R-R interval pattern recognition for arrhythmia identification

Abstract

Described herein are methods, devices, and systems that improve the specificity of detection of arrhythmia seizures, such as, but not limited to, the specificity of detection of atrial fibrillation (AF) seizures. Such a method may include obtaining an ordered list of R-R intervals within a window until before detection of a potential arrhythmic episode; determining a measurement of a predominantly repeating R-R interval pattern within the window; and comparing the measurement of the predominantly repeating R-R interval pattern to a pattern threshold. If the measurement of the predominantly repeating R-R interval pattern is below a pattern threshold, it indicates that a regularly irregular pattern is present, and it is determined that the detection of the potential arrhythmic episode does not correspond to the actual arrhythmic episode. Such embodiments may be beneficially used to significantly reduce the number of false positive arrhythmia detections.

Description

[0001] This invention is a divisional application of the following invention patent application: Application number: 202110491005.9, Application date: May 6, 2021, Invention name: RR interval pattern recognition for arrhythmia identification.

[0002] Priority declaration

[0003] This application claims priority to U.S. non-provisional patent application No. 17 / 226,915 filed on April 9, 2021 and U.S. provisional patent application No. 63 / 019,550 filed on May 4, 2020. Technical Field

[0004] The embodiments described herein generally relate to methods, systems and devices for performing arrhythmia identification, and more specifically, to methods, systems and devices for improving the specificity of arrhythmia episode detection (such as but not limited to atrial fibrillation (AF) episode detection specificity). Background Art

[0005] Various types of implantable medical devices (IMDs) are used to monitor cardiac arrhythmias. Certain types of IMDs, such as implantable pacemakers and implantable cardioverter-defibrillators (ICDs), are capable of providing appropriate therapy in response to detected arrhythmias of the heart. Other types of IMDs, such as insertable cardiac monitors (ICMs), are used for diagnostic purposes. ICMs have been increasingly used to diagnose cardiac arrhythmias, such as atrial fibrillation (AF).

[0006] Atrial fibrillation (AF) is a very common type of supraventricular tachycardia (SVT), which causes approximately one-fifth of all strokes and is the leading risk factor for ischemic stroke. However, AF is usually asymptomatic and intermittent, which often leads to failure to make appropriate diagnosis and / or treatment in a timely manner. To overcome this, many cardiac devices, such as ICMs, now monitor AF by obtaining an electrogram (Electrogram, EGM) signal and measuring the RR interval variability based on the EGM signal. For example, an ICM or other IMD can compare the measurement of the RR interval variability with a variability threshold to automatically detect AF when the variability threshold is exceeded. In fact, ICMs identify AF primarily by quantifying the variability of the RR interval (i.e., by quantifying the variability in the timing of ventricular contraction).

[0007] There are several commercially available ICMs, including Confrim Rx manufactured by Abbott Laboratories in Chicago, Illinois. TM ICM, Reveal LINQ, manufactured by Medtronic, Minneapolis, Minnesota TM ICM and BioMonitor manufactured by Biotronik SE&Co.KG, Berlin, Germany TM 2 (AF and S versions). When the ICM detects an episode of an arrhythmia (such as AF), information about the episode can be recorded and a corresponding EGM segment (and / or other information) can be transmitted from the ICM to the patient care network for clinician review. False-positive arrhythmia detections are highly undesirable because the burden of triaging a large number of episodes of clinically irrelevant arrhythmias is both time consuming and expensive. In addition, false-positive arrhythmia detections may result in additional expenses for technician review, patients may suffer unnecessary anxiety, and false-positive arrhythmia detections may result in expensive, uncomfortable, and inconvenient follow-up testing.

[0008] One source of false-positive arrhythmia detections, such as false-positive AF detections, is caused by regularly irregular "patterns" of RR intervals. For example, multiple active atrial foci may activate the ventricles in turn. While this can result in high heart rate variability (i.e., RR interval variability) and trigger AF detection, the heart rate patterns repeat themselves. This phenomenon is not due to the "irregularly irregular" chaotic rhythms of true AF, and these episodes should be rejected. However, current IMDs, such as ICMs, typically trigger false-positive AF detections from regularly irregular "patterns" of these RR intervals. Therefore, there remains a need for improved techniques for distinguishing such patterns from actual chaotic rhythms resulting from AF, and more generally, a need for improved techniques for performing improved arrhythmia identification.

[0009] Heartbeat rhythms can be (1) regular, (2) regularly irregular, or (3) irregularly irregular. An irregularly irregular rhythm exhibits no pattern in the heartbeat intervals and is often associated with AF. In contrast, a regularly irregular rhythm that exhibits an underlying or hidden pattern (which may not be apparent using conventional automatic AF detection methods) is often not associated with AF, but may result in false-positive AF detections. In particular, a healthy heart rhythm that is found to be regularly irregular is unlikely to be associated with AF.

[0010] Automatic methods for detecting AF episodes that rely on detecting irregularities in sequential heartbeat intervals (e.g., RR intervals) to determine the likelihood that the rhythm is AF can be fooled by regularly irregular patterns in the heartbeat signal. This can cause the automatic methods to incorrectly identify AF, i.e., to cause false positive AF detections. Summary of the invention

[0011] Certain embodiments of the present technology are directed to methods, devices, and systems that can be used to improve arrhythmia episode detection specificity, such as, but not limited to, AF episode detection specificity.

[0012] According to certain embodiments, an implantable medical device (IMD) comprises: a plurality of electrodes; a sensing circuit coupled to one or more electrodes and configured to obtain a signal indicative of electrical activity of a patient's heart; a therapy circuit coupled to one or more electrodes and configured to deliver therapy to the patient's heart; at least one of a processor or a controller; and a battery configured to provide power to the sensing circuit, the therapy circuit, and at least one of the processor or the controller; at least one of the processor or the controller being configured to: detect a potential arrhythmia attack; obtain an ordered list of RR intervals within a window leading to detection of a potential arrhythmia attack, wherein each RR interval has a respective duration; based on the ordered list of RR intervals within the window, determine pairwise differences between pairs of RR intervals separated by various different intervals, thereby generating multiple groups of pairwise differences; based on the multiple groups of pairwise differences, determine a measurement of a predominantly repeated RR interval pattern; compare the measurement of the predominantly repeated RR interval pattern with a pattern threshold; and in response to the measurement of the predominantly repeated RR interval pattern being lower than the pattern threshold, classify the detection of the potential arrhythmia attack as a false positive.

[0013] In accordance with certain embodiments, a method for use by a device includes: obtaining an ordered list of RR intervals within a window that leads to detection of a potential arrhythmia attack, wherein each RR interval has a respective duration; based on the ordered list of RR intervals within the window, determining pairwise differences between pairs of RR intervals separated by various different intervals, thereby generating multiple sets of pairwise differences; based on the multiple sets of pairwise differences, determining a measurement of a predominantly repetitive RR interval pattern; comparing the measurement of the predominantly repetitive RR interval pattern with a pattern threshold; and in response to the measurement of the predominantly repetitive RR interval pattern being below the pattern threshold, classifying the detection of the potential arrhythmia attack as a false positive, and in response, preventing delivery of therapy to the patient's heart.

[0014] According to certain embodiments, a method for improving specificity of arrhythmic episode detection includes obtaining an ordered list of RR intervals within a window up to and including a detection of a potential arrhythmic episode, wherein each of the RR intervals has a respective duration, and determining (based on the ordered list) a measure of a predominantly repeating RR interval pattern within the window. The method also includes comparing the measure of the predominantly repeating RR interval pattern to a pattern threshold, and in response to the measure of the predominantly repeating RR interval pattern being below the pattern threshold, determining that the detection of the potential arrhythmic episode does not correspond to an actual arrhythmic episode.

[0015] According to certain embodiments, determining a measurement of a primarily repetitive RR interval pattern within a window includes: for each k from k=1 to M, where M is an integer of at least 4, determining an indicator of a median RR interval difference between pairs of RR intervals separated by k intervals within the window, thereby generating M indicators of median RR interval differences; identifying a minimum value among the M indicators of median RR interval differences; and determining the measurement of a primarily repetitive RR interval pattern within the window as the minimum value among the M indicators of median RR interval differences.

[0016] In accordance with certain embodiments, the M indicators of median RR interval differences that are determined include M median percentage difference values, each of which is determined using the following equation: for interval n=1:N and interval spacing k=1:M, for k, median percentage difference=MEDIAN(100*|RR(n)–RR(n+k)| / RR(n)), where RR(n) is the nth RR interval in the window, RR(n+k) is the (n+k)th RR interval in the window, |RR(n)-RR(n+k)| is the absolute value of RR(n)–RR(n+k), and N is the total number of RR intervals in the window.

[0017] According to certain embodiments, the method is for use by an implantable medical device (IMD) that is configured to perform at least one of storing data corresponding to one or more arrhythmic episodes detected by the IMD within a memory of the IMD or transmitting to an external device so that the data is available at a later time for further analysis. In response to the IMD detecting a potential arrhythmic episode, the steps outlined above are performed by the IMD. After the IMD determines that the detection of the arrhythmic episode does not correspond to an actual arrhythmic episode, the IMD performs at least one of allowing the data corresponding to the potential arrhythmic episode to be overwritten or preventing the data corresponding to the potential arrhythmic episode from being transmitted to the external device.

[0018] In accordance with certain embodiments, the window includes a specified time period, and obtaining an ordered list of RR intervals includes: obtaining an electrogram (EGM) or electrocardiogram (ECG) segment corresponding to the specified time period until before detection of a potential arrhythmia attack; identifying R waves in the EGM or ECG segment; and determining intervals between consecutive ones of these R waves.

