stability measure, which is an indication of points of change that have occurred in the probability distribution of a time series formed by continuous values ​​of respiratory features extracted from a patient's respiratory signals across multiple monitoring phases. The stability measure represents an indication of points of change that have occurred in the statistical distribution of respiratory features during that monitoring phase. If a point of change in the distribution is detected, an alert can be generated. The calculation method can be retrospective or online.

[0047] According to a first aspect of the present technology, a method for monitoring the chronic disease state of a patient is provided. The method includes: using a processor, for each of a plurality of monitoring phases, extracting respiratory features from a respiratory signal indicative of the patient's breathing during the monitoring phase, the respiratory signal originating from at least one sensor; and using the processor, calculating a stability measurement of the patient during the monitoring phase, the stability measurement representing an indication of a point of change in the statistical distribution of the respiratory features that has occurred during the monitoring phase.

[0048] According to a second aspect, a chronic disease monitoring device is provided, comprising: a sensor configured to generate a respiratory signal indicative of a patient’s breathing during a monitoring phase; and a processor configured to implement the method according to the first aspect.

[0049] According to a third aspect, a method for monitoring a patient's chronic disease state is provided. The method, implemented using one or more processors, includes: for each of a plurality of monitoring phases, extracting respiratory features from a respiratory signal indicative of the patient's respiration, the respiratory signal originating from at least one sensor during the monitoring phase; forming a time series from continuous values ​​of the respiratory features; and calculating a stability measurement of the patient during the monitoring phase. The stability measurement represents a measure of the dissimilarity of the probability distributions of two sets of subsequences of the time series. The two sets comprise subsequences substantially composed of samples of the time series taken before and after the monitoring phase, respectively.

[0050] According to a fourth aspect, a chronic disease monitoring device is provided, comprising: a sensor configured to generate a respiratory signal indicating a patient's breathing during a monitoring phase; and a processor configured to implement the method according to a third aspect.

[0051] According to a fifth aspect, a method for monitoring a patient's chronic disease state is provided. The method, implemented using one or more processors, includes: for each of a plurality of monitoring phases, extracting respiratory features from a respiratory signal indicative of the patient's breathing, the respiratory signal originating from at least one sensor during the monitoring phase; and calculating a stability measure for the patient during the monitoring phase. The stability measure represents the probability that a point of change in the statistical distribution of the respiratory features has occurred during that monitoring phase. The calculation includes: if there is an upper limit to the values ​​of the respiratory features included in the monitoring phase, then calculating a posterior distribution of the travel length specific to the monitoring phase; and calculating the sum of the values ​​of the posterior distribution of the travel length.

[0052] According to a sixth aspect, a chronic disease monitoring device is provided, comprising: a sensor configured to generate a respiratory signal indicating a patient's breathing during a monitoring phase; and a processor configured to implement the method according to a fifth aspect.

[0053] Of course, parts of the aforementioned aspects can form subordinate aspects of this technology. Furthermore, subordinate aspects and / or various different aspects of the aspects can be combined in various different ways, and also constitute additional or subordinate aspects of this technology.

[0054] Other features of this technology will become apparent from the following detailed description, abstract, brief description of the drawings, and information contained in the claims. Attached Figure Description

[0055] This technique is illustrated by way of example and not by way of limitation in the accompanying drawings, wherein the same element symbols refer to similar elements, including:

[0056] Treatment System

[0057] Figure 1 An exemplary treatment system according to one form of the present technology is shown. A patient 1000 wearing a patient interface 3000 receives a positive pressure air supply from an RPT device 4000. The air from the RPT device 4000 is humidified in a humidifier 5000 and delivered to the patient 1000 along an air loop 4170.

[0058] Respiratory system and facial anatomy

[0059] Figure 2 It shows an overview of the human respiratory system, including the nose and mouth, larynx, vocal cords, esophagus, trachea, bronchi, lungs, alveolar sacs, heart, and diaphragm.

[0060] Patient Interface

[0061] Figure 3 A patient interface in the form of a nasal mask according to the present technology is shown.

[0062] Respiratory Pressure Therapy (RPT) device

[0063] Figure 4A An RPT device of one form according to the present technology is shown.

[0064] Figure 4B One form according to the present technology is shown. Figure 4A A schematic diagram of the pneumatic path of the RPT device shown. Indicate the upstream and downstream directions.

[0065] Figure 4C One aspect of the present technology is shown. Figure 4A A schematic diagram of the electrical components of the RPT device shown.

[0066] Humidifier

[0067] Figure 5 An isometric view of a humidifier according to one aspect of the present technology is shown.

[0068] respiratory waveform

[0069] Figure 6A This shows a typical breathing waveform model during human sleep.

[0070] Figure 6B The data shows sleep multidimensional graphs of patients during a period of approximately 90 seconds, which is typically non-REM sleep apnea.

[0071] Figure 6C This shows sleep multidimensional data from patients with obstructive sleep apnea.

[0072] Figure 6D This shows patient flow rate data for patients experiencing a series of overall obstructive respiratory arrest conditions.

[0073] Figure 6E This shows sleep multidimensional data of patients with Cheyne-Stokes respiratory syndrome.

[0074] Monitoring equipment

[0075] Figure 7A This invention illustrates a device for monitoring a patient's sleep according to the present technology.

[0076] Figure 7B For more detailed illustrations Figure 7A A block diagram of the monitoring equipment.

[0077] Figure 7C As illustrated, this is one form of the technology. Figure 7B The flowchart describes a method for monitoring a patient's chronic disease status using monitoring equipment.

[0078] Figure 7D The illustration is for implementation according to one form of the present technology. Figure 7C A block diagram illustrating the feature extraction step in the method shown.

[0079] Figure 7E The illustration shows one form of the technology used in implementing it online. Figure 7C A block diagram illustrating the stability measurement calculation steps in the method shown.

[0080] Figure 7F This illustration shows one form of the technology used for implementation based on a traceability method. Figure 7C A block diagram illustrating the stability measurement calculation steps in the method shown.

[0081] Figure 8 Includes showing the use Figure 7C The method shown is from Figure 7A The chart shows the results obtained by the monitoring equipment. Detailed Implementation

[0082] Before describing the technology in more detail, it should be understood that the technology is not limited to the specific instances described herein, and these specific instances may vary. It should also be understood that the terminology used in this invention is merely for describing the specific instances discussed herein and is not intended to be limiting.

[0083] The devices and methods described below are particularly suitable for monitoring cardiopulmonary health, and are described using these terms. However, the devices and methods described may also be applied to monitoring other chronic diseases that affect a patient's breathing.

[0084] Monitoring equipment and methods

[0085] Non-blocking monitoring equipment

[0086] Figure 7A The illustration shows a form of non-invasive monitoring device 7000 according to the present technology. The monitoring device 7000 is located adjacent to and quite close to the sleeping patient 1000 (e.g., on a bedside table).

[0087] Figure 7B For more detailed illustration, one form according to this technology Figure 7A A block diagram of the components of the monitoring device 7000 is shown. In the monitoring device 7000, the non-contact sensor unit 1200 includes a non-contact motion sensor 7010 typically for a patient 1000. The motion sensor 7010 is configured to generate one or more signals representing body movements of the patient 1000, from which one or more respiratory motion signals representing respiratory movements of the patient can be derived.

[0088] The sensor unit 1200 may also include a microcontroller unit (MCU) 7001 and a memory 7002 (e.g., a memory card) for recording data. In one embodiment, the sensor unit 1200 may include communication circuitry 7004 configured to transmit data to an external computing device 7005, such as a local general-purpose computer or a remote server, via a connection 7008. The connection 7008 may be wired or wireless; in this case, the communication circuitry 7004 is wirelessly capable and can communicate directly or indirectly via a local area network or a wide area network (not shown) (such as the Internet).

[0089] The sensor unit 1200 includes a processor 7006 configured to process signals generated by the motion sensor 7010, as described in detail below.

[0090] Sensor unit 1200 includes a display device 7015 configured to provide visual feedback to a user. In one embodiment, display device 7015 includes one or more warning lights (e.g., one or more light-emitting diodes). Display device 7015 may also be implemented as a display screen, such as an LCD or a touch-sensitive display. Operation of display device 7015 is controlled by processor 7006 based on an assessment of the patient's cardiopulmonary health. Display device 7015 is operable to display data to users of monitoring device 7000, such as patient 1000, or physicians or other clinicians. Display device 7015 may also display a graphical user interface for operation of monitoring device 7000.

[0091] The sensor unit 1200 may also include an audio output 7017 configured to provide auditory feedback to the user under the control of the processor 7006, such as a tone whose frequency changes with breathing, or a warning that sounds in accordance with a specific situation.

[0092] User control of the operation of the monitoring device 7000 can be based on control operations (not shown) sensed by the processor 7006 of the monitoring device 7000.

