Automated control for detection of flow limitation

Abstract

To provide methods and apparatus for the detection, diagnosis and / or treatment of respiratory conditions such as conditions related to sleep apnea hypopnea syndrome (OSAHS) or obstructive sleep apnea (OSA).SOLUTION: A respiratory flow limitation detection device, which can include an airway pressure treatment generator, determines a flow limitation measure based on one or more shape indices for detecting partial obstruction and a measure of a patient's ventilation or respiratory duty cycle. The shape indices are based on function(s) that ascertain the likelihood of the presence of M-shaped breathing patterns and / or chair-shaped breathing patterns. The measure of ventilation may be based on analysis of current and prior tidal volumes to detect less than normal patient ventilation. The duty cycle measure may be a ratio of current and prior measures of inspiratory time to respiratory cycle time to detect an increase in the patient's inspiratory cycle time relative to the respiratory cycle time.SELECTED DRAWING: Figure 5

Description

[Technical field]

[0001] [CROSS REFERENCE TO RELATED APPLICATIONS] This application was filed on May 11, 2007, the disclosure of which is incorporated herein by reference. Australian Provisional Patent Application No. 2007902561 filed on August 17, 2007 Benefit of the filing date of U.S. Provisional Patent Application No. 60 / 965,172 is claimed.

[0002] [Field of the Invention] The proposed technology is aimed at treating Obstructive Sleep Apnea Hypopnea Syndrome (OSAHS) or Obstructive Sleep Apnea Hypopnea Syndrome (OSAHS). For the detection, diagnosis, and / or treatment of respiratory conditions, including conditions related to obstructive sleep apnea (OSA) The present invention relates to a method and an apparatus. [Background technology]

[0003] Sullivan U.S. Patent No. 5,199,424 issued April 6, 1993 As described by Lynch et al., the application of continuous positive airway pressure (CPAP) may cause obstruction. It is used as a means of treating the occurrence of obstructive sleep apnea. Patients are fitted with a nasal mask or The prongs connect to a positive pressure air source. The air breathed by the patient is at atmospheric pressure. The application of continuous positive airway pressure supports and stabilizes the upper airway, thus preventing upper airway obstruction. It has been found that this results in what can be described as an "air splint" that eliminates the occurrence of blockages. The application of targeted positive airway pressure is effective in eliminating both snoring and obstructive sleep apnea, and many In many cases, it is effective in treating central and mixed apnea.

[0004] U.S. Patent No. 5,549, issued August 27, 1996 to Gruenke Issue 106 includes a device that adjusts a patient's breathing to treat mixed sleep apnea and obstructive sleep apnea. The device is designed to reduce the airway pressure of the patient just before inspiration. Increase the intranasal air pressure delivered to the nasal cavity and then decrease the pressure to ease the expiratory effort. Can.

[0005] In U.S. Pat. No. 5,245,995, Sullivan describes a pre-occlusion stroke or Snoring and abnormal breathing patterns that can be early indicators of other forms of respiratory disease We investigated how this could be detected by measuring inspiratory and expiratory pressure during sleep. In particular, patterns of respiratory parameters are monitored and CPAP pressure is adjusted to ideally meet the obstructive To provide increased airway pressure to eliminate the occurrence of stroke and other forms of respiratory illness. is raised upon detection of a predefined pattern.

[0006] Berthon U.S. Patent No. 5,704,345, issued January 6, 1998. - Detects abnormal breathing patterns indicative of obstructive breathing, as described by Jones Various techniques for detecting and detecting apnea are known. Berthon-Jones and describes a method based on detecting events such as respiratory flow flattening. The treatment pressure may be automatically adjusted in response to the detected condition.

[0007] International Patent Application PCT / AU01 / 01948 (Publication No. WO02 / 18002) As described by Wickham, the flow flattening decision is subject to various weighting factors. Weighting factors may be based on the following to increase sensitivity to different types of respiratory obstruction: Applied to the area of ​​airflow.

[0008] Other methods of detecting occlusions are also in use. See, e.g., U.S. Patent No. 5,490 In US Pat. Nos. 5,803,066 and 5,502, Rapoport was developed to treat a patient's respiratory tract. The air flow from the flow generator is adjusted while trying to ensure that no flow restrictions appear in the A method and apparatus for optimizing the controlled positive pressure to minimize Controlled positive pressure to the patient's airway is achieved by detecting flow limitation from the shape of the inspiratory flow waveform. The pressure setting is adjusted depending on whether a flow restriction is detected and whether the system is Thus, depending on the previous treatment, it may increase, decrease or remain the same.

[0009] In U.S. Patent No. 5,645,053, Remmers et al. A system that automatically and continuously adjusts the nasal pressure level to an optimal value during apnea treatment. The time profile of the inhalation flow, such as the degree of roundness and flatness of the inhalation profile, is described. Parameters related to the shape of the profile are determined. OSA treatment is then performed by adjusting the applied pressure. Automatically re-evaluates force and continuously adjusts to the minimum pressure required to adequately inflate the patient's pharyngeal airway. This is carried out by searching systematically.

[0010] Despite the availability of such devices to treat OSA, some sleep obstructive events remain It has not yet been treated by the use of certain devices. Therefore, automatic detection and New methods of treatment and management are desirable. Summary of the Invention

[0011] In accordance with aspects of the present technology, improved automated detection and / or treatment of sleep disordered breathing or flow limitation is provided. Apparatus and methods for providing automated therapy are provided.

[0012] In accordance with another aspect of the present technology, improved detection and / or analysis of flow waveforms indicative of partial occlusions is provided. Treatment is provided.

[0013] In accordance with yet another aspect of the present technology, a more rapid response to indications of partial occlusion or flow limitation is provided. An apparatus and method for is provided.

[0014] In accordance with yet another aspect of the present technology, measurements of the airway are performed to more accurately detect obstructive events. or a flow-limiting waveform or its shape index due to a detected secondary condition. Alternatively, an apparatus and method for optimizing an indication of partial occlusion is provided.

[0015] Aspects of the present technology include determining a respiratory flow index and estimating flow limitation from the respiratory flow index. determining a shape index representative of the pattern of the respiratory flow; and calculating a ventilation index or a respiratory rate index from the respiratory flow index. determining a duty cycle index and determining the determined shape index and the determined ventilation index. deriving a flow limitation index as a function of one or both of the load cycle index and the load limit index; The present invention relates to a method for detecting flow limitation, which may include:

[0016] In certain embodiments, the shape index is a flattening index, an "M" shaping index, a chair shape index, Indices such as roundness and / or roundness, or other indices that may represent partial occlusion, The ventilation index is a measure of tidal volume, such as the ratio of the current tidal volume to the previous tidal volume. The breath duty cycle index may be a ratio of the current breath inspiration time to the breath cycle time. , and ratios such as the ratio of the average previous inspiration time to the breath cycle time.

[0017] These methods include detecting the flow limitation by a flow limitation detector and / or by a flow limiting pressure therapy device. For example, these methods may be implemented using a shape index that is similar to the M-shape in respiratory airflow. indicates the presence of aspiration and the ventilation index is sufficiently reduced to indicate less than normal ventilation. Adjusting the treatment pressure of a pressure therapy device, such as by increasing the pressure as a condition for In some cases, the therapeutic pressure value is determined by the shape index being the M in the respiratory airflow. Indicating the presence of a shape breath and adjusting or increasing as a condition of an increase in the duty cycle index This may be the case.

[0018] In one embodiment of the present technology, a device for detecting flow limitation comprises a patient a patient interface coupled to the patient interface and configured to receive a patient a flow sensor for generating a flow signal representative of the flow of the intake gas; and a controller for processing the shape index, the air permeability index, and and a flow limitation index based on one or both of the breathing duty cycle index and the flow limitation index. The apparatus is configured to control a method for controlling the flow of the fluid from the flow restriction index, the method comprising: and calculate the pressure demand as described herein. The controller and patient interface may be configured to set the The system may further include a flow generator coupled to the gas.

[0019] In one embodiment of the present technology, a system for detection of flow limitation comprises: and an interface means connected to the interface means. a flow measuring means for generating a flow signal representative of the flow of respiratory gas through the and processing means for processing the flow signal, the processing means comprising: a shape index and a ventilation index. and deriving a flow limitation index based on either a target or a breath duty cycle index. The system is configured to process the method of the present invention. The processing means is coupled to the processing means and configured to calculate the pressure demand as a function of the flow restriction index. generating a controlled flow of breathing gas through the interface means so that and a flow means for configuring the flow generator in accordance with a pressure requirement.

[0020] In another embodiment, the present technology provides a process capable of controlling a device for detecting respiratory flow limitation. The information bearing medium may have processor-readable information. Determine the respiratory flow index and calculate the shape index that represents the pattern of flow limitation from the respiratory flow index. determining a ventilation index and / or a breathing duty cycle index from the respiratory flow index; and determining The shape index is determined based on one or both of the determined ventilation index and the determined breathing duty cycle index. The processor may include control instructions for deriving a flow limitation index as a function of the flow rate. Such information is used to calculate the pressure requirement as a function of the flow limitation index, as described herein. The control may further include instructions to adjust the pressure as follows:

[0021] Another aspect of the present technology is to provide a pressure or flow control system that is proportionally derived as a function of the measure of flow limitation. Related to the generation of control information or signals. For example, a controller or processor may: A device for controlling or determining pressure or flow settings for respiratory therapy equipment. It may be configured to indicate the degree of obstruction or flow limitation based on a measure of respiratory flow. The obstruction index or shape index may be determined or calculated by the device. In some cases, the device may also determine a ventilation index that indicates the degree of change in ventilation from the respiratory flow index. Therapeutic pressure or flow settings may be adjusted based on: (1) the presence of an obstruction or as a proportional function of either or both of (1) the degree of flow restriction and (2) the degree of change in ventilation. It may be derived by the device.

[0022] Further embodiments of the present technology will be apparent from the disclosure that follows.

[0023] The present technology is illustrated, by way of example only, in the accompanying drawing figures, in which like reference numerals refer to like elements and in which: It is not intended to be limiting, but is described for illustrative purposes. [Brief description of the drawings]