[0019] Certain embodiments of the present technology are directed to a device comprising one or more electrodes, a sensing circuit, and at least one of a processor or a controller. The sensing circuit is coupled to the one or more electrodes and is configured to obtain a signal indicative of electrical activity in the patient's heart. At least one of the processors or controllers is configured to obtain an ordered list of RR intervals within a window until the detection of a potential arrhythmic episode, wherein each of these RR intervals has a respective duration; determine (based on the ordered list of RR intervals within the window) a measure of a predominantly repeated RR interval pattern within the window; and compare the measure of the predominantly repeated RR interval pattern with a pattern threshold. At least one of the processors or controllers is also configured to determine that the detection of a potential arrhythmic episode does not correspond to an actual arrhythmic episode if the measure of the predominantly repeated RR interval pattern is below the pattern threshold. In accordance with certain embodiments, at least one of the processors or controllers is configured to determine a measurement of a primarily repeating RR interval pattern within a window by: for each k from k=1 to M, where M is an integer of at least 4, determining an indicator of a median RR interval difference between pairs of RR intervals separated by k intervals within the window, thereby generating M indicators of the median RR interval difference; and identifying a minimum value among the M indicators of the median RR interval difference; and determining the measurement of the primarily repeating RR interval pattern within the window as the minimum value among the M indicators of the median RR interval difference.

[0020] In accordance with certain embodiments, the M indicators of median RR interval difference include M median percentage difference values, each of which is determined by at least one of the processor or controller using the following equation: for interval n=1:N and interval spacing k=1:M, for k, median percentage difference=MEDIAN(100*|RR(n)–RR(n+k)| / RR(N)), where RR(n) is the nth RR interval in the window, RR(n+k) is the (n+k)th RR interval in the window, |RR(n)–RR(n+k)| is the absolute value of RR(n)–RR(n+k), and N is the total number of RR intervals in the window.

[0021] According to certain embodiments, the device includes one of: a user-wearable device; an insertable cardiac monitor (ICM); a pacemaker connected to one or more leads; a leadless cardiac pacemaker (LCP); or an implantable cardioverter defibrillator (ICD).

[0022] According to certain embodiments, the device is an IMD and includes a transceiver configured to wirelessly communicate with an external device. In addition, the IMD is configured to perform at least one of storing data corresponding to one or more arrhythmic episodes detected by the IMD in a memory of the IMD or transmitting to the external device so that the data is available at a later time for further analysis. In certain such embodiments, after the IMD determines that the detection of an arrhythmic episode does not correspond to an actual arrhythmic episode, the IMD allows the data corresponding to the potential arrhythmic episode to be overwritten and / or prevents the data corresponding to the potential arrhythmic episode from being transmitted to the external device using the transceiver.

[0023] According to certain embodiments, a method for improving specificity of arrhythmic episode detection includes obtaining an ordered list of RR intervals within a window until detection of a potential arrhythmic episode, wherein each of the RR intervals has a respective duration. The method further includes, for a value of k selected from a range of values ​​of k, determining a corresponding indicator of a median RR interval difference between pairs of RR intervals separated by k intervals within the window; comparing the indicator of the median RR interval difference to a threshold; and determining that the detection of a potential arrhythmic episode does not correspond to an actual arrhythmic episode if the indicator of the median RR interval difference is less than the threshold. The method further includes, if the indicator of the median RR interval difference is not less than the threshold, repeating the above steps for another selected value of k until it is determined that the detection of a potential arrhythmic episode does not correspond to an actual arrhythmic episode, or until there are no additional values ​​within the range of values ​​of k for selection for k. According to certain embodiments, the range of values ​​of k is from 1 to M, where M is an integer of at least 4. In some such embodiments, the initial value of k selected from the range of values ​​of k is 1, and each time the above steps are repeated for another selected value of k, the other selected value of k is generated by incrementing k so that k = k + 1. According to other embodiments, the initial value of k selected from the range of values ​​of k is M, and each time the above steps are repeated for another selected value of k, the other selected value of k is generated by decrementing k so that k = k - 1. In accordance with certain embodiments, the corresponding indicator of the median RR interval difference between a pair of RR intervals determined for the value of k includes a median percentage difference value determined using the following equation: for interval n=1:N, median percentage difference=MEDIAN(100*|RR(n)–RR(n+k)| / RR(n)), where RR(n) is the nth RR interval in the window, RR(n+k) is the (n+k)th RR interval in the window, |RR(n)–RR(n+k)| is the absolute value of RR(n)–RR(n+k), and N is the total number of RR intervals in the window.

[0024] Certain embodiments of the present technology are directed to a device comprising one or more electrodes, a sensing circuit, and at least one of a processor or a controller. The sensing circuit is coupled to the one or more electrodes and is configured to obtain a signal indicative of electrical activity in the patient's heart. At least one of the processors or controllers is configured to obtain an ordered list of RR intervals within a window until the detection of a potential arrhythmic episode, wherein each of these RR intervals has a respective duration. At least one of the processors or controllers is also configured to, for a value of k selected from a value range of k, determine a corresponding indicator of a median RR interval difference between pairs of RR intervals separated by k intervals within the window; compare the indicator of the median RR interval difference with a threshold; and if the indicator of the median RR interval difference is less than the threshold, determine that the detection of a potential arrhythmic episode does not correspond to an actual arrhythmic episode. According to some such embodiments, if at least one of the processor or controller determines that the detection of a potential arrhythmic episode does not correspond to an actual arrhythmic episode for a value of k, then determining an indicator of a median RR interval for one or more further values ​​of k within the range of values ​​until the detection of a potential arrhythmic episode is determined to not correspond to an actual arrhythmic episode, or until no additional values ​​of k are available for selection for k within the range of values ​​of k. According to some such embodiments, the range of values ​​of k is from 1 to M, where M is an integer of at least 4, and wherein at least one of the processor or controller is configured such that: an initial value of k selected from the range of values ​​of k is 1, and each time there is another selected value of k, the another selected value of k is generated by incrementing k such that k=k+1; or an initial value of k selected from the range of values ​​of k is M, and each time there is another selected value of k, the another selected value of k is generated by decrementing k such that k=k-1. In accordance with certain embodiments, the corresponding indicator of the median RR interval difference between a pair of RR intervals determined for the value of k includes a median percentage difference value determined using the following equation: for interval n=1:N, median percentage difference=MEDIAN(100*|RR(n)–RR(n+k)| / RR(n)), where RR(n) is the nth RR interval in the window, RR(n+k) is the (n+k)th RR interval in the window, |RR(n)–RR(n+k)| is the absolute value of RR(n)–RR(n+k), and N is the total number of RR intervals in the window.

[0025] In some embodiments, the paired differences between the RR interval pairs separated by k intervals are expressed as percentages, and the median of the Nk paired difference percentages is expressed as a percentage, i.e., the median percentage difference for k as summarized above. In other embodiments, the paired differences between the RR interval pairs separated by k intervals are expressed as simple differences (i.e., RR(n)–RR(n+k)), also known as δ, which are not percentages, and the median of the Nk paired difference percentages is also expressed as a simple difference that is not a percentage. In other embodiments, the paired differences between the RR interval pairs separated by k intervals are expressed as ratios, the median of which is also expressed as a ratio. For example, the difference can be expressed by the ratio of RR(n) / RR(n+k) or RR(n+k) / RR(n), or by the higher of the two RR intervals compared to the lower of the two RR intervals or vice versa, rather than by calculating RR(n)–RR(n+k) to determine the difference between the nth and (n+k)th RR intervals. The closer such a ratio is to unity (i.e., 1), the more likely it is that a regularly irregular pattern exists, which is an indicator of a non-arrhythmic event. Where a ratio is used, the threshold used may be a threshold range, for example, between 0.8 and 1.2, but not limited thereto. Other variations are possible within the scope of the embodiments described herein.

[0026] This summary is not intended to be a complete description of embodiments of the present technology. Other features and advantages of embodiments of the present technology will become apparent from the following description, in which preferred embodiments have been set forth in detail in conjunction with the accompanying drawings and claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 Example EGM segments within a thirty second window up to and including the detection of a potential AF episode.

[0028] Figure 2 Shown included in Figure 1 A table showing a list of forty RR intervals within a thirty second window is shown in FIG.

[0029] Figure 3 is a graph showing forty RR intervals (in Figure 1 Within the thirty second window shown and Figure 2 ) versus time.

[0030] Figure 4A , 4B and 4C are high-level flow charts outlining various different methods of the present technology for improving the specificity of arrhythmia onset detection.

[0031] Figure 5A block diagram of one embodiment of an IMD implanted in a patient and that can be used to implement certain embodiments of the present technology is shown.

[0032] Figure 6 Shown for use with Figure 5 0048] Block diagram of one embodiment of an external device for communicating with and / or programming the IMD described in , which external device can be used to implement certain embodiments of the present technology. DETAILED DESCRIPTION

[0033] As is well known, each cardiac cycle represented in an electrogram (EGM) or electrocardiogram (ECG) generally includes a P wave, followed by a QRS complex, followed by a T wave, and the QRS complex includes a Q wave, an R wave, and an S wave. The P wave is caused by the depolarization of the atrium. This is followed by atrial contraction, which is indicated by a slight rise in atrial pressure, which helps to further fill the ventricle. Atrial contraction is followed by ventricular depolarization, as indicated by the QRS complex, which initiates ventricular contraction, causing the ventricular pressure to rise until the ventricular pressure exceeds the pulmonary and aortic diastolic pressures, causing forward flow when blood is ejected from the ventricle. Ventricular repolarization occurs thereafter, as indicated by the T wave, which is related to the beginning of ventricular diastole, in which the forward flow stops, the pressure in the ventricle drops below the pressure in the atrium, at which time the mitral and tricuspid valves open, and passive filling of the ventricle begins during diastole. The terms EGM, EGM signal, and EGM waveform are used interchangeably herein. Similarly, the terms ECG, ECG signal, and ECG waveform are used interchangeably herein. Both ECG and EGM signals are signals indicative of the electrical activity of a patient's heart, which may also be referred to as cardiac electrical signals, etc.