[0093] An example of sensor unit 1200 is the SleepMinder device manufactured by ResMed Sensor Technologies Ltd, which includes a non-contact Doppler radio frequency (RF) motion sensor 7010.

[0094] In one form of this technology, such as when the SleepMinder device is used as sensor unit 1200, the motion sensor 7010 includes a radio frequency transmitter 7020 configured to transmit a radio frequency signal 7060. The transmitted signal 7060 has, for example, the following form:

[0095] s(t)=u(t)cos(2πfc t+θ)(Equation 1)

[0096] In Equation 1, the carrier frequency is f. c (Typically in the 100MHz to 100GHz range, e.g., 3GHz to 12GHz, e.g., 5.8GHz or 10.5GHz), where t is time, θ is an arbitrary phase angle, and u(t) is the pulse shape. In a continuous waveform system, the amplitude of u(t) can be uniform and can be omitted from Equation 1. More generally, the pulse u(t) can be defined as follows: Equation 2:

[0097]

[0098] Where T is the period width, and T p It is the pulse width. Where T p << T, this becomes a pulse continuous waveform system. In one case, when T p When the frequency becomes very small, the spectrum of the transmitted signal becomes very wide, and the system is called ultra-wideband (UWB) radar or pulse radar. Alternatively, the carrier frequency of the radio frequency transmission signal 7060 can be changed (chirped pulse) to produce a so-called frequency modulated continuous waveform (FMCW) system.

[0099] The radio frequency (RF) signal 7060 can be generated by the transmitter 7020 using a local oscillator 7040 coupled with a pulse-gated circuit. In the case of FMCW, a voltage-controlled oscillator is used in conjunction with a voltage-to-frequency converter to generate the RF signal 7060 for transmission. The coupling between the air and the transmitted RF signal 7060 can be achieved using an antenna 7050. The antenna 7050 can be omnidirectional (transmitting the same or less power in all directions) or directional (transmitting stronger power in a specific direction). It is advantageous to use a directional antenna 7050 in the monitoring device 7000 so that the transmitted and reflected energy mainly comes from one direction. In one embodiment of the monitoring device 7000, a single antenna 7050 with a single carrier frequency can be used for both the transmitter 7020 and the receiver 7030. Alternatively, multiple receiving and transmitting antennas 7050 with multiple carrier frequencies can be used.

[0100] Device 7000 is compatible with various specific embodiments using different types of antennas 7050, such as simple dipole antennas, planar antennas, and helical antennas, and certain factors may influence antenna selection, such as required directivity, size, shape, or cost. It should be noted that monitoring device 7000 can be operated in a manner safe for human use. Device 7000 has been fully validated with an average transmit power of 1 mW (milliwatts) (0 dBm) or lower. The recommended safety level for radio frequency (RF) exposure is 1 mW / cm². 2(milliwatts per square centimeter). At a distance of 1 meter from a 0 dBm transmission system, the equivalent power density will be at least 100 times less than the recommended limit.

[0101] In use, the transmitted radio frequency (RF) signal 7060 is a reflective object that reflects radio waves (such as the air interface of patient 1000), and some of the reflected signals 7070 are received by a receiver 7030 employing a so-called "bistatic" structure, which may be configured with or separate from the transmitter 7020. The received signal 7070 and the transmitted signal 7060 can be multiplied in a mixer 7080 (in analog or digital mode). This mixer 7080 can be in the form of a multiplier (as shown below (in Equation 3)) or a circuit that approximates the effect of a multiplier (e.g., a wave-blocking detector circuit with a sinusoidal waveform added). For example, in the CW case, the mixed signal will be equal to the following equation:

[0102] m(t)=γcos(2πf c t)cos(2πf c t+φ(t)) (Equation 3)

[0103] in It is the phase term caused by the path difference between the transmitted signal 7060 and the received signal 7070 (in the case where reflection is controlled by a single reflecting object), and γ is the attenuation experienced by the reflected signal 7070. If the reflecting object is fixed, then Fixed. In device 7000, reflective objects (e.g., the chest of patient 1000) typically move, and It will change over time. To give a simple example, if the chest moves at a frequency f due to breathing... m Given a sinusoidal motion, the mixed signal m(t) includes f m Components (and centering at 2f) c The component may only be removed by the low-pass filter. The output signal of the mixed low-pass filter is called the motion signal or demodulated sensor motion signal 7003, and includes information about whole-body (non-respiratory) motion and respiratory motion.

[0104] The amplitude of the demodulated sensor motion signal 7003 is affected by the average path distance of the reflected signals, causing the motion sensor 7010 to experience detection nulls and peaks (i.e., areas of excessive or low sensitivity for the motion sensor 7010). This effect can be reduced using orthogonal techniques, where the transmitter 7020 simultaneously transmits single, orthogonal 90-degree phase signals 7060 as shown in Equation 1. This results in two reflected signals, which can be mixed and low-pass filtered by the mixer 7080, resulting in two demodulated sensor signals, referred to as the "I signal" and "Q signal" in their respective I″ and Q″ channels. The motion signal 7003 may include one or both of these signals.

[0105] In an ultra-wideband (UWB) implementation, an alternative method to acquiring the motion signal 7003 can be used. The path distance to the most critical airborne interface can be determined by measuring the delay between the transmitted pulse and the peak reflection signal. For example, if the pulse width is 1 ns (nanoseconds) and the distance from the motion sensor 7010 to the body is 0.5 meters, then the delay before the peak reflection of the pulse reaches the receiver 7030 will be 1 / (3 x 10^6)^2. 8 s (seconds) = 3.33 ns (nanoseconds). By transmitting a large number of pulses (e.g., 1 ns (nanosecond) pulse per 1 μs (microsecond)) and assuming that the path distance changes slowly over a specific time period, the motion signal 7003 can be calculated as an average of the time delay over that period.

[0106] Thus, the motion sensor 7010 (e.g., a radio frequency sensor) can assess respiratory movements of the chest wall; or, more generally, the monitoring device 7000 is monitoring the movements of 1000 body parts of the patient.

[0107] As mentioned above, the received signal 7070 may include large motion noise, such as the result of whole-body movement. This is because reflected signals from the body may include more than one reflective path and result in complex signals (e.g., if a hand is moving toward the sensor and the chest is moving away). Receiving such signals is useful when they indicate that the upper body is in motion, and is helpful in determining sleep status.

[0108] To improve the quality of respiratory motion signals, and more generally, body motion signals, the actual amount of reflected energy collected by sensor unit 1200 may be limited by various methods. For example, sensor unit 1200 may be made "directionally selective" (i.e., transmitting more energy in a certain direction), such as the antenna of receiver 7030. Directional selectivity can be achieved using a directional antenna 7050 or a multiple radio frequency (RF) transmitter 7020. In alternative forms of this technology, a continuous waveform, FMCW, or UWB radar can be used to obtain similar signals. A technique called "time-domain gating" can be used only to measure the reflected signal 7070 generated from a specific actual distance from sensor unit 1200. Frequency-domain gating (filtering) can be used to ignore the motion of the reflecting object above a specific frequency.

[0109] In the device 7000 embodiment using multiple frequencies (e.g., 500 MHz and 5 GHz), the lower frequencies can be used to accurately determine large movements without phase ambiguity, which can then be subtracted from the higher frequency sensor signal (more suitable for measuring small movements). Using this sensor unit 1200, the device 7000 can collect information from the patient 1000 and use this information to determine respiratory movements, and more generally, to collect body movement information.

[0110] The motion signal 7003 may be stored in the memory 7002 of the sensor unit 1200 for each monitoring phase, and / or transmitted via a link (e.g., connection 7008) and stored in an external computing device 7005. In one embodiment, the duration of each monitoring phase is one night.

[0111] The processor of sensor unit 1200 or external computing device 7005 can process the stored motion signal 7003 according to the monitoring process, as described in detail below. Instructions describing the process can be stored on a computer-readable storage medium, such as the memory 7002 of sensor unit 1200, and interpreted and executed by a processor, such as the processor 7006 of sensor unit 1200.

[0112] Alternative monitoring equipment

[0113] In other forms of this technology, such as Figure 1 As shown, the RPT device 4000, which is configured to provide respiratory pressure therapy to the patient 1000 via the air loop 4170 through the patient interface 3000, can also be configured as a monitoring device.

[0114] The patient interface 3000 may include the following functional aspects: a sealing-forming structure 3100, an inflation chamber 3200, a positioning and stabilizing structure 3300, an air exchange port 3400, a connection port 3600 for connecting to an air loop 4170, and a forehead support 3700. In some forms, the functional aspects may be provided via one or more physical components. In some forms, a single physical component may provide one or more functional aspects. In use, the sealing-forming structure 3100 is configured to surround the inlet of the patient's airway to facilitate the supply of positive pressure air to the airway.