[0024] [Figure 1] The figures show first the airflow in a normally breathing patient, then in a patient experiencing airway narrowing, and finally in an awake patient. [Diagram 2] FIG. 13 illustrates the Flattening Index for a patient experiencing obstruction as indicated by an "M" shaped breathing pattern. [Diagram 3] FIG. 13 illustrates the Flattening Index with and without multi-breath averaging for a patient experiencing an obstruction indicated by an "M" shaped breathing pattern leading to a breathing pattern indicative of the patient arousing from sleep. [Figure 4] 1A-1C illustrate examples of components of a device for detection and / or treatment of flow limitation or partial occlusion. [Diagram 5] FIG. 13 illustrates the steps involved in a detection or treatment device in detecting a flow restriction or partial obstruction. [Figure 6] FIG. 13 illustrates an example of a fuzzy membership function for a variable based on a single breath flattened shape index. [Figure 7] FIG. 13 illustrates an example of a fuzzy membership function for variables based on the m-shape breath shape index. [Figure 8] FIG. 13 illustrates an example of a fuzzy membership function for a variable based on the air permeability index. [Figure 9] FIG. 13 illustrates an example of a fuzzy membership function for a variable based on breath duty cycle ratio. [Figure 10] FIG. 13 illustrates an example of an output membership function for flow limitation measurement. [Figure 11] FIG. 13 illustrates an example of a defuzzification operation in determining flow limitation or partial obstruction index using the centroid method. [Figure 12] FIG. 13 illustrates an example of a function that modifies the detection ability of a flow limitation or partial obstruction indicator based on an indicator of ventilation. [Figure 13] FIG. 13 is a graphical comparison of the calculated flow limitation index FFL of the techniques described herein and the conventional flattening index determined from the same flow signal. [Figure 14] FIG. 13 shows a flow signal derived from nasal flow and an output metric of a continuous RERA detector that quantifies the likelihood of respiratory arousal. [Figure 15] FIG. 13 shows the ratio of peak expiratory flow to mean expiratory flow, which may be an index for sleep-wake-awake detection. [Figure 16] FIG. 1 shows two superimposed flow signals, a filtered and an unfiltered flow signal recording an obstructive breath sequence during low frequency snoring. [Figure 17] FIG. 13 illustrates the response of an embodiment of an FIR filter useful for filtering an airflow signal. [Figure 18] FIG. 13 plots an example of a function that serves as basis vectors B1 and B2 for detecting M-shape breathing. [Figure 19] 1 is a graph of a typical M-shaped breath flow signal. [Figure 20] 13 is a graph of a flow signal showing M-shape breathing enhanced by simple chair shaping. [Figure 21] 13 is a graph of a membership function of a variable based on the ratio of peak inspiratory flow to average inspiratory flow useful for detecting an M-shaped breath having a chair geometry. [Figure 22] 13 is a graph of a membership function of a variable based on inspiration peak location useful for detecting an M-shaped breath having a chair shape. [Diagram 23] FIG. 2 shows a graph of a flow signal, a graph of a determined flatness index, and a graph of a determined chair shape index. [Figure 24] 1 is a graph of the flow signal and corresponding single breath flattening index for a patient experiencing upper airway resistance with increasing respiratory effort that terminates during a sleep arousal at t=420. [Diagram 25] 25 is a graph of an example of a flow limitation index, a ventilation index, and a duty cycle index based on the flow signal of FIG. 24. [Figure 26] 25 is a graph of a sustained obstruction index or a sustained flattening index based on the breath sequence shown in the flow signal of FIG. 24. [Figure 27] FIG. 13 illustrates a function of de-weighting factors that can be used to modify the flow limitation index. [Figure 28] FIG. 2 shows a graph of a respiratory flow signal, a graph of an instantaneous snore signal, and a graph of three snore indices. [Figure 29] FIG. 1 illustrates a classification of snoring treatments based on multiple snoring indices. [Diagram 30] 1 is a graph of an M-shaped breath with correct and incorrect trim results marked; [Diagram 31] FIG. 1 includes a graph of a flow signal versus time including an M-shaped breath and a graph of inspiration time determined by two methods. [Diagram 32] FIG. 13 shows a graph of a function useful for reducing the weight of a flow limitation index in response to a leak condition. [Diagram 33] FIG. 13 is a diagram of a further function useful for reducing the weight of a flow limitation index or partial obstruction index (eg, FFL) based on a leak condition. [Diagram 34]13 is a graph of a function suitable for de-weighting flow limitation or partial obstruction indicators with level of mask pressure. [Diagram 35] 13 is a graph of a function suitable for reducing the weighting of a flow limitation or partial obstruction index by an index of obstruction. [Diagram 36] 13 is a graph of a function suitable for reducing the weight of the snoring index as a function of treatment pressure. [Figure 37] 13 is a graph of a function suitable for de-weighting a snore index based on a ventilation index representing a large breath; [Figure 38] 13 is a histogram of normalized expiratory peak location for various patient groups. [Figure 39] 13 is a graph of a function of a de-weighting factor to be applied to a flow limitation or partial obstruction index based on a normalized expiratory peak location value. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0025] Patients with OSA experience periods of apnea or breathing pauses during sleep that end only for the patient to awaken. These frequent events are the cause of sleep fragmentation and stimulation of the sympathetic nervous system. This can cause patients to experience daytime sleepiness (which may be associated with car accidents), The consequences can be serious, including unexpected psychiatric effects, memory problems, depression, and high blood pressure. Patients with A may also snore loudly, disrupting the sleep of their partner at the same time. The best form of treatment for patients with OSA is the use of a connected hose and Continuous positive airway pressure (CPAP) delivered by a compressor through a mask Positive pressure prevents the patient's airway from narrowing during inspiration, thus reducing frequent apneas or hypopneas. and to prevent these aftereffects.

[0026] Typically, patients are fitted with a number of sensors from which a recording known as a polysomnogram is taken. Patients underwent two sleep studies monitored using a sleep apnea monitor. The first study (without treatment) confirmed the diagnosis of OSA. Once confirmed, a second probe is used to titrate the patient to the correct therapeutic pressure. Pressure requirements change throughout the night due to changes in position, posture and sleep state. We propose a pressure (titration pressure) that can cover any contingency. An alternative to airway therapy is an automated machine that adapts pressure to the patient's needs. This is known as APAP or positive pressure airway pressure. The benefits of APAP therapy are that it can be used in a variety of situations. The advantage of this is the adaptability to pressure requirements that may vary over time scales. For example: Pressure requirements change during the night with posture and sleep state, and possibly with alcohol consumption. This can vary over the weekend due to the beneficial effects of the therapy itself, and over the course of several months due to weight loss or may change over the years due to increased Pressure is increased according to the patient's condition, allowing the patient to feel comfortable at a pressure well below any therapeutic maximum. Finally, APAP machines operate at potentially lower pressures, so This tends to improve the effect of mask leakage somewhat.

[0027] Known algorithms used to automatically set patient pressure in APAP machines The ResMed AutoSet is called ResMed AutoSet. Overall, the AutoSet device and its The algorithm is excellent for treating OSA patients. The algorithm responds to three conditions: flow limitation, snoring (audible noise), and apnea. Automated Pressure Setting is incorporated herein by reference in its entirety. The flow restriction is also described in U.S. Pat. No. 5,704,345. These are the fluid dynamics of the so-called "pushing tube". The pharynx of a patient with OSA is Well, rather than the simple passive tubes that are being investigated on the benchtop, we have extruded muscle tubes. In essence, the flow restriction is the In a constricted tube, the flow is proportional to the downstream pressure (assuming the upstream pressure is held constant). A state in which the force is not increased any further by decreasing the force (i.e., increasing the flow-driving differential pressure). In patients with a narrowed upper airway, this condition occurs when the patient is no longer receiving adequate ventilation. This is the same as the situation where the inspiratory effort is no longer increased and an increase in the inspiratory flow rate is no longer increased. A patient like this may first breathe normally at 102, then develop airway narrowing at 104, and Despite the larger negative pressure deviation (an indicator of respiratory effort), Ultimately, the patient is able to open the airway and increase blood flow. Awaken at 106 to take several big breaths to restore gas homeostasis.

[0028] ResMed AutoSet algorithms monitor patient flow and provide flow restriction or When snoring is detected, pressure is increased. Normally, apnea is preceded by flow limitation (called partial obstruction). It is rare to encounter apnea, as it is preceded by periods of snoring or snoring. As a back-up measure, pressure is also increased if apnea is detected. If no disturbance is measured, the pressure is allowed to decrease slowly, hopefully allowing the patient to An equilibrium pressure is achieved that allows you to sleep without waking up. Since rhythms respond proportionally, a metric is used for each condition to which this algorithm responds. The metrics, namely, the flattening index for flow limitation and the flattening during inspiration for snoring, A calibrated RMS measure of the equalized sounds and the length of the detected apnea are used.

[0029] The flattening index is a dimensionless feature (e.g., a real number) that is calculated using the patient's inspiration waveform. The Flattening Index essentially measures how flat the tops of a waveform are. The flow restriction is characterized by a condition in which the downstream pressure is low enough to keep the tube in a constricted state. On the other hand, the flow rate remains almost constant regardless of changes in the driving pressure. In patients with flow-limited breathing, this results in an inspiratory waveform that contains a flat apex ( i.e., constant inspiratory flow rate).

[0030] For example, depending on the chosen scale of the index, normal breathing has a flattening index of about 0.2. whereas significantly flattened waveforms may produce flattening exponents of approximately 0.1 or less. In APAP therapy, limitations are typically set (e.g., certain (0.19 for this machine), and the pressure is increased in proportion to the extent to which the flattening index falls below the threshold. To reduce the effects of noise and increase specificity, a typical pressure setting algorithm is used. Sometimes a moving average over five breath points is used. Possible disadvantages of the five breath average requires five abnormal breaths to be averaged before the plateau drops to its final nadir. The aim of the heuristic weighting is to slow down the detection of flow limitations as the leak increases. As the therapeutic pressure increases, and evidence of valve-like breathing begins to As the pressure increases, the flattening increases to the extent that a greater reduction in flattening is required. These rules of thumb may be applied to the threshold that causes the No.) nevertheless, it helps contain a potential pressure surge.

[0031] Although the Flattening Index is a good indicator of flow limitation, it is designed to detect certain conditions. However, some specific implementations have been found to not cover some rare situations. Areas where such observations have been made include:

[0032] A 1.5 breath moving average slows down the detection of flow limitation. This is shown in Figure 2. In 2, the upper trace shows a plot of the conventional five-breath moving average flattened index. The lower trace shows the respiratory flow index. The patient begins to gradually obstruct and flatten. The trace has a step-like drop at 202 due to the five-breath average. As the blockage becomes more severe, it gradually begins to take on more of an "M" shape. As such, the flattening index eventually begins to reverse direction and increase rather than decrease as the obstruction worsens. Add. 2. Because the various intake shapes may average out to give a completely new shape The five-breath moving average can be important. This is illustrated in Figure 3. , 302, the so-called M-shaped obstructed breath sequence, 304, the awakening The end shows a sequence followed by some reasonably normal recovery breathing at 306. As shown in the graph, both single-breath flattening and conventional flattening have a series of M breaths. (the former reaches its maximum value more quickly), and after awakening, the conventional flattening actually reaches 0 The M breath falls to below .1, but this is not because the breath is flattened, but rather because the M breath is normal. This is because it gets averaged with the breath, producing a pseudo-flat shape. 3. The flattening index is not designed to detect M breaths. In fact, the flattening index is The shape becomes higher when breathing occurs. This is shown in Figures 2 and 3. 4. The flattening index is the pressure drop independent of the current ventilation or sleep state of the patient-user of the device. This may result in increased force. 5. The heuristics applied to reduce the flattening weights may lead to undertreatment of some patients. There is a possibility that this will happen. 6. Flattening index is any algorithm that uses flattening index to detect flow limitations. The quantification is subject to normal random variations that have significance for the sensitivity and specificity of the assay.

[0033] Although M-shaped breaths may be rare, they may be difficult to detect and / or tolerate due to existing methods. It remains desirable to develop further methods and instruments to improve the present invention. It is.

[0034] Referring to FIG. 4, the present technology may include a flow generator such as a servo-controlled blower 402. This device is typically a mask. 406 and a patient interface for delivering a flow of air or respiratory gas to a patient and / or and an air delivery conduit 408 for conveying the air from the patient. The exhaust gas may be coupled to the conduit 408 and the mask 406. In some cases, the flow sensor 404f and / or the pressure sensor 404b may be used to ventilate the blood. A force sensor 404p may also be used. For example, mask flow may be measured using a pneumotachograph, and a differential pressure transducer or a tube or It may be measured using similar equipment such as a bundle of ducts. The mask pressure is expressed by the pressure signal P mask A pressure transducer is used to derive (t). The pressure sensor 404f and the flow sensor 404p may be used to measure the pressure at the pressure hole. It is only symbolically represented in FIG. 4 because other structures and other equipment may be It will be understood that the present invention may be implemented to measure flow and pressure. signal F(t) and pressure signal P mask (t) is the pressure request signal P request (t) One or more analog-to-digital (A / D) converters / samplers (not shown) to derive The signal may be transmitted to the controller or microprocessor 415 via

[0035] Alternatively, the flow signal f(t) and / or the pressure signal P mask (t) is as described above. The current ( By monitoring the motor speed (S) and / or rotation speed (w), the air flow is In some cases, the blower motor speed is estimated or calculated. The pressure inside the mask can be kept fairly constant, and the airflow to the mask can be variably controlled by the pressure changes. This is accomplished by controlling the opening of a servo valve that may bypass / vent or deliver. Additionally, other types of patient interfaces may be used in place of a mask. The flow and / or pressure sensor may be selective with respect to the patient interface. Flow and pressure may be measured at various locations.

[0036] The controller or processor 415 controls the methods described in more detail herein. or an algorithm, and includes an integrated chip, memory, and / or or other control instructions, data, or information storage media. For example, instructions programmed with the control method are encoded on an integrated chip in the memory of the device. Such instructions may be implemented in software on a suitable medium. Or it may be loaded as firmware.

[0037] With such a controller, the device can either: Examples of the embodiment described herein that are used to operate the vent using a release valve Investigating pressure delivery systems that may be based on pressure delivery-related detection methods such as those shown in By adjusting the pressure, it can be used for many different pressure therapy procedures. Pressure therapy is a method for diagnosing, detecting, and / or quantifying the presence of flow limitation using a device. In equipment that does not contain a pressure therapy component, as may be used in It may be carried out.

[0038] The present invention provides a method for detecting and / or treating breathing patterns indicative of partial obstruction. The method or algorithm described uses a controller as a fuzzy logic control system. In general, fuzzy logic is used to generate ideas about system behavior. Convert the data into computer code that is based on real numbers for the parameters of the control system. However, the problem at hand is fundamentally one of statistical pattern recognition. and using other techniques, such as the alternatives discussed in more detail herein. Therefore, the following technique is sometimes implemented as a fuzzy logic control system. It has been proposed from the viewpoint of control system design based on desired control input and typical output functions. It will be appreciated that there are other techniques for introducing the

[0039] To this end, embodiments of the control system of the present technology utilize a fuzzy flow limit (FFL) index This is done according to the following general approach using partial obstruction or flow limitation indicators such as: The use of such an index is a step forward in overcoming traditional problems as system control variables. An approach that may improve upon some of the observations made above regarding the use of the flattening index The flow limitation index (FFL) is a measure of the severity of flow limitation (or, more precisely, partial upper airway obstruction). Multiple feature pattern recognition to improve the sensitivity and specificity of the algorithm's response to In an embodiment, a general approach using the flow limitation measure FFL is as follows: It may include some or all of the steps.