[0034] The R wave is the largest wave in the QRS complex, and it is usually identified by comparing a sample of the EGM or ECG with an R wave threshold. Various measurements can be obtained based on the EGM or ECG waveform, including the measurement of the RR interval, where the RR interval is the duration between a pair of consecutive R waves. As described above, in the background art, the common technique for detecting AF is based on the measurement of RR interval variability. However, for detecting AF, relying solely on the RR interval variability can lead to many false positive AF detections. This is because relying solely on the RR interval variability will not be able to detect the regularly irregular pattern of the RR interval, which does not indicate AF, but can be caused if multiple active atrial foci activate the ventricle in turn. As will be described in further detail below, certain embodiments of the present technology described herein provide improved AF identification by distinguishing between the regularly irregular pattern of the RR interval that does not indicate AF and the irregular irregular pattern that does indicate AF. Beneficially, such an embodiment can be used to reduce false positive AF detections, and thereby increase the specificity of AF detection. More generally, embodiments of the present technology can be used to increase the specificity of arrhythmia detection.

[0035] According to certain embodiments of the present technology, an ordered list of RR intervals within a window until the detection of a potential arrhythmia attack is obtained, wherein each of these RR intervals has a respective duration. A potential arrhythmia attack, which may also be referred to as an arrhythmia trigger, may occur because the RR interval variability exceeds a specified threshold for at least a specified time period, and / or using some other technique for detecting a potential arrhythmia attack. Such a window may be defined, for example, as a specified number N of RR intervals until the detection of a potential arrhythmia attack. Such a window may alternatively be defined, for example, as a specified number N of seconds until the detection of a potential arrhythmia attack. For a more specific example, N may be equal to thirty (30), in which case the aforementioned window may include thirty RR intervals until the potential arrhythmia attack, or RR intervals within a thirty second window until the detection of a potential arrhythmia attack. The use of smaller or larger values ​​for N is also within the scope of the embodiments described herein.

[0036] In certain embodiments, a measurement of a predominantly repeating RR interval pattern within a window is determined and compared to a pattern threshold. If the measurement of the predominantly repeating RR interval pattern is below the pattern threshold, then it is determined that the detection of a potential arrhythmia attack does not correspond to an actual arrhythmia attack. Conversely, if the measurement of the predominantly repeating RR interval pattern is above the pattern threshold, then there is an increased confidence that the detection of a potential arrhythmia attack corresponds to an actual arrhythmia attack, i.e., there is an increased confidence that the detected potential arrhythmia attack corresponds to a true positive detection. Although embodiments of the present technology can be used to increase the specificity of arrhythmia detection for a variety of different types of arrhythmias, such embodiments are particularly useful for increasing the specificity of AF detection. Therefore, in the discussion that follows, the type of arrhythmia discussed will typically be AF. However, it should be understood that embodiments of the present technology may also be used to increase the specificity of arrhythmia detection for various other types of arrhythmias besides AF, such as, but not limited to, bradycardia, asystole, atrial tachycardia (AT), ventricular tachycardia (VT), and ventricular fibrillation (VF). In addition, it is noted that the terms "arrhythmia attack", "attack of arrhythmia", "attack of arrhythmia", etc. are considered interchangeable.

[0037] Figure 1 An example EGM segment is included within a thirty (30) second window up to the detection of a potential AF episode, wherein there are forty (40) RR intervals within the window. The detection of a potential AF episode may also be referred to herein as potential AF detection, AF triggering, etc. In order to fit the EGM segments on a single page, Figure 1 The EGM clip in is divided into three separate ten-second panels, which are shown one above the other. Figure 1 Example RR intervals are labeled in the top panel of the three ten-second panels. The RR intervals in the top panel of the three ten-second panels are labeled RR1, RR2, RR3, ... RR12, where RR1 corresponds to RR interval number 1, RR2 corresponds to RR interval number 2, RR3 corresponds to RR interval number 3, ... RR12 corresponds to RR interval number 12. With the exception of RR40, which corresponds to RR interval number 40, the RR intervals in the other two ten-second panels are not labeled in order to minimize Figure 1 The chaos in .

[0038] Figure 2 The tables in Figure 1 A list of forty RR intervals within a thirty second window is shown in . Figure 2Shown in the rightmost column of is an example of an ordered list of RR intervals within the window until detection of a potential AF episode, where each of these RR intervals has a respective duration. Figure 2 , the duration of RR1 is 734.4 milliseconds (ms), the duration of RR2 is 632.8ms, the duration of RR3 is 921.9ms, ... the duration of RR40 is 625.0ms. Figure 2 It can be appreciated that the minimum RR interval duration shown therein is 275.8 ms, and the maximum RR interval duration shown therein is 1085.9 ms. Figure 2 As can be appreciated, there is a very high RR interval variability which is responsible for the detection of potential AF episodes.

[0039] Figure 3 is a graph showing forty RR intervals (in Figure 1 within the thirty second window shown in Figure 2 The minimum RR interval duration (275.8 ms) and the maximum RR interval duration (1085.9 ms) are indicated by arrows, which are also Figure 2 From Figure 3 It can also be appreciated that there is a very high RR interval variability which is responsible for the detection of potential AF episodes.

[0040] According to certain embodiments of the present technology, an ordered list of RR intervals within a window up to the detection of a potential AF episode (or other type of potential arrhythmia episode) is obtained and analyzed to determine whether there is a regularly irregular pattern hidden therein, which indicates that the detection of the potential AF episode is a false positive detection. More specifically, in certain embodiments, in order to determine whether there is a regularly irregular pattern of RR intervals within the window, pairwise differences are determined between pairs of RR intervals that are one interval apart, two intervals apart, three intervals apart, ... and M intervals apart. This results in M ​​groups of pairwise differences. For example, assuming there are forty RR intervals within the window (as described above with reference to Figure 1 and 2The first set of pairwise differences may include the difference between the first and second RR intervals, the difference between the second and third RR intervals, the difference between the third and fourth RR intervals, ... the difference between the thirty-ninth and fortieth RR intervals. The second set of pairwise differences may include the difference between the first and third RR intervals, the difference between the second and fourth RR intervals, the difference between the third and fifth RR intervals, ... the difference between the thirty-eighth and fortieth RR intervals. The third set of pairwise differences may include the difference between the first and fourth RR intervals, the difference between the second and fifth RR intervals, the difference between the third and sixth RR intervals, ... the difference between the thirty-seventh and fortieth RR intervals. The Mth set of pairwise differences may include the difference between the first and Mth RR intervals, the difference between the second and (M+1)th RR intervals, the difference between the third and (M+2)th RR intervals, ... the difference between the (40-M)th and 40th RR intervals. For each of the M groups of paired intervals, the median (or an indicator of the median) of the group is determined, where a relatively low median (or its indicator) indicates the presence of a hidden pattern (most likely meaning that AF did not occur) and a relatively high median (or its indicator) indicates the absence of a regularly irregular pattern (most likely meaning that AF did in fact occur). In some embodiments, the minimum median (or its indicator) of the M groups is identified and compared to a pattern threshold, where the minimum median (or its indicator) is an example of a measure of a pattern that is predominantly repeated within a window. If the minimum median (or its indicator) is below the pattern threshold, then it is determined that the detection of a potential AF episode does not correspond to an actual AF episode. Conversely, if the minimum median (or its indicator) is above the pattern threshold, then there is an increased confidence that the detected potential AF episode corresponds to an actual AF episode, i.e., there is an increased confidence that the detected potential AF episode corresponds to a true positive detection. Refer below to Figure 4A The high-level flow chart in describes additional details of the embodiment outlined above. More specifically, Figure 4A Certain methods of the present technology for improving AF episode detection specificity are outlined. Such methods can be triggered in response to detection of a potential AF episode.

[0041] refer to Figure 4A, step 402 involves obtaining an ordered list of RR intervals within a window until detection of a potential arrhythmia attack (e.g., a potential AF attack), wherein each of these RR intervals has a respective duration. For example, an ordered list of RR intervals may be obtained by identifying R waves within an EGM or ECG segment and determining intervals between consecutive R waves among these R waves, thereby generating an ordered list of RR intervals. Such R waves within an EGM or ECG segment may be identified by comparing the EGM or ECG segment or a sample thereof with an R wave sensing threshold and identifying the R wave when the R wave sensing threshold is reached or exceeded. Other variations are possible within the scope of the embodiments described herein. For example, an R wave or QRS complex morphology template may be used alternatively or additionally to identify the R wave.

[0042] As mentioned above, in Figure 2 An example of an ordered list of RR intervals within a window until detection of a potential AF episode is shown in . The ordered list of RR intervals obtained at step 402 will preferably include only true RR intervals. However, due to oversensing of T waves and / or P waves, the ordered list of intervals obtained at step 402 may also include one or more oversensed RR intervals. In other words, the ordered list of RR intervals included in the window until detection of a potential arrhythmia episode, such as a potential AF episode (also referred to as an "AF trigger"), in addition to true RR intervals, may also include one or more oversensed RR intervals that may exist, for example, if one or more P waves and / or T waves are mistakenly identified as R waves. In order to maximize the reference Figure 4A Specific to the outlined method, one or more techniques for identifying and removing over-sensed RR intervals may be performed prior to step 402 , as part of step 402 , or between step 402 and the next step 403 .