[0115] The RPT device 4000 may include mechanical and pneumatic components 4100, electrical components 4200, and is configured to execute one or more algorithms. The RPT device preferably has a housing 4010, preferably formed in two parts, namely an upper portion 4012 and a lower portion 4014. Furthermore, the housing 4010 may include one or more panels 4015. Preferably, the RPT device 4000 includes a chassis 4016 for supporting one or more internal components of the RPT device 4000. The RPT device 4000 may include a handle 4018.

[0116] The pneumatic path of the RPT device 4000 preferably includes one or more air path articles, such as an inlet filter 4112; an inlet silencer 4122; a pressure generator 4140 capable of supplying positive pressure air (preferably a blower 4142); an outlet silencer 4124; and one or more converters 4270, such as a pressure sensor 4272 and a flow rate sensor 4274.

[0117] One or more air path items may be located within a removable single structure, referred to as pneumatic stop block 4020. Pneumatic stop block 4020 may be located within housing 4010. In one form, pneumatic stop block 4020 is supported by or forms part of chassis 4016.

[0118] The RPT device 4000 preferably includes a power supply 4210, one or more input devices 4220, a central controller 4230, a treatment device controller 4240, a pressure generator 4140, one or more protection circuits 4250, a memory 4260, a converter 4270, a data communication interface 4280, and one or more output devices 4290. Electrical components 4200 may be mounted on a single printed circuit board assembly (PCBA) 4202. Alternatively, the RPT device 4000 may include more than one printed circuit board assembly 4202.

[0119] In one form of this technology, the central controller 4230 is one or more processors suitable for controlling the RPT device 4000. Suitable processors may include x86 Intel processors, which are based on processors provided by ARM Holdings. -M processors, such as the STM32 series microcontrollers from STMicroelectronics. In certain alternative forms of this technology, 32-bit RISC CPUs (such as the STR9 series microcontrollers from STMicroelectronics) or 16-bit RISC CPUs (such as processors in the MSP430 series microcontrollers manufactured by Texas Instruments) are also suitable.

[0120] In one embodiment of this technology, the central controller 4230 is a dedicated electronic circuit. In another embodiment, the central controller 4230 is a dedicated integrated circuit. In yet another embodiment, the central controller 4230 includes discrete electronic components.

[0121] The central controller 4230 may be configured to receive input signals from one or more converters 4270 and one or more input devices 4220.

[0122] The central controller 4230 may be configured to provide output signals to one or more of the output device 4290, the treatment device controller 4240, the data communication interface 4280, and the humidifier 5000.

[0123] In some forms of this technology, the central controller 4230 is configured to implement one or more programs as described herein, represented as computer programs stored in a non-transitory computer-readable storage medium (such as memory 4260).

[0124] Data communication interface 4280 can be connected to a remote external communication network 4282 and / or a local external communication network 4284. Remote external communication network 4282 can be connected to a remote external device 4286. Local external communication network 4284 can be connected to a local external device 4288. Data communication interface 4280 can connect to the Internet using wired communication (e.g., via Ethernet or fiber optic) or wireless protocols (e.g., CDMA, GSM, LTE).

[0125] In one embodiment, the local external communication network 4284 utilizes one or more communication standards (such as Bluetooth) or consumer infrared protocols. The local external device 4288 may be a personal computer, mobile phone, tablet computer, or remote control device.

[0126] In one form, the remote external communication network 4282 is the Internet. In another form, the remote external device 4286 is one or more computers, such as a cluster of computers connected to a network. In one form, the remote external device 4286 may be a virtual computer rather than a physical computer. In any case, such a remote external device 4286 may be accessed by appropriately certified personnel, such as a clinician.

[0127] Output device 4290 may take the form of one or more visual, audio, and touch units. The visual display may be a liquid crystal display (LCD) or a light-emitting diode (LED) display. Display driver 4292 receives characters, symbols, or images displayed on display 4294 as input and converts them into commands that cause display 4294 to display these characters, symbols, or images. Display 4294 is configured to visually display characters, symbols, or images in response to commands received from display driver 4292.

[0128] Monitoring process

[0129] In one aspect of this technology, the monitoring device enables the monitoring process to monitor the cardiopulmonary health of a patient from respiratory signals indicating 1000 breaths.

[0130] In this technical form, the monitoring device is Figure 7B The non-invasive device 7000 shown has a respiratory motion signal derived from motion signal 7003. This monitoring process can be implemented by the processor 7006 of the non-contact sensor unit 1200, configured to utilize instructions stored on a computer-readable storage medium (such as memory 7002). Alternatively, the processor of the external computing device 7005 can implement all or part of the described monitoring process, acquiring necessary data (raw or partially processed) from the sensor unit 1200 of the monitoring device 7000 and any other sensors via the aforementioned connection 7008. In this implementation, the visual display 7015 and audio output 7017 of the monitoring device 7000, as described above, are equally applicable to the equivalent elements of the external computing device 7005. In one example, the external computing device 7005 is a clinician-accessible device, such as a multi-patient monitoring device, allowing clinicians to review data from multiple remote patient data recording devices (such as the monitoring device 7000). In these systems, a database is available for recording patient monitoring data. Through this external computing device 7005, clinicians can receive reports or alerts about special patients who may require detailed observation or should go to the hospital for treatment.

[0131] In this embodiment, where the monitoring device is an RPT device 4000 and the respiratory signal is a signal representing a patient respiratory flow rate Qr of 1000 originating from one or more transducers 4270, the monitoring process can be implemented by a central controller 4230 of the RPT device 4000, which is configured to utilize instructions stored on a computer-readable storage medium (such as memory 4260). Alternatively, a local external device 4288 or a remote external device 4286 can implement all or part of the described process, which can obtain the necessary data (raw or partially processed) from the RPT device 4000 via a data communication interface 4280, as described above. In such embodiments, the output function of the output device 4290 of the RPT device 4000 is implemented by a peer element of the local external device 4288 or the remote external device 4286.

[0132] Figure 7C This is a flowchart illustrating a method 7100 for implementing a monitoring process according to the present technology. Method 7100 can be implemented at the end of each monitoring phase that includes storing respiratory signals within a monitoring stage.

[0133] Method 7100 begins with step 7110, in which the respiratory signal undergoes preprocessing. Preprocessing step 7110 (e.g.) Figure 7C The dashed line (shown) is optional and can be omitted from method 7100. In the next step 7120, the respiratory signal is (possibly preprocessed) analyzed to extract one or more respiratory features. The extracted respiratory features may be stored in a memory, such as the memory 7002 of the sensor unit 1200 or the external computing device 7005.

[0134] Then, in step 7130, method 7100 uses extracted respiratory features from the just-completed monitoring phase, along with possible respiratory features from one or more previous monitoring phases, to calculate a stability measurement. The stability measurements thus established, or a history of continuously calculated stability measurements based on phase-by-phase calculations, can be stored in one or more memories, such as the memory 7002 of sensor unit 1200 or external computing device 7005, or other memories of the processor that calculates the stability measurements. In step 7130, the calculated stability measurement can serve as a predictor of possible clinical events, as changes in the stability measurement (e.g., an increase) may indicate a deterioration in the patient's condition, which can be a precursor to a clinical event. Changes in the stability measurement (e.g., an increase) when the patient's condition improves are also events of concern for monitoring chronic diseases.

[0135] Then, in step 7140, the stability measurement is evaluated to determine whether it meets criteria, such as by comparing one or more threshold values. For example, in step 7140, a processor may be used to compare the stability measurement with the threshold values. If the stability measurement exceeds the threshold value, for example, yes ("Y"), a change point is detected, and an alert may be generated in step 7150. If no ("N"), then method 7100 terminates in step 7160. When performing the monitoring process on training data, the selection of the threshold value affects the sensitivity and characteristics of the monitoring process used to detect possible clinical events, and is selected based on the desired level of sensitivity and characteristics. In some embodiments, the threshold value may be adjusted between monitoring phases based on observed false positive and false negative detections. In step 7140, other assessments may determine whether the stability measurement is within a specific range, such as by comparing one or more threshold values ​​belonging to one or more ranges. Thus, automated monitoring processes can efficiently transform respiratory signal data (which may appear harmless) into patient monitoring tools, i.e., stability measurements, through processing, thereby improving monitoring equipment and enhancing the ability of field clinicians to more effectively monitor their patients, such as making necessary treatment changes in a timely manner.

[0136] In this embodiment, the monitoring device is a non-invasive monitoring device 7000, and the non-contact motion sensor 7010 is a Doppler radio frequency (RF) motion sensor. As mentioned above, in this embodiment, the motion signal 7003 may include two signals, labeled I and Q, each typically indicating body movement, but generally 90 degrees out of phase with each other.

[0137] When motion signal 7003 includes I and Q signals, several approaches are feasible. In the "parallel" approach, steps 7110 and 7120 are performed for each of the parallel I and Q signals, and the individually acquired features are combined at the end of feature extraction step 7120. In one embodiment of the parallel approach, preprocessing step 7110 is omitted. In the "combined" approach, the I and Q signals are combined as part of preprocessing step 7110, and processing steps 7120 to 7130 are performed for the combined motion signal. Compared to the parallel approach, the combined approach may have the advantage of lower computational complexity at the expense of lower accuracy.