[0040] 1. Monitor the patient's respiratory flow. 2. Extract the inspiratory and expiratory waveforms from the flow signal. 3. Compute features from each component of the waveform or, where appropriate, from the composite waveform. features are sometimes thought to constitute a pattern. 4. Calculate the fuzzy input variables from the raw variables. This step creates a new fuzzy set of input variables. Data mining of the appropriate population of patients is done by This allows for good estimation. 5. Fuzzy variables derived using fuzzy logic to generate fuzzy output This step is formulated using matrices that allow intuitive understanding by laymen. It is possible to do so. 6. To generate a crisp value (a real number) to be used by the pressure control algorithm Then, the fuzzy output is de-fuzzified. 7. Combining fuzzy output from multiple sources. 8.Adjust treatment pressure based on FFL results.

[0041] According to this method, an embodiment of the system uses fuzzy logic to calculate widely varying input parameters. To map the variances to variables that can be easily interpreted using theoretical principles, e.g., the smoothing index Shape indices are mapped to fuzzy variables such as high flatness, medium flatness, low flatness, etc. Each of these variables is determined by a mathematical function that is related to the variable. It depends on the value of the flattening index as determined by the inputs relevant to the flattening decision, such as respiratory airflow. Based on the data, it takes a value between zero and one. Similarly, the other fuzzy variables are Based on mathematical functions and input data related to determining ventilation, high ventilation, normal ventilation, It is possible to map ventilation indicators such as low ventilation, etc. Fuzzy variables also have low T i -on-T tot , Normal T i -on-T tot , or high T i -on-T tot Respiratory duty cycle indices such as Time (T tot ) versus inspiration time (T i It is possible to map the ratio of These variables are then used to generate the fuzzy output variables. The fuzzy output associated with the flow limitation index may be further mapped to, for example, Other fuzzy variables may be mild, moderate and severe, as in the examples described above. For example, a fuzzy logic controller might have the following statements: The conditions expressed by the embodiment may be implemented. If (a) (low flattening and low aeration), the flow restriction is severe. (b) (High flatness and high T i -on-T tot ), the flow limitation is mild. do.

[0042] The defuzzified output, which can be used as a flow restriction indicator, is zero (no flow restriction). Such a system can be used to vary the score between 0 (severe) and 1 (dead). The force space is nonlinear, without any loss of real-number precision, and with rules that are easily interpretable. can be classified by meaning.

[0043] In fact, the accuracy of such flow limitation indices, which express the degree of obstruction on a continuous scale from 0 to 1, is Using this degree, the indicator is used to determine the response that is suitable for handling the condition detected using this indicator. Proportionality can be implemented to derive pressure or flow control settings. For example, such indicators may be compared to determine whether a pressure adjustment should be made. The indicators can be used to determine the respiratory treatment device settings or Be more direct in determining treatment settings for such devices For example, the obstruction index may be directly applied to the amount of pressure or flow adjustment. may serve as some proportional function (e.g., index x adjustment amount). The amount of adjustment may then be applied to the treatment settings of the respiratory treatment device. As explained in more detail in the paper, the "FFL" indicator is These may be implemented to act as proportional treatment setting functions such as:

[0044] As further shown in FIG. 5, in such a system, the flow restriction is: (a) Step A shape index that represents the pattern of flow limitation is determined from the respiratory flow indices shown in step 502. (b) calculating a ventilation index or (c) determining a breath duty cycle index as shown in step 506; the determined shape index and either the determined ventilation index or the determined respiratory cycle index, and deriving a flow limitation index as a function of both. These methods may be used to provide a high degree of control over the therapeutic pressure. For example, to treat a flow restriction, an increase in (a) shape index indicates the presence of an M-shape or flattened breath in the respiratory airstream; and (b) the ventilatory finger. It was determined that both the suction pressure and the suction pressure had decreased sufficiently to indicate less than normal ventilation. This may be based on the condition that if a pressure change occurs, a flow restriction may occur. Further increases to treat (a) the shape index indicates the presence of M-shaped breath in the respiratory airstream; and (b) an increase in the duty cycle index is detected. If the pressure is increased, it may be based on other conditions such as the possibility of a pressure change. Using such criteria, appropriate treatment responses can be determined using shape index, ventilation index, or This appears when the respiratory duty cycle index does not accurately predict the level of flow limitation. This can be given without unnecessary therapeutic changes.

[0045] Therefore, in one embodiment, the system includes a flow limitation measure FFL for deriving It is characterized by the use of multiple (e.g., four) input features. The input features are (1 ) Single breath flattening (SBF) shape index, (2) M shape index (MS), (3) ventilation rate (V (4) the ratio of inspiration time to total breath time (T i -on -T tot ) may also be included as a respiratory duty cycle indicator.

[0046] The embodiment of the shape index implemented as a single breath flattening index SBF is preferably 5 The method of the present invention is described in detail in US Pat. No. 5,704,345, except that it is implemented without the use of a moving average of respiration. The method may be similar to the approach disclosed in the above publication. Implementation details are disclosed in Section A herein.

[0047] An exemplary implementation of the determination of the shape index of the M-shape is described in Section B herein. The illustrated calculation of the M-shape characteristic is performed for zero (i.e., the intake flow waveform is M-shaped). (i.e., the inhalation flow waveform is definitely M-shaped) to 1 (i.e., the inhalation flow waveform is definitely M-shaped). The M shape index can sometimes be expressed in other ways, such as a simple "chair shape" index. The chair shape index (e.g., the fuzzy "chair neck") may be augmented by the chair shape index. A typical "Patent Law" decision is also detailed in Section B.

[0048] An embodiment of the determination of a suitable ventilation index may be based on tidal volume calculations. The ventilation index is based on the current tidal volume so that the index may represent changes in ventilation or tidal volume. It may be a ratio including the inspiration volume and the previous tidal volume. In one suitable embodiment, the indicator is The ventilation rate VR may be calculated as follows:

[0049] 1. Obtain the absolute value of the respiratory flow and filter it using a simple low-pass filter (time constant = 3 minutes). Filter the absolute value of 2. Divide the filter output by 2 to yield a "3 minute vent". 3. Calculate the respiratory tidal volume (basically the average of the inhaled and exhaled volumes). 4. Divide the tidal volume by the length of the breath, i.e., to produce the average breath flow rate, and then is divided by the 3 minute breath to yield an indication of the degree to which this breath matches the previous breath.

[0050] Normally, VR oscillates in a narrow range around 1. In cases of severe upper airway obstruction, VR is much higher than 1. In the case of complete obstruction, it may decrease to zero. A VR that is much greater than 1 (e.g., 1.5 to 2). Further details on the calculation of the ventilation rate VR are contained in Section G herein.

[0051] As described above, the system may optionally include, during the derivation of the flow limitation measure FFL: A respiratory duty cycle index is available. This index represents the relative This metric may be based on the percentage of the current duty cycle that is being Exemplary details for the calculation of the respiratory duty cycle ratio TTR are provided herein. Further described in Section G. For example, a suitable breath duty cycle indicator is T i -o nT tot The variables selected may be based on the calculation of zero (non-inhalation) and and 1 (non-breathable) restrictions. i -on-T tot teeth T relative to the previous average value, as this may vary from patient to patient. i -on-T tot of The ratio, i.e., T i -on-T tot The ratio expressed as a percentage of It is calculated. i -on-T tot The average value immediately before each calculated T i -on-T tot It can be calculated by feeding the value into a simple low-pass filter. i -o nT tot When increases compared to the previous history, the TTR exceeds 1. i -on- T tot The TTR has a value of around 0.4, and the TTR is usually 1, which indicates that the patient has an obstructed upper airway. Despite the fact that , greater than 1.

[0052] As described above, these input features are mapped to several variables by mathematical functions. Therefore, the shape index SBF is calculated by, for example, six fuzzy variables, namely, B_L OW_FLATTENING, EXTRA_LOW_FLATTENING, VERY_ LOW_FLATTENING, LOW_FLATTENING, NORMAL_FLA May be mapped to TTENING and HIGH_FLATTENING. These fuzzy memberships are graphed in Figure 6 and are listed above. The following example functions are presented in left-to-right order with each variable being As an example of the application of the membership function, the shape index SBF is calculated and has a value of 0.05. If the result is that the variable B_LO W_FLATTENING results in a value of 1.0, which is the same as all other graphs in Figure 6. The value zero will be used for all fuzzy variables. Other suitable functions and variables may be selected as required. may be selected.

[0053] Similarly, an example of membership functions and variables based on the calculation of the M-shape shape index is shown in FIG. 7. This graph shows the LOW_M-SHAPE, NORMAL_M-SHAPE PE, HIGH_M-SHAPE, VERY_HIGH_M-SHAPE, and B_ The appropriate function for HIGH_M-SHAPE is shown. Based on the calculation of the air permeability index An example of the membership functions and variables is shown in FIG. 8. This graph shows _LOW_VENTILATION, LOW_VENTILATION, NORMAL_ VENTILATION, HIGH_VENTILATION, and VERY_HI Here is an appropriate function for GH_VENTILATION. Finally, the breathing load cycle An example of the membership functions and variables based on which the rate is calculated is shown in FIG. The graph is T i -on-T tot _LOW, T i -on-T tot _NORMAL, T i -on-T tot _HIGH, T i -on-T tot _VERY_HIGH and T i -on-T tot The appropriate function for _EXTRA_HIGH is shown below.

[0054] Based on some or all of these mapped variables, the system then Rules based on combinations of variables may be applied or evaluated during the derivation of the restriction indicator FFL. For example, the fuzzy rules are shown in the table below based on the pre-specified fuzzy variables. These matrices further express in natural language the idea of ​​what might constitute a partial occlusion. It expresses.

[0055] Table A Flattening and ventilation index [Table 1]

[0056] Table A shows the fuzzy variables associated with the single-breath flattening shape index and the ventilation rate index. The paper outlines a set of fuzzy rules that can be used to apply the The “Very High” column for the fuzzy variable of the index and the “ The bold ("light to medium" i.e. light to medium) rules in the "low" line are It represents the G-rule.

[0057] (Very high flattening and low ventilation), then the FFL is "mild to moderate."

[0058] This rule and any other rules shown in the table can be evaluated, and "AND" means Fuzzy AND. A common output for each output response (e.g., "mild to moderate") The results are then fuzzy ORed together. For example, the output "medium" (e.g., the two rules with indicators of moderate flow restriction) will produce a final “medium” output. These are fuzzy ORed together to produce:

[0059] Moderate = (VERY_HIGH_FLATTENING AND VERY_LOW_ VENTILATION) OR (EXTRA_HIGH_FLATTENING A ND LOW_VENTILATION)

[0060] Based on the application of these rules during the derivation of the flow restriction index, at least three things can be observed: terms, i.e., 1) as the flattening severity increases (e.g., as the shape exponent decreases) 2) When flattening appears and aeration decreases, the FFL index increases. The indicators are more severe and 3) ventilation is high (e.g., recovery (big) breath, etc.). It was found that during rule application, the response to the flattened shape index was adjusted when It is white.

[0061] Flow limitation is not (at least initially) accompanied by a reduction in tidal volume (e.g., ventilation rate) In these cases, the patient may be advised to increase the duty cycle, i.e. Maintain your own tidal volume by lengthening your inhalation time as a percentage of the total inhalation-exhalation time. This tendency is seen in the present embodiment when the duty cycle becomes greater than 1 as the inhalation time increases. It can be measured or determined using an indicator (eg, TTR).

[0062] The system uses a method similar to the method in which the flattening shape index information is combined with a ventilation index (e.g., VR) and further combines the flattening shape index information with the inspiratory load cycle index (e.g., TTR) using the rules expressed by Table B below. In the same way as, using the rules expressed by Table B below.

[0063] Table B Flattening and Inspiratory Load Cycle Index [Table 2]

[0064] Similar to Table A, Table B represents the fuzzy rules of flow blockage based on the index of the inspiratory load cycle and the flattening shape index and the corresponding natural language description. For example, the bold rules (from medium to heavy, i.e., medium to severe) in the "Extra High" column and the "Ultra High" row are as follows. If (EXTRA_HIGH_FLATTENING AND T -on-T _V

[0065] ERY_HIGH), then FFL is from medium to heavy. i -on-T tot _V ERY_HIGH), then FFL is from medium to heavy.

