Techniques for extracting respiratory parameters from noisy short duration chest impedance measurements

Abstract

One example is a method of extracting respiratory parameters from a short duration thoracic impedance ("TI") signal, the method comprising pre-processing a TI measurement signal to obtain a respiratory signal therefrom; evaluating the respiratory signal for at least one of signal quality and signal integrity; implementing at least one of an autocorrelation algorithm and a time domain zero crossing algorithm on the respiratory signal to extract at least one respiratory parameter therefrom, the at least one respiratory parameter comprising at least one of respiratory rate ("RR") and tidal volume ("TV").

Description

[0001] Related applications

[0002] This disclosure claims priority to U.S. Provisional Patent Application No. 63 / 115762, filed November 19, 2020, entitled “Technique for Extracting Respiratory Parameters from Short-Duration Noise-Induced Chest Impedance Measurements,” the disclosure of which is incorporated herein by reference in its entirety. Technical Field

[0003] This disclosure generally relates to techniques for detecting respiratory parameters from chest impedance measurements, and more specifically, to techniques for extracting such parameters from short-duration chest impedance measurements with noise. Attached Figure Description

[0004] To provide a more complete understanding of this disclosure and its features and advantages, reference is made to the following description in conjunction with the accompanying drawings, wherein like reference numerals denote like parts, wherein:

[0005] Figure 1 An example environment is shown, illustrating a system for deriving respiratory parameters from noisy, short-duration chest impedance measurements, according to some examples of this disclosure.

[0006] Figure 2 These are examples illustrating some aspects of this disclosure. Figure 1 A block diagram of exemplary functional components of the system;

[0007] Figure 3A This is a flowchart illustrating the operation of a method for extracting respiratory parameters from a noisy, short-duration chest impedance measurement, according to some examples of this disclosure;

[0008] Figure 3B This is a flowchart illustrating the operation of a method for extracting respiratory rate from a noisy, short-duration thoracic impedance measurement, according to some examples of this disclosure;

[0009] Figure 4A and 4B Examples of using autocorrelation-based algorithms according to this disclosure are shown respectively. Figure 4A ) and time-domain based (zero-crossing) algorithm ( Figure 4B Respiratory parameters are extracted from the respiratory-modulated thoracic impedance signal.

[0010] Figure 5 This is a flowchart illustrating a method for performing an impedance-specific signal quality check according to some examples of this disclosure;

[0011] Figure 6 This is a graph illustrating the effects of noise and motion artifacts in chest impedance signals according to some examples of this disclosure;

[0012] Figure 7 This is a flowchart illustrating a method for performing an artifact detection signal quality check according to some examples of this disclosure;

[0013] Figures 8A-8E Methods for removing detected artifacts from chest impedance signals according to some examples of this disclosure are shown;

[0014] Figure 9 This is a flowchart illustrating a method for evaluating the signal quality of the longest well-preprocessed segment of a chest impedance signal, according to some examples of this disclosure;

[0015] Figures 10A-10C These are diagrams illustrating the operation of an autocorrelation-based algorithm for extracting RR from a chest impedance signal according to some examples of this disclosure;

[0016] Figure 11A and 11B Together, flowcharts illustrating operations of autocorrelation-based algorithms according to some examples of this disclosure are shown;

[0017] Figure 12 It is a chart illustrating the physiological significance of chest impedance signals and their derivatives, serving as substitutes for chest volume (in liters) and flow rate (in liters per minute), respectively.

[0018] Figures 13A-13B Together, flowcharts illustrating the operation of time-domain-based zero-crossing algorithms according to some examples of this disclosure are shown;

[0019] Figure 14A-14F Including joint explanation Figures 13A-13B The diagram shows the operation of the zero-crossing algorithm based on the time domain. Detailed Implementation

[0020] Chest impedance measurements obtained using electrodes placed on a patient's chest offer an indirect, non-invasive way to collect respiratory parameters of interest, as modulation of lung air levels due to respiration is reflected in proportional modulation of chest impedance. However, such measurements are susceptible to extremely high levels of noise artifacts, such as those caused by movement, coughing, and / or inappropriate skin electrode contact, making it challenging to extract parameters such as respiratory rate (RR) and tidal volume (TV) from the measurements. Furthermore, certain clinical conditions requiring the extraction of events such as shallow breathing, apnea, and / or periodic or oscillatory breathing present even more challenging conditions in the presence of these artifacts. Examples described herein include two approaches to address these challenges: a time-domain-based approach and an autocorrelation-based approach. Both approaches closely follow the physiological aspects of the respiratory cycle and keep heuristic rules to a minimum, enabling the extraction of most of these parameters from a single 60-second chest impedance measurement with errors limited to within two breaths per minute (BPM), including quantification errors.

[0021] Abnormal respiratory activity is an early indicator of respiratory, cardiac, and / or neurological disorders. Clinically, respiratory rate (RR) is reported by counting the number of chest wall shifts during inspiration and expiration. This method is often erroneous, depending on the nurse's skill level. Clinical methods for extracting air volume (inhaled and exhaled air volume) involve breathing through a tube into the mouth using a nasal clip, and are therefore not suitable for home monitoring.

[0022] As previously mentioned, chest impedance monitoring via electrodes placed on the chest provides an indirect, non-invasive method for extracting respiratory parameters such as RR and TV. However, the accuracy of this technique is compromised by one or more of the following: very low-frequency baseline drift due to improper electrode contact with the skin; high-frequency physiological interferences such as cardiac activity; broadband circuit noise; and motion artifacts caused by coughing, hiccups, or body movements. Furthermore, certain physiological conditions exhibit different characteristics in signal morphology, making it even more difficult to extract respiratory parameters with high confidence.

[0023] Traditional time-domain based methods, such as peak detection / counting and frequency-domain based methods, struggle to extract parameters of interest from chest impedance measurement signals (or simple chest impedance signals) due to the non-stationarity of the signal itself and the noise embedded in the signal.

[0024] The examples described in this paper provide solutions to these problems and offer techniques for reliably extracting respiratory parameters from chest impedance signals by applying two different methods (time-domain based and autocorrelation-based) in the presence of varying physiological signal morphologies and artifact conditions. A novel method for evaluating the signal quality of chest impedance signals using the input signal, accelerometer data, and filtered noise is also proposed.

[0025] Time-domain-based methods can be used to report RR in cases of low-confidence RR estimation (rather than signal quality) based on autocorrelation techniques, as well as to estimate RR and calculate TV in cases of apnea.

