Detection of cardiac signal qt interval

Abstract

An example device for detecting one or more parameters of a cardiac signal is disclosed herein. The device includes one or more electrodes and a sensing circuit configured to sense a cardiac signal via the one or more electrodes. The apparatus further includes processing circuitry configured to determine an R-wave of the cardiac signal and determine whether the R-wave is noisy. Based on the R-wave being noisy, the processing circuitry is configured to determine whether the cardiac signal around a determined T-wave is noisy. Based on the cardiac signal around the determined T-wave not being noisy, the processing circuitry is configured to determine a QT interval or a corrected QT interval based on the determined T-wave and the determined R-wave.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to cardiac monitoring, and more specifically to detection of a QT interval or a corrected QT interval (QTc) in a cardiac signal. BACKGROUND

[0002] Cardiac signal analysis can be performed by various devices, such as implantable medical devices (IMDs), insertable cardiac monitors (ICMs), and external devices (e.g., smartwatches, fitness monitors, mobile devices, Holter monitors, wearable defibrillators, etc.). For example, a device can be configured to process a cardiac signal (e.g., a cardiac electrogram (EGM) and electrocardiogram (ECG)) sensed by one or more electrodes. Features of a cardiac signal can include P-waves, Q-waves, R-waves, S-waves, QRS complexes, and T-waves. A QT interval is the time from the beginning of a QRS complex to the end of a T-wave. A QTc interval is a QT interval that has been normalized or corrected with respect to heart rate using a formula. Accurate detection and delineation of features in a cardiac signal, such as a QT interval or a QTc interval, can be important for monitoring a patient’s health condition, such as risk of sudden cardiac death. SUMMARY

[0003] Generally, the present disclosure is directed to devices and techniques for identifying one or more features of a patient’s cardiac signal (e.g., an EGM and / or an ECG) and / or determining one or more parameters of the cardiac signal. For example, the present disclosure describes techniques for identifying a QT interval or a QTc interval, which can enable prediction of whether a patient is experiencing or will experience a tachyarrhythmia or other abnormal heart rhythm that can lead to sudden cardiac death. In some examples, an IMD can deliver therapy to a patient to terminate or prevent a predicted tachyarrhythmia.

[0004] In one example, a device includes one or more electrodes; sensing circuitry configured to sense a cardiac signal via the one or more electrodes; and processing circuitry configured to: determine an R-wave of the cardiac signal; determine whether the R-wave is noisy; based on the R-wave not being noisy, determine whether the cardiac signal around a determined T-wave is noisy; and based on the cardiac signal around the determined T-wave not being noisy, determine a QT interval or a corrected QT interval based on the determined T-wave and the determined R-wave.

[0005] In another example, a method includes: determining, by processing circuitry, an R-wave of a cardiac signal; determining, by the processing circuitry, whether the R-wave is noisy; determining, by the processing circuitry, based on the R-wave not being noisy, whether the cardiac signal around a determined T-wave is noisy; and determining, by the processing circuitry, based on the cardiac signal around the determined T-wave not being noisy, a QT interval or a corrected QT interval based on the determined T-wave and the determined R-wave.

[0006] In another example, a non-transitory computer-readable storage medium storing a set of instructions, which when executed, cause a system to: determine an R-wave of a cardiac signal; determine whether the R-wave is noisy; based on the R-wave not being noisy, determine whether the cardiac signal around a determined T-wave is noisy; and based on the cardiac signal around the determined T-wave not being noisy, determine a QT interval or a corrected QT interval based on the determined T-wave and the determined R-wave.

[0007] This summary is intended to provide an overview of the subject matter described in this disclosure. It is not intended to provide an exclusive or exhaustive explanation of the systems, devices, and methods described in detail within the accompanying drawings and description below. Further details of one or more examples of this disclosure are set forth in the accompanying drawings and description below. Other features, objects, and advantages will be apparent from the description and drawings, and from the claims. BRIEF DESCRIPTION OF DRAWINGS

[0008] Figure 1 An environment in which an example medical system is shown in connection with a patient.