[0066] This rule expresses that if the waveform of the inspiratory breathing pattern tends to flatten severely and the inspiration is extended moderately considering the inspiratory load cycle index, the flow restriction index FFL is from medium to heavy. As described for Table A, the results of the rules for the common output function of this table can be combined using fuzzy logical sum operations. This rule expresses that if the waveform of the inspiratory breathing pattern tends to flatten severely and the inspiration is extended moderately considering the inspiratory load cycle index, the flow restriction index FFL is from medium to heavy. As described for Table A, the results of the rules for the common output function of this table can be combined using fuzzy logical sum operations. This rule expresses that if the waveform of the inspiratory breathing pattern tends to flatten severely and the inspiration is extended moderately considering the inspiratory load cycle index, the flow restriction index FFL is from medium to heavy. As described for Table A, the results of the rules for the common output function of this table can be combined using fuzzy logical sum operations. The results of the rules for the common output function of this table can be combined using fuzzy logical sum operations.

[0067] The flow restriction is in the same way as the method in which the flow restriction index is derived using the flattening shape index information and one or both of the ventilation index and the inspiratory load cycle index, and the flow restriction is M-shaped and / or using the same method as the method in which the flow restriction index is derived using the flattening shape index information and one or both of the ventilation index and the inspiratory load cycle index, the flow restriction is M-shaped and / or Both the Enhanced M-Shaped Detection Index and either the Ventilation Index or the Breath Duty Cycle Index or both and this derivation is shown in Tables C and D.

[0068] Table C M-shape or enhanced M-shape and ventilation index [Table 3]

[0069] Table C relates to fuzzy variables based on M-shape shape index and / or chair shape index. This table expresses the same rules as the above table, except that the output function selection As is evident from the quality, the rule is that the patient should have normal or above average ventilation. The purpose of the device is to block responses to a detected M-shape breathing pattern when Therefore, the flow limitation index indicates whether the system is capable of regulating certain types of wakefulness and REM sleep. "Behavioral" M-shaped breathing patterns that may occur during breathing but do not represent current flow limitation. For example, the top row of the last column in Table C The rule expressed as follows is that when a good ventilation index (e.g., "very high") exists, the output function is "zero" for the indices associated with the poor "very high" M shape index.

[0070] Furthermore, most M-shaped breaths represent moderate to severe obstruction and show reduced ventilation. Therefore, the other rules in Table C are based on this detected M-shape and the reduction in ventilation. Flow restriction detection for system pressure changes to address conditions such as low airflow index This makes it possible.

[0071] As described for Table A, the results of the rules for the common output functions in this table are They can be combined using the fuzzy OR operation.

[0072] Table D presents rules regarding breath duty cycle indices.

[0073] Table D M-Shaped or Enhanced M-Shaped and Duty Cycle Indices [Table 4]

[0074] Table D is derived from the breath duty cycle index and M-shaping and / or enhanced M-shaping determinations. The specific fuzzy rules for the flow blockage or flow limitation index and the corresponding natural language In this embodiment, the matrix represents M shape information (rather than air permeability) and T i -on-T tot The matrix is ​​"combined" with the ratio, under the tendency of increasing the breathing duty cycle. It is also thought that the degree of M-shaped pattern is related to the breath that shows M-characteristics. This combination allows the patient to breathe more or less "normal" and allows the patient to take a longer breath. If the patient is "compensating" by

[0075] As described with reference to Table A, the results of the rules for the common output functions in this table are They can be combined using a fuzzy OR operation.

[0076] The results of the applied rules based on tables A, B, C and D are then applied to the output function. Figure 10 shows the appropriate fuzzy output that may be applied with this result. The output membership functions are negative, zero, slight, and slight. The severity ranges from mild to moderate, moderate, moderate to severe, and severe.

[0077] As an example, consider Table D. Once the individual fuzzy rules in the table are calculated (e.g. , (EXTRA_HIGH_MSHAPE AND T i -on-T tot _VERY_ If FFL is moderate to severe, the outputs are collected and fuzzy For example, three moderate to severe and If this is done for all the outputs, the inverse fuzzy set in Figure 10 is The function is fed into the quantization function and the centroid method is used to give a crisp output as in Figure 11. This results in a real number between 0 and 1.25. If multiple outputs are available, the highest output may be used for the FFL value.

[0078] In determining a single "crisp" index in the defuzzification step just mentioned, , a centroid method or other such defuzzification operation may be performed. The crisp value of the output variable is found by finding the value associated with the centroid of the output function values. An example of such a method is shown in FIG. Calculated for fuzzy inputs: 0.1, mild to moderate = 0.5, moderate = 0.8 The result is 0.61064. to understand other ways to calculate crisp metrics.

[0079] In one embodiment of the system, the flow limitation indicator is used to avoid treatment of other potential detected conditions. For example, the awakening breath can be flat or M-shaped. , and thus may represent a flow restriction based on the determination using the shape index. However, this does not represent the actual flow limitation in the patient, since the breath actually has a longer inspiratory time. Therefore, the flow limitation index does not treat such breaths as flow limited. One way to perform such a derivation is to use the functions shown in FIG. Tables A, B, C and D are used to adjust the combination of the breathing duty cycle metric TTR. The goal is to modify the fuzzy output of the fuzzy rules from D and D. The rules based on a combination of the TTR indicators (for example, the rules in Tables B and D) The fuzzy output to be obtained is multiplied by the output of the function in Figure 12 before defuzzification. Therefore, if the ventilation rate is approximately 1 or less, the outcome of each specific rule is unchanged. With increasing ventilation rate index VR, which indicates an increase in the possibility of According to the function embodiment of FIG. 12, and through the multiplication operation, the ventilation rate index VR approaches 1.5. The weights are gradually decreased until the weights of the affected rules are Set the output to zero.

[0080] The above building blocks (e.g., shape indices, fuzzy membership functions and variables, Based on the basis of fuzzy rules, the flow restriction index F in the flow restriction detection system is The calculation of an embodiment of FL can be further summarized using the following exemplary steps.

[0081] 1. Breath is used to identify and extract data representing the inspiratory and expiratory waveforms from the flow signal. A typical method may be framed from the patient flow signal. 2. The inspiration waveform is trimmed, in some cases, using a trimming method that takes into account the specific M-shape. Any forward pauses may be removed. Such methods are described in Section It is described in F. 3. M-shape determination and fuzzy chairness determination are performed using the typical This may be based on calculations. 4. The augmented M-shape may, in some cases, be determined by the M-shape shape index and the chair shape index (e.g., It may be calculated as a fuzzy logical OR with fuzzy chairness. 5. Filtered Single Breath Flattening (SBF) Shape Indices Presented in Section A. The calculation may be based on typical calculations. 6.Ventilation rate (VR) and T i -on-T tot The ratio (TTR) is shown in Section G. The determination may be based on typical calculations performed. 7. SBF, VR, TTR and augmented M-shape are typical membership functions described above. They may be fuzzified into variables such as numbers. 8. The fuzzy rules are then applied to tables A, B, C and / or D as described above. It may be applied according to the matrix 9. Fuzzy rules from Tables B and D using the breath duty cycle indicator TTR In some cases, this may be modified by the "Limit Arousal Breath" function described above with respect to FIG. do. 10. Each fuzzy rule matrix is ​​collected and de-fuzzy as described above separately. It will be turned into. 11. As a result, all fuzzy sets are used to generate the flow limitation index FFL as follows: - A logically ORed fuzzy output is obtained. FFL=fuzzy logical sum (SBF-VR,SBF-TTR,M-shape-VR,M -shape-TTR) where SBF-VR is the defuzzification result based on the rules in Table A; SBF-TTR is the defuzzification result based on the rules in Table B. M-shape-VR is the defuzzification result based on the rules in Table C. M-shape-TTR is the defuzzification result based on the rules in Table D. 12. Next, the Fuzzy Persistent Flattening Index (FPF) is calculated, if necessary, as follows: It may be fuzzy ORed with the flow limitation index FFL according to the formula below: An exemplary calculation of F is provided in Section C of this specification. The FFL index is adjusted as follows: This can be done by the following formula: FFL = fuzzy OR (FFL, FPF) 13. The FFL may, in turn, utilize, for example, the calculations set forth in Section D herein. As a result, breath framing (e.g., detection of inspiration or expiration waveforms) may be incorrect. The resulting inaccurate breath framing coefficients are then corrected according to the degree to which they are true. FFL may be multiplied. FFL = FFL x Bad Breath Framing Coefficient 14. This final FFL value may be a ring of length, say 3, in some cases. It may be used in the buffer and ultimately by the pressure setting algorithm. The FFL value used may be based on a moving average of the most recent FFL values ​​in the buffer. An appropriate formula for the calculation of may be the following formula:

number

[0082] FIG. 13 shows the flow limitation index F derived by the conventional flattening index and the steps summarized above. A graphical comparison between the FL and the index is provided. The index and the indicator are the same flow shown in the lower trace. The flow signal in the lower trace is followed by a fairly normal recovery breath. It includes a series of blocked breaths that end in awakening, and further obstructions. In the upper trace, the conventional flattening exponent first drops off and then gradually becomes M-shaped. , then rises, and finally drops sharply during the recovery breath. The flow limitation index FFL of the middle trace is , which rises steadily due to both the flattening of the breath and the M-shaped shape, and then the breath in flow signal becomes more It will drop to zero when it normalizes.

[0083] The flow limitation index may also be implemented under the control of a pressure therapy device. In form, the indicators may be implemented using any one or more of the following:

[0084] 1. Measure the flow velocity with a flow generator (FG). 2.Measure the pressure with FG. 3. Calculate the pressure drop between the FG and the mask using the flow rate from step 1. 4. The pressure at the FG minus the pressure drop calculated in step 3 is taken as the pressure at the Calculate the pressure at the disk. 5. Use pressure on the mask to measure flow through the vents in the mask (sometimes intentional leaks) ) is calculated. 6. The sum of patient flow (breathing flow) and unintentional (mask or mouth) leak is calculated. To account for this, the vent flow is subtracted from the flow measured at the FG. 7. Filter the signal from step 6 to extract the DC (unintentional leakage) component. Understand. 8. To generate the patient flow (respiratory flow), the DC component calculated in the previous step is Subtract from the flow calculated in step 6. 9. Eliminate unnecessary high The patient flow is lightly filtered to remove high frequencies. 10. Frame the breath using patient flow in the usual manner. 11. Any forward port may be The nozzle is cut from the front of the intake. 12. Once a complete breath (inhale + exhale) is framed, calculate the following features: The length of any apnea that precedes the breath. b. Inspiratory Snoring Index (eg, the average of the snore signal during the currently analyzed inspiration). c. Currently analyzed (e.g., by the methods set forth in Section A herein) Filtered Single Breath Flattening Index (SBF) values ​​for the current breath. d. For the current intake, for example, by the method set forth in Section B herein. The M shape index value and the fuzzy chairness value for the current inspiration. The fuzzy chairness values ​​are fuzzy ORed to yield the augmented M shape feature. e. For the breath currently being analyzed, according to the method described in Section G herein. The value of the ventilation rate (VR) and T i -on-T tot Ratio (TTR). f. FFL as described in detail above using SBF, VR, TTR and Augmented M Calculate. g. Calculate valve-like leak rate using the currently analyzed breath. h. Measure unintentional leak at the end of the currently analyzed exhalation. i. Check the current setting for mask pressure (i.e., EPAP). j. Using the currently analyzed breath, for example, as described in Section K herein The normalized expiratory peak position NEPL is calculated by the method described above. k. Check the previous peak disturbance value. 13. The weighting factor is calculated as follows: Examples of individual weighting functions that may be used are given in the following sections. a.deweight = 1.0 b.deweight *= FFL_function_of_Leak(leak ) c.valve-like leak *= valve_like_leak_fu action_of_leak(leak) d.dweight *= FFL_function_of_valve_like _leak (valve-like leak) e.deweight *= FFL_function_of_pressure (EPAP) f.deweight *= FFL_function_of_NEPL (NEP L) g.deweight *= FFL_function_of_Jamming ( jamming) 14. Current threshold required for pressure rise (current_crit_FFL) is then calculated using the following formula: Current_crit_FFL = 1.0 - deweight * (1. 0 - crit_FFL) 15. The standard value of uncorrected crit_FFL is 0.05. Therefore, the total weight reduction After the reduction is applied, if deweight=1.0, current_cri t_FFL=0.05. Alternatively, if deweight=0.0, then cu rrent_crit_FFL=1.0. 16. The value of the pressure rise associated with the FFL (the FFL “module”) is now calculated. It is calculable and can be "predetermined." a.dp = 1.0 * (FFL - current_crit_FFL) b. if (dp > 0.0) then { max_dp = (EPAP-range-max - EPAP) if (dp > max_dp) then dp = max_dp if (dp > 0.0) then FFL_prescriptio n += dp } else FFL_prescription = decay (FFL_prescr iption, Ti+Te, 20) The final step is to increase the pressure over the duration of the breath, preferably with a time constant of about 20 minutes. It simply decays exponentially. 17. Now the pressure rise due to snoring can be calculated and "predetermined" and this A separate weighting function may be used as described in Section J of the specification. . a.Current_crit_snore = Snore_function_o f_Pressure (EPAP) b.Insp_snore *= Snore_function_of_VR(VR ) c.if (Insp_snore > Current_crit_snore) then { dp = 1.5 * (Insp_snore - Current_c rit_snore) max_dp1 = ttot * 0.2 if (dp > max_dp1) then dp = max_dp 1 max_dp2 = (EPAP_range_max - EPAP) if (dp > max_dp2) then dp = max_dp 2 if (dp > 0.0) then snore_prescript ion += dp } else snore_prescription = decay (sno re_prescription,Ti+Te,20) 18.Now the pressure rise due to apnea may be calculated. if ((apnoea_airway_closed && (apnoea_d uration > 10)) then { Head_room = (Max_Pressure_Apnoea - E PAP) if (headroom > 0.0) then { new_epap = Max_Pressure_Apnoea - h ead_room * exp(-ExpRiseTimeApnoea * apno ea_duration) if (new_epap > EPAP_range_max) the n new_epap = EPAP_range_max apnoea_prescription += (new-epap-E PAP) } } else apnoea_prescription = decay (ap noea_prescription,Ti+Te,20) 19. A new EPAP setting may then be calculated. a.EPAP = EPAP_range_min + apnoea_prescr iption + snore_prescription + FFL_prescr iption b. The new EPAP setting is then adjusted to a maximum mask pressure of 1.0 cmH2O per second. This is achieved by increasing the FG pressure at a rate that approaches the new EPAP value. Similarly, the therapeutic pressure is preferably increased only while the patient is inhaling. .