[0026] Figure 1 Example environment 100 is depicted, providing an illustrative example of a system 102 for deriving and monitoring respiratory parameters (e.g., RR and TV) of human subjects using noisy, short-duration chest impedance measurements, according to some examples of this disclosure. Monitoring can be performed in a continuous or periodic manner. Figure 1 As shown, according to an exemplary example, system 102 includes a chest impedance measurement module 112 and a plurality of surface electrodes / sensors 114a-114d (e.g., four (4) surface electrodes or sensors, or any other suitable number of surface electrodes / sensors). For example, one or more surface electrodes may be implemented as solid gel surface electrodes, or any other suitable surface electrodes. System 102 may be configured as a generally triangular device, or any other suitable shape device, operable to contact one or more of the torso, upper chest, and neck regions, or any other suitable part or region of the body of a human subject 104 via at least the plurality of surface electrodes / sensors 114a-114d.

[0027] In various implementations, system 102 may be configured to allow it to be implemented as multiple patch-like devices or any other suitable structure or device within a wearable vest-like structure. In a possible environment, such as environment 100, system 102 may operate for bidirectional communication with smartphone 106 via wireless communication path 116, and smartphone 106 may in turn operate for bidirectional communication with communication network 108 (e.g., the Internet) via wireless communication path 118. Alternatively, a direct link to cloud 110 may be provided without hopping through base stations or cellular phones. Smartphone 106 may also operate via communication network 108 to bidirectionally communicate with cloud 110 via wireless communication path 120, which may include resources for cloud computing, data processing, data analysis, data trending, data reduction, data fusion, data storage, and other functions. System 102 may also operate for direct bidirectional communication with cloud 110 via wireless communication path 122.

[0028] Figure 2 Example block diagrams of system 102, according to some examples of this disclosure, are depicted for deriving and monitoring respiratory parameters, such as RR and TV, in human subjects using noisy, short-duration chest impedance measurements. Figure 2 As shown, the system includes a chest impedance measurement module 112, a processor 202 and associated memory 208, a data memory 206 for storing chest impedance measurement data, and a transmitter / receiver 204. The transmitter / receiver 204 can be configured to perform Bluetooth communication, Wi-Fi communication, or any other suitable short-range communication for communication with a smartphone 106 via wireless communication path 116. Figure 1 The transmitter / receiver 204 can be further configured to perform cellular communication or any other suitable long-range communication for communication with the cloud 110 via wireless communication path 122. Figure 1 Communication. In some examples, the chest impedance measurement module 112 may also include an electrode / sensor connection switching circuit 224 for communication with... Figure 1 The multiple surface electrodes / sensors 114a-114d shown are switchably connected.

[0029] Processor 202 may include multiple processing modules, such as data analyzer 226 and data fusion / decision engine 228. Transmitter / receiver 204 may include at least one antenna 210 operable to transmit / receive wireless signals (e.g., Bluetooth or Wi-Fi signals) to / from smartphone 106 via wireless communication path 116. Smartphone 106 may be a Bluetooth or Wi-Fi enabled smartphone or any other suitable smartphone. Antenna 210 may also be operable to transmit / receive wireless signals, such as cellular signals, to / from cloud 110 via wireless communication path 122.

[0030] Processor 202 may also include an autocorrelation module 230 and a time-domain module 232 for implementing autocorrelation-based and time-domain-based techniques, respectively, for deriving respiratory parameters from the chest impedance signal, as described herein. Processor 202 may also include a signal quality assessment module 234 for performing signal quality checks related to the chest impedance signal, as described herein.

[0031] Transmitter / receiver 204 may include at least one antenna 210, operable to transmit / receive wireless signals (e.g., Bluetooth or Wi-Fi signals) to / from smartphone 106 via wireless communication path 116. Smartphone 106 may be a Bluetooth or Wi-Fi enabled smartphone or any other suitable smartphone. Antenna 210 may also be operable to transmit / receive wireless signals, such as cellular signals, to / from cloud 110 via wireless communication path 122.

[0032] Please refer to the following illustrative examples and Figure 1 and Figure 2 Further understanding is provided regarding the operation of system 102 for deriving and monitoring respiratory parameters (e.g., RR and TV) of human subjects using noisy, short-duration chest impedance measurements, according to some examples. In this illustrative example, while the human subject 104 is in a supine or upright position, at fixed times each day (e.g., twice daily) or for a predetermined number of consecutive days, the human subject or human assistant is positioned in system 102 configured as a generally triangular device (or any other suitable shape) such that it is in contact with one or more of the subject's torso and upper chest and neck regions (or any other suitable part or area of ​​the body) via a plurality of surface electrodes / sensors 114a-114d.

[0033] After positioning system 102 in contact with the torso and / or upper chest and / or neck region of a human subject, chest impedance measurement module 112 can be activated to collect, gather, sense, measure, or otherwise acquire chest impedance data from human subject device 104 and generate a signal indicative of that data. In some examples, the chest impedance data obtained using illustrative methods is noisy and short in duration.

[0034] The chest impedance measurement module 112 can perform chest impedance measurements using some or all of a plurality of surface electrodes 114a-114d that are in contact with the skin of the human subject 104 on the torso, upper chest, and / or neck region. Based on the characteristics of the examples described herein, as will be described in more detail below, respiratory parameters, such as respiratory rate and tidal volume, can be derived from noisy, short-duration chest impedance data from the chest impedance measurement module 112.

[0035] In some examples, chest impedance data from chest impedance measurement module 112 can be provided to data analyzer 226 for at least partial data analysis, data trend analysis, and / or data reduction. In one example, chest impedance measurement data, combined with other metadata such as medical history, demographic information, and other test patterns, can also be analyzed, trended, and / or reduced "in the cloud," and preset alerts can be provided in cloud-based data storage 110 for clinical interventions at various levels regarding respiratory parameters.

[0036] Data analyzer 226 can provide at least partially analyzed chest impedance data to data fusion / decision engine 228, which can efficiently fuse or combine the chest impedance data with other sensed data, at least partially, according to one or more algorithms and / or decision criteria, for subsequent use in making one or more inferences about human subject 104. Processor 202 can then provide the at least partially combined chest impedance and other sensed data to transmitter / receiver 204, which can directly transmit the combined chest impedance and sensed data to cloud 110 via wireless communication path 122, or to smartphone 106 via wireless communication path 116. Smartphone 106 can then transmit the combined chest impedance and sensed data to cloud 110 via wireless communication paths 118, 120 through communication network 108, where it can be further analyzed, trended, reduced, and / or fused. It will be appreciated that, as described above, communication data can be transmitted directly to cloud 110 without involving smartphones / cellular phones or base stations.