[0138] Alternatively, the non-contact motion sensor 7010 can provide a single motion signal 7003. This is processed in a manner known as "single-channel".

[0139] The following sections describe the process in more detail. Figure 7C The implementation of the steps of the monitoring method 7100 shown.

[0140] The implementation of steps 7110 and 7120 is described from the perspective of the form of this technology, wherein the monitoring device is Figure 7BThe monitoring equipment shown is 7000. The monitoring equipment is... Figure 4A In the form of the RPT device 4000 shown in this technology, the preprocessing step 7110 can be omitted, and the respiratory feature extraction step 7120 can be performed in a conventional manner for the respiratory flow signal Qr.

[0141] The description of steps 7130 to 7150 is applicable to the aforementioned forms of this technology.

[0142] Preprocessing

[0143] In the combination method, preprocessing step 7110 begins by combining the combined I and Q signals into a combined motion signal c using appropriate geometry. In one embodiment, the combination sub-step comprises three stages applicable to a window sliding along the I and Q signals (e.g., processing the amount of signal data (window size) progressively over time). In one embodiment, the window is a 10-second duration with 50% overlap.

[0144] a. Use correlation to check if the signal is 180 degrees out of phase, and if so, turn it back to the same quadrant.

[0145] b. When the vector (I, Q) forms a point cloud patch around the quasi-circular arc, subtract the average value of the point cloud up to the arc center at (0, 0), and set a minimum value m in both directions. IQ The center point cloud, and calculate with respect to m IQ The length m of each vector (I, Q) is .

[0146] m IQ =(m I m Q )=(min[I- ],min[Q- <q>(Equation 4)

[0147]

[0148] c. Subtract the average value of m to generate a (one-dimensional) combined signal c.

[0149] c = m - <m>(Equation 6)

[0150] The combined motion signal c is then (selectively) de-trending to remove baseline drift. In one implementation, de-trending is performed using a third-order polynomial:

[0151] c1 = DT poly，3 [c] (Equation 7)

[0152] In another implementation, a dual-pass median filter is used to perform detrending.

[0153] The detrending signal c1 is a bandpass (selectively) filtered using a Butterworth passband filter, the range of which is set to the frequency range of respiratory operation, which is implemented in one embodiment [0.1 Hz, 0.8 Hz] (corresponding to 6 to 48 breaths per minute).

[0154] A further (optional) sub-step of preprocessing step 7110 is noise reduction. In one embodiment, particularly suitable for signals from a Doppler radio frequency (RF) motion sensor 7010 (which is non-stationary), the noise reduction sub-step is implemented in the wavelet transform domain for de-stressing the combined motion signal c2 (bandpass filtering):

[0155] c3 = W -1 MWc2 (Equation 8)

[0156] Where W represents wavelet transform, such as the 30-coefficient "symmlet" wavelet reaching the fifth double-valued level, and M is a masking matrix that passes through specific wavelet coefficients and rejects other wavelet coefficients considered "disruptive".

[0157] The steps for performing action M are as follows:

[0158] a. Select wavelet coefficients whose "pseudo-signal" (see below) exceeds the first critical value T. A A two-valued proportional quantity;

[0159] b. From this set of two-valued proportions, perform hard-critical evaluation of the wavelet coefficients based on the standard deviation (with a critical value T). c ).

[0160] A certain proportion of "pseudosignal" determines the extent to which the pseudosignal affects the signal at that proportion. The pseudosignal can include a slope measurement of a signal that is unlikely to have high amplitude values. The pseudosignal of signal x can be calculated using the following formula:

[0161]

[0162] Where σ x It is the standard deviation of the signal x. Furthermore, Art(x) starts from 1, and the larger the artifact signal, the greater the deviation.

[0163] In parallel mode, preprocessing step 7110 omits the combination sub-step and performs any or all of the subsequent sub-steps (detrending, filtering, and noise reduction) in parallel for each of the I and Q signals.

[0164] In single-channel mode, any or all of the detrending, filtering, and noise reduction sub-steps are performed on the motion signal 7003.

[0165] In the following description, the input to feature extraction step 7120 is referred to as the (preprocessed) motion signal to reflect the selective nature of preprocessing step 7110.

[0166] Respiratory feature extraction

[0167] Figure 7D This illustration shows one form of the use of this technology for implementation. Figure 7C The block diagram of method 7200 for feature extraction step 7120 of the method shown.

[0168] In method 7200, the activity assessment and motion detection module 7210 generates an activity count signal and a series of motion flags from the (preprocessed) motion signal. (In combined or single-channel mode, this is only the (preprocessed) motion signal.) The positive / negative detection module 7220 generates a series of positive / negative flags and a series of motion flags from the (preprocessed) motion signal. The sleep / wake analysis module 7230 calculates a sleep graph from the positive / negative flag series, the series of motion flags, and the activity count signal. The respiratory rate assessment module 7240 generates a series of assessments of the patient's respiratory rate from the (preprocessed) motion signal and the sleep graph. The signal selection module 7250 uses the series of motion flags and the sleep graph to select portions of the (preprocessed) motion signal.

[0169] The modulation cycle metric calculation module 7255 generates an assessment of the modulation cycle length of the patient's breathing from a selected portion of the (preprocessed) motion signal. The packet generation module 7260 uses the assessed respiratory rate to generate a packet from the selected portion of the (preprocessed) motion signal. The sleep apnea (SDB) event detection module 7265 uses the assessed modulation cycle length to generate a desired SDB event from the selected portion of the (preprocessed) motion signal. The sleep apnea (SDB) event confirmation module 7270 uses the assessed modulation cycle length to determine the desired SDB event generated by the sleep apnea (SDB) event detection module 7265. Finally, the feature calculation module 7280 calculates respiratory feature values ​​from the confirmed SDB events.

[0170] In parallel mode, modules 7210 to 7270 of method 7200 are simply copied to independently process the two (preprocessed) motion signals 7003. A revised version of feature calculation module 7280 combines SDB events from the two parallel processing streams to calculate a single respiratory feature set of the two (preprocessed) motion signals.

[0171] Modules 7210 to 7280 of method 7200 have been described in detail by ResMed Sensor Technologies Limited in co-pending PCT application No. PCT / AU2013 / 000564, entitled "Methods and apparatus for monitoring cardiopulmonary health", the entire contents of which are incorporated herein by reference.

[0172] In one implementation, feature extraction step 7120 extracts four respiratory features for each monitoring phase:

[0173] • Total number of SDB events;

[0174] • The 50th percentile (median) of respiratory rate;

[0175] • The 75th percentile of respiratory rate; and

[0176] • 75th percentile of the duration of the Chen-Sch respiratory (CSR) cycle.

[0177] Calculate stability measurement

[0178] Clinical event prediction from respiratory characteristics is an example of highly imbalanced data sets with a very small number of events in many stable phases, necessitating a robust approach to minimize the number of false positive predictions. The assumption under this technique is that when a patient is stable, respiratory characteristics follow a statistical distribution, and at certain points before clinical events, through "points of change," they follow a different distribution. The stability measure is therefore calculated such that a change in distribution at the t-indicating monitoring phase results in an increase in the stability measure at or near the t-indicating monitoring phase. In other words, the stability measure for a monitoring phase is an indication of points of change in the respiratory characteristic distribution that have occurred at that phase. Step 7130 of this technique is therefore inherently distribution-based. Compared to a classification-based approach, a distribution-based approach to calculating stability measures results in low false positives (high specificity).

[0179] In step 7130, two methods for calculating the stability measurement are described below. In principle, any change in the respiratory characteristic distribution can be detected online or continuously. That is, sufficient data can be obtained after monitoring phase t is completed to calculate the stability measurement at monitoring phase t. The online method is best suited for improving or avoiding acute clinical events, i.e., the one-week delay between the point of change in the distribution and the occurrence of a clinical event.

[0180] Retrospective approaches can detect changes in the distribution of respiratory characteristics over a time delay of one to two weeks, depending on the parameters selected. Retrospective approaches are best suited for improving or preventing precipitating clinical events, which are defined as a two- to three-week delay between the point of change in the distribution and the occurrence of a clinical event.

[0181] Each method generates and analyzes a continuous value or time series of a sampled respiratory feature, denoted by the number of monitoring phases t (integer), {y(t)} or {y... t The (integer) monitoring phase designation t is sometimes simplified to "time t". It should be understood that time is measured in units of the monitoring phase.

[0182] Online or in sequence

[0183] The online method is based on Bayesian Online Change Point Detection (BOPCD). The online method includes a quantity called the travel length r at time t, which uses r... t It is expressed as the number of samplings y from the last point of change in the input sampling distribution. t If there are all samples y with an upper limit of time t and including time t. t In the online mode, step 7130 calculates the travel length r at time t. t The posterior distribution of . These samples can be taken using y 1：t To simplify the representation, the posterior travel length distribution to be calculated can be represented as p(r t |y 1：t This is represented by ) . It belongs to the current series of samples y t-rt+1：t The simplified form yt can be used. (r) express.