[0085] In yet another embodiment of the present technology, the flow limitation measure FFL is optionally calculated based on the respiratory effort. It is sometimes implemented as part of the Related Awakenings (RERA) detector. The Task Force defined RERA as follows: "Characterized by increased respiratory effort resulting in arousal from sleep, but not apnea or are sequences of breaths that do not meet the criteria for hypopnea. These events must meet both of the following criteria: is necessary. 1. Gradually improves the condition by ending with a sudden change in pressure and awakening to a smaller negative level. A pattern of greater negative esophageal pressure. 2. The event lasts for more than 10 seconds."

[0086] In 2000, at the New York University School of Pharmacy, Sleep, vol. 23, no. The study, "Non-Invasive Detecti on of Respiratory Effort-Related Arousal s(RERAs)by a Nasal Cannula / Pressure Tran The "Nasal Cannula / Pressure Transducer System" is a It was demonstrated to be suitable and reliable for the detection of ERA.

[0087] By utilizing the techniques described herein, the RERA detector detects flow generation It may be based on real flow signals obtained from a device, for example, any of the methods described above. The flow limitation index may be determined based on the flow signal. It can therefore be derived as a function of the flow restriction index and a further function of the index of increased ventilation. There is a saying.

[0088] Thus, in one embodiment, the RERA detector may be based on the following method. There is a previous flow restriction (for example, FFL is greater than 0 (FFL>0)), If a ventilation step change follows (eg, a big breath), then RERA is detected.

[0089] Preferably, the index is a continuous variable so that adjustments of thresholds based on experimental data can be made. This is done by setting a Boolean threshold for each input parameter that causes a loss of information. (This may be an alternative.) The following algorithm may be used:

[0090] 1. Keep track of the three most recent FFL values ​​in a rolling buffer. 2. Track the 3 most recent ventilation rates (VR is the ratio of average tidal volume to the current 3-minute ventilation) (It is.) 3. Sum the 3 most recent FFL values ​​and clamp the result to the range [0.0:1.0]. 4. Calculate the difference between the two most recent VRs (i.e., VR n -VR n-1 and VR n -VR n-2 ). 5. Take the maximum value from step 4 and simultaneously limit this maximum to a value greater than or equal to zero. 6. Multiply the result of step 3 by the result of step 5 to produce the RERA result. 7. The result of step 6 may be normalized by taking the square root of the result. 8. A threshold is set such that if the result of step 7 exceeds the threshold, a RERA is recorded. This may be the case. 9. Having a RERA score below the threshold during a breath recorded as RERA (multiple There is breath. 10. In the extreme case of UARS, there are only three detectable breaths between awakenings. Therefore, the buffer size is chosen as 3.

[0091] As shown in FIG. 14, the lower signal represents the continuous RERA detector signature based on the present method. The lower trace represents the respiratory effort-related signal based on arousal of the upper trace, which represents the flow signal. 1 illustrates the detection of arousal.

[0092] The RERA detector may also be implemented as part of the Respiratory Obstruction Index. The detector is further used to calculate the RDI (Respiratory Obstruction Index) in diagnostic mode according to the following formula: This may be the case. RDI = RERAs + apnea + hypopnea per hour

[0093] Alternatively, RERAs may be reported by the treatment device as an indication of treatment efficacy. Finally, the RERA index described herein is a useful tool for assessing the relevance of treatment algorithms. For example, the RERA index may be used as an input for The results may be determined over a certain time frame, and the results may include flow restrictions that increase pressure. Used in the "outer loop controller" to set the threshold (or gain) of the limiting exponent The outer loop controller embodiment may be used in conjunction with a and WO2005 / 051470 (PCT / AU2004 / 001652).

[0094] The ratio of peak expiratory flow to mean expiratory flow is another type of index for sleep-wake-wake detection. The ratio of peak expiratory flow to mean expiratory flow is sometimes used as the sleep-wake / FIG. 15 shows a hierarchical diagram comparing the wake standard to two sleep data sets. The waking breath tends to be "flattened", i.e. the peaks are By comparing the ratio with a threshold value such as 0.5, RERA may be thought to be detected.

[0095] Section A – Single Breath Flattening One embodiment of the Single Breath Flattening Index may be calculated by the following method: be.

[0096] 1. Frame your breath. 2. Extract the inhalation part of the breath. 3. Cut out any backward or forward poses as needed. 4. Interpolate the resulting intakes over a standard grid of N points (usually N=65) . 5. Divide the value of point y by a factor such that the breath area is normalized to 1 at unity base length do. 6. Calculate the rms value of the difference of the middle half of the points from 1.

[0097] The flow signal used to calculate the flattening is sometimes called (among other things) the "patient flow." This flow signal is filtered using a filter such as a "10Hz" filter to obtain the raw flow signal. The filter type may vary, but may be the result of a filtering process. Preferably, the present invention is adapted to reduce unwanted signal content, such as that caused by turbulence and snoring. The filter may also slightly attenuate cardiac content. It is a trade-off between accurate detection of points on the one hand and rejection of unwanted signals on the other. If the filter is made more aggressive to reject all cardiac vibrations, breath detection will become inaccurate. However, once breath detection is complete, further filtering can be performed if necessary. This is sometimes done.

[0098] In some cases, the filtered single breath flattening is determined as follows: There is.

[0099] 1. The raw flow is filtered in the usual way to give the patient flow. 2. Patient flow is used to frame the breath as usual. 3. The patient flow is continuously fed into an FIR filter which further attenuates the unwanted signal components. do. 4. FIR filters have a constant phase delay with respect to frequency, so simply It is possible to select the points required for the planarization calculation by taking into account the spread of the lattice. 5. Calculate flattening in the usual way.

[0100] Figure 16 shows that the normal flattening calculation is (in this case) adversely affected by low frequency snore. Figure 1 shows two superimposed flow signals recording a potentially obstructive breath sequence. One signal is filtered (and delayed) so that snoring is not evident. The filtered signal is based on another signal that clearly shows snoring. This gives a suitably lower value of flattening calculated based on the signal amplitude compared to a faster calculation.

[0101] In the illustrated signal of FIG. 16, the last breath is ) 0.21, and (based on the filtered signal) 0.13. The latter value indicates the required pressure rise, while the former indicates the pressure rise. As an alternative to such filtering, the average over the five traditional points is The filter can also achieve this filtered result, but at the expense of significant response delay. be.

[0102] A typical (and not computationally intensive) filter used is the boxcar FIR filter. A boxcar filter of length 12 is The response shown in Figure 17 is obtained by dividing the middle half of each intake by the middle half of the intake. If so, the impulse response of the filter is not important.

[0103] Section B - M Shape Index and Enhanced M Shape A suitable method for determining blockage indices such as the M-shape index is achieved using the following algorithm: This index detects the presence of an "M" shaped breath pattern. The index is sometimes commonly thought of as an indicator of a "u" shaped breathing pattern. may be augmented or modified by additional obstruction indices. As described in detail in US Pat. No. 6,999,463, the obstruction index is a first obstruction index, such as an "M" shape index. It is derived as a function of the obstruction index and a second obstruction index, such as "h" or chair shape index. This may happen.

[0104] For example, in some embodiments, each inspiration is interpolated over a grid of N points, such as N=65. In this embodiment, two basis functions are calculated as follows: t=i / (N-1), where i runs from 0 to N-1. B1 = sin(πt) B2 = sin(3πt) These basis functions are then used with all subsequent calculations of the M shaped exponents. Therefore, it may be memorized.

[0105] Each inspiration is then sampled and interpolated over a grid of N points. Two coefficients are then added: It is calculated as follows: F1 = sum(B1 fs) F2 = sum(B2 fs) where fs represents the interpolated intake point and · represents the dot product operator.

[0106] The final shape value is obtained by normalization as follows:

number

[0107] This shape factor is then made to vary between zero (purely sinusoidal) and 1 (super-M shaped). Figure 18 shows suitable functions that can serve as basis vectors B1 and B2. is plotted.

[0108] A typical M-shaped breath flow signal is plotted in the graph of Figure 19. Based on this, the calculation of the plotted breath is as follows: F1=4.6082 F2=2.6538 Shape index=0.50 A typical non-flow limiting breath may have an M shape exponent of only about 0.2.

[0109] As mentioned above, M-shape breathing can be enhanced by simple chair geometry. This is illustrated in Figure 1, which plots the flow signal with a flow restriction that produces an enhanced shape. The flow-limited inspiration shown in FIG. 20 (from t=0 to approximately t=2 seconds) , has a typical "chair" shape. Such intakes have two characteristics: 1) a piston A high ratio of peak to mean inspiratory flow, and 2)0 (as shown in Figure 20) Close to either 1 (left side chair back) or 2 (right side chair back not shown) In one embodiment, these indices may be characterized by a normalized peak position. can be calculated as follows: Normalized inspiratory peak location (NormPeakLoc):

number

[0110] During the ceremony, t0 is the time when inspiration begins, t end is the time of end of inspiration, t peak is the time of peak inspiratory flow. Ratio of Peak Inspiratory Flow to Average Inspiratory Flow (RPMIF): RPMIF=Q peak / Q' During the ceremony, Q peak is the maximum flow velocity during inspiration, Q' is the average flow velocity during inspiration.

[0111] During restful sleep, normal inspiration has a normalized peak position of approximately 0.5 and a peak inspiration of 1.35. The ratio of the mean inspiratory flow to the mean inspiratory flow is fuzzified to measure "chairness." These features are utilized based on the graphs in Figures 21 and 22. It is applied to the mathematical functions of each graph.

[0112] Once these fuzzy variables are calculated, the final fuzzy chairness index is calculated using the fuzzy The results are as follows: Fuzzy chairness = fuzzy logical product (fuzzy peak vs. average, fuzzy peak position)

[0113] Based on this determination, the intake of FIG. 20 has a fuzzy chairness of 0.82. The scaling of chairness features is M-shaped because the fuzzy chairness features are combined shape indices. The scaling is such that it can be directly fuzzy-ORed with the shape features. High fuzzy chairness means flow restriction. Fuzzy chairness is expressed in Japanese as follows: "The ratio of inhalation peak to inhalation average is high, and the peak inhalation flow location is at the beginning or end of inspiration. If the intake is close to completion, the intake is chair shaped. A typical flow sequence is shown in which the intake geometry is initially flat such that the F-exponent is approximately less than 0.2. However, the shape is such that the SBF index (SB flattening) is higher than about 0.2, The chair shape index is approximately 0.3 or higher. Change the shape.

[0114] Section C - Sustained Flattening The embodiment of the flow limitation index FFL outlined above provides a highly sensitive and accurate method for measuring flow limitation in a timely manner. It is designed to respond to flow limitation with both specificity and specificity to increased upper airway resistance. Increased respiratory effort at t=420 may prevent arousal. Diagram of a patient experiencing upper airway resistance ending with arousal from sleep (breathing in upper panel). Consider the trace shown in Figure 24. The lower panel of Figure 24 shows the single-breath flattening index. is.

[0115] Figure 25 explains why the FFL flow restriction index did not result in an increase in pressure in this series. The patient is then able to measure his or her inspiration as a function of total breath time (e.g., a TTR of approximately 1). Maintain your tidal volume (e.g., a VR of approximately 1) without extending the time Therefore, despite achieving low values ​​of the single-breath flattening index, FFL flow limitation The indicators do not reach the level required to adequately increase pressure.