[0043] Still reference Figure 4A , at step 403, values ​​of various variables are specified. More specifically, at step 403, there is a setting of k=1, where k represents the interval spacing between the RR intervals in the RR interval pair. In addition, a value of N is specified, where N is the total number of RR intervals within the window until the detection of a potential AF episode. In the above reference Figure 1 and Figure 2 In the example discussed, N=40. Additionally, in step 403, M is set to the maximum interval spacing to be analyzed. According to some embodiments, M is an integer at least equal to 4. In some examples described herein, M is set to 10, ie, in such examples, M=10.

[0044] Step 403 involves setting n=1, where n is the interval indicator.

[0045] Step 406 involves determining the pairwise difference between the duration of the nth interval and the duration of the n+kth interval. Assuming n=1 and k=1 the first time step 406 is performed, then step 406 involves determining the pairwise difference between the duration of the first interval and the duration of the second interval.

[0046] At step 408, the value of n is incremented by setting n = n + 1. Thus, the first time step 408 is performed, n = 1 + 1, thereby setting n = 2.

[0047] At step 410, there is a determination as to whether n+k is greater than N. Assuming N=40, the first time step 410 is performed, n+k will equal 3, which is less than N (i.e., less than 40), resulting in a "no" answer to the determination at step 410. If the answer to the determination at step 410 is "no," then the flow returns to step 406. This will result in Nk pairwise differences being determined before the answer to the determination at step 410 is ultimately "yes," at which point the flow proceeds to step 412. Thus, for n=1, iterations of steps 406 to 410 will result in a pairwise difference between the first and second RR intervals, a pairwise difference between the second and third RR intervals, a pairwise difference between the third and fourth RR intervals, ..., and a pairwise difference between the thirty-ninth and fortieth RR intervals, resulting in a total of 39 pairwise differences being determined for k=1. More generally, N k pairwise differences are determined prior to a “yes” answer to the determination at step 410 , which results in flow proceeding to step 412 .

[0048] At step 412, there is a determination of the median of the Nk pairwise differences determined for the value of k. According to certain embodiments, the median value identified at step 412 is at least temporarily saved so that it can be compared with a later determined median value determined in a later instance of step 412. To find the median of the Nk pairwise differences for the value of k, the pairwise differences may be ordered from smallest to largest, with the median being the value halfway through the set, i.e., the middle value. If there is an even number of values ​​in the data set, then the median may be determined by determining the mean (average) of the two middlemost numbers, or either of the two middlemost numbers may be selected depending on the particular implementation.

[0049] At step 414, the value of k is incremented by setting k = k + 1. Thus, the first time step 414 is performed, k = 1 + 1, thereby setting k = 2.

[0050] At step 416, there is a determination whether k is greater than M. If the answer to the determination at step 416 is "no," then the flow returns to step 405, where n is reset to one.

[0051] Steps 406 to 410 are then repeated for k=2, which results in the determination of the pairwise differences between the first and third RR intervals, the pairwise differences between the second and fourth RR intervals, the pairwise differences between the third and fifth RR intervals, ... and the pairwise differences between the thirty-eighth and fortieth RR intervals, resulting in a total of thirty-eight pairwise differences. Then, in the next instance at step 412, there is a determination of the median of the thirty-eight pairwise differences determined for the value of k=2.

[0052] At the next instance of step 414, the value of k is incremented by setting k = k + 1. Thus, the second time step 414 is performed, k = 2 + 1, thereby setting k = 3.

[0053] At the next instance of step 416, there is a determination whether k (now equal to 3) is greater than M. Assuming M=10, the answer to the determination at step 416 is "no" and the flow returns to step 405 where n is reset to 1.

[0054] Steps 406 to 410 are then repeated for k=3, which results in the determination of the pairwise differences between the first and fourth RR intervals, the pairwise differences between the second and fifth RR intervals, the pairwise differences between the third and sixth RR intervals, ... and the pairwise differences between the thirty-seventh and fortieth RR intervals, resulting in a total of thirty-seven pairwise differences. Then, in the next instance at step 412, there is a determination of the median of the thirty-seven pairwise differences determined for the value of k=3.

[0055] Assuming M=10, ten instances of step 412 will be performed before the answer is determined to be "yes" at step 416, resulting in the determination of 10 median pairwise differences (one for each k of k=1, k=2, k=3, ... and k=10) before the process proceeds to step 418. At step 418, the minimum of the median of the pairwise differences is identified, where the value of k corresponding to the minimum of the median of the pairwise differences corresponds to a predominantly repeating RR-interval pattern. For example, if the minimum of the 10 medians of the pairwise differences corresponds to k=3, then it can be said that the most predominant pattern within the window is an RR-interval pattern that repeats every three RR intervals. This happens to be Figure 1 and Figure 2 Briefly refer back to Figure 2It can be understood that the duration of the first RR interval is 734.4ms, the duration of the fourth RR interval is 773.4ms, the duration of the seventh RR interval is 773.4ms, the duration of the tenth RR interval is 757.8ms, the duration of the thirteenth RR interval is 757.8ms, … the duration of the thirty-sixth RR interval is 765.6ms, and the duration of the thirty-ninth RR interval is 820.3ms.

[0056] Back to Figure 4A As discussed above, after identifying the minimum of the M median pairwise differences at step 418, the process proceeds to step 420.

[0057] At step 420, a determination is made as to whether the minimum of the M median pairwise differences (identified at step 418) is less than a corresponding threshold. If the answer to the determination at step 420 is "no," this indicates that no hidden pattern has been identified, and the process proceeds to step 422. At step 422, a potential arrhythmia episode (also referred to as an arrhythmia trigger) is classified as a true positive, or further arrhythmia identification is performed. If the answer to the determination at step 420 is "yes," this indicates that a hidden regularly irregular pattern is being identified, and the process proceeds to step 424. At step 424, a potential arrhythmia episode (also referred to as an arrhythmia trigger) is classified as a false positive, or further arrhythmia identification is performed. In a specific embodiment for AF identification, at step 422, a potential AF episode (also referred to as an AF trigger) is classified as a true positive, or further AF identification is performed. If the answer to the determination at step 420 is "yes," this indicates that a hidden regularly irregular pattern is being identified, and the process proceeds to step 424; and at step 424, the potential arrhythmia episode (also called an arrhythmia trigger) is classified as a false positive, or further arrhythmia identification is performed.

[0058] refer to Figure 4AThe method of overview essentially performs pattern recognition, wherein the method first calculates the median RR interval difference between each interval and the next interval (interval 1 to interval 2, interval 2 to interval 3, interval 3 to interval 4, etc.), i.e., between intervals separated by 1 interval. Repeat the calculation for 2 intervals separated by 2 intervals (interval 1 to interval 3, interval 2 to interval 4, etc.), 3 intervals separated by 3 intervals, and at most 10 intervals separated by 10 intervals (or for some other values ​​of M, wherein M is an integer of at least 4). This results in 10 (or more generally, M) median paired differences across all intervals in the window before the AF trigger mark. Then, the minimum value of these paired differences across all intervals is used to identify the interval period of the main pattern. For example, a pattern that repeats itself every 3 intervals will result in the minimum median difference calculated between each interval and 3 intervals later. If the “Minimum Median Pairwise Difference” is below a specified threshold, the interval variability that causes an arrhythmia trigger (e.g., AF trigger) can be interpreted as actually associated with a repetitive pattern (e.g., false positive AF) and the arrhythmia trigger (e.g., AF trigger) can be rejected.

[0059] An example algorithm that may be used to determine the median of N-1 pairwise difference values ​​is shown below, where the median is expressed as a median percentage difference.

[0060] For interval n = 1:N, median percentage difference = MEDIAN(100*|RR(n)–RR(n+1)| / RR(n)),

[0061] in

[0062] RR(n) is the nth RR interval in the window,

[0063] RR(n+1) is the (n+1)th RR interval in the window,

[0064] |RR(n)–RR(n+1)| is the absolute value of RR(n)–RR(n+1), and

[0065] N is the total number of RR intervals in the window.

[0066] The above calculation compares intervals separated by 1 interval, i.e., 1 interval apart. The calculation is then repeated for interval spacings from 2 to 10, or more generally for intervals from 2 to M. The calculation is extended for all interval spacings k = 1:10 or more generally k = 1:M using the following equation:

[0067] For interval n=1:N and interval spacing k=1:M, for k, median percentage difference=MEADIAN(100*|RR(n)–RR(n+k)| / RR(n)).

[0068] This resulted in 10 median percentage differences, each comparing intervals separated by 1 to 10 intervals. The minimum of these pairwise differences across all intervals was then used to identify the interval periodicity of the main pattern (i.e., how many intervals it took for the pattern to repeat).

[0069] As mentioned above, Figure 2 The table in includes a list of 40 RR intervals within a thirty second window until the detection of a potential AF episode (eg Figure 1 , 10 intervals, the median percentage difference of the intervals separated by 1 to 10 intervals calculated using the above equation is as follows: Median percentage difference = [24.4%, 30.5%, 3.0%, 18.8%, 31.3%, 9.6%, 16.5%, 29.1%, 17.3%, 13.8%]. In this example, when each interval is compared to the interval 3 intervals later, the minimum percentage difference (3.0%) occurs (i.e., the 3rd of the 10 median percentage differences). Assuming a threshold of 6%, it can be understood that in this example, the minimum median percentage difference (i.e., 3.0%) is less than the threshold of 6%, and therefore it can be concluded that the interval variability that causes AF triggering is actually associated with a repetitive pattern (i.e., false positive AF) and can be rejected.