[0184] Posterior travel length distribution p(r) t |y 1：t This can be achieved by adjusting the stroke length r. t Joint probability p(r) t y 1：t )Normalization calculation:

[0185]

[0186] Joint probability p(r) t y 1：t (Simplified to γ) t ) can be achieved by using γ t Represented as p(r) t r t-1 y 1：t And from its previous value γ t-1 To calculate, it can be expanded to r t-1 Marginalization:

[0187]

[0188] The first factor p(r) in the summation term of Equation 11 t |r t-1 The prior condition is the length of the journey, also known as the change point prior. In one implementation, the change point prior p(r) is... t |r t-1 There are only two non-zero values:

[0189]

[0190] The first non-zero value (for r) t =0) is in the length of the journey r t-1 The probability H(r) of the subsequent change point t-1 +1), and another non-zero value (for r) t =r t-1 +1) is its complement, r t The probability is greater than r t-1 The probability, i.e., the current travel length for each stage of the extension. Function H(r) t-1 +1) is called the "hazard function". In one implementation, the hazard function H(r) is... t-1 +1) This can be set to a predetermined constant value h (known as the "hazard rate"), which is related to r. t-1 This is unrelated (so-called "memoryless" process), resulting in a geometric distribution of travel length with a time ratio of 1 / h. In one implementation, the hazard rate h is set to 1 / 90, meaning there is a point of variation in the time ratio every 90 monitoring phases.

[0191] The second factor P(y) in the summation term of Equation 11 t |r t-1 y 1：t-1 The given value is the posterior predicted probability, because it is the probability of observing the current sample y. t The probability that if all previously sampled y... 1：t-1 and the previous journey length r t-1 Posterior prediction probability P(y) t |r t-1 y 1：t-1 It can be abbreviated as Because it only depends on previous samples belonging to the current travel length.

[0192] The basic prediction model (UPM) is used to calculate the posterior prediction probability. Time series {y t The model is based on autonomous and identically distributed Gaussian sampling. In one online implementation, the basic prediction model is based on Gaussian sampling.

[0193] y t ~N(μ, σ 2 (Equation 13)

[0194] Where the mean and variance σ 2 It will vary at each point of change. In one implementation, the mean and variance σ... 2 Derived from the normal and inverse gamma distributions respectively:

[0195]

[0196] σ -2 ~Γ(α,β) (Equation 14)

[0197] Therefore, this basic prediction model is called the Normal Inverse Gamma (NIG) model. The parameters μ0, K0, α0, and β0 of the Normal Inverse Gamma (NIG) model are determined by adapting the Normal Inverse Gamma model to the training data set.

[0198] In order to calculate the posterior prediction probability at time t, the parameters of the basic prediction model are first updated for all time periods from 1 to the present time t:

[0199]

[0200] κ 1：t =[κ0 1+κ 1：t-1 ]

[0201] α 1：t =[α0 α 1：t-1 +0.5]

[0202]

[0203] The variance σ of the normal inverse gamma model 2 Then calculate according to the following formula:

[0204]

[0205] Finally, the posterior predicted probability Calculate using the following formula:

[0206]

[0207] Additionally, if all samples are taken... 1：t The posterior travel length distribution p(r) at time t t |y 1：t This can be used to calculate the next sample y. t+1 Marginal prediction distribution p(y) t+1 |y 1：t Marginal prediction distribution p(y) t+1 |y 1：t One method to calculate this is by using the travel length r t Posterior prediction probability Marginalization to maintain stroke length r t Uncertainty:

[0208]

[0209] For the travel length r t Posterior predicted probability for each value It can be calculated from the basic prediction model as described above.

[0210] Stability measurement in online mode S t It can be calculated using the probability at time t, which is the time since the last change point was detected. This probability is derived from the posterior travel length distribution p(r) of all possible travel lengths since the time since the detection of the previous change point. t |y 1：t The sum is calculated as the sum of the values ​​of ). That is, the sum is the travel length r from 0 to the current time t, which is less than the time of the previous point of change. t Calculate using all values:

[0211]

[0212] The previous change point before time t was detected at time t0 (the previous warning time t0 was initialized to zero before any change point was detected).

[0213] In one form of this technology, as described below, based on a clinician's respiratory characteristic test received or calculated by an external computing device 7005, the clinician can issue a manual alert to the monitoring device 7000 through the user interface of the associated external computing device 7005. If such a manual alert is issued, the value of the previous alert time t0 is updated to the time when the manual alert was issued.

[0214] Figure 7E Including illustrations of one form of the present technology used for implementation in an online manner. Figure 7C The flowchart of method 7300 is shown as step 7130 of the stability measurement calculation in method 7100. Method 7300 is repeated (executed once) after each monitoring phase.

[0215] Method 7300 begins with step 7310, where the joint probability p(r) t y 1：t (that is, γ) t In one implementation, the process is initialized (i.e., the specified value at t=0) to 1. Step 7310 is only implemented in the first repetition of method 7300, therefore it is... Figure 7E As shown by the dashed line. In step 7320, the current time t will increment, and the current sample y will be received. t Next is step 7330, where method 7300 uses the basic prediction model (UPM) to calculate the current posterior prediction probability. And use equations 15 to 17 to calculate the current sample y. t Then step 7340 uses Equation 11 to obtain the previously combined probability γ. t-1 and current posterior predicted probability To calculate the current joint probability p(r) t y 1：t )=γ t For the travel length r t For a sum equal to zero, there exists a term t in Equation 11. However, for a travel length r greater than zero... t Regarding the value of , there is only one term in the sum of Equation 11, because only r... t-1 A value of r, where the prior at the point of change (Equation 12) is nonzero, i.e., r. t-1 =r t-1 This actually improves computational efficiency for online methods.

[0216] In the next step 7350, method 7300 makes the current joint probability p(r) as shown in Equation 10. t y 1：t Normalization is used to calculate the current posterior travel length distribution p(r) t |y 1：t Step 7360 then uses Equation 19 to obtain the current posterior travel length distribution p(r) t |y 1：t To calculate the current stability measurement S t Method 7300 then ends.

[0217] Traceability

[0218] The retrospective approach in step 7130 operates by comparing the probability distributions of subsequences of the time series {y(t)} before and after a specific time. Step 7130 under the retrospective approach is a stability measure when calculating the dissimilarity between the two distributions.

[0219] Suppose that the subsequence Y(t) of length k of the time series {y(t)} is defined as follows:

[0220]

[0221] Where k is a parameter of the tracing method. Each subsequence Y(t) is treated as a k-component vector, which is a sample from the basic k-dimensional joint distribution. The set Ψ(t) of n consecutive subsequences Y(t) at time t is defined as follows:

[0222]

[0223] Where n is a further parameter of the tracing method. The probability distribution of the set Ψ(t) of n subsequences Y(t) can be obtained using P. t To express.

[0224] The tracing method calculates the distribution P of the set Ψ(t+n) of n subsequences. t and distribution P t+n Symmetrical dissimilarity between D S Sampling n is slower than sampling the set Ψ(t). In one implementation, the symmetric dissimilarity D S Measure D using the dissimilarity between two distributions P and P', whose divergence is known to be f. f (P||P'), and defined as follows:

[0225]

[0226] Where f is a convex function such that f(1) = 0, and p(Y) and P′(Y) are the probability density functions (densities) of distributions p and p′, respectively. Symmetrical dissimilarity D S Distribution P can be used t and allocation P t+n The divergence between f and D f (P t ||P t+n ), and distribution P t+n and distribution P t The divergence between f and D f (P t+n ||P t Calculate by the sum of )

[0227] D s (P t ||P t+n ) = D f (P t ||P t+n )+D f (P t+n ||P t (Equation 23)

[0228] Symmetrical dissimilarity D S This is used to represent it, because P t and P t+n Symmetrical dissimilarity between D S It is the same as P t+n and P t Symmetrical dissimilarity between D S In general, the f-divergence D of equation 22 f In this sense, it is not symmetrical. When P t and P t+n The divergence between f and D f or P t+n and P t The divergence between f and D f When the value is high, the symmetry dissimilarity D calculated according to Equation 23 is... S High values ​​are used. Symmetrical anisotropy D S Therefore, the sensitivity at the point of change exceeds that at P. t and P t+n The divergence between f and D f (Alone) or in P t+n and P t The divergence between f and D f (Alone). P t and P t+n Symmetrical dissimilarity between D S Therefore, it can be used as a stability measure S of patient 1000 at time t+n. t+n To use. In P t (The probability distribution of the set Ψ(t), which includes a subsequence Y(t) consisting of samples of the time series {y(t)} before the monitoring phase t+n), and P t+n The symmetric dissimilarity D among the probability distributions of the set Ψ(t+n), which includes the subsequence Y(t+n) consisting of samples of the time series {y(t)} after the monitoring phase Ψ(t+n). S A higher value indicates whether the change point of the time series {y(t)} may be between the monitoring stage (t+n-1) and (t+n).