[0116] To deal with this situation, the system may introduce an additional blockage indicator. For example, a first occlusion index may be used to filter a second occlusion index, such as a flattening index. This may be derived by processing the device to determine whether it has some kind of occlusion associated with a previous occlusion. In some cases, the historical values ​​may be based on history or filtering. The measured obstruction index is an indicator of continuous or persistent past obstruction. For example, it may be reset if non-blocking is achieved.

[0117] For example, "fuzzy persistent flattening" may be implemented. In other words, this fuzzy persistent flattening index responds essentially to consistently low values ​​of the flattening index. This index is furthermore intended to be used without interfering with the flow limitation index FFL and without overtreating the patient. The flow limitation index is designed to respond slowly in relation to the FFL flow limitation index so as not to cause any serious problems. Therefore, the system preferably filters the single breath flattening index in the following manner: Filter processing.

[0118] A simple first-order autoregressive digital filter is given by: Y n =y n-1 +G(x n -y n-1 ) where G is the gain of the filter. A time constant for the rate of breath may be determined. So, for example, initially, the breath lasts for 4 seconds. If it is considered to be the length, Time constant = τ = 5 × 4 / 2 = 10 To take into account that a breath may not be 4 seconds long, the breath detection algorithm The current breathing rate (RR, breaths per minute) can be determined as the average of the last five detected breaths. This may be used to obtain the number of breaths per minute. τ=5×60 / 2RR It is possible to implement this as an adaptive time constant given by: If the typical RR is 15 breaths per minute, the time constant can be calculated as It becomes 10 as expected.

[0119] The appropriate gain for the filter is simply G=1 / τ is given by:

[0120] The value to be filtered is the (filtered) single breath flattening index. To prevent similarity values ​​from adversely affecting the filtering process, the system Therefore, the input value may be "maximum limited". The high-level flattening (HLF) index may be determined as follows: If (SBF index>0.3), HLF=0.3 Otherwise, HLF = SBF index.

[0121] The output of the filter is a persistent flattening factor (PF) that does not initially affect treatment. As such, it is possible to initialize it to an arbitrarily high value, such as PF=1.

[0122] Another characteristic of the filter is that the gain is increased for inputs greater than the current value of PF as follows: The advantage of this approach is that a different gain is used for inputs less than the current value of PF. HLF <PF ならば、 PF=PF+G(HLF-PF) Otherwise, PF=PF+3*G(HLF-PF) This asymmetry causes the filter to fall off slowly for any sustained low input, This means that the flow limitation index FFL is reset relatively quickly. The noise threshold is set so that there is a response only to values ​​of flattening that are consistently low over a short time frame. Help to improve.

[0123] The value of PF is then prevented from deviating from a reasonable value as follows: If (PF>0.2), then PF=0.2 Otherwise, PF=PF

[0124] Finally, PF is a fuzzy variable, e.g., by using the following formula: It may be mapped to the permanent flattening (FPF). FPF = (if PF>0.19, then 0.0, otherwise (0.19-PF) / (0 .19-0.05) FIG. 26 shows the FPF response to the same sequence of breaths shown in FIG.

[0125] Section D - Poor Breathing Framing In some cases, breath detection algorithms may frame breaths inaccurately, and The patient's coughing or swallowing is a respiratory event that provides little information about the current state of the airway. Selective heuristics may not be related to breath weight based on inspiration time. Therefore, the derived occlusion index may be inaccurately framed. If a breath pattern representative of a rhythmic breath is detected, it may be adjusted. ,Figure 27 shows the weighting factor that may be multiplied by the ,flow limitation index (e.g., FFL). The function for a small number of coefficients is shown. i >2.5 seconds) or short (T i <0.7 seconds) are gradually de-weighted.

[0126] Section E - Snoring Entropy Traditionally, snoring has been measured using a calibrated inspiratory snoring index, which can be considered an indicator of obstructive snoring. Figure 28 shows a typical snoring sequence captured during a patient's sleep. The upper panel shows the respiratory flow rate, and the middle panel shows the instantaneous snoring (e.g., The top panel shows the snoring index (an index of sound power in the frequency range), and the bottom panel shows three snoring indices. Calibrated Inhalation Snoring Index (named "Snoring Index" in Figure 28) is calculated as follows: Inspiration is framed using the respiratory flow signal. The instantaneous snore signal is a background signal caused by current conditions (e.g., set pressure, turbine speed, etc.) It is adjusted to take noise into account. The snoring index is the average of the instantaneous snoring signal during each inspiration (i.e., approximately the average per inspiration). The noise level is calculated as the acoustic power.

[0127] This technology works by increasing mask pressure in proportion to the measured snoring index. It is reliable when used to treat snoring. However, this technique requires calibration constants that flow Each combination of transducer and mask must be measured and stored in the device's non-volatile memory. Such calibration is time consuming and therefore costly, and is subject to aging effects. may be subject to

[0128] Matthew Alder et al., PCT Application No. PCT / AU2007 In issue / 000002, an alternative, named Delta Snore in Figure 28, , Delta snore is calculated by subtracting the average of the instantaneous snore signal during inspiration from the average of the instantaneous snore signal during expiration. This delta snore is essentially calculated as the difference between the background noise and the It is a self-calibration index that assumes that the sound source changes little between inspiration and expiration. However, This assumption is not always true.

[0129] In another alternative method, an inhalation snore entropy method is implemented. The obstructive index is the frequency of snoring-related frequencies (e.g., 30 to 300 Hz). The filter in this frequency range is determined by filtering the respiratory flow index. The power or energy of the processed signal is then calculated by the power signal being To assess whether the suction shape represents the actual intake shape or is simply random noise, For example, the Shannon entropy function may be used. If the function shows that the energy signal is nothing but random information or noise, then the true Since snoring may not be occurring, the snoring index is adjusted or deweighted. It can be reduced.

[0130] Such methods use information contained in the instantaneous snore signal during inspiration. The index also does not require calibration. For example, the occlusion index may be calculated as follows: Construct a vector S of length n that contains the instantaneous snore values ​​for the inspiration of interest. If n<2 or n>any large value, abort and return zero. Subtract the lower limit of the snore vector and add 1. S = S - min(S) + 1 Calculate the area A.

number

[0131] If A is less than any very small value, stop and return zero. Normalize the snore vector. S=S / A Calculate the Shannon entropy (se) of the normalized snore vector.

number

[0132] In some cases, the results are scaled to approximate the inspiratory snore index as follows: Inhalation snore entropy = 10.0*se

[0133] Both delta snore and snore entropy are plotted in the lower panel of Figure 28. As shown, the “snoring entropy” trace shows low mean snoring values ​​(t= 100) tend to capture sharp snores. Delta snore and snore entropy A smart combination of both can accurately classify snoring in almost all cases. OK, the graph in Figure 29 shows one such sequence. The plotted line The classifier shown by If delta snoring is negative, then there is a gradually increasing negative delta snoring that needs to be addressed. It requires snoring entropy.'

[0134] Section F - Improved Forward Pose Trim Some pressure therapy devices include a feature known as "forward pause trim." The function detects the peak and extrapolates back to the appropriate zero crossing which will represent the start of inspiration. This adjusts the current frame to remove any expiratory pauses continuing from the previous exhalation. The intake duct was designed to cut off the front of the modularized intake. This function is shown in Figure 30. As such, it may not be useful for M-shaped breaths. 4A-4B show an imprecise trim (relative to the front peak) and an accurate trim (relative to the front peak).

[0135] This method involves finding a boundary where the peak of inspiration is found, which can be assisted with an M-shape breath. Therefore, for an M-shaped breath, the start of inspiration occurs at the beginning of inspiration. Based on the boundaries of the range, the outlier from the first peak but not the second peak of the inhalation part of the breath For example, this boundary might be calculated as follows:

[0136] 1. Calculate the total intake.

number

number

[0137] The plot of FIG. 31 shows a flow signal in which the breath frequency is similar to that shown in FIG. The upper panel of FIG. 31 shows a sequence of restricted breathing. Flow rate plotted against time. The lower panel shows the calculated inspiratory time (T i ) The results of the current algorithm and the new algorithm (the "corrected" T i ) result Both are shown. The current algorithm cuts some breath in half, making it artificial and fraudulent. Certainly a low value of T i The new algorithm is a consistent form of "forward pass". We will implement a "Zutrim" program.

[0138] Section G - Ventilation and Duty Cycle Indicators (1) Calculation of ventilation rate (VR) Ventilation indices such as ventilation rate (VR) relate the current breath to the previous mid-breath (V3). In this example, the mid-term has a time constant τ of 3 minutes. However, other time constants may be used. It may also be appropriate. The filter used may be a simple first-order autoregressive filter. The time constant of the filter is quite long, so the filter takes a certain amount of time to rise from zero. The index thus changes gently between a reasonable ventilation value and the filter output during the time [t0:3×τ]. VR can be calculated as follows.

[0139] 1. Set the gain of the filter to G = 1 / f s τ where f s is the sampling frequency and τ is the time constant of the filter in seconds. For example, if the sampling frequency is 50 Hz and the time constant is 3 minutes, G is G = 1 / (50×180) calculated by. 2. Initialize the filter to a reasonable ventilation value such as 0.2 liters / second. 3. Subtract the vent flow with any mask leakage subtracted from the total flow to obtain the properly filtered patient respiratory flow (Q p ). 4. Calculate the mid-term ventilation. V3 = V3 + G(Q p - V3) 5. During the transition period 0 < t ≤ 3τ V3 = (3τ - t) / 3τ × 0.2 + t / 3τ × V3 6. Frame the breaths in the normal way. 7. When inspiration is confirmed, perform the following calculations. Inspiratory volume: V i Inspiratory time: T i Average inspiratory flow rate:

Number

Number

number

number

[0140] (2) Calculation of duty cycle indicators (e.g., TTR) The duty cycle index is introduced to derive the obstruction index as described above. For example, from the respiratory flow index, the respiratory cycle as a function of the respiratory flow index can be calculated as follows: is the second ratio of the duration of the inspiratory portion of the respiratory cycle to the duration of the A second such indicator may be determined which may follow the first indicator in due course. An index of may then be derived as a function of the first ratio and the second ratio.

[0141] T i -on-T tot A suitable duty cycle index, such as the TTR, is i -on-T tot Current (breath) T for value i -on-T tot It is determined as a ratio of the values The mid-term may be filtered using a 5-minute time constant or other suitable time constant. The filter can be a simple first-order autoregressive filter. Good. TTR may be calculated as follows:

[0142] 1. Filter gain G=1 / f s τ where f s is the sampling frequency and τ is the filter time constant in seconds For example, if the sampling frequency is 0.25 Hz (approximately the breath frequency) and the time If the number is 5 minutes, G is G = 1 / ((1 / 4) x 300) It is calculated by: 2. Initialize the filter to a reasonable value, such as 0.4. 3. Total flow minus vent flow, which is subtracted for any mask leakage, and properly flushed. Filtered patient respiratory flow (Q p ) to calculate 4.Medium T i -on-T tot is calculated as follows: T i T tot(5) =T i T tot(5) +G(T i T tot -T i T tot(5) )

[0143] Calculate TTR as follows: TTR=TiTtot / TiTtot5

[0144] Section H - Weight Reduction Functions The functions graphed in Figures 32 and 33 are indices of flow limitation (e.g., FFL) or This function can be used to reduce the weight of the influence of the snoring indicator. What indicators (e.g. FFL or snoring) need to be exceeded to trigger a pressure increase? It is possible to apply a weighting function to increase the threshold. For example, The valve-like leakage index is a variable value between 0 and 5. It is possible to calculate in the usual way to make the leakage smaller. The system is fully sealed to prevent pseudo valve-like leaks from impeding pressure build-up. This is done to reduce the weight of the valve-like leak value to the extent that no leak is present. Therefore, the algorithm for determining the weight reduction due to valve-like leakage is as follows: It is.

[0145] 1. Calculate valve-like leakage from the previous intake. 2. Determine the value of the perfect leak (value at the end of expiration). 3. The valve-like leakage value is fitted with a complete function like that shown in the left graph of Figure 32. Multiply the output of the total leak function. For example, if the total leak is less than (<) 0.025, If so, valve-like leakage is set to zero. Alternatively, if complete leakage is greater than 0.05 ( >), the valve-like leakage value remains unchanged and is between 0.025 and 0.05. In the section, the value of valve-like leakage decreases linearly. 4. To output the weight reduction coefficient, use the function shown in the right graph of Figure 32. Use the valve-like leakage value together with the leakage function. For example, set the value to be less than (<) 4. For values ​​between 4 and 5, the output is 1, and for values ​​greater than (>) 5, the output is zero. In the interval, the output decreases linearly from 1 to zero.