[0009] Figure 2 is a functional block diagram of an example configuration of an implantable cardiac monitor (ICM) of the medical system of Figure 1

[0010] Figure 3 is a conceptual side view of an example configuration of the ICM of Figure 1 and 2

[0011] Figure 4 is a functional block diagram of an example configuration of an external device of Figure 1

[0012] Figure 5A and 5B are conceptual diagrams illustrating example primary and secondary sensing channels for an R-wave and T-wave detector in accordance with the techniques of this disclosure.

[0013] Figure 6 is a conceptual diagram illustrating an example segment of an EGM signal and a corresponding rectified waveform.

[0014] Figure 7 is a conceptual diagram illustrating an example segment of an EGM signal in accordance with the techniques of this disclosure, which depicts different parameters that can be computed by a QT detection algorithm.

[0015] Figure 8 is a conceptual diagram illustrating an example segment of an EGM signal in accordance with the techniques of this disclosure, which depicts the computation of window_start and window_end parameters.

[0016] Figure 9 ​​​Figure 1 -D is a conceptual diagram illustrating an example segment of an EGM signal showing the calculation of the window start and window end parameters in the case where the difference between the current RR interval and the previous RR interval is greater than 500 milliseconds (ms) in accordance with the techniques of this disclosure.

[0017] Figure 10 Figure 1 -A is a conceptual diagram illustrating an example MATLAB GUI for assisting in manually annotating a dataset.

[0018] Figure 11 Figure 1 -B is a conceptual diagram illustrating an example of manual annotation.

[0019] Figure 12 Figure 1 -D is a conceptual diagram illustrating a histogram of the difference between the manually annotated QT interval and the QT interval detection results based on the techniques of this disclosure for each beat in a development dataset.

[0020] Figure 13 Figure 1 -E is a conceptual diagram illustrating a histogram of the difference between the manually annotated QT corrected (QTc) interval and the QTc interval detection results based on the techniques of this disclosure for each beat in a development dataset.

[0021] Figure 14 Figure 1 -F is a conceptual diagram illustrating a histogram of the mean difference between the manually annotated QTc interval and the QTc interval detection results based on the techniques of this disclosure for 46 unique devices in a development dataset.

[0022] Figure 15A Figure 1 -D is a conceptual diagram illustrating an example EGM strip from a device where the mean of (QTc interval (true) - QTc interval (algorithm detected)) is greater than 25 ms in accordance with the techniques of this disclosure, depicting both manual annotation and detection.

[0023] Figure 16 Figure 1 -E is a conceptual diagram illustrating an example EGM strip from a development dataset with T waves that have different morphologies and orientations at different RR intervals in accordance with the techniques of this disclosure, depicting both manual annotation and detection.

[0024] Figure 17A Figure 1 -B is a conceptual diagram depicting an example EGM strip from a development dataset in accordance with the techniques of this disclosure, depicting beat-to-beat changes in QTc interval detected by both manual annotation and detection.

[0025] Figure 18 Figure 2 is a flow diagram illustrating an example noise determination technique of this disclosure.

[0026] Figure 19A Figure 2 -B is a flow diagram illustrating an example technique of this disclosure.

[0027] Figure 20 is a flowchart illustrating example techniques of the present disclosure.

[0028] Figure 21 is a flowchart illustrating example techniques of the present disclosure. DETAILED DESCRIPTION

[0029] The present disclosure describes techniques for identifying one or more parameters of a cardiac signal, such as a QT interval. The parameters can be used, for example, to detect or predict cardiac arrhythmias, to assess a patient for other conditions such as changes in electrolytes, changes in diabetic status, fluid overload, or dehydration, or to configure and / or assess a therapy such as a pharmacological therapy.

[0030] The T-wave represents ventricular repolarization. Repolarization of the ventricles begins at the epicardial surface of the ventricles and proceeds inward through the ventricular wall to the endocardial surface. The T-wave occurs at the last stage of ventricular contraction. The onset of the T-wave is the first or abrupt or gradual deviation from the S-T segment. The point at which the T-wave returns to the baseline marks the end of the T-wave.