[0229] The stability measurement S at monitoring stage t+n is evaluated using the definition in Equation 23. t+n A sample y(t) needs to be taken from a time t+2n+k-2 that includes the time t+2n+k-2. In other words, when sample y(T) is received at time T, the stability measurement S according to the retrospective method can be calculated at time Tn-k+2. Therefore, according to the retrospective method, at time T, the increase in the stability measurement S calculated using the most recent sample y(T) indicates that the point of change occurred approximately n+k-2 monitoring stages prior to time T. The retrospective method can therefore be viewed as a delay with n+k-2 samples.

[0230] In one implementation, the convex function f used in the definition of the divergence (Equation 22) is the Kullback-Liebler divergence defined as f(t) = t log(t). In another implementation, the convex function f is the Pearson divergence defined as a quadratic function.

[0231]

[0232] Replacing the Pearson divergence f(t) in Equation 24 with Equation 22 provides the Pearson dissimilarity D. PE .

[0233] Due to distribution P t and P t+n The densities p(Y) and p'(Y) are unknown, making it impossible to directly calculate the symmetry dissimilarity D. s One implementation of the retrospective approach uses conventional methods to evaluate the densities p(Y) and p'(Y) from the sets Ψ(t) and Ψ(t+n) respectively, and then applies equations 22 and 23 to calculate the symmetry dissimilarity D from the evaluated densities. s However, conventional density assessment methods tend to become less accurate as the dimension (k in this case) increases.

[0234] Alternative implementations of the traceability method can assess the ratio between densities p(Y) and p'(Y). Compared to accuracy, density ratio assessment is easier for assessing individual densities p(Y) and p'(Y).

[0235] Density ratio g(Y) ＝p The approximate value of (Y) / p'(Y) can be obtained by weighted summation of the core basis functions:

[0236]

[0237] The core basis function K is a Gaussian function:

[0238]

[0239] The core width σ is determined based on cross-validation, and the weighted OR coefficient θ l These are elements of the parameter vector θ. Core center Y l (l=1,...,n) are n subsequences Y(t),...,Y(t+n-1) that constitute the set Ψ(t).

[0240] The optimal parameter vector for the density ratio g(Y) given in Equation 25 Under squared loss, the true density ratio can be adjusted by using an approximation value g. And it was found that this is equivalent to minimizing the following objective function O with respect to the parameter vector θ:

[0241]

[0242] Where H is an n-by-n matrix of the (l, l')th element H(l, l') given by the following formula.

[0243]

[0244] Where Y' j (j=1,...,n) are n subsequences Y(t+n,...,Y(t+2n-1) that constitute the set Ψ(t+n).

[0245] Vector h is an n-vector of element v, given by the following formula.

[0246]

[0247] The last term in the objective function of Equation 27 is the penalty term, which includes the regularization objective, where λ is used as the regularization parameter.

[0248] The objective function in Equation 27 is the parameter vector given by the following formula. And minimize:

[0249]

[0250] Pearson dissimilarity D in Equation 22 OE (P t ||P t+n An approximate value can be obtained using the following formula:

[0251]

[0252] Pearson anisotropy D PE (P t+n ||P t This can be achieved by swapping subsequences Y. i (i=1,...,n) and Y' j (j = 1, ..., n) and adopting a similar approach An approximate value is obtained using this method. The resulting approximate value... Then add an approximation. To obtain symmetric dissimilarity D s (P t ||P t+n ), which is the stability measurement S under traceability mode. t+n .

[0253] Figure 7F Including illustrations of one form of the present technology for implementation in a traceability manner Figure 7C The flowchart of method 7400 shows the stability measurement calculation step 7130 of method 7100.

[0254] Method 7400 begins with step 7410, which, according to equation 21, forms two sets of subsequences Y that constitute sets Ψ(t) and Ψ(t+n) respectively. i (i=1,...,n) and Y' j (j = 1, ..., n). Because of the definition of these subsequences, method 7400 is implemented at time t + 2n + k - 2 or thereafter. Following step 7420, the matrix H and vector h are computed using equations 28 and 29, along with the core definition equation 26. In the next step 7430, equation 30 is used to compute the parameter vector. Following step 7440, where method 7400 utilizes equations 31 and 25 using parameter vectors. To calculate Pearson dissimilarity D PE (P t ||P t+n The approximate value of ).

[0255] In step 7450, subsequence Y i (i=1,...,n) and Y' j (j = 1, ..., n) are swapped. Steps 7460, 7470, and 7480 involve swapping the subsequence Y. j (j = 1, ..., n) and Y' i (i = 1, ..., n) Repeat steps 7420, 7430, and 440 to obtain the Pearson dissimilarity D. PE (P t+n ||P t An approximation of ). Finally, in step 7490, D PE (P t ||P t+n ) and D PE (P t+n ||P t The approximate values ​​of ) are added together to obtain symmetric dissimilarity D. S (P t ||P t+n ), which is the stability measurement S at time t+n according to the traceback method. t+n .

[0256] In one implementation of method 7400, parameters n and k are 10 and 5, respectively, so the delay of the tracing method is n + k - 2 = 13 samples. The core width σ and the regularization parameter λ are obtained using complex kernel regularization, because the values ​​come from two discontinuous possible sets, which make the result for the subsequence Y i and Y' j The objective function in Equation 27 is minimized for multiple arbitrarily selected subsets of the sets Ψ(t) and Ψ(t+n).

[0257] Combination method

[0258] Step 7130 combines the online stability measurement and traceability stability measurement as described above to generate an alert. In one embodiment of the combination, the alert is generated when the online stability measurement meets a first criterion (e.g., one or more critical value comparisons) and the traceability stability measurement also meets a second criterion (e.g., one or more critical value comparisons). In another embodiment of the combination, the alert is generated when either the online stability measurement meets the first criterion (e.g., one or more critical value comparisons) or the traceability stability measurement also meets the second criterion (e.g., one or more critical value comparisons).

[0259] Warning generated

[0260] The clinical alert generated in step 7150 may include warning or alert messages in many forms. For example, processor 7006, used to generate a clinical alert for patient 1000, may activate a status light (e.g., an LED or an icon on display device 7015) on monitoring device 7000. More detailed information regarding the assessment of the indicator may also be displayed to patient 1000 on display device 7015. Where appropriate, processor 7006 may also (or alternatively) transmit the alert message to an external computing device 7005 of the relevant clinician via connection 7008. This message may be in the form of wired or wireless communication. For example, processor 7006 may generate the alert message via a paging system, such as by automatically dialing a paging system. Processor 7006 may also be configured to generate an automated voice telephone message. Processor 7006 may also transmit the alert message via fax transmission. In some specific embodiments, processor 7006 may also transmit the alert message via Internet transport protocols (such as email messages) or via any other Internet data file transfer protocol. The alert message may even be encrypted to maintain patient information privacy. Typical alert messages may identify the patient. This message may also include data recorded by the monitoring device 7000, or any other recorded patient data. Depending on the circumstances, in some specific embodiments, the alert message may even indicate whether the patient requires additional treatment, hospitalization, or evaluation due to the detection of a potential clinical event.

[0261] While alert messages can be transmitted to the patient by the processor 7006 via the display device 7015 of the monitoring device 7000 and to the clinician via the connection 7008, in some specific embodiments, the alert messages can be transmitted more selectively. For example, if no alert is displayed on the display device 7015, the first alert message can be transmitted only to the clinician via the connection 7008, transmitting the alert message only to the external computing device 7005. However, a second alert message, which may be more urgent, can then be transmitted to the external computing device 7005 while simultaneously being actively displayed on the display device 7015. Selective audible alerts controlled by the processor 7006 can also be implemented. The audible alerts can be used depending on the urgency of the alert message.

[0262] In one form of this technology, a clinician can issue a manual alert to the monitoring device 7000 through a user interface of the external computing device 7005 based on a clinician examination of respiratory characteristics received on the external computing device 7005.

[0263] ask

[0264] In another form of this technology, processor 7006 can limit alerts based on the patient's response to inquiries, which can be used to avoid unnecessary alerts. In a variation of method 7100, instead of immediately generating an alert upon meeting a stability measurement criterion (step 7140), as in step 7150, processor 7006 can prompt the patient 1000 to take action, such as taking a prescription medication, or to provide a response to an inquiry. Display device 7015, under the control of processor 7006, can ask the patient 1000 inquiries, prompting the patient 1000 to input a response via a user interface. The questions or inquiries can be selected from a database or other question data structure, such as a data structure in the memory 7002 of monitoring device 7000. Processor 7006 can then evaluate the response to the inquiry. Based on this evaluation, processor 7006 can generate an alert (as shown in step 7150), abort an alert, and / or delay the generation of an alert in response to one or more additional inquiries. Such additional inquiries can be triggered after a specific time, after further detection of a change point, or after further use of monitoring device 7000. In the form of this technology, where step 7150 includes transmitting an alert message to the clinician to the external computing device 7005, the received query response can instead be forwarded to the external computing device 7005 for manual evaluation by the clinician. The clinician can then decide whether to maintain or cancel the alert based on their evaluation.