[0070] To reduce processing time, the algorithm can stop calculating the median percentage difference as soon as any median RR interval difference, or indicator thereof (e.g., median percentage difference) below a threshold is observed. In other words, and more generally, instead of determining all M indicators, then identifying a minimum, and then comparing that minimum to a threshold, the following operations can instead be performed. Each time a new median RR interval difference, or indicator thereof (e.g., median percentage difference) is determined for a value of k, the median RR interval difference can be compared to the threshold, and then as soon as there is a determination that the median is below the threshold, there can be a conclusion that the detection of a potential arrhythmia episode is a false positive arrhythmia detection (e.g., a potential AF episode is a false positive AF detection). If the determined median is not less than the threshold, another value of k is selected and analyzed until there is a conclusion of a false positive arrhythmia detection or there are no longer new values ​​for k to select and analyze. Using the above reference Figure 1 and Figure 2 For the example windows described, only three median percentage differences need to be calculated (i.e., for values ​​of k = 1, 2, and 3) to reject an AF detection as a false positive. Figure 4B Examples of such alternative embodiments are described.

[0071] More specifically, Figure 4BFurther methods of the present technology for improving specificity of arrhythmic episode detection are outlined, where such methods may be triggered in response to detection of a potential arrhythmic episode (eg, a potential AF episode). Figure 4B References Figure 4A Steps that are identical to the steps discussed are labeled identically and do not require detailed description, as further details for such steps can be found in Figure 4A For reference. Figure 4B , step 402, step 403, step 405, step 406, step 408, step 410 and step 412 are Figure 4A The steps numbered together are the same as those in . The first time step 412 is performed, i.e., for k=1, there is a determination of the median of the Nk (i.e., N-1, e.g., if N=40, then Nk=39 when k=1) pairwise differences determined for the value of k=1, and at step 413, the median of the N-1 pairwise differences (or an indicator thereof, such as a median percentage difference) is compared to a threshold value (e.g., 6%). If the median is less than the threshold value (i.e., if the answer to the determination at step 413 is "yes"), the process proceeds to step 424. At step 424, the potential arrhythmia episode (also known as an arrhythmia trigger) is classified as a false positive, or further arrhythmia identification is performed. If the median is not less than the threshold value (i.e., if the answer to the determination at step 413 is "no"), the process proceeds to step 414.

[0072] At step 414, the value of k is incremented by setting k = k + 1. Thus, the first time step 414 is performed, k = 1 + 1, thereby setting k = 2.

[0073] At step 416 , there is a determination as to whether k is greater than M. If the result of the determination at step 416 is “no,” then the flow returns to step 405 , where n is reset to 1.

[0074] Steps 406 to 410 are then repeated for k=2, which will result in the determination of the pairwise differences between the first and third RR intervals, the pairwise differences between the second and fourth RR intervals, the pairwise differences between the third and fifth RR intervals, ... and the pairwise differences between the thirty-eighth and fortieth RR intervals, resulting in a total of 38 pairwise differences. Then, in the next instance at step 412, there is a determination of the median of the thirty-eight pairwise differences determined for the value of k=2. Then, in the next instance at step 413, the median of the N-2 pairwise differences for k=2 (or an indicator thereof, such as a median percentage difference) is compared to a threshold value (e.g., 6%). If the median is less than the threshold value (i.e., if the answer to the determination at step 413 is "yes"), the flow proceeds to step 424, otherwise the flow proceeds to step 414.

[0075] Assumption reference Figure 4B The method outlined is to use the reference Figure 1 and Figure 2 If the median percentage difference is performed for the RR intervals within the described example window (up to and including the detection of a potential AF episode), then for k=3, the median percentage difference would be calculated to be 3.0%, resulting in a "yes" answer to the determination at step 413, and the detection of a potential AF episode being classified as a false positive at step 424. It will be appreciated, therefore, that for this example, the median percentage difference only needs to be performed for three values ​​of k (i.e., k=1, 2, and 3) before the method reaches a false positive conclusion.

[0076] In reference Figure 4B In the method outlined, the median of the pairwise differences for k=1 (or an indicator thereof) is first determined and compared to a threshold. If the median for k=1 is not less than the threshold, the median of the pairwise differences for k=2 (or an indicator thereof) is determined and compared to the threshold. If the median for k=2 is not less than the threshold, the median of the pairwise differences for k=3 (or an indicator thereof) is determined and compared to the threshold. This process is repeated, then, until the median (or an indicator thereof) is greater than the threshold, or until there are no more values ​​of k to analyze within the range of values ​​of k (e.g., until k=M). Thus, in Figure 4A, the initial value of k selected from the range of values ​​of k is 1, and the median (or an indicator thereof) of the pairwise differences is determined for k=1. Then, if the median is not less than the threshold, 405-412 is repeated for another selected value of k, wherein another selected value of k is generated by incrementing k so that k=k+1. In other words, different values ​​of k are tested in increasing numerical order (starting from k=1), where k specifies the number of intervals separated by RR intervals in each pairwise difference. In an alternative embodiment, different values ​​of k may be tested in decreasing numerical order starting from k=M, where M is the maximum interval spacing to be tested, for example, M=10, or more generally, M is an integer of at least 4. More generally, the values ​​of k may be tested in any order, including ascending, descending, random, or some predetermined order. Examples of predetermined orders for testing values ​​of k in the range of 1 to 10 may be 6, 5, 7, 4, 8, 3, 9, 2, 1, 10. Figure 4C The flowchart shown in summarizes how this can be achieved by selecting an initial (or another) value for k at each instance of step 404 (where k need not start with k=1, but may start with k=1), and by determining at each instance of step 419 whether there is at least one additional value of k for selection within the range of values ​​of k from 1 to M. Figure 4B The method outlined is actually a reference Figure 4C A special case of the outlined method in which different values ​​of k are tested in ascending order starting from k=1. Figure 4C Step 403' and Figure 4B Unlike step 403 in , k=1 is not initially set. Instead, any initial value of k (within the range of values ​​from 1 to M) is selected at the first instance of step 404. Figure 4C Zhongyu Figure 4A and Figure 4B The other steps with the same number are the same as above. Figure 4A and Figure 4B Those steps described are the same and do not need to be described again.

[0077] In some embodiments, the paired differences between the interval pairs separated by k intervals are expressed as percentages, and the median of the Nk paired difference percentages is expressed as a percentage, i.e., the median percentage difference for k, as described above. In other embodiments, the paired differences between the interval pairs separated by k intervals are expressed as simple differences (i.e., |RR(n)–RR(n+k)|), i.e., δ, which are not percentages, and the median of the Nk paired difference percentages is also expressed as a simple difference that is not a percentage. In other embodiments, the paired differences between the interval pairs separated by k intervals are expressed as ratios, as are the medians. For example, the difference can be expressed by the ratio of RR(n) / RR(n+k) or RR(n+k) / RR(n), or by the higher of the two RR intervals compared to the lower of the two RR intervals or vice versa, rather than by calculating RR(n)–RR(n+k) to determine the difference between the nth and (n+k)th RR intervals. The closer the ratio is to unity (i.e., a value of 1), the more likely it is that there is a regularly irregular pattern, which is an indicator of a non-AF event. Where a ratio is used, the threshold used may be a threshold range, such as, but not limited to, between 0.8 and 1.2. Other variations are possible within the scope of the embodiments described herein.

[0078] According to certain embodiments, in response to detecting a potential arrhythmia episode (eg, a potential AF episode), the IMD may perform the steps described above with reference to Figure 4A-4C One of the methods described. As described above, detection of a potential arrhythmia episode may also be referred to as an arrhythmia trigger (e.g., an AF trigger). Such an IMD may be configured to transmit data corresponding to the potential arrhythmia episode detected by the IMD to an external device communicatively coupled to the patient care network. In some such embodiments, the IMD does not (is prevented from) sending data corresponding to the potential arrhythmia episode detected by the IMD but which is subsequently determined by the IMD to be a false positive detection (to an external device communicatively coupled to the patient care network).

[0079] The present technology described above was tested to determine whether and to what extent the present technology could be used to reduce reporting of false positive AF detections. Such testing demonstrated that embodiments of the present technology could be used to identify and thereby reject more than 30% of false positive AF detections compared to if AF detection were based solely on comparing measurements of RR interval variability to a variability threshold. Beneficially, testing also showed that embodiments of the present technology incorrectly flagged less than 1% of AF detections as false positives, where they were actually true positives. Thus, embodiments of the present technology can be used to significantly increase AF detection specificity without significantly reducing AF detection sensitivity.

[0080] Embodiments of the present technology may be used with other types of techniques that may be used to identify false positive arrhythmia detections, or more generally, may be used with other techniques for improving arrhythmia detection and / or discrimination. Figure 4A , Figure 4B and Figure 4C Prior to or as part of one or more instances of step 402 in , R-wave oversensing and / or R-wave understanding may be detected and accounted for and / or corrected within an ordered list of sensing intervals. For another example, the ordered list of sensing intervals until an arrhythmia trigger (e.g., AF trigger) obtained at step 402, which would have been obtained after a potential R-wave detection and / or potential RR interval, has been subjected to another type of arrhythmia discriminator (e.g., AF discriminator) such as a sudden AF discriminator that scans the heartbeats within a window until the AF detection to characterize heart rate changes and rejects the AF detection as a false positive if there is no significant change in the heart rate until the AF detection. Additionally or alternatively, there may be a determination of whether R-wave undersensing exceeding a corresponding threshold or intermittent AV conduction block within the window until the arrhythmia detection (e.g., AF detection) occurs, and if at least one of them does occur, the arrhythmia detection (e.g., AF detection) may be rejected as a false positive. Additionally or alternatively, irregular RR intervals due to T-wave and / or P-wave oversensing in the window up to AF detection may be analyzed, and if their extent exceeds corresponding thresholds, arrhythmia detection (e.g., AF detection) may be rejected as a false positive. These are just a few examples of other types of time-based arrhythmia identification techniques that may be used in conjunction with the present invention. Figure 1-4C In addition or alternatively, the present invention may be used in conjunction with, or in conjunction with, the embodiments of the present invention described herein. Figure 1-4C Embodiments of the technology described herein may be preceded or followed by the use of one or more morphology-based arrhythmia discriminators (e.g., AF discriminators). For example, there may be a determination as to whether P wave oversensing exceeding a corresponding threshold within a window up to and prior to AF detection has occurred (based on the DGM morphology), and if so, the AF detection may be rejected as a false positive. Additionally or alternatively, there may be a determination as to whether an actual P wave exceeding a specified threshold within a window up to and prior to AF detection exists (since P waves should not be present during an episode of AF), and if so, the AF detection may be rejected as a false positive. The foregoing AF discriminators may be applied in a variety of different orders, as described herein with reference to the accompanying drawings. Figure 1-4C At least one of the described embodiments of the present technology is used as only one of a plurality of AF discriminators. Other variations are possible and within the scope of the embodiments of the present technology described herein.