[0037] Hospital clinicians can then remotely download the resulting carefully curated combination of sensor data for risk scoring / stratification, monitoring, and / or tracking purposes.

[0038] Figure 3A This is a flowchart illustrating the operation of a method 300 for extracting respiratory parameters from a noisy, short-duration chest impedance measurement or signal, according to some examples of this disclosure.

[0039] In step 302, the test subject is placed in a human subject (e.g., human test subject 104). Figure 1 Electrodes on (e.g., electrode 114) Figure 2 The obtained short-duration (e.g., 60 seconds) chest impedance signal is preprocessed to obtain the respiratory signal. In a particular example, step 302 can be performed using a low-pass filter with a frequency cutoff (Fc) of 0.65 Hz.

[0040] In step 304, the respiratory signal is evaluated for signal quality and signal integrity (e.g., via module 234). Figure 2 In a specific example, signal quality assessment may include calculating specific chest impedance measures of the signal, such as electrode contact impedance and whole-body impedance, and comparing them to thresholds established based on physiological limits. Signal integrity assessment may include examining signal characteristics to detect and remove large artifacts or interference in the signal. Accelerometer data may be used to detect motion artifacts in the signal, which may also be removed in step 304.

[0041] Some examples of methods for extracting RR and TV from noisy, short-duration chest impedance signals, as will be described in more detail below, such as method 300, utilize the fact that the chest impedance signal is a substitute for the lung volume of a human subject and that the derivative of the chest impedance signal is a substitute for the airflow velocity inside and outside the lungs of a human subject.

[0042] Typically, autocorrelation algorithms derive the inherent periodicity of noisy, non-stationary signals. (See again...) Figure 3A In step 306, an autocorrelation-based technique is used (e.g., by module 230). Figure 2 This method utilizes the autocorrelation of the respiratory signal to process the respiratory signal. Specifically, in step 306, the respiratory signal is autocorrelated to determine its second-order average. In most autocorrelation-based algorithms used to extract periodicity, the dominant peak or local maxima of the autocorrelated signal are considered in isolation; however, this is prone to error due to large high / low frequency noise. Based on the characteristics of the example described herein, the autocorrelation-based technique implemented in step 306 leverages the entire autocorrelated respiratory signal to better understand the hidden periodicity and its variations within the signal. Therefore, in the example shown, in step 306, the expected value is calculated based on the time lag between peaks in the autocorrelation signal to derive the estimated RR for the autocorrelation algorithm. This technique is well-suited for respiratory signals with anomalous morphology, periodic oscillating breathing patterns, and circuit noise.

[0043] In step 308, a time-domain based technique (e.g., by module 232) is used. Figure 2 This method processes the respiratory signal. Specifically, as will be described in more detail below, in step 308, the respiratory signal is divided into inspiratory and expiratory cycles by calculating the zero-crossing points on the first derivative of the respiratory signal. Heuristic rules based on physiological limitations, such as invalid RR (e.g., more than 40 breaths per minute, less than 6 breaths per minute, etc.) and / or invalid inspiratory to expiratory ratios (e.g., 1:4 or 4:1), are applied to identify valid breaths and eliminate invalid breaths. According to the feature examples described herein, the time-domain-based algorithm calculates RR by interval counting and TV based on the median of peak chest impedance values. The time-domain-based algorithm described herein is well-suited for respiratory signals with frequency and amplitude modulation, including both respiratory and apnea signals.

[0044] In step 310, estimates from autocorrelation-based and time-domain-based algorithms can be selected based on certain signal characteristics representing certain clinical conditions. For example, in the case of apnea, i.e., no breathing for several seconds, the number of inhalations and exhalations is best described by a time-domain-based algorithm, while an autocorrelation-based algorithm precisely specifies the subject's respiratory rate (i.e., RR) before or after the apnea event. Conversely, in the case of oscillatory breathing, RR is specified by an autocorrelation-based algorithm, and oscillations in the TV are specified by a time-domain-based "zero-crossing" algorithm.

[0045] In step 312, confidence assessment can be performed on the estimates from autocorrelation-based algorithms and time-domain algorithms, as described below.

[0046] In step 314, select and report and / or record RR and TV estimates as needed.

[0047] It will be recognized that for chest impedance signals (e.g., signals with a lot of noise), frequency domain methods, such as autocorrelation methods, will be more useful in deriving the RR from the chest impedance signal, while for other chest impedance signals, such as chest impedance signals that are not particularly periodic, time domain-based methods will be more useful in deriving the RR from the chest impedance signal. The examples described in this paper take advantage of the relative strengths of the two methods, derive the RR using both methods, and then select the method that may be more accurate in this case.

[0048] Figure 3B This is a flowchart illustrating the operation of a method 320 for detecting RR from a noisy, short-duration chest impedance measurement or signal, according to some examples of this disclosure.

[0049] In step 322, the test subject is placed in a human subject (e.g., human test subject 104). Figure 1 Electrodes on (e.g., electrode 114) Figure 2 The obtained short-duration (e.g., 60 seconds) chest impedance signal is preprocessed to obtain the respiratory signal. In a particular example, step 322 can be performed using a low-pass filter with a frequency cutoff (Fc) of 0.65 Hz. In some implementations, the chest impedance signal can be filtered to a bandwidth of interest between 0.1 Hz and 0.75 Hz.

[0050] In step 324, and in step 308, the respiratory signal is processed using a time-domain-based technique to generate an estimated time-domain RR (“TD_RR”). Furthermore, if apnea is detected in the respiratory signal during the execution of the time-domain-based technique, the FLAG_APNEA_DETECTED flag is set.

[0051] In step 326, the derivative of the respiratory signal is calculated, and in step 328, an autocorrelation-based technique is used to process the respiratory signal and / or its derivative to generate an estimated autocorrelation RR (“AC_RR”) and a confidence metric for the estimated AC_RR. In some examples, the confidence metric (CM) is equal to the ratio of signal power (corresponding to respiratory volume per minute (BPM) within ±5 bpm of the estimated AC_RR) to noise power (corresponding to BPM outside ±5 pm of the estimated AC_RR).

[0052] In step 328, it is determined whether the quality of the chest impedance signal (as determined by one or more signal quality checks described below) is good. If the quality of the chest impedance signal is poor, the process proceeds to step 332, where it is determined that no RR (Resistance Rate) should be reported because the signal is unreliable / unusable.