[0265] Such inquiries can be used to reduce false positives (e.g., when a clinical intervention and patient alert are raised, but it is later found that the intervention was unnecessary). Some false positives may be due to changes in patient behavior that may not be corrected by clinical intervention. This behavior may include missing or incorrect medication dosages, failure to follow dietary recommendations and / or rest needs, etc. Inquiry questions may be based on patient requests for medication and / or lifestyle compliance (e.g., getting patients to take prescription medications and / or follow physician treatment recommendations, etc.). Where appropriate, in some cases, one or more questions may address the operational integrity of the monitoring device 7000 to ensure the validity of received respiratory signals. Where appropriate, the processor 7006 may conduct a series of inquiries within a predetermined time frame (e.g., one or more monitoring phases), and generate an alert only after the predetermined time frame has elapsed.

[0266] In an online mode, if the processor 7006 aborts the alert response at the point of change detected, or if the clinician subsequently manually cancels the generated alert, the processor 7006 can restore the previous alert time t0 (used in Equation 19) to the second-to-last time of the alert that was generated.

[0267] Example Results

[0268] Figure 8 This includes a graph 8000 showing example results obtained from monitoring device 7000 using method 7100. The upper waveform 8010 shows one of the aforementioned respiratory characteristics, namely the 75th percentile of respiratory rate at stage t (over 400 stages). The gray band 8015 shows 28 symmetrical stage intervals around the ADHF event experienced by the patient at approximately stage number 182 (indicated by upward arrow 8020). The lower waveform 8050 shows the peak at stage t, where the stability measurement S calculated retrospectively... t The threshold value was exceeded, so step 7150 generates an alert. Specifically, the double peak 8060 matches the ADHF event. Other peaks (e.g., 8070) do not match the ADHF event and therefore represent "false positives".

[0269] Vocabulary

[0270] For the purposes of this technical disclosure, one or more of the following definitions may be applied in certain forms of this technology. Other definitions may be applied in other forms of this technology.

[0271] General Rules

[0272] Air: In certain forms of this technology, air means the atmosphere, and in other forms of this technology, air means certain other combinations of breathable gases, such as an oxygen-rich atmosphere.

[0273] Continuous Positive Airway Pressure (CPAP): CPAP therapy refers to the application of a continuous positive pressure supply of air relative to atmospheric pressure to the airway inlet, preferably remaining substantially constant throughout the patient's respiratory cycle. In some forms, the pressure at the airway inlet will be slightly higher during expiration and slightly lower during inspiration. In other forms, the pressure will vary between different respiratory cycles, for example, increasing with indications of local upper airway obstruction and decreasing with the absence of indications of upper airway obstruction.

[0274] Aspects of the respiratory cycle

[0275] Respiratory arrest: Preferably, respiratory arrest occurs when the flow rate falls below a predetermined threshold for a period of time, such as 10 seconds. Interference-related respiratory arrest occurs when, despite the patient's efforts, some obstruction in the airway prevents airflow. Central respiratory arrest occurs when respiratory arrest is detected despite a patent airway, due to reduced or absent respiratory effort. Mixed respiratory arrest occurs when reduced or absent respiratory effort coincides with an obstructive airway.

[0276] Respiratory rate: The patient's natural respiratory rate, usually measured in breaths per minute.

[0277] Working cycle: The ratio of inhalation time Ti to total breathing time Ttot.

[0278] Effort (breathing): Preferably, the effort exerted by breathing should be the amount of work done by a natural breather attempting to breathe.

[0279] The expiratory phase of the respiratory cycle: the period from the start of the expiratory flow to the start of the inspiratory flow.

[0280] Flow rate limitation: Preferably, flow rate limitation is a state that affects the patient's breathing, wherein greater effort by the patient does not cause a relative increase in flow rate. When flow rate limitation occurs during the inspiratory portion of the respiratory cycle, it is described as inspiratory flow rate limitation. When flow rate limitation occurs during the expiratory portion of the respiratory cycle, it is described as expiratory flow rate limitation.

[0281] Shallow breathing: Preferably, shallow breathing reduces flow but does not stop flow. In one form, shallow breathing should occur when the flow rate is reduced below a critical value for persistence. Central shallow breathing should occur when shallow breathing is detected due to a reduction in respiratory effort.

[0282] Rapid breathing: The flow rate increases to a level higher than normal.

[0283] Inspiratory phase of the respiratory cycle: Preferably, the period from the start of the inspiratory flow to the start of the expiratory flow should be the inspiratory phase of the respiratory cycle.

[0284] Airway patency: The degree to which the airway is open, or the extent to which the airway is open. An open airway is one that is open. Airway patency can be quantified, for example, a value of one (1) indicates patency, while a value of zero (0) indicates closure (obstruction).

[0285] Positive end-expiratory pressure ventilation: The pressure inside the lungs exceeds atmospheric pressure at the end of expiration.

[0286] Peak flow rate (Q peak): The maximum flow rate during the inspiratory portion of the respiratory flow waveform.

[0287] Respiratory flow rate, airflow, patient airflow, and respiratory airflow (Qr): These synonyms can be viewed as the RPT device's assessment of respiratory airflow, rather than "actual respiratory flow rate" or "actual respiratory airflow," which is the correct respiratory flow rate experienced by the patient, usually expressed in liters per minute.

[0288] Tidal volume (Vt): The amount of air inhaled or exhaled during normal breathing without additional effort.

[0289] (Inspiratory) time (Ti): The duration of the inspiratory portion of the respiratory flow waveform.

[0290] (Exhalation) Time (Te): The duration of the expiratory portion of the respiratory flow waveform.

[0291] Total Time (Ttot): The total duration between the start of the inspiratory portion of one respiratory flow waveform and the start of the inspiratory portion of the next respiratory flow waveform.

[0292] Typical recent ventilation: ventilation values ​​where the recent values ​​tend to be concentrated in certain predetermined time periods, i.e., the median number of recent ventilation values.

[0293] Upper airway obstruction (UAO): This includes partial and total upper airway obstruction. This may be a state of flow restriction, where the flow rate is only slightly increased, or may even decrease when the pressure gradient across the upper airway increases (Starling resistance behavior).

[0294] Ventilation: A measurement of the total amount of gases exchanged by a patient's respiratory system, including the inspiratory and expiratory flow rates per unit time. When expressed as ventilation per minute, this amount is often referred to as "minute ventilation." Minute ventilation is sometimes expressed simply as ventilation volume, i.e., ventilation per minute.

[0295] Parameters of Positive Airway Pressure (RPT) device

[0296] Flow rate (or velocity): The instantaneous volume (or mass) of air transferred per unit time. Flow rate is the flow rate measured over a very short time when the velocity has the same volume or mass as ventilation per unit time. In some cases, the reference flow rate can be considered a pure quantity, i.e., a quantity with only magnitude. In other cases, the flow rate can be considered a vector reference, i.e., a quantity with both magnitude and direction. In the case of symbolic values, the flow rate during the inspiratory portion of the patient's respiratory cycle is a nominal positive value; therefore, the flow rate during the expiratory portion of the patient's respiratory cycle is a negative value. Flow rate is denoted by the symbol Q. Total flow rate Qt is the flow rate of air leaving the RPT device. Ventilation port flow rate Qv is the flow rate of air leaving the ventilation port to expel exhaled gas. Leakage flow rate Ql is the unintentional leakage flow rate from the patient interface system. Respiratory flow rate Qr is the flow rate of air entering the patient's respiratory system.

[0297] Leakage: Preferably, the term leakage is used to refer to air flowing into the environment. Leakage is intentional (e.g.) to allow the exhaled CO2 to escape. Leakage is unintentional, for example, and therefore, it is an incomplete seal between the mask and the patient's face. In one instance, leakage may occur in the rotary joint.

[0298] Other supplementary notes

[0299] The disclosure of this patent document includes material subject to copyright protection. The copyright holder has no objection to the reproduction of any part of the patent document or patent disclosure, as it appears in the patent archives or records of the Patent and Trademark Office, but retains all copyright in all other respects.

[0300] Unless otherwise specified and provided in this specification, it should be understood that each intermediate value (to one-tenth of the lower limit unit) between the upper and lower limits of the range, and any other specified or intermediate value within the range, are included in this technique. The upper and lower limits of these intermediate ranges (which may be included independently in the intermediate range) are also included in this technique (subject to any specific exclusions within the range). Where the range includes one or both limitations, the ranges excluding either or both of those limitations are also included in this technique.