[0081] The specific thresholds (e.g., 6%) and the thresholds mentioned above for testing the performance of embodiments of the present technology can be optimized more systematically for a wider patient population or for individual patients. In addition, the maximum number M of interval spacings analyzed can also be similarly optimized to determine whether a regular irregular pattern exists. Therefore, the embodiments of the present technology described herein should not be limited to use with the exemplary thresholds of other values ​​described herein.

[0082] Embodiments of the present technology described herein can be used for various types of IMDs, including but not limited to insertable cardiac monitors (ICMs), pacemakers to which one or more leads are attached, leadless pacemakers (LCPs), or implantable cardioverter-defibrillators (ICDs). Such an ICD can be a transvascular ICD, or a non-vascular ICD, wherein the non-vascular ICD can be a subcutaneous (SubQ) ICD. In the case where an embodiment of the present technology is implemented by an ICM, such an embodiment can be used, for example, to reduce the number of false-positive AF detections transmitted from the ICM to a patient care network for review by a clinician. This is beneficial because false-positive arrhythmia detections (e.g., false-positive AF detections) are highly undesirable because the burden of classifying a large number of clinically irrelevant AF episodes and / or other types of arrhythmias can be time-consuming and expensive. In the case where an embodiment of the present technology is used by an ICD or an IMD that communicates with an ICD, such an embodiment can reduce the frequency of delivering defibrillation shocks in response to false-positive AF detections. This is beneficial because defibrillation shocks are often painful, and delivering such shocks in response to a false positive AF detection subjects the patient to an unnecessary shock and may prematurely deplete energy stored in the battery. Figure 4A-4C It is also possible that one or more of the described methods are performed by an external device (e.g., a home monitor or external programmer) that obtains the EGM segments prior to potential AF detection from the IMD, or receives an ordered list of RR intervals (within the window until before the onset of potential AF) from the IMD. It is also possible that Figure 4A-4C Some of the steps in are performed by the IMD, while other steps are performed by external devices. Other variations are also possible and within the scope of the embodiments described herein.

[0083] Figure 5A block diagram of an embodiment of an IMD implanted in a patient according to an embodiment of the present technology is shown. IMD 501 can be implemented as a fully functional biventricular pacemaker, equipped with sensing and pacing circuits for both the atria and ventricles, for four-chamber sensing and stimulation therapy (including both pacing and shock therapy). Optionally, IMD 501 can provide fully functional cardiac resynchronization therapy. Alternatively, IMD 501 can be implemented with a simplified set of functions and components. For example, if the IMD is an ICM, the IMD can be implemented without pacing. IMD 501 can be coupled to one or more leads for single-chamber or multi-chamber pacing and / or sensing. Alternatively, IMD 501 can be an LCP including electrodes located on or very close to the housing 500 of IMD 501.

[0084] The IMD 501 has a housing 500 that houses electronic / computing components. The housing 500 (which is often referred to as a "can," "shell," "package," or "shell electrode") can be programmably selected to act as a return electrode for certain stimulation modes. The housing 500 may also include a connector (not shown) having a plurality of terminals 502, 504, 506, 508, and 510. The terminals can be connected to electrodes located at different locations on the housing 500, or to electrodes located on the leads. The IMD 501 includes a programmable microcontroller 520 that controls various operations of the IMD 501, including cardiac monitoring and / or stimulation therapy. The microcontroller 520 includes a microprocessor (or equivalent control circuitry), RAM and / or ROM memory, logic and timing circuits, state machine circuits, and I / O circuits.

[0085] IMD 501 also includes a pulse generator 522 that generates stimulation pulses and communication pulses for delivery by one or more electrodes coupled thereto. Pulse generator 522 is controlled by microcontroller 520 via control signal 524. Pulse generator 522 can be coupled to (multiple) selected electrodes via electrode configuration switch 526, which includes multiple switches for connecting desired electrodes to appropriate I / O circuits, thereby facilitating electrode programmability. Switch 526 is controlled by control signal 528 from microcontroller 520.

[0086] exist Figure 5 In the embodiment of FIG. 5 , a single pulse generator 522 is shown. Alternatively, the IMD may include multiple pulse generators similar to pulse generator 522, wherein each pulse generator is coupled to one or more electrodes and controlled by microcontroller 520 to deliver selected stimulation pulses to the corresponding one or more electrodes.

[0087] The microcontroller 520 is shown as including a timing control circuit 532 to control the timing of stimulation pulses (e.g., pacing rate, atrio-ventricular (AV) delay, atrial conduction (AA) delay, or interventricular conduction (VV) delay, etc.). The timing control circuit 532 can also be used for timing of refractory periods, blanking intervals, noise detection windows, evoked response windows, alarm intervals, marker channel timing, etc. The microcontroller 520 also has an arrhythmia detector 534 and a morphology detector 536 for detecting arrhythmia conditions. The arrhythmia detector 534 can, for example, detect potential onset of AF and / or other types of arrhythmias. Although not shown, the microcontroller 520 can also include other dedicated circuits and / or firmware / software components that assist in monitoring various conditions of the patient's heart and managing pacing therapy. The microcontroller 520 is also shown as including an arrhythmia identifier 540, which can be used to perform the above-referenced Figure 1-4C Embodiments of the present technology are described, for example, to determine whether one or more potential episodes of AF and / or other types of arrhythmias are false positives. Arrhythmia identifier 540 can be implemented more generally using hardware, software, firmware, and / or a combination thereof. The microcontroller can include a processor. The microcontroller and / or its processor can be used to perform the methods of the present technology described herein.

[0088] The IMD 501 may also be equipped with a communication modem (modulator / demodulator) to enable wireless communication with a remote slave pacing unit. The modem may include one or more transmitters and two or more receivers. In one embodiment, the modem may use low-frequency or high-frequency modulation. As an example, the modem may transmit implant-to-implant (i2i) messages and other signals through conducted communication between electrode pairs. Such a modem may be implemented in hardware as part of the microcontroller 520, or as software / firmware instructions that are programmable into the microcontroller 520 and executed by the microcontroller 520. Alternatively, the modem may be placed separately from the microcontroller as an independent component.

[0089] IMD 501 includes a sensing circuit 544 selectively coupled to one or more electrodes performing sensing operations to detect the presence of cardiac activity in the right chamber of the heart through switch 526. Sensing circuit 544 may include a dedicated sensing amplifier, a multiplexed amplifier, or a shared amplifier. It may also employ one or more low power precision amplifiers with programmable gain and / or automatic gain control, bandpass filtering, and threshold detection circuits to selectively sense cardiac signals of interest. Automatic gain control enables the unit to sense low amplitude signals characteristic of atrial fibrillation. Switch 526 determines the sensing polarity of the cardiac signal by selectively closing the appropriate switch. In this way, the clinician can program the sensing polarity independently of the stimulation polarity.

[0090] The output of the sensing circuit 544 is connected to the microcontroller 520, which in turn triggers or inhibits the pulse generator 522 in response to the presence or absence of cardiac activity. The sensing circuit 544 receives control signals 546 from the microcontroller 520 for controlling the gain, threshold, polarization charge removal circuit (not shown), and the timing of any blocking circuit (not shown) coupled to the sensing circuit input.

[0091] exist Figure 5 In the embodiment of FIG. 5 , a single sensing circuit 544 is shown. Alternatively, the IMD may include multiple sensing circuits similar to sensing circuit 544, wherein each sensing circuit is coupled to one or more electrodes and controlled by microcontroller 520 to sense electrical activity detected at the corresponding one or more electrodes. Sensing circuit 544 may operate in a unipolar sensing configuration or a bipolar sensing configuration.

[0092] IMD 501 also includes an analog-to-digital (A / D) data acquisition system 550 coupled to one or more electrodes via switch 526 to sample cardiac signals across any desired electrode pair. Data acquisition system 550 is configured to acquire intracardiac electrogram signals, convert raw analog data into digital data, and store the digital data for later processing and / or telemetry transmission to an external device 554 (e.g., a programmer, a local transceiver, or a diagnostic system analyzer). Data acquisition system 550 is controlled by control signals 556 from microcontroller 520.

[0093] Microcontroller 520 is coupled to memory 560 via a suitable data / address bus. Programmable operating parameters used by microcontroller 520 are stored in memory 560 and are used to customize the operation of IMD 501 to suit the needs of a particular patient. Such operating parameters define, for example, pacing pulse amplitude, pulse duration, electrode polarity, rate, sensitivity, automatic features, arrhythmia detection criteria, and the amplitude, waveform, and vector of each shock pulse to be delivered to the patient's heart within each corresponding layer of therapy.

[0094] The operating parameters of IMD 501 can be non-invasively programmed into memory 560 by telemetry circuitry 564 that communicates telemetrically with external device 554 via communication link 503. Telemetry circuitry 564 allows intracardiac electrograms and status information related to the operation of IMD 501 (such as contained in microcontroller 520 or memory 560) to be sent to external device 554 via communication link 503. Telemetry circuitry 564 may also be referred to as transceiver 564.