[0053] If the chest impedance signal quality is determined to be good in step 328, the process proceeds to step 334, where it is determined whether the confidence metric is less than a predetermined threshold (e.g., 1). If the confidence metric is determined to be less than the predetermined threshold in step 334, the process proceeds to step 336, where the estimated TD_RR is output as RR. If the confidence metric is determined to be not less than the predetermined threshold in step 334, the process proceeds to step 338, where the estimated AC_RR is output as RR.

[0054] In some examples, RR estimates (e.g., AC_RR or TD_RR) can be used to adjust the filter used for TV extraction. For example, if an RR of 10 ppm is found, the center frequency Fc and bandwidth of the low-pass filter can be selected to be 10 ppm + / - 3 ppm to improve TV extraction. It will be noted that TV information is one of the determining factors in the RR confidence metric (CM_RR) report. For example, very low TV (possibly due to poor contact) or very high TV (perhaps due to contact impedance modulation) will reduce the confidence of the RR report. Furthermore, combining RR and TV information can provide important clinical insights. For example, minute ventilation is defined as the amount of air breathed per minute and is the product of RR and TV (e.g., typically 5-8 liters / minute). Additionally, although TV cannot be directly detected in liters, by comparing the estimated TV to baseline readings, a significant decrease / increase in minute ventilation can flag possible hypoventilation / hyperventilation.

[0055] Figure 4A and 4B This illustrates some examples of the use of autocorrelation-based algorithms according to this disclosure. Figure 4A ) and time-domain based (zero-crossing) algorithm ( Figure 4BA diagram showing the extraction of respiratory parameters from oscillatory respiratory signals.

[0056] Figure 5 An example is shown according to the description herein (e.g., as performed in step 304). Figure 3A )) A flowchart of method 500 for performing impedance-specific signal quality checks. (See attached flowchart.) Figure 5 As shown, in step 502, the original chest impedance (“TI”) signal is examined to determine if it is within the effective impedance range (e.g., greater than 30 ohms and less than 250 ohms). If the chest impedance signal is determined to be outside the effective impedance range, execution proceeds to step 504, where an error code is generated to indicate that the chest impedance signal is out of range, and the signal quality is rated as -1 (“no confidence”). Furthermore, in step 504, the value of the parameter valid_RR (which is set to indicate whether the reported RR is valid) is set to 0 (i.e., the reported RR is invalid). If the chest impedance signal is determined to be within the effective impedance range in step 502, execution proceeds to step 506, where the value of valid_RR is set to 1 (i.e., the reported RR is valid).

[0057] In step 508, it is determined whether the stability deviation of the chest impedance signal is less than a specified percentage (e.g., 10%). As used herein, "settlement deviation" refers to the change in chest impedance over the measurement duration. For example, if the chest impedance changes by more than 10%, the electrode contact may be unstable. If it is determined that the stability deviation of the chest impedance signal is not less than the specified percentage, execution proceeds to step 510, where an error code is generated to indicate that the chest impedance stability deviation is too large. Furthermore, in step 510, the value of parameter gSQM_valid_TV is set to 0, and the value of parameter gSQM_valid_RR is set to 0. If it is determined in step 508 that the stability deviation of the chest impedance signal is less than the specified percentage, execution proceeds to step 512, where the value of gSQM_valid_TV is set to 1. It will be appreciated that gSQM_valid_TV and gSQM_valid_RR are signal quality measures of TV and RR, respectively, with a value of "1" indicating good signal quality and a value of "0" indicating poor signal quality.

[0058] In step 514, it is determined whether the contact impedance mismatch is less than a specific value (e.g., 2000 ohms). If the contact impedance mismatch is determined to be not less than the specific value, execution proceeds to step 516, where an error code is generated to indicate that the contact impedance mismatch is too high. Additionally, in step 516, the value of gSQM_valid_RR is set to 0. If the contact impedance mismatch is determined to be less than the specific value in step 514, execution proceeds to step 518.

[0059] In step 518, it is determined whether the contact impedance is less than a specific value (e.g., 3000 ohms). If the contact impedance is determined to be not less than the specific value, execution proceeds to step 520, where an error code is generated to indicate that the contact impedance is too high. Additionally, in step 520, the value of gSQM_valid_RR is set to 0. If the contact impedance is determined to be less than the specific value in step 518, execution proceeds to step 522.

[0060] In step 522, the signal is considered to have passed the impedance-specific signal quality check, and the value of gSQM_valid_RR is set to 1.

[0061] Now for reference Figure 6 It will be recognized that if a perturbation (e.g., an artifact) 600 is present in the chest impedance signal 602, the sample distribution 604 of the signal may tail in one direction due to large / very small numbers. Simply put, the presence of artifacts increases the deviation of the signal from the mean. To assess this, the coefficient of variation (CoV) can be considered, which increases with increasing noise in the chest impedance signal. It can be recognized that the standard deviation (std) of the chest impedance signal also reflects this effect, but it is difficult to define an optimal threshold to achieve this. In contrast, CoV defines the ratio of noise to signal (std(signal) / mean(signal)). CoV greater than 1 indicates that the sample distribution is super-exponential, while CoV less than 0.4 indicates that the sample distribution tails in the opposite direction.

[0062] Figure 7 A flowchart is shown of a method 700 for performing an artifact detection signal quality check according to the example described herein (e.g., as performed in step 304). Figure 3A )). refer to Figure 7 In step 702, the preprocessed chest impedance signal and the corresponding accelerometer data are normalized to the range [0-1] to eliminate the influence of DC in the mean, and the CoV of the chest impedance signal is calculated. Simultaneously, in step 704, the preprocessed chest impedance signal and accelerometer data are normalized to zero mean and unit variance, and the kurtosis is calculated.

[0063] In step 706, it is determined whether (1) CoV is greater than 1, or (2) CoV is less than 0.4 and kurtosis is greater than 7, for the chest impedance signal or accelerometer data. If either of these conditions is true for either signal, an artifact is detected in step 708. If neither of the conditions for either signal in step 706 is met, execution proceeds to step 710.