[0031] The QT interval on an electrocardiogram (ECG) is measured from the onset of the QRS complex to the end of the T-wave. The QT interval represents the time it takes for the ventricles to depolarize and repolarize, or contract and relax. This electrical activity of the heart is mediated by channels, which are complex molecular structures within the membrane of the myocardial cell that regulate the flow of ions in and out of the myocardial cell. See, e.g., Viskin S. Long QT Syndrome and Torsades de Pointes, The Lancet, vol. 62(13), pp. 1625-1633, 1999 (hereinafter Viskin). Rapid inflow of positively charged ions (sodium and calcium) causes normal myocardial depolarization. When the outflow of potassium ions exceeds this inflow, myocardial repolarization occurs. Malfunction of ion channels results in an excess of positively charged ions within the cell through insufficient outflow of potassium ions or excess inflow of sodium ions. This excess of positively charged ions within the cell prolongs ventricular repolarization and results in a prolonged QT interval. See, e.g., Al-Khatib SM, et al., "The QT Interval: What the Clinician Needs to Know", JAMA, vol. 289(16), pp. 2120-2127, 2003 (hereinafter Al-Khatib).

[0032] Abnormally long or abnormally short QT intervals are associated with an increased risk of developing abnormal heart rhythms and sudden cardiac death. QT interval abnormalities can be caused by genetic conditions (e.g., long QT syndrome), certain medications (e.g., sotalol or piroxicam), disturbances in the concentration of certain salts in the blood (e.g., hypokalemia and hypomagnesemia), or hormonal imbalances (e.g., hypothyroidism) or changes in blood glucose. Normal QT intervals vary with age and gender, and are typically about 0.36 to 0.44 seconds. Any value greater than or equal to 0.50 seconds can be considered dangerous for any age or gender. See, e.g., Cox, Natalie K., "QT Interval: How Long Is Too Long?", Nursing Made Incredibly Easy!, Vol. 9(2), pp. 17-21, 2011. The QT interval of an ECG has gained clinical importance primarily because prolongation of this interval can predispose a person to potentially fatal ventricular arrhythmias, known as torsades de pointes, which can lead to sudden cardiac death as discussed in Viskin.

[0033] In a clinical setting, it is now widely recognized that typical measurements of the QT interval have significant variability that can affect interpretation. This variability in QT interval measurement is caused by biological factors, such as diurnal effects, autonomic tone, electrolyte and drug differences; technical factors, including environment, recording processing and ECG recording acquisition; and observer within and between observer variability caused by T wave morphology changes, noise baseline and presence of U waves. Observer between variability also stems from lack of agreement among experts on standardized methods for measuring the QT interval. See, e.g., Al-Khatib; Morganroth J, et al., "Variability of QT Measurements in Healthy Men, Implications for Selecting Abnormal QT Values Predictive of Drug Toxicity and Arrhythmogenicity", Am. J. Cardiol., Vol. 67(8), pp. 774-776; 1991; and Molnar J, et al., "Diurnal Pattern of QTc Interval: How Prolonged Is Too Prolonged? Possible Relationship to Circadian Triggering of Cardiovascular Events", Am. J. Cardiol., Vol. 27(1), pp. 76-83, 1996.

[0034] The QT interval of an ECG has gained clinical importance primarily because prolongation of this interval can predispose a person to potentially fatal ventricular arrhythmias, known as torsades de pointes, which can lead to sudden cardiac death. A variety of factors have been shown to be associated with causing QT prolongation and torsades de pointes. Among these, an important risk factor for long QT syndrome is the use of QT prolonging medications. As discussed in Al-Khatib, a QT interval greater than 500 ms has been shown to be associated with a higher risk of torsades de pointes.

[0035] Morganroth J et al., "Use of electrocardiogram assessments and management of cardiac safety in oncology clinical trials: focus on cardiac repolarization (QTc interval)", Clinical Pharmacology and Therapeutics, vol. 87(2), pp. 166-74, 2010, suggest that 10-20 ms QT corrected (QTc) changes are considered clinically relevant and patients in this range, especially those with QT-related risk factors, should be protected during treatment by careful ECG assessments. Based on a thorough risk-benefit assessment, the authors suggest that oncology drugs can accept higher tolerable limits of QTc prolongation effects as they meet the patients' special medical needs.