[0095] IMD 501 may save in memory 560 and / or transmit to an external device (e.g., 502) data corresponding to one or more arrhythmic episodes detected by IMD 501 so that the data is available at a later time for further analysis. If IMD 501 determines that the detection of an arrhythmic episode does not correspond to an actual arrhythmic episode, the IMD may allow the data corresponding to the potential arrhythmic episode to be overwritten in memory 560 and / or may prevent the data corresponding to the potential arrhythmic episode from being transmitted to an external device using transceiver 564.

[0096] IMD 501 may also include a magnet detection circuit (not shown) coupled to microcontroller 520 to detect when a magnet is placed on the unit. The clinician may use the magnet to perform various test functions of IMD 501 and / or to signal microcontroller 520 that external device 554 is in place to receive data or send data to microcontroller 520 via telemetry circuit 564.

[0097] IMD 501 may also include one or more physiological sensors 570. Such sensors are often referred to as "rate-responsive" sensors because they are often used to adjust the pacing stimulation rate according to the patient's motion state. However, (multiple) physiological sensors 570 may be further used to detect changes in cardiac output, changes in the physiological condition of the heart, or diurnal changes in activity (e.g., detecting sleep and wakefulness). The signals generated by (multiple) physiological sensors 570 are transmitted to the microcontroller 520 for analysis. The microcontroller 520 responds by adjusting various pacing parameters (such as rate, AV delay, VV delay, etc.) with which it manages atrial and ventricular pacing pulses. Although shown as being included in IMD 501, one or more physiological sensors 570 may be outside of IMD 501, but still implanted in the patient or carried by the patient. Examples of physiological sensors include, for example, sensors that sense respiratory rate, blood pH, ventricular gradient, activity, position / posture, minute ventilation, etc.

[0098] Battery 572 provides operating power for all components in IMD 501. Battery 572 is preferably capable of operating at low current consumption for long periods of time and capable of providing high current pulses (for capacitor charging) when a patient requires a shock pulse (e.g., greater than 2A, voltage greater than 2V, for 10 seconds or more). Battery 572 also desirably has predictable discharge characteristics so that selective replacement time can be detected. As an example, IMD 501 employs a lithium / silver vanadium oxide battery.

[0099] IMD 501 also includes impedance measurement circuitry 574, which can be used for many things, including: lead impedance monitoring during acute and chronic stages for proper lead positioning or displacement; detecting operable electrodes and automatically switching to operable electrode pairs if displacement occurs; measuring respiration or minute ventilation; measuring chest impedance to determine shock thresholds; detecting when the device has been implanted; measuring stroke volume; and detecting the opening of a heart valve; etc. Impedance measurement circuitry 574 is coupled to switch 526 so that any desired electrode can be used. In this embodiment, IMD 501 also includes shock circuitry 580 coupled to microcontroller 520 via data / address bus 582.

[0100] The embodiments of the present technology described above are mainly described as being used with an implantable medical device or system that monitors HR and / or one or more types of arrhythmia attacks based on sensing intervals (e.g., RR intervals). Such embodiments of the present technology may alternatively be used with a non-implantable device or system (also referred to as an external device or system) that includes at least two electrodes in contact with a person's skin and is used to monitor HR and / or one or more types of arrhythmia attacks based on sensing intervals. More specifically, such embodiments may alternatively be used with a user wearable device or implemented by a user wearable device (such as a wrist-worn device, or a user wearable device designed to be worn on one or more other parts of a person's body other than the wrist, such as an ankle, upper arm, or chest, but not limited thereto). Such a user wearable device may include electrodes configured to contact a person's skin, a sensing circuit (also referred to as a sensing circuit) coupled to the electrodes and configured to obtain a signal indicating the patient's cardiac electrical activity, and at least one of a processor or controller configured to execute one or more of the algorithms described above. Such a user-wearable device (or more generally, an external device or system) can monitor AF and / or (multiple) other types of arrhythmias and determine when a false positive detection exists. The user-wearable device can both obtain a signal indicative of the patient's cardiac electrical activity and monitor the person's HR and / or (multiple) arrhythmias based on the RR intervals obtained from the obtained signal. Alternatively, the user-wearable device can be communicatively coupled to another external device, such as a smartphone or tablet, and the other external device can obtain a signal from the user-wearable device and monitor the person's HR and / or (multiple) arrhythmias based on the RR intervals, etc. The user-wearable device or other external device or system can determine when a false positive arrhythmia detection may exist. Other implementations of such an external device or system are also possible and are within the scope of the embodiments described herein.

[0101] Figure 6An example assembly of an example external device 554 for communicating with and / or programming an IMD 501 is shown. In a certain embodiment, an external device 502 can be used to analyze EGM segments obtained and stored by an IMD 501. More generally, an external device 502 can allow a doctor or other authorized user to program the operation of an IMD 501, and retrieve and display information received from the IMD 501, such as EGM data and device diagnostic data. Additionally, an external device 502 can receive and display ECG data from a separate external ECG lead that may be attached to a patient. In addition, the external device 502 can enable the IMD to perform functions necessary to complete certain algorithms of the present invention. Depending on the specific programming of the programmer, the external device 502 can also process and analyze data received from the IMD 501 and ECG leads 632, such as to make a preliminary diagnosis of the patient's medical condition or to the operation of the IMD 501. Such leads 632 can also be used to obtain an actual surface ECG, from which an ordered list of RR intervals before potential arrhythmia detection can be obtained. Additionally, the external device 502 can accept various user inputs accepted according to the embodiments of the present invention described above.

[0102] Now, by referring to Figure 6 Considering the components of the external device 502, the operation of the external device 502 may be controlled by a CPU 602, which may be a generally programmable microprocessor or microcontroller, or may be a dedicated processing device, such as an Application Specific Circuit (ASIC). The software instructions executed by the CPU may be accessible from a read-only memory (ROM) 606 and a random access memory (RAM) 630 via an internal bus 604. Additional software may be accessed from a hard drive 608, a floppy disk drive 610, and a CD ROM drive 612, or other suitable permanent mass storage devices. Depending on the specific implementation, a Basic Input Output System (BIOS) is retrieved from the ROM by the CPU at power-up. Based on the instructions provided in the BIOS, the CPU "boots" the entire system according to well-established computer processing techniques.

[0103] Once operated, the CPU displays a menu of programming options to the user via the LCD display 614 or another suitable computer display device. To this end, for example, the CPU may display a menu of specific programming parameters of the IMD 501 to be programmed, or may display a menu of diagnostic data types to be retrieved and displayed. In response, the physician enters various commands via a touch screen 616 overlaid on the LCD display 614 or through a standard keyboard 618 supplemented by additional custom keys 620 (such as an Emergency VVI (EVVI) key). The EVVI key sets the IMD 501 to a safe VVI mode with high pacing output. This ensures life-sustaining pacing operation in almost all scenarios, but it is by no means desirable to have the cardiac stimulation device 100 in EVVI mode at all times.

[0104] Typically, the physician initially controls the external device 502 to retrieve data stored within the implanted medical device and also retrieve data from the ECG leads (see above) coupled to the patient's myocardium. Figure 1The telemetry subsystem 622 may include a separate CPU 624 for coordinating the operation of the telemetry subsystem 622. The main CPU 602 of the external device 502 communicates with the telemetry subsystem CPU 624 via an internal bus 604. The telemetry subsystem 622 additionally includes a telemetry circuit 626 connected to a telemetry stick 628, which cooperates with the telemetry circuit 626 to receive and send electromagnetic signals from the telemetry circuit 564 of the IMD 501. The telemetry stick 628 is placed above the patient's chest near the IMD 501 to allow reliable transmission of data between the telemetry stick and the IMD 501 via the telemetry link 503. Typically, at the beginning of a programming session, the external programming device controls the IMD 501 via appropriate signals generated by the telemetry wand 628 to output all previously recorded patient diagnostic information and device diagnostic information. The patient diagnostic information includes, for example, measured physiological variable data, recorded IEGM data, and statistical patient data, such as the percentage of paced and sensed heartbeats. The device diagnostic data includes, for example, information representing the operation of the IMD 501, such as lead impedance, battery voltage, battery recommended replacement time (Recommended Replacement Time, RRT) information, etc. The data retrieved from the IMD 501 is stored by the external device 502 in a random access memory (Random Access Memory, RAM) 630, a hard disk drive 608, a floppy disk placed in a floppy disk drive 610, etc. Additionally, or alternatively, if the entire system is configured with a drive for recording data to a digital media disk, such as a Write Once Read Many (WORM) drive, the data can be permanently or semi-permanently stored in a compact disk (CD) or other digital media disk.

[0105] Patient and device diagnostic data stored within IMD 501 may be communicated to external device 502. In addition, IMD 501 may be instructed to execute the electrode algorithm of the present invention, details of which are provided above.

[0106] The external devices 502 may also include a network interface card ("NIC") 660 to allow transmission of data to and from other computer systems via a router 662 and a wide area network ("WAN") 664. Alternatively, the external devices 502 may include a modem for communicating via a Public Switched Telephone Network (PSTN). Depending on the implementation, the modem may be directly connected to the internal bus 604 and may be connected to the internal bus via a parallel port 640 or a serial port 642. Data transmitted from other computer systems may include, for example, data regarding medications prescribed, administered, or sold to patients.

[0107] The CPU 602 may include an arrhythmia identification module 650, which may control the arrhythmia identification module 650 described above. Figure 4A-4C The IMD 501 may perform the described steps or a subset thereof, and / or may instruct the IMD 501 to perform certain such steps.