[0064] In step 710, it is determined that gSQM_valid_RR = 1 and gSQM_valid_TV = 1 (as in method 500). Figure 5 The conditions are as determined in step 710, and whether the signal length is greater than 30 seconds. If all these conditions are true, execution proceeds to step 712, where the signal is considered to have high-quality data confidence (data_quality = 1). If one or more of the conditions in step 710 are not true, execution proceeds to step 714, where the signal is considered to have low-quality data confidence (data_quality = 0). As used herein, data_quality represents the final combined signal quality metric. For an artifact-free signal of sufficient duration (>30 seconds), it is a logical AND of gSQM_valid_TV and gSQM_valid_RR, and therefore can be 1 or 0, depending on the SQM of TV and RR. If the signal has artifacts or is insufficient in length, it is set to -1 (indicating a bad / unusable signal).

[0065] Figures 8A-8E A method for removing detected artifacts from a chest impedance signal to generate a signal from which RR and TV can be derived according to the example described herein is shown. Figure 8A The original chest impedance signal 800, including artifact 802, is shown. The original chest impedance signal 800 is normalized to zero mean and unit variance. Furthermore, the Shannon energy envelope of the chest impedance signal is calculated, and a threshold is applied. It will be recognized that the Shannon energy exhibits better discrimination than the signal energy alone, and, compared to the limit, it assigns weight to artifacts in the intermediate range. Figure 8B Waveform 810 is shown, representing the Shannon energy of the original chest impedance signal 800 and the artifact 802.

[0066] like Figure 8C As shown, from waveform 810 ( Figure 8B A mask 820 is formed to identify segments of the chest impedance signal that include artifact 802. Now refer to... Figure 8D The chest impedance signal was segmented into bad segment 830 (including artifacts) and good segment 832. Figure 8D In the example shown, the longest good segment includes all good segments 832, which are identified and preprocessed to create the longest preprocessed good segment. Figure 8E The figure below is denoted by reference numeral 840. Then, as described herein, the longest preprocessed segment 840 is used to derive RR and TV. Figure 9 As shown, the signal quality of the longest preprocessed segment 840 is evaluated.

[0067] Figure 9A method 900 for evaluating the signal quality of the longest preprocessed segment is shown, for example, segment 840 ( Figure 8E In step 902, the CoV and kurtosis of the segment are calculated. In step 904, it is determined whether the CoV is less than 1, or whether the CoV is greater than 0.4 and the kurtosis is less than 7. If a negative determination is made in step 904, execution proceeds to step 906, in which a no-quality confidence value is assigned to the segment, and the data_quality parameter of the segment is set to -1.

[0068] If a positive determination is made in step 904, execution proceeds to step 908, where it is determined whether gSQM_valid_RR equals 1, gSQM_valid_TV equals 1, and the signal length is less than 30 seconds. If all the conditions specified in step 908 are met, execution proceeds to step 910, where a high-quality confidence value is assigned to the segment, and the data_quality parameter is set to 1.

[0069] If one of the conditions specified in step 908 is not met, execution proceeds to step 912, where it is determined whether gSQM_valid_RR is equal to 1, gSQM_valid_TV is equal to 1, and the signal length is less than 15 seconds. If all the conditions specified in step 912 are met, execution proceeds to step 914, where a low-quality confidence value is assigned to the segment, and the data_quality parameter is set to 0.

[0070] If one of the conditions specified in step 912 is not met, the execution proceeds to step 916, in which a quality-free confidence value is assigned to the segment and the data_quality parameter is set to -1.

[0071] Depending on the details of a specific example, the autocorrelation-based algorithm described herein derives the inherent periodicity of respiratory signals (which need not be strictly periodic and / or stationary) without the influence of external noise. As will be described, using an autocorrelation-based algorithm to extract RR involves depolarizing the preprocessed signal to derive a trend-stationary signal (zero mean) from the autocorrelation signal, the autocorrelation of the signal, and heuristic-based RR calculation. Furthermore, the signal-to-noise ratio (SNR) and TV can be calculated from the autocorrelation signal. Figures 10A-10C The operation of an autocorrelation-based algorithm for extracting RR from the chest impedance signal 1000 is shown. Figure 10A The calculation of the expected value of RR will be described in more detail below. Figure 10B ), and uses a relative threshold to limit the peak value to effective noise ( Figure 10C ).

[0072] Figure 11A and 11B This is a flowchart 1100 illustrating the operation of an autocorrelation-based module according to the example described herein. In step 1102, a 60-second chest impedance signal (e.g., a chest impedance signal segment) is input to the autocorrelation-based module. In step 1104, the input chest impedance signal segment is low-pass filtered to remove high-frequency noise. Specifically, the low-pass filter may be a finite impulse response (FIR) filter with an Fc of 0.65 Hz and a length of 3 taps.

[0073] In step 1106, a first derivative is performed to remove baseline drift (if the signal is not stationary), thereby generating a difference signal (Δamplitude / Δtime).

[0074] In step 1108, the correlation is calculated for the difference signal with its own time-lag version (the lag of one sample) to generate an autocorrelation signal ((Δamplitude / Δtime)). 2 ).

[0075] In step 1110, all local maxima or peaks are identified in the autocorrelation signal.

[0076] In step 1112, if the intensity of the peak is negatively correlated and if the amplitude of the peak is less than 40% of the amplitude of the adjacent peak, then the peak is discarded.

[0077] In step 1114, the relative amplitude and relative time lag between peaks are calculated to generate a relative time lag array and a relative amplitude or signal power array equivalent to the harmonic period.

[0078] In step 1116, an array of breaths per minute (BPM) is calculated using an array with a relative time lag (e.g., 60 / Δtime / sampling rate).

[0079] In step 1118, if (1) the BPM value is greater than 44 or less than 6, or (2) the difference between the value and its adjacent BPM value is greater than or equal to 10, then the BPM value can be excluded from the BPM array calculated in step 1116. The result is an array of valid relative BPM values.

[0080] In step 1120, the average of the effective relative BPM values ​​is calculated and used as the estimated average RR.

[0081] In step 1122, the highest peak value in the autocorrelation signal corresponding to the estimated average RR is identified. This is the estimated principal RR. The change in tidal impedance is equal to the square root of the highest signal peak value.

[0082] In step 1124, the RR corresponding to the highest peak value from the origin is calculated and regarded as the estimated principal RR.

[0083] In step 1126, the permissible deviation of the instantaneous BPM is calculated (e.g., the estimated mean RR ± 5).

[0084] In step 1128, the SNR is calculated by summing all the relative signal powers falling within the signal band (signal) and outside the band (noise).