[0036] Chouchoulis k et al., "Effect of QT interval prolongation after antiarrhythmic drug therapy on left ventricular function", Future Cardiology, vol. 13(1), 2016, assessed whether antiarrhythmic drug-induced QT interval prolongation affects left ventricular function. The study population included 54 patients with recent symptomatic atrial fibrillation who spontaneously cardioverted to sinus rhythm. After sotalol (from 424 ± 40 ms to 460 ± 57 ms and from 446 ± 35 ms to 474 ± 48 ms, respectively, both p < 0.01) and amiodarone (from 437 ± 41 ms to 504 ± 39 ms and from 469 ± 35 ms to 527 ± 50 ms, respectively, both p < 0.01) were taken, QTmax and QTc interval were found to be significantly prolonged. Thus, it was noted in this study that QTc was significantly prolonged after taking antiarrhythmic drugs such as sotalol and amiodarone.

[0037] Several studies also showed a correlation between QT changes and diabetes. Type 1 diabetes can increase the risk of death compared to the general population, mainly due to an increased risk of cardiovascular disease. Almost half of type 1 diabetic patients have prolonged QTc intervals (> 440 ms). According to Rossing P et al., "Prolonged QTc interval predicts mortality in type 1 diabetic patients", Diabetes Care, vol. 18(3), pp. 199-205, 2001, prolonged QTc interval in diabetes was associated with a 29% mortality rate within 10 years, while the mortality rate was 19% for normal QTc interval.

[0038] One study has shown that QTc dispersion is an important predictor of cardiac mortality. In the Rotterdam study (de Bruyne MC, et al., "QTc Dispersion Predicts Heart- Rate Mortality in the Elderly," Circulation, Vol. 97(5), pp. 467-472, 1998), people in the highest tertile of QTc dispersion (>60 ms) had a 2-fold increased risk of cardiac mortality relative to those in the lowest tertile (<39 ms). The Rotterdam study also showed that QTc dispersion is greater in diabetic patients than in non-diabetic patients, according to Marfella et al., "QTc Dispersion, Hyperglycemia, and Hyperinsulinemia," Circulation, Vol. 100, 1999 (hereinafter "Marfella").

[0039] Marfella evaluated the effects of acute hyperglycemia on QTc duration and QTc dispersion in 27 normal subjects. Plasma glucose values were stabilized at 15 mmol / L during glucose clamp administration to the subjects, and plasma insulin showed a biphasic response pattern, with an early rise at 10 minutes (327 ± 89 pmol / L) followed by a gradual, sustained rise (456 ± 120 pmol / L). QTc increased from 413 ± 26 to 442 ± 29 ms (P < 0.05) and QTc dispersion increased from 32 ± 9 to 55 ± 12 ms (P < 0.01) at the end of clamp administration. This indicates that acute hyperglycemia in normal subjects resulted in significant increases in QTc and QTc dispersion.

[0040] Lee SP et al., "Effect of Autonomic Neuropathy on QTc Interval Prolongation During Hypoglycemia in Type 1 Diabetes," Diabetes, Vol. 53(6), pp. 1535-42, 2004, discusses a study of 28 adult patients with type 1 diabetes and 8 non-diabetic control subjects. QTc was then measured during controlled hypoglycemia (2.5 mmol / l) using a high insulin clamp. The mean (+ / - SE) QTc of the diabetic participants (BRS+ subjects) was prolonged from 377 + / - 9 ms (baseline) to a maximum of 439 + / - 13 ms during hypoglycemia, and the mean (+ / - SE) QTc of the control subjects was prolonged from 378 + / - 5 to 439 + / - 10 ms. This study indicates that hypoglycemia causes QTc prolongation on the electrocardiogram, which is a predictor of arrhythmia risk and sudden death.

[0041] Accordingly, continuous monitoring of the QT interval can identify a long QT interval, which can indicate a need for medical intervention. In some examples, continuous monitoring of the QT interval can be performed using an implantable cardiac monitor (ICM). The present disclosure describes example algorithms that can be used with an ICM, such as the LINQ® ICM produced by Medtronic, PLC, of Dublin, Ireland, to monitor the QT interval. TM ICM.