[0108] The external device 502 receives data from the IMD 501, including parameters representing the current programming state of the IMD 501. The external device 502 may also receive an IEGM, a sample thereof, and / or an indication of its date from the IMD 501. Under the control of the physician, the external device 502 displays the current programming parameters and allows the physician to reprogram the parameters. To this end, the physician enters appropriate commands via any of the aforementioned input devices, and under the control of the CPU 602, the programming commands are converted into specific programming parameters for transmission to the IMD 501 via the telemetry stick 628, thereby reprogramming the IMD 501. Before reprogramming the specific parameters, the physician may control the external programmer to display any or all of the data retrieved from the IMD 501, including the display of the ECG, the display of electrodes as cathode and / or anode candidates, and statistical patient information. Any or all of the information displayed by the external device 502 may also be printed using the printer 636.

[0109] A speaker 644 is included to provide audible tones to the user, such as warning beeps in the event of inappropriate input provided by a physician. The telemetry subsystem 622 may additionally include an input / output circuit 646 that may control the transmission of analog output signals (such as ECG signals output to an ECG machine or chart recorder). Other peripheral devices may also be connected to the external device 502 via the parallel port 640 or the serial port 642. Although one of each is shown, multiple input-output (IO) ports may be provided.

[0110] Using external device 502 configured as shown, a physician or other authorized user can retrieve, process, and display a wide range of information received from IMD 501, and reprogram IMD 501, including configuration of CRT pacing parameters, if desired. Figure 6 The description provided is intended only to provide an overview of the operation of the example external device 502 and is not intended to describe in detail every feature of the device's hardware and software and is not intended to provide an exhaustive list of functions performed by the device.

[0111] It should be understood that the subject matter described herein is not limited in its application to the construction details and component arrangements set forth in the description herein or shown in its accompanying drawings. The subject matter described herein is capable of other embodiments and can be practiced or executed in various ways. In addition, it should be understood that the words and terms used herein are for descriptive purposes and should not be considered as limiting. The use of "including", "comprising" or "having" and its variations herein means to include the items listed thereafter and their equivalents as well as additional items. In addition, it should be noted that the term "based on" as used herein, unless otherwise specified, should be interpreted to mean at least partially based on, meaning that there may be one or more additional factors, about which decisions are made, etc. For example, if a decision is based on the result of a comparison, the decision may be based on one or more other factors in addition to the result of the comparison.

[0112] Embodiments of the present technology have been described above with the aid of functional building blocks that illustrate the performance of specified functions and their relationships. For ease of description, the boundaries of these functional building blocks have often been defined herein. Alternative boundaries may be defined so long as the specified functions and their relationships are appropriately performed. Therefore, any such alternative boundaries are within the scope and spirit of the claimed invention. For example, combination or separation Figure 4A , Figure 4B and Figure 4C Some of the steps shown in will be possible. For another example, change Figure 5 and Figure 6 Boundaries of some of the blocks shown in are possible.

[0113] It should be understood that the above description is intended to be illustrative, not restrictive. For example, the embodiments described above (and / or their various aspects) can be used in combination with each other. In addition, without departing from the scope of the present technology, many modifications can be made to adapt specific situations or materials to the teachings of the embodiments of the present technology. Although the dimensions, material types and coatings described herein are intended to define the parameters of the embodiments of the present technology, they are by no means restrictive, but exemplary embodiments. After reviewing the above description, many other embodiments will be obvious to those skilled in the art. Therefore, the scope of the embodiments of the present technology should be determined with reference to the attached claims and the full range of equivalents enjoyed by such claims. In the attached claims, the terms "including" and "wherein" are used as the concise Chinese equivalents of the corresponding terms "including" and "wherein". In addition, in the claims, the terms "first", "second" and "third" etc. are used only as labels and are not intended to impose numerical requirements on their objects. Furthermore, the claim limitations are not written in a “means-plus-function” format and are not intended to be interpreted under 35 U.S.C. §112(f) unless and until such claim limitations expressly use the phrase “means for” preceded by a statement of function without further structure.

Claims

1. An implantable medical device (IMD), comprising: Multiple electrodes; a sensing circuit coupled to the one or more electrodes and configured to obtain a signal indicative of electrical activity of the patient's heart; a therapy circuit coupled to the one or more electrodes and configured to deliver therapy to the patient's heart; at least one of a processor or a controller; and a battery configured to provide power to the sensing circuit, the treatment circuit, and at least one of the processor or the controller; At least one of the processor or the controller is configured to: Detect potential arrhythmia episodes; obtaining an ordered list of RR intervals within a window leading to detection of a potential arrhythmia episode, wherein each RR interval has a respective duration; determining pairwise differences between pairs of RR intervals separated by various intervals based on the ordered list of RR intervals within the window, thereby generating a plurality of sets of pairwise differences; determining a measure of the predominantly repeated RR interval pattern based on the plurality of pairwise differences; comparing a measure of the RR interval pattern of the primary repeat to a pattern threshold; and In response to the measurement of the predominantly repeating RR interval pattern being below the pattern threshold, the detection of the potential arrhythmic episode is classified as a false positive.

2. The IMD of claim 1 , wherein at least one of the processor or the controller is configured to: In response to detection of the potential arrhythmic episode being classified as a false positive, the therapy circuit is prevented from delivering therapy to the patient's heart.

3. The IMD of claim 1 , wherein at least one of the processor or the controller is configured to: For each set of pairwise differences in the plurality of sets of pairwise differences, determining a median or an indicator of the median for the set, thereby producing a plurality of medians or indicators of the median; and A measure of the predominantly repeating RR spacing pattern within the window is determined by determining a minimum of the plurality of medians or indicators of medians.

4. The IMD according to claim 1, wherein: The potential arrhythmia episodes include potential atrial fibrillation (AF) episodes.

5. The IMD of claim 1 , wherein: The IMD includes an implantable cardioverter defibrillator ICD; The therapeutic circuit includes an electric shock circuit; and At least one of the processors or controllers is configured to, in response to a measurement of the predominantly repeating RR interval pattern being below the pattern threshold, classify the detection of the potential arrhythmic episode as a false positive and, in response, prevent the shock circuit from delivering a shock to the patient's heart.

6. The IMD according to claim 1, wherein: The IMD includes a pacemaker.

7. The IMD of claim 1 , wherein at least one of the processor or the controller is configured to: for each of k=1 to M, generating the plurality of sets of pairwise differences by determining pairwise differences between pairs of RR intervals that are k intervals apart within the window, where M is an integer of at least 4, thereby generating M indicators of median RR interval differences; and A measure of the predominantly repeating RR interval pattern within the window is determined by determining the minimum of the M indicators of median RR interval differences.

8. The IMD of claim 7 , wherein the determined M indicators of median RR interval difference comprise M median percentage difference values, each median percentage difference value being determined using the following equation: Median percentage difference of k = MEDIAN(100*|RR(n)–RR(n+k)| / RR(n)), For interval n=1:N and interval spacing k=1:M, in RR(n) is the nth RR interval in the window, RR(n+k) is the (n+k)th RR interval in the window, |RR(n)–RR(n+k)| is the absolute value of RR(n)–RR(n+k), and N is the total number of RR intervals within the window.

9. A method for use by a device, the method comprising: obtaining an ordered list of RR intervals within a window leading to detection of a potential arrhythmia episode, wherein each RR interval has a respective duration; determining pairwise differences between pairs of RR intervals separated by various intervals based on the ordered list of RR intervals within the window, thereby generating multiple sets of pairwise differences; determining a measure of the predominantly repeated RR interval pattern based on the plurality of pairwise differences; comparing a measure of the RR interval pattern of the primary repeat to a pattern threshold; and In response to the measurement of the predominantly repeating RR interval pattern being below a pattern threshold, the detection of the potential arrhythmic episode is classified as a false positive, and in response, delivery of therapy to the patient's heart is prevented.

10. The method according to claim 9, further comprising: In response to classifying the detection of the potential arrhythmic episode as a false positive, delivery of therapy to the patient's heart is prevented.

11. The method of claim 9, wherein the measurements to determine the primarily repeated RR interval pattern comprise: For each set of pairwise differences in the plurality of sets of pairwise differences, determining a median or an indicator of a median for the set, thereby producing a plurality of medians or indicators of medians; and A measure of the predominantly repeating RR spacing pattern within the window is determined by determining a minimum of the plurality of medians or indicators of medians.

12. The method of claim 9, wherein the potential arrhythmia episode comprises a potential atrial fibrillation (AF) episode.

13. The method according to claim 9, wherein: Devices using the method include implantable cardioverter defibrillators (ICDs); and The treatment includes an electrical shock to the patient's heart.

14. The method of claim 9, wherein: Devices using the method include pacemakers.

15. The method of claim 10, wherein the determining pairwise differences based on the ordered list of RR intervals within the window comprises: For each of k=1 to M, pairwise differences between pairs of RR intervals that are k intervals apart within the window are determined, thereby producing M indicators of median RR interval differences, where M is an integer of at least four.

16. The method of claim 15, wherein the determining of the measurements of the predominantly repeated RR interval pattern based on the multiple sets of pairwise differences comprises: A measure of the predominantly repeating RR interval pattern within the window is determined by determining a minimum of the M indicators of the median RR interval differences.

17. The method according to claim 16, wherein: The M indices of median RR interval differences determined include M median percentage differences, each of which is determined using the following equation: Median percentage difference of k = MEDIAN(100*|RR(n)–RR(n+k)| / RR(n)), For interval n=1:N and interval spacing k=1:M, in RR(n) is the nth RR interval in the window, RR(n+k) is the (n+k)th RR interval in the window, |RR(n)–RR(n+k)| is the absolute value of RR(n)–RR(n+k), and N is the total number of RR intervals within the window.

Images