[0085] The expected values ​​for all relative time lags represent the RR, which is affected by harmonics of high-period sequences in the chest impedance signal, frequency increases / decreases between periods, low-frequency artifacts, and uneven signal amplitude (e.g., due to shallow breathing, apnea). At any time lag with a finite number of overlapping signals, only the relevant data is represented as a peak, and all irrelevant data are canceled out. The algorithm is not entirely dependent on the signal amplitude, so large artifacts have little effect. To identify valid peaks in the autocorrelation plot, a relative threshold is applied instead of a global threshold.

[0086] Based on the features of the examples described herein, a time-domain-based method is also provided, which can be used to report RR in cases of low-confidence RR estimation (rather than signal quality) based on autocorrelation techniques, and to estimate RR and calculate TV in cases of apnea.

[0087] Figure 12 The physiological significance of the thoracic impedance signal and its derivative is explained, serving as substitutes for the thoracic volume (in liters) shown in curve 1200 and the flow rate (in liters per minute) shown in curve 1202, respectively. Figures 13A-13B Operation of a time-domain-based zero-crossing method 1300, characterized according to the example described herein, is illustrated. A time-domain-based algorithm is required to report RR based on time-domain counts in the case of low-confidence RR estimation using an autocorrelation-based algorithm, to estimate RR in the case of apnea, and to calculate TV, as will be described.

[0088] refer to Figure 13A In step 1302, short-duration chest impedance signals (such as...) Figure 14A The input signal (as shown) is input to the time-domain module. In step 1304, the input signal is preprocessed through a low-pass filter (e.g., at 0.65 Hz) to produce a filtered signal, such as... Figure 14B As shown. In step 1306, the derivative of the preprocessed input signal is generated ( Figure 14C In step 1308, the zero-crossing points in the derivative signal are identified. Figure 14DIn step 1310, peaks and troughs are identified in the derivative signal. Heuristic rules applied to identify valid peaks may include rejecting peaks that are less than a minimum threshold of the valid impedance peak (e.g., 5% of the highest peak after artifact removal), rejecting inhalation peaks that are less than 1.5 s (40 bpm) apart from adjacent inhalation peaks, and / or rejecting peaks where the inhalation and expiratory peaks differ by 90% (e.g., an inhalation peak of 40 milliohms and an expiratory peak of 400 milliohms).

[0089] In step 1312, the shallow breathing threshold (in) is applied. Figure 13B (Described in more detail below). In step 1314, the median chest impedance is calculated ( Figure 14E This is used to generate TV estimates. In step 1316, the effective peak values ​​for each valley are counted. Figure 14F ), to produce RR(“RRt_estimate”) estimated using time-domain methods.

[0090] Figure 13B This is an application flowchart of the shallow breathing threshold method 1350 based on the example described in this article. (See attached flowchart.) Figure 13B As shown, the application of the shallow breathing threshold involves integrating one cycle of inspiration and expiration (step 1352) and then determining whether the peak value is greater than 20 molhms (step 1 ohm or 5% of the maximum peak value) (step 1354). If the peak value is not greater than 20 molhms or not greater than 5% of the maximum peak value, the inspiration / expiration cycle is excluded from the count (step 1356). If the peak value is greater than 20 molhms or 5% of the maximum peak value, the inspiration / expiration cycle is included in the count (step 1358). These aforementioned steps are repeated for all inspiration / expiration cycles (step 1360).

[0091] Example 1 provides a method for extracting respiratory parameters of a human subject from a chest impedance (TI) measurement signal, the method comprising performing a signal quality check on the TI measurement signal; and implementing at least one of an autocorrelation algorithm and a time-domain zero-crossing algorithm on at least a portion of the TI measurement signal to extract at least one respiratory parameter of the human subject from the at least a portion of the TI measurement signal, wherein the at least one respiratory parameter includes at least one of respiratory rate (“RR”) and tidal volume (“TV”).

[0092] Example 2 provides the method of Example 1, further including low-pass filtering of the TI measurement signal before execution and implementation.

[0093] Example 3 provides the method of Example 2, wherein the cutoff frequency of the filter used to perform the low-pass filtering is 0.65 Hz.

[0094] Example 4 provides a method from any of Examples 1-3, where the signal quality check includes an impedance-specific signal quality check.

[0095] Example 5 provides the method of Example 4, wherein impedance-specific signal quality checking includes checking at least one of electrode contact impedance and whole-body impedance with reference to a threshold based on physiological limits.

[0096] Example 6 provides a method for any of Examples 1-5, wherein the signal quality check includes identifying at least one signal artifact in the TI measurement signal.

[0097] Example 7 provides the method of Example 6, which further includes removing the at least one artifact from the TI measurement signal to produce at least a portion of the TI measurement signal.

[0098] Example 8 provides the method of Example 6, wherein at least one artifact includes noise.

[0099] Example 9 provides the method of Example 6, wherein at least one artifact is the result of movement of a human subject.

[0100] Example 10 provides a method of any one of Examples 1-9, wherein performing at least one of an autocorrelation algorithm and a time-domain zero-crossing algorithm on the TI measurement signal further includes autocorrelation of the TI measurement signal to determine a second-order average value of the TI measurement signal; and calculating an expected value based on a time lag between peaks in the autocorrelated TI measurement signal to derive an estimated respiratory rate (“RR”).

[0101] Example 11 provides the method of Example 10, and further includes deriving the signal-to-noise ratio (SNR) of the TI measurement signal from the autocorrelation TI measurement signal.

[0102] Example 12 provides a method for any of Examples 10-11, and further includes calculating a confidence measure of the estimated RR.

[0103] Example 13 provides a method of any one of Examples 1-12, wherein performing at least one of an autocorrelation algorithm and a time-domain zero-crossing algorithm on the TI measurement signal further includes counting zero-crossings on the first derivative of the TI measurement signal to divide the TI signal into inspiratory and expiratory cycles to calculate respiratory rate (RR); and calculating tidal volume (“TV”) based on the median of the TI peak value.

[0104] Example 14 provides the method of Example 13, further including applying a shallow breathing threshold to the first derivative before calculating RR and TV.

[0105] Example 15 provides a method of any of Examples 1-14, further comprising selecting an estimate generated by at least one of the autocorrelation algorithm and the time-domain zero-crossing algorithm based on signal features indicative of a clinical condition.

[0106] Example 16 provides a method of any one of Examples 1-15, further comprising selecting an estimate generated by at least one of the autocorrelation algorithm and the time-domain zero-crossing algorithm based on signal characteristics indicative of a clinical condition.

[0107] Example 17 provides a method of any of Examples 1-16, wherein the duration of the TI measurement signal is less than 60 seconds.