[0146] De-weighting of flow limitation index (e.g., FFL) based on leakage (e.g., L / s) This can be done using the function shown in the graph of Figure 33. Leaks less than (<) 0.5 For this value, there is no weighting reduction and the output is 1. For leak values ​​greater than (>) 0.7, For leakage values ​​between 0.5 and 0.6, there is a complete weight reduction and the output is zero. There is a linear decrease in output from

[0147] Deweighting of flow limitation indices by pressure, such as mask CPAP levels in cmH2O may be implemented using a function such as the example graphed in FIG. In the example, there is no weight reduction for pressure levels below 10cmH2O, and above 15cmH2O For high pressures there is a partial weighting reduction and the output is 0.8. For pressures between 10 and 15 There is a linear decrease in output power from 1 to 0.8 as the force increases. The output power then decreases to the flow limit. It may be multiplied by the sign.

[0148] Finally, the influence of flow limitation or obstruction indices may be de-weighted with increasing obstruction. The disturbance is now indicated by a fuzzy range where inspiration or expiration can continue for too long. For example, interference is disclosed in U.S. Pat. High interference may be determined by the method described in US Pat. No. 6,484,719. For example, when the patient opens their mouth or when the patient moves to the bed and the mask is placed on the face. This represents the transient change in leakage when changing the position of the valve. When the level of obstruction is high, the flow estimate It is possible that the leak constant may be reduced to help improve the flow estimation. While this is happening, it is prudent to avoid a pressure rise until the event has settled. The avoidance can be realized by the function example of Figure 35. When the disturbance reaches 0.25, the system The system can start deweighting until the indicator is fully deweighted at a disturbance level of 0.5. The output of the function may be multiplied by a measure of flow limitation to achieve weighting. .

[0149] Similarly, the snoring indicator may be de-weighted based on various conditions of the system. For example, the snoring value is de-weighted by the function of pressure shown in FIG. This de-weighting may be used to estimate the effect of the snoring index for the purpose of generating a pressure rise. This reduces the sensitivity of the system to the treatment pressure. increases with increasing pressure.

[0150] The inhalation snoring index may be weighted down by ventilation indices such as the "big breath" index. A big breath often induces noise simply due to the high peak flows that are achieved. Using the function example of FIG. 37, ventilation is considered to be a full breath. This may be used to generate weighted output coefficients depending on whether In some cases, ventilation indices such as ventilation rate (VR) are used as an index of breath. Values ​​above (>) the threshold VR may be interpreted as indicating greater than mid-ventilation.1 For values ​​of VR less than (<) .2, the system may reduce the weight of any snoring indicator. For values ​​of VR greater than (>) 1.5, the system may avoid the effect of snoring. For VR values ​​between 1.2 and 1.5, the output The weighting factor may decrease linearly from one to zero.

[0151] The effects of flow restriction or snoring based on leakage values, pressure, obstruction and / or ventilation indices Other corrective calculations may also be used.

[0152] Section I Normalized Expiratory Peak Position Normalized expiratory peak position (NEPL) is the time course of awakening from sleep or obstructive sleep flow waveforms. waking up or other unnatural, unusual, or unrecognizable symptoms such as significant nausea or vomiting The peak expiration time is a good indicator of the transition to a waveform that represents a non-expiratory event. One may introduce indicators of arousal based on flow, e.g., the duration of the expiratory portion of the respiratory cycle. Depending on the location of the peak expiratory flow or normalized peak expiratory flow within the range, arousal is assessed. In such an embodiment, the duration of the exhalation portion may be defined by a range. The appearance of a peak within a defined range defines an index representing arousal. , if the expiratory peak occurs during the latter part of the respiratory cycle or during the expiratory part of the respiratory cycle, , can represent awakening.

[0153] Such an index may be calculated as follows:

[0154] 1. Frame your breath. 2. Isolate the exhalation portion of each breath. 3. Find the time when peak expiratory flow occurs. 4. Divide the time from the beginning of the exhalation to the peak by the total exhalation time, for example, the exponent is [0 .0:1.0] range.

[0155] Typically, while a patient is sleeping, the NEPL lies in the range of zero to 0.3. For example, consider the histogram in Figure 38, which shows a comparison of three data sets corresponding to the NEPL. do. (1) Essen - OSA patients undergoing treatment (2) Concord - Patients receiving gradually increasing doses (3) Awake - AutoSet Sprit Airway available from ResMed Person breathing through pressure equipment.

[0156] The graph on the right of Figure 38 shows that awake breathers breathe faster than sleeping patients who take fewer breaths. This shows that the system takes a much larger number of breaths than the 0.5 breaths. ,The ,detection ,of ,unusual, ,unnatural, ,or ,unexpected ,events ,related ,to ,sleep ,based ,on ,flow ,limitation ,indicators. This can be accomplished by using a weighting function for the normalized absorption coefficient. A suitable function based on the peak position values ​​is shown in the graph of FIG. If it becomes more unnatural, unusual, or unrecognizable (e.g., bizarre) to increase the magnitude of the flow limitation index required to induce an increase in therapeutic pressure. It is used for this purpose.

[0157] In the above description and in the accompanying drawings, certain terms and drawing symbols have been used to provide a thorough understanding of the present technology. In some cases, the terms and symbols are used to In addition, the technology herein may refer to specific details that are not required for the purposes of the present invention. Although the present invention has been described with reference to certain embodiments, these embodiments are intended to be illustrative of the principles and applications of the present technology. It should be understood that the present invention is merely an example of the present invention. Thus, numerous variations are possible in the exemplary implementation. Other arrangements are contemplated without departing from the spirit and scope of the present technology. It should be understood that the above-mentioned can be obtained.

[0158] For example, a desirable flow limitation detection and / or therapy system would capture each inspiration and Interpolate over a grid of as many points as needed to preserve the desired frequency information. It could be based on other techniques for pattern recognition, such as: may be interpolated on a grid of 65 points. A large set of training data is It may be constructed using various types of occlusion waveforms that can be recorded based on the These waveforms may be classified as mild, moderate, severe, or other similar categories, as appropriate. The 65 points for each waveform are then used for learning purposes. (such as neural networks, support vector machines, or other) The input to the classifier is a suitable algorithm, such as a genetic algorithm. Using the algorithm, any boundary can be found in the 65-dimensional space between the various categories of occlusion. Such a system can then be compared against the classifier metrics. The inspiratory wave is measured in the patient and used in the control system for the treatment of the patient. Ability to classify shapes.

[0159] Potential issues to consider with this approach include:

[0160] 1. The "difference in dimensionality" is the exponential growth of hypervolumes as a function of dimensionality. In other words, as the dimensionality of the input space increases, the The number of training vectors required to "make a 65-dimensional Covering can present problems with the speed of this technique. 2. A priori classification of waveforms with respect to (and degree of) obstruction is cumbersome and There is probably high interobserver variability. In the absence of an increase in CPAP pressure, it is difficult to determine whether an obstruction actually exists. A reliable indicator of effort, e.g., esophageal pressure, may be useful in solving the detection problem. However, the task of sorting through such data in decision making can remain intractable. For example, REM sleep can produce a variety of unexpected waveforms. 3. The training phase requires significant numerical resources and many processor cycles. do. 4. The resulting classifier is difficult for anyone to interpret (e.g. If you use a neural network, depending on how many neurons you end up with) Running an embedded system can be numerically intensive. 5. To test the resulting classifier, we test it on the sleep patterns of every person on Earth. There may be a need to provide sleep studies.

[0161] However, one way to reduce the complexity of a system is to label the information fed into the system as "relevant The calculation of characteristic features such as the smoothing index is precisely This restriction acts as a form of compression of the signal.

Claims

1. 1. A method for detecting a partial occlusion, comprising: determining a measure of respiratory flow; determining a shape index representative of a pattern of partial obstruction from the respiratory flow indices; determining a ventilation index from the respiratory flow index; A method for deriving the flow restriction index as a function of the determined shape index and the determined ventilation index. Tep and A method comprising:

2. The method of claim 1 , wherein the pattern of partial occlusion is an M-shaped flow detection function.

3. 3. The method of claim 2, wherein the ventilation index is a ratio comprising a current tidal volume and a previous tidal volume. How to.

4. 4. The method of claim 3, further comprising the step of calculating a pressure demand as a function of the flow limitation index. Method of posting.

5. The a priori pressure demand is determined based on (a) the shape index indicating the presence of an M-shape breath in the respiratory airstream; and and (b) an increase as a condition for the aeration index to decrease sufficiently to represent less than normal aeration. The method according to claim 4 .

6. A second shape index representative of a chair pattern of partial obstruction determined from the respiratory flow index is determined. wherein the derived flow restriction index is a further function of the second shape index. The method according to claim 5 .

7. 1. A method for detecting a partial occlusion, comprising: determining a measure of respiratory flow; determining a shape index representative of a pattern of partial obstruction from the respiratory flow indices; determining a duty cycle index from the respiratory flow index; The flow restriction index is derived as a function of the determined shape index and the determined duty cycle index. The step of issuing; A method comprising:

8. The method of claim 7 , wherein the pattern of partial occlusion is an M-shaped flow detection function.

9. Duty cycle indicator is a function of the ratio of current inspiration time to cycle time and cycle time. and a ratio of the average prior inspiration time to the average prior inspiration time.

10. 10. The method of claim 9, further comprising the step of calculating a pressure demand as a function of the flow limitation index. Method of posting.

11. The prior pressure demand is determined based on the shape index indicating the presence of an M-shape breath in the respiratory airstream, and 11. The method of claim 10, wherein the load cycle index is increased as a condition of an increase.

12. A second shape index representative of a chair pattern of partial obstruction determined from the respiratory flow index is determined. wherein the derived flow restriction index is a further function of the second shape index. The method of claim 11 .

13. 1. An apparatus for detecting a partial occlusion, comprising: a patient interface for communicating a flow of respiratory gas; A flow rate of respiratory gas through the patient interface is controlled by the patient interface. a flow sensor generating a flow signal representative of the a controller coupled to the flow sensor for processing the flow signal; Equipped with The controller (a) determining a shape index representative of a pattern of partial obstruction from the flow signal; (b) determining aeration index from the flow signal; and (c) Derivation of the flow restriction index as a function of the determined shape index and the determined ventilation index. Out, The apparatus is configured to control.

14. The apparatus of claim 13 , wherein the pattern of partial occlusion is an M-shaped flow detection function.

15. 15. The method of claim 14, wherein the ventilation index is a ratio including a current tidal volume and a previous tidal volume. The apparatus described.

16. a flow generator coupled to the controller and the patient interface; A controller (a) calculates a pressure demand as a function of the flow limitation index; and (b) calculates a pressure demand. and further configured to configure the flow generator according to 14. The apparatus of claim 13.

17. The controller determines: (a) that the shape index indicates the presence of an M-shape breath in the respiratory airstream; and (b) The condition for the pre-pressure to be sufficiently reduced to indicate less than normal ventilation.

17. The apparatus of claim 16, further comprising: increasing force demand.

18. The controller determines a second shape representing a chair pattern of partial occlusion from the flow signal. and further configured to control the determination of the index, the derived flow restriction index being a second shape index. The apparatus of claim 13 , wherein the θ is a further function of

19. 1. An apparatus for detecting a partial occlusion, comprising: a patient interface for communicating a flow of respiratory gas; A patient interface is connected to the patient interface to represent the flow of respiratory gas through the patient interface. a flow sensor for generating a flow signal; a controller coupled to the flow sensor for processing the flow signal; Equipped with The controller determining a shape index representative of a pattern of partial obstruction from the flow signal; Determining a duty cycle index from the flow signal; and of the flow restriction index as a function of the determined shape index and the determined duty cycle index Derivation and The apparatus is configured to control.

20. 20. The apparatus of claim 19, wherein the pattern of partial occlusion is an M-shaped flow volume detection function.

21. Duty cycle index is the ratio of current inspiration time to cycle time and previous average inspiration time 21. The apparatus of claim 20, wherein the ratio comprises a ratio of the amount of heat generated to the amount of cycle time.

22. a flow generator coupled to the controller and the patient interface; A controller (a) calculates a pressure demand as a function of the flow limitation index; and (b) calculates a pressure demand. and further configured to configure the flow generator according to 20. The apparatus of claim 19.

23. The controller determines whether or not the shape index is indicative of the presence of an M-shaped breath in the flow signal; and (b) increasing the pre-pressure demand as a condition of an increase in the duty cycle index. The apparatus described in

24. The controller determines a second shape representing a chair pattern of partial occlusion from the flow signal. and further configured to control the determination of the index, the derived flow restriction index being a second shape index.

20. The apparatus of claim 19, wherein the metric is a further function of

25. 1. A system for detection of a partial occlusion, comprising: an interface means for communicating a flow of respiratory gas; A device coupled to the interface means and representing the flow of breathing gas through the interface means. a flow measurement means for generating a flow signal; a processing means coupled to the flow measurement means for processing the flow signal; Equipped with The processing means: (a) determining a shape index representative of a pattern of partial obstruction from the flow signal; (b) determining aeration index from the flow signal; and (c) Derivation of the flow restriction index as a function of the determined shape index and the determined ventilation index. Out, A system that is configured to process

26. 26. The system of claim 25, wherein the pattern of partial occlusion is an M-shaped flow detection function.