[0301] Furthermore, where one or more values ​​described herein are implemented as part of this technology, it should be understood that, unless otherwise stated, these values ​​may be approximate and may be used with any appropriate significant figures permitted or required by the actual technical implementation.

[0302] 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. While similar or equivalent methods and materials described herein may be used to practice or experiment with this technique, a limited number of exemplary methods and materials are described herein.

[0303] When a particular material is considered preferred for constructing a component, it is obvious that alternative materials with similar properties may be used instead. Furthermore, unless otherwise specified herein, any and all components described herein should be understood as manufacturable, both together and individually.

[0304] It must be noted that, as used herein and in the appended claims, unless expressly stated otherwise, the singular forms “a”, “an” and “said” include a plurality of their equivalents.

[0305] All publications mentioned herein are incorporated by reference into the disclosures and describe the methods and / or materials of the subject matter of those patents. The publications discussed herein are provided only for disclosures prior to the filing date of this application. They should not be construed as an admission that the technology did not precede these publications simply because it was a prior invention. Furthermore, the publication dates provided may differ from the actual publication dates and may require separate verification.

[0306] Furthermore, in interpreting this disclosure, all terms should be interpreted in the manner most reasonably appropriate to the context. In particular, the term "comprising" should be interpreted in a non-exclusive manner as referring to an element, component, or step, indicating that the referenced element, component, or step may be presented, utilized, or combined with other elements, components, or steps not expressly referenced.

[0307] The headings used in the detailed description are included only for the reader's convenience and should not be used to limit the subject matter used in the disclosure or claims. The headings should not be used to constitute a limitation of the claims or the scope of the claims.

[0308] While this technology is described herein with reference to specific embodiments, it should be understood that these embodiments are merely illustrative of the principles and applications of this technology. In some cases, terms and symbols may denote specific details that are not required to implement this technology. For example, although the terms "first" and "second" may be used, they do not imply any order unless specifically specified, but are used to distinguish different elements. Furthermore, although the processing steps in a method may be described or illustrated in a certain order, this order is not necessary. Those skilled in the art will understand that this order can be modified and / or aspects of it can be performed simultaneously or even concurrently.

[0309] Therefore, it should be understood that many modifications can be made to the illustrative embodiments and other arrangements can be designed without departing from the spirit and scope of this technology.

[0310] Component Symbol Explanation

[0311] 1000 patients

[0312] 1200 sensor units

[0313] 3000 Patient Interface

[0314] 3100 Sealing Formation Structure

[0315] 3200 Inflation Chamber

[0316] 3300 stable structure

[0317] 3400 ventilation port

[0318] 3600 connection port

[0319] 3700 Forehead Support

[0320] 4000 Respiratory Pressure Therapy (RPT) Device

[0321] 4010 Outer Shell

[0322] 4012 upper part

[0323] 4014 lower part

[0324] 4015 Panel

[0325] 4016 chassis

[0326] 4018 handle

[0327] 4020 Pneumatic Block

[0328] 4100 Pneumatic Components

[0329] 4112 Inlet Filter

[0330] 4122 Imported Muffler

[0331] 4124 Export Silencer

[0332] 4140 Pressure Generator

[0333] 4142 Blower

[0334] 4170 Air Loop

[0335] 4200 Electrical Components

[0336] 4202 Printed Circuit Board Assembly

[0337] 4210 Power Supply

[0338] 4220 Input Device

[0339] 4230 Central Controller

[0340] 4240 Treatment Device Controller

[0341] 4250 Protection Circuit

[0342] 4260 Memory

[0343] 4270 Converter

[0344] 4272 Pressure Sensor

[0345] 4274 Flow Sensor

[0346] 4280 Data Communication Interface

[0347] 4282 Remote External Communication Network

[0348] 4284 Local external communication network

[0349] 4286 Remote External Device

[0350] 4288 Local External Device

[0351] 4290 Output Device

[0352] 4292 Display Driver

[0353] 4294 Monitor

[0354] 4300 Algorithm

[0355] 5000 Humidifier

[0356] 7000 monitoring equipment

[0357] 7001 Microcontroller Unit

[0358] 7002 Memory

[0359] 7003 Motion Signal

[0360] 7004 Communication Circuit

[0361] 7005 External Computing Device

[0362] 7006 processor

[0363] 7008 connection

[0364] 7010 Motion Sensor

[0365] 7015 Display Device

[0366] 7017 Audio Output

[0367] 7020 transmitter

[0368] 7030 Receiver

[0369] 7040 Local Oscillator

[0370] 7050 antenna

[0371] 7060 radio frequency signal

[0372] 7070 reflected signal

[0373] 7080 Mixer

[0374] 7100 Monitoring Method

[0375] 7110 Steps

[0376] 7120 steps

[0377] 7130 Steps

[0378] 7140 steps

[0379] 7150 steps

[0380] 7160 steps

[0381] 7210 Motion Detection Module

[0382] 7220 Positive / Negative Detection Module

[0383] 7230 Sleep / Wakefulness Analysis Module

[0384] 7240 Respiratory Rate Assessment Module

[0385] 7250 Signal Selection Module

[0386] 7255 Modulation Period Metric Calculation Module

[0387] 7260 Packet Generation Module

[0388] 7265 Sleep Disorder Breathing (SDB) Event Detection Module

[0389] 7270 Sleep Disorder Breathing (SDB) Event Confirmation Module

[0390] 7280 Feature Calculation Module

[0391] 7300 Method

[0392] 7310 Steps

[0393] 7320 steps

[0394] 7330 Steps

[0395] 7340 steps

[0396] 7350 steps

[0397] 7360 steps

[0398] 7400 Method

[0399] 7410 Steps

[0400] 7420 steps

[0401] 7430 Steps

[0402] 7440 steps

[0403] 7450 steps

[0404] 7460 steps

[0405] 7470 steps

[0406] 7480 steps

[0407] 7490 steps

[0408] 8000 curve chart

[0409] 8010 waveform curve

[0410] 8015 Gray Band

[0411] 8020 Upward Arrow

[0412] 8050 waveform curve

[0413] 8060 peak

[0414] 8070 peak< / m> < / q>

Claims

1. A chronic disease monitoring device, comprising: A sensor configured to generate a respiratory signal indicating the patient's breathing during the monitoring phase; and The processor is configured as follows: For each of the multiple monitoring phases, respiratory features are extracted from the respiratory signals; as well as Calculate a stability measurement for the patient during the monitoring phase, the stability measurement representing the probability that a point of change in the statistical distribution of the respiratory characteristic has occurred during the monitoring phase, wherein the calculation includes: If there is an upper limit to the value of the respiratory feature of the monitoring phase and includes the value of the monitoring phase, then calculate the posterior distribution of the journey length of the monitoring phase; as well as Calculate the sum of the values ​​of the posterior distribution of the journey length.

2. The chronic disease monitoring device of claim 1, wherein the processor is further configured to control the generation of an alert based on a comparison of the stability measurement with a threshold value.

3. The chronic disease monitoring device according to claim 2, further comprising an external computing device, wherein generating an alert includes sending an alert message to the external computing device.

4. The chronic disease monitoring device according to any one of claims 1 to 3, wherein the sensor is a non-contact motion sensor, and the respiratory signal is a signal representing the patient's respiratory movements.

5. The chronic disease monitoring device according to any one of claims 1 to 3, wherein the device is a respiratory pressure therapy device, and the respiratory signal is a signal representing the respiratory flow rate of a patient from the stage of using the respiratory pressure therapy device.

6. A computer-readable storage medium comprising processor control instructions, which, when executed by a processor of a chronic disease monitoring device including sensors configured to generate respiratory signals indicative of patient respiration during a monitoring phase, cause the processor to: For each of the multiple monitoring phases, respiratory features are extracted from the respiratory signals; as well as Calculate a stability measurement for the patient during the monitoring phase, the stability measurement representing the probability that a point of change in the statistical distribution of the respiratory characteristic has occurred during the monitoring phase, wherein the calculation includes: If there is an upper limit to the value of the respiratory feature of the monitoring phase and includes the value of the monitoring phase, then calculate the posterior distribution of the journey length of the monitoring phase; as well as Calculate the sum of the values ​​of the posterior distribution of the journey length.

7. The computer-readable storage medium of claim 6, wherein the processor control instructions cause the processor to control the generation of an alert based on a comparison of the stability measurement with a threshold value.

8. The computer-readable storage medium of claim 7, wherein the processor control instructions cause the processor to send an alert message to an external computing device in order to generate the alert.

9. The computer-readable storage medium according to any one of claims 6 to 8, wherein the sensor is a non-contact motion sensor and the respiratory signal is a signal representing the patient's respiratory movements.

10. The computer-readable storage medium according to any one of claims 6 to 8, wherein the device is a respiratory pressure therapy device, and the respiratory signal is a signal representing the respiratory flow rate of a patient from the stage of using the respiratory pressure therapy device.

Images