[0108] Example 18 provides a method of any of Examples 1-17, wherein the duration of the TI measurement signal is less than 30 seconds.

[0109] Example 19 provides a method for determining the respiratory rate (RR) of a human subject based on a chest impedance (TI) measurement signal. The method includes preprocessing the TI measurement signal to generate a respiratory signal; performing a signal quality check on the respiratory signal; applying a time-domain zero-crossing algorithm to at least a portion of the respiratory signal to determine an estimated time-domain RR (TD_RR); applying an autocorrelation algorithm to at least a portion of the respiratory signal to determine an estimated autocorrelation RR (AC_RR) and a confidence metric of the estimated AC_RR; selecting one of the estimated TD_RR and the estimated AC_RR based on the confidence metric; and outputting the selected one of the estimated TD_RR and the estimated AC_RR as the final RR.

[0110] Example 20 provides the method of Example 19, wherein selecting one of the estimated TD_RR and the estimated AC_RR based on the confidence metric includes selecting the estimated AC_RR if the confidence metric is greater than or equal to a threshold; and selecting the estimated TD_RR if the confidence metric is less than the threshold.

[0111] Example 21 provides a method of any of Examples 19-20, further comprising, if the result of the signal quality check is poor, suppressing the output of one of the estimated TD_RR and the estimated AC_RR as the final RR.

[0112] Example 22 provides a method from any of Examples 19-21, where preprocessing includes filtering the TI measurement signal using a low-pass filter.

[0113] Example 23 provides a method from any of Examples 19-22, where the signal quality check includes an impedance-specific signal quality check.

[0114] Example 24 provides the method of Example 23, wherein impedance-specific signal quality checking includes checking at least one of electrode contact impedance and whole-body impedance with reference to a physiological limit-based threshold.

[0115] Example 25 provides a method for any of Examples 19-24, wherein the signal quality check includes identifying at least one signal artifact in the respiratory signal.

[0116] Example 26 provides the method of Example 25, further comprising removing the at least one artifact from the respiratory signal to generate at least a portion of the respiratory signal.

[0117] Example 27 provides a method for any of Examples 25-26, wherein the at least one artifact includes noise.

[0118] Example 28 provides a method for any of Examples 25-27, wherein at least one artifact is the result of movement of a human subject.

[0119] Example 29 provides a method of any of Examples 19-28, wherein performing an autocorrelation algorithm on the at least portion of the respiratory signal further includes autocorrelation of the at least portion of the respiratory signal to determine an autocorrelation signal; and calculating an expected value based on a time lag between peaks in the autocorrelation signal to derive an estimated respiratory rate (“RR”).

[0120] Example 30 provides the method of Example 29, wherein the confidence metric is the ratio of the signal power to the noise power of the autocorrelation signal.

[0121] Example 31 provides a method of any of Examples 19-30, wherein performing a time-domain zero-crossing algorithm on at least a portion of the respiratory signal further includes counting the number of zero-crossings of the first derivative signal of at least a portion of the respiratory signal, wherein the number of zero-crossings corresponds to the estimated TD_RR.

[0122] Example 32 provides the method of Example 31, wherein performing the time-domain zero-crossing algorithm on the at least portion of the breathing signal further includes combining the at least portion of the breathing signal to mark the apnea condition.

[0123] Example 33 provides a method of any one of Examples 31-32, wherein performing a time-domain zero-crossing algorithm on at least a portion of the breathing signal further includes combining at least a portion of the breathing signal to mark shallow breathing conditions.

[0124] Example 34 provides a method for determining tidal volume (TV) of a human subject from a chest impedance (TI) measurement signal, the method comprising: preprocessing the TI measurement signal to generate a respiratory signal; performing a signal quality check on the respiratory signal; applying a time-domain zero-crossing algorithm to at least a portion of the respiratory signal to determine an estimated TV; and selectively reporting the estimated TV based on the result of the signal quality check.

[0125] Example 35 provides the method of Example 34, further including suppressing the reporting of the estimated TV if the result of the signal quality check is poor.

[0126] Example 36 provides a method from any of Examples 34-35, where preprocessing includes filtering the TI measurement signal using a low-pass filter.

[0127] Example 37 provides a method from any of Examples 34-36, where the signal quality check includes an impedance-specific signal quality check.

[0128] Example 38 provides the method of Example 37, wherein impedance-specific signal quality inspection includes checking at least one of electrode contact impedance and whole-body impedance with reference to a threshold based on physiological limits.

[0129] Example 39 provides a method for any of Examples 34-39, wherein the signal quality check includes identifying at least one signal artifact in the respiratory signal.

[0130] Example 40 provides the method of Example 39, which further includes removing the at least one artifact from the respiratory signal to produce at least a portion of the respiratory signal.

[0131] Example 41 provides a method of any of Examples 34-40, wherein performing a time-domain zero-crossing algorithm on at least a portion of the respiratory signal further includes estimating the TV from the median of the peak TI values.

[0132] It should be noted that all specifications, dimensions, and relationships (e.g., number of elements, operations, steps, etc.) outlined herein are for illustrative and educational purposes only. Such information may be significantly altered without departing from the spirit of this disclosure or the scope of the appended claims. This specification applies only to a non-limiting example and should therefore be interpreted as a non-limiting instance. Exemplary examples have been described with reference to specific arrangements of components in the foregoing description. Various modifications and changes may be made to such examples without departing from the scope of the appended claims. Therefore, the specification and drawings should be considered illustrative rather than restrictive.

[0133] Note that in the numerous examples provided herein, interactions may be described based on two, three, four, or more electrical components. However, this is done merely for clarity and illustration. It should be understood that the system can be combined in any suitable manner. Any components, modules, and elements shown in the figures can be combined into a wide variety of possible configurations, all of which are clearly within the broad scope of this specification, based on similar design alternatives. In some cases, it may be easier to describe one or more functions of a given set of flows by referring only to a limited number of electrical components. It should be understood that the figures and the circuits they teach are readily expandable and can accommodate a large number of parts as well as more complex / complex arrangements and configurations. Therefore, the examples provided should not limit the scope of the circuits or inhibit the broad teaching of the circuits, as the circuits can be applied to countless other architectures.

[0134] It should also be noted that, in this specification, references to various features (e.g., elements, structures, modules, components, steps, operations, characteristics, etc.) included in “one example,” “exemplary example,” “example,” and “another example” are intended to indicate that any such feature is included in one or more examples of this disclosure, but may or may not be combined in the same example.