27. 27. The method of claim 26, wherein the ventilation index is a ratio including a current tidal volume and a previous tidal volume. The system described.

28. A respiratory gas supply is coupled to the processing means and the interface means and passes through the interface means. a flow means for generating a controlled flow of gas; A processing means (a) calculates a pressure demand as a function of the flow limitation index; and (b) responds to the pressure demand. and further configured to configure the flow generator accordingly.

26. The system of claim 25.

29. The processing means determines: (a) that the shape index is indicative of the presence of an M-shaped breath in the respiratory airstream; and (b) ) the pre-pressure requirement as a condition for the ventilation index to decrease sufficiently to represent less than normal ventilation.

30. The system of claim 28, further comprising: increasing the value of the requested value.

30. The controller determines a second shape representing a chair pattern of partial occlusion from the flow signal. and further configured to control the determination of the index, the derived flow restriction index being a second shape index.

26. The system of claim 25, wherein the value is a further function of:

31. 1. A system for detection of a partial occlusion, comprising: an interface means for communicating a flow of respiratory gas; a respiratory flow of respiratory gas through the interface means; a flow measurement means for generating a flow signal representative of the a processing means coupled to the flow measurement means for processing the flow signal; Equipped with The processing means: determining a shape index representative of a pattern of partial obstruction from the flow signal; Determining a duty cycle index from the flow signal; and of the flow restriction index as a function of the determined shape index and the determined duty cycle index Derivation and A system configured to control

32. 32. The system of claim 31, wherein the pattern of partial occlusion is an M-shaped flow detection function.

33. Duty cycle index is the ratio of current inspiration time to cycle time and previous average inspiration time 33. The system of claim 32, wherein the ratio comprises a ratio of the time to cycle time.

34. A respiratory gas supply is coupled to the processing means and the interface means and passes through the interface means. a flow means for generating a controlled flow of gas; A processing means (a) calculates a pressure demand as a function of the flow limitation index; and (b) responds to the pressure demand. and further configured to configure the flow generator accordingly.

32. The system of claim 31.

35. The processing means determines: (a) that the shape index indicates the presence of an M-shaped breath in the flow signal; and (b) ) further configured to increase the pre-pressure demand value as a condition of an increase in the duty cycle index. The system of claim 34 .

36. The processing means determines a second shape index representative of a chair pattern of partial obstruction from the flow signal. and further configured to control processing the determination of the derived flow limitation index as a second shape index.

20. The apparatus of claim 19, wherein the metric is a further function of

37. having processor readable information for controlling a device for detecting partial respiratory obstruction. An information-bearing medium, the processor-readable information comprising: determining a measure of respiratory flow; determining a shape index representative of a pattern of partial obstruction from the respiratory flow indices; determining a ventilation index from the respiratory flow index; A method for deriving the flow restriction index as a function of the determined shape index and the determined ventilation index. Tep and An information-holding medium comprising:

38. 38. The information bearing medium of claim 37, wherein the pattern of partial occlusion is an M-shaped flow detection function.

39. 39. The method of claim 38, wherein the ventilation index is a ratio including a current tidal volume and a previous tidal volume. The information bearing medium.

40. The processor readable information includes calculating a pressure demand as a function of a flow restriction index.

40. The information bearing medium of claim 37, further comprising:

41. The a priori pressure demand is determined based on (a) the shape index indicating the presence of an M-shape breath in the respiratory airstream; and and (b) an increase as a condition for the aeration index to decrease sufficiently to represent less than normal aeration.

41. The information bearing medium of claim 40.

42. The processor readable information indicates a chair-part of a partial obstruction determined from a respiratory flow index. determining a second shape index representative of a turn; 38. The information bearing medium of claim 37, wherein: is a further function of the second shape exponent.

43. having processor readable information for controlling a device for detecting partial respiratory obstruction. An information-bearing medium, the processor-readable information comprising: determining a measure of respiratory flow; determining a shape index representative of a pattern of partial obstruction from the respiratory flow indices; determining a duty cycle index from the respiratory flow index; The flow restriction index is derived as a function of the determined shape index and the determined duty cycle index. The step of issuing; An information-holding medium comprising:

44. 44. The information bearing medium of claim 43, wherein the pattern of partial occlusion is an M-shaped flow detection function.

45. Duty cycle indicator is a function of the ratio of current inspiration time to cycle time and cycle time. and a ratio of the average previous inhalation time to the average previous inhalation time. body.

46. The processor readable information includes calculating a pressure demand as a function of a flow restriction index.

44. The information bearing medium of claim 43, further comprising:

47. The prior pressure demand is determined based on the shape index indicating the presence of an M-shape breath in the respiratory airstream, and 47. An information bearing medium according to claim 46, increased as a condition of an increase in the duty cycle index.

48. The processor readable information indicates a chair-part of a partial obstruction determined from a respiratory flow index. determining a second shape index representative of a turn; 44. The information bearing medium of claim 43, wherein: is a further function of the second shape exponent.

49. 1. A method for determining treatment settings for a respiratory treatment device, comprising: determining a measure of respiratory flow; determining a shape index representative of the degree of partial obstruction from the respiratory flow index; determining a ventilation index indicative of a degree of change in ventilation from the respiratory flow index; Treatment settings were calculated as a proportional function of both the degree of partial obstruction and the degree of change in ventilation. and A method comprising:

50. 50. The method of claim 49, wherein the shape index is determined from a number of breaths in the range of 1 to 3 breaths. Method of posting.

51. 50. The method of claim 49, wherein the shape index is an M-shape index.

52. 50. The method of claim 49, wherein the therapy setting is applied to set an auto-titration respiratory therapy device. The method described.

53. 1. An apparatus for determining settings for respiratory therapy, comprising: a sensor for determining a measure of respiratory flow; (a) determining a shape index representing the degree of partial obstruction from an index of respiratory flow; and (b) (c) determining an aeration index representative of the degree of change in aeration from the indices; and A process configured to derive treatment settings as a proportional function of both the degree of change. And, An apparatus comprising:

54. A shape index is determined by the processor from multiple breaths ranging from one to three breaths.

54. The apparatus of claim 53.

55. 54. The apparatus of claim 53, wherein the shape index is an M shape index.

56. The patient interface further includes a flow generator configured to provide respiratory pressure therapy to the patient interface. An apparatus for detecting a temperature difference, comprising: The processor is configured to control therapy to the patient interface at a therapy setting. It was made, the apparatus comprises an auto-titration respiratory therapy device; 54. The apparatus of claim 53.

57. 1. An apparatus for determining treatment settings for respiratory therapy, comprising: a means for determining a respiratory flow index; means for determining a shape index representative of the degree of partial obstruction from the respiratory flow index; means for determining a ventilation index indicative of a degree of change in ventilation from the respiratory flow index; The treatment settings were derived as a proportional function of both the degree of partial obstruction and the degree of change in ventilation. The means to issue An apparatus comprising:

58. A shape index is determined by the processor from multiple breaths ranging from one to three breaths.

58. The apparatus of claim 57.

59. means for generating and delivering respiratory pressure therapy to the patient in response to the derived therapy settings. The apparatus further comprises an auto-titration respiratory therapy device.

58. The apparatus of claim 57.

60. 1. A method for determining a change in therapy in a device for treating sleep disordered breathing, comprising: determining a first indicator of the extent of the partial obstruction from the indicator of respiratory flow; determining a second index of the respiratory parameter from the index of respiratory flow; The first index is used to determine the degree of partial obstruction based on the respiratory parameters. adjusting the target; A method comprising:

61. Adjusting the first metric upon detection of a breath pattern representing an incorrectly framed breath.

61. The method of claim 60, further comprising the step of:

62. generating treatment settings as a proportional function of the adjusted first index; 61. The method of claim 60 comprising:

63. 63. The method of claim 62, wherein the first index is a shape index.

64. The second metric is the ratio of the duration of the inhalation portion of the respiratory cycle to the duration of the respiratory cycle.

64. The method of claim 63, wherein the degree of change in rate is expressed.

65. 65. The method of claim 64, wherein the second index is a function of the relative duration of inspiration and expiration.

66. The method further comprises adjusting the adjusted first index as a function of peak flow. The method of claim 65.

67. 67. The method of claim 66, wherein the peak flow is a peak expiratory flow location.

68. 61. The method of claim 60, wherein the first index is a shape index determined from a single breath.

69. 69. The method of claim 68, wherein the shape index is a single breath flattening index.

70. Breath framing is performed prior to filtering to remove frequencies associated with the patient's snoring.

69. The method of claim 68,

71. The flattening is determined prior to filtering to remove frequencies associated with the patient's snoring.

69. The method of claim 68.

72. In a device for treating sleep-disordered breathing, the change in the therapeutic pressure to be delivered by the device is A method for automatically determining, determining a first indication of the extent to which patient inspiratory effort correlates with patient inspiratory flow; determining a second measure of patient ventilation; determining a change in therapeutic pressure as a first function of said first index and said second index; Steps and A method comprising:

73. 73. The method of claim 72, wherein the first indicator is an index of partial obstruction.

74. 73. The method of claim 72, wherein the second indication of patient ventilation is an indication of the degree to which patient ventilation is adequate. The method described.

75. 73. The method of claim 72, wherein the first metric is a metric of the relative duration of inspiration and expiration. Law.

76. A device for treating sleep-disordered breathing automatically determines an index of partial obstruction of a patient's airway. The method comprises: determining a measure of respiratory flow; determining a first partial obstruction measure from the respiratory flow measure; determining a second partial obstruction measure from the respiratory flow measure; determining a third partial obstruction index as a function of the first partial obstruction index and the second partial obstruction index; and A method comprising:

77. 77. The method of claim 76, wherein the first obstruction index is an index representative of an M-shape breathing pattern.

78. 78. The method of claim 77, wherein the second obstruction index is an index representative of a chair shape breathing pattern.

79. A device for treating sleep-disordered breathing automatically determines an index of partial obstruction of a patient's airway. The method comprises: determining a measure of respiratory flow; determining a first partial obstruction measure from the respiratory flow measure; filtering the first partial occlusion indicator to derive a historical indicator of partial occlusion; P and resetting the filter upon detection of a non-obstructed breath pattern from the respiratory flow index. and, A method comprising:

80. 80. The method of claim 79, wherein the first partial occlusion index is a flattening index.

81. A device for treating sleep-disordered breathing automatically determines an index of partial obstruction of a patient's airway. The method comprises: determining a measure of respiratory flow; filtering the respiratory flow measure in a frequency range representative of snoring; of the frequency domain as a function of time to assess whether the power signal represents noise. Analysing the magnitude of the power; A method comprising:

82. The analyzing step applies a Shannon entropy function to the power magnitude in the frequency domain.

82. The method of claim 81, comprising the step of:

83. 1. A method for automatically determining the start of inspiration in an apparatus for treating sleep disordered breathing, comprising: determining a measure of respiratory flow; detecting a first peak from the respiratory flow; determining the start of a respiratory cycle as a function of the first peak; Equipped with The detection of the first peak is limited to avoid detection of the second peak during the respiratory cycle. There is a way.

84. A device for treating sleep-disordered breathing automatically determines an index of partial obstruction of a patient's airway. The method comprises: determining a measure of respiratory flow; Inhalation portion of the respiratory cycle versus duration of the respiratory cycle as a function of respiratory flow index determining a first ratio of minute durations; The duration of the inspiratory portion of the respiratory cycle relative to the duration of the respiratory cycle as a function of respiratory flow. determining a second ratio of the duration; deriving an indication of partial occlusion as a function of the first ratio and the second ratio; A method comprising:

85. A device for treating sleep-disordered breathing automatically determines an index of partial obstruction of a patient's airway. The method comprises: determining a measure of respiratory flow; determining an index of partial obstruction from the index of respiratory flow; determining an index of obstruction from the respiratory flow, the index being indicative of transient changes in leakage; , adjusting the measure of partial occlusion as a function of the measure of obstruction; A method comprising:

86. 86. The method of claim 85, wherein the indication of partial obstruction is a function of detection of an M-shaped pattern breath.

87. 1. A method for automatically determining an indicator of arousal in an apparatus for treating sleep disordered breathing, comprising: determining a measure of respiratory flow; determining a peak index of the expiratory portion of the respiratory flow index; determining an index of arousal as a function of the peak index; A method comprising:

88. 88. The method of claim 87, wherein the function of the peak index is a function of position within the expiratory portion.

89. If the peak index occurs in the latter half of the expiratory portion of the respiratory flow index, then the arousal index 90. The method of claim 88, representing wakefulness.

Images