[0042] Various types of medical devices sense cardiac EGM. Some medical devices for sensing cardiac EGM are non-invasive, such as those using multiple electrodes placed in contact with external parts of the patient, such as at various locations on the patient's skin. As an example, electrodes used for monitoring cardiac EGM in these non-invasive procedures can be attached to the patient using adhesive, tape, belt, or vest, and electrically coupled to a monitoring device, such as an electrocardiograph, Holter monitor, or other electronic device. The electrodes are configured to sense electrical signals associated with the electrical activity of the patient's heart or other cardiac tissue, and these sensed electrical signals are provided to electronic devices for further processing and / or display of the electrical signals. Non-invasive devices and methods can be used on a temporary basis, such as to monitor the patient during a clinical visit, such as during a physician appointment, or, for example, within a predetermined time period, such as a day (24 hours), or a period of several days.

[0043] External devices that can be used for noninvasive sensing and monitoring of cardiac EGM include wearable devices, such as patches, watches, or necklaces, with electrodes configured to contact a patient's skin. An example of a wearable physiological monitor configured to sense cardiac EGM is the SEEQ, commercially available from Medtronic plc in Dublin, Ireland. TM Mobile cardiac telemetry systems. These external devices facilitate relatively long-term monitoring of patients during normal daily activities and can periodically transmit collected data to network services, such as Medtronic's Carelink. TM network.

[0044] Some implantable medical devices (IMDs) also sense and monitor cardiac EGM. Electrodes used by the IMD to sense cardiac EGM are typically integrated into the IMD housing and / or coupled to the IMD via one or more thin leads. Examples of IMDs for monitoring cardiac EGM include pacemakers and implantable cardioverter defibrillators that can be coupled to intravascular or extravascular leads, and pacemakers with a housing configured for implantation in the heart, which may be leadless. An example of a pacemaker configured for intracardiac implantation is the Micra from Medtronic. TM Transcatheter pacing systems. Some implantable device-controlled devices (IMDs) do not provide treatment, such as implantable patient monitors that sense cardiac EGM. One example of this IMD is the subcutaneously insertable LINQ. TM ICM. This type of IMD facilitates relatively long-term monitoring of patients during normal daily activities and allows for the periodic transmission of collected data to network services, such as Medtronic's Carelink. TM network.

[0045] Although this disclosure discusses techniques for measuring QT intervals via example ICM, any medical device configured to sense cardiac EGM via implanted or external electrodes (including the examples mentioned herein) can implement the techniques of this disclosure for measuring QT intervals. These techniques involve assessing cardiac EGM using criteria configured to provide the desired sensitivity and specificity for QT interval detection, despite noise and depolarization morphological variations due to different electrode locations. The techniques of this disclosure for identifying QT intervals can facilitate the determination of cardiac health and risk of sudden cardiac death, and can enable clinical interventions to suppress the risk of sudden cardiac death.

[0046] Figure 1 An example medical system 2 for patient 4 is illustrated, combining one or more technologies according to this disclosure. The example technologies can be used with an ICM 10, which can be integrated with an external device 12 and... Figure 1 At least one of the other devices not depicted communicates wirelessly. In some instances, the ICM 10 is implanted outside the chest cavity of patient 4 (e.g., subcutaneously). Figure 1 The chest position is shown in the diagram. The ICM 10 can be positioned near the sternum, close to or just below the patient's heart level, for example, at least partially within the heart's outline. The ICM 10 contains multiple electrodes ( Figure 1 (Not shown in the image), and is configured to sense cardiac EGM via the plurality of electrodes. In some instances, the ICM 10 employs LINQ. TM In the form of ICM, or similar to, for example, LINQ TM Another form of ICM, either a version or a modified version.

[0047] External device 12 may be a computing device having a user-viewable display and an interface (i.e., a user input mechanism) for providing input to external device 12. In some instances, external device 12 may be a laptop, tablet, workstation, one or more servers, cellular phone, personal digital assistant, or another computing device that can run an application that enables the computing device to interact with ICM 10. External device 12 is configured to communicate wirelessly with ICM 10 and optionally another computing device. Figure 1 (Not shown in the image). For example, the external device 12 can communicate via near-field communication technology (e.g., inductive coupling, NFC, or other communication technology that can operate within a range of less than 10-20 cm) and far-field communication technology (e.