[0135] It should also be noted that the functions related to the circuit architecture are only shown as some possible circuit architecture functions that can be performed by the system shown in the figure or within the system described in the figure. Some of these operations may be deleted or removed, or significantly modified or changed, as appropriate, without departing from the scope of this disclosure. Furthermore, the timing of these operations may vary considerably. The preceding operational flows are provided for illustrative and discussion purposes. The examples described herein offer considerable flexibility, as any suitable arrangement, timeline, configuration, and timing mechanism can be provided without departing from the teachings of this disclosure.

[0136] Those skilled in the art can identify many other changes, substitutions, variations, and modifications, and this disclosure includes all such changes, substitutions, variations, and modifications that fall within the scope of the appended claims.

[0137] Note that all optional features of the devices and systems described above can also be implemented relative to the methods or processes described herein, and the details in the examples can be used anywhere in one or more examples. In these cases, the “means” (as described above) may include (but are not limited to) the use of any suitable components discussed herein, as well as any suitable software, circuits, hubs, computer code, logic, algorithms, hardware, controllers, interfaces, links, buses, communication paths, etc.

[0138] Note that the interactions can be described using two, three, or four network elements for the examples provided above, as well as many other examples presented herein. However, this is done merely for clarity and illustration. In some cases, it may be easier to describe one or more functions of a given set of flows by referring to only a limited number of network elements. It should be understood that the topologies shown in and described with reference to the accompanying drawings (and their teachings) are readily extensible and can accommodate a large number of components, as well as more complex / complex arrangements and configurations. Therefore, the examples provided should not limit the scope of the illustrated topologies or restrict their broad teachings, as they can be applied to countless other architectures.

[0139] It is equally important to note that the steps in the preceding flowcharts illustrate only some possible signaling scenarios and patterns that can be performed by or within the communication system shown in the figures. Some of these steps may be omitted or removed where appropriate, or these steps may be significantly modified or altered, without departing from the scope of this disclosure. Furthermore, many of these operations have been described as being performed simultaneously or in parallel with one or more additional operations. However, the timing of these operations can vary considerably. The preceding operational flowcharts are provided for illustrative and discussion purposes. The communication system shown in the figures offers considerable flexibility, as any suitable arrangement, timeline, configuration, and timing mechanism can be provided without departing from the teachings of this disclosure.

[0140] Although this disclosure has been described in detail with reference to specific arrangements and configurations, these exemplary configurations and setups may be significantly modified without departing from the scope of this disclosure. For example, although this disclosure has been described with reference to a specific communication switch, the examples described herein can be applied to other architectures.

[0141] Many other changes, substitutions, variations, and modifications can be identified by those skilled in the art, and this disclosure includes all such changes, substitutions, variations, and modifications that fall within the scope of the appended claims. In order to assist the United States Patent and Trademark Office (USPTO) and any reader of any patent published in this application in interpreting the appended claims, the applicant wishes to draw attention to the fact that the applicant: (a) does not intend to invoke Section 142, paragraph 6 of 35 U.S.SC, which existed as of the filing date of this application, in the appended claims unless the word “means” or “step” is specifically used in a particular claim; and (b) does not intend to limit this disclosure by any statement in the specification in any way not otherwise reflected in the appended claims.

Claims

1. A method of determining a respiratory rate, RR, of a human subject from a thoracic impedance, TI, measurement signal, the method comprising: pre-processing the TI measurement signal to generate a respiration signal; performing a signal quality check on the respiration signal; implementing a time-domain based algorithm on at least a portion of the respiration signal to determine an estimated time-domain RR, TD_RR, including counting zero-crossings on a first derivative of the respiration signal to divide the respiration signal into inspiration and expiration periods to calculate RR, and applying heuristic rules to identify valid respiration periods and eliminate invalid respiration periods, including identifying invalid RRs and invalid inspiration-to-expiration ratios, to determine the estimated TD_RR; implementing an autocorrelation algorithm on at least a portion of the respiration signal to determine an estimated autocorrelation RR, AC_RR, and to determine a confidence metric of the estimated AC_RR, including autocorrelating the respiration signal to determine an autocorrelation respiration signal, and calculating an expectation value based on a time lag between peaks of the autocorrelation respiration signal to derive the estimated AC_RR; selecting one of the estimated TD_RR and the estimated AC_RR based on the confidence metric; and outputting the selected one of the estimated TD_RR and the estimated AC_RR as a final RR. selecting one of the estimated TD_RR and the estimated AC_RR based on the confidence metric comprises:

2. The method of claim 1, wherein, selecting the estimated AC_RR if the confidence metric is greater than or equal to a threshold value; and selecting the estimated TD_RR if the confidence metric is less than the threshold value.

3. The method of claim 1 or 2, further comprising, in response to a result of the signal quality check being poor, refraining from outputting the selected one of the estimated TD_RR and the estimated AC_RR as the final RR. the pre-processing comprises filtering the TI measurement signal using a low-pass filter.

4. The method of any one of claims 1 to 2, wherein, the signal quality check comprises an impedance-specific signal quality check.

5. The method of any one of claims 1 to 2, wherein, the impedance-specific signal quality check comprises checking at least one of an electrode contact impedance and a whole-body impedance against a threshold based on a physiological limit.

6. The method of claim 5, wherein, the signal quality check comprises identifying at least one signal artifact in the respiration signal.

7. The method of any one of claims 1 to 2, wherein, 8. The method of claim 7, further comprising removing the at least one signal artifact from the respiration signal to produce the at least a portion of the respiration signal. the at least one signal artifact comprises noise.

9. The method of claim 7, wherein, the at least one signal artifact is a result of human subject motion.

10. The method of claim 7, wherein, implementing the autocorrelation algorithm on at least a portion of the respiration signal further comprises:

11. The method of any one of claims 1 to 2, wherein, autocorrelating at least a portion of the respiration signal to determine an autocorrelation signal; and calculating an expectation value based on a time lag between peaks in the autocorrelation signal to derive an estimated respiratory rate, RR. the confidence metric is a ratio of a signal power to a noise power of the autocorrelation signal.

12. The method of claim 11, wherein, ​ 13. The method of claim 1, wherein implementing a time-domain based algorithm on at least a portion of the respiratory signal further comprises flagging an apnea condition in conjunction with at least a portion of the respiratory signal.

14. The method of claim 1 or 13, wherein, implementing a time-domain based algorithm on at least a portion of the respiratory signal further comprises flagging a hypopnea condition in conjunction with at least a portion of the respiratory signal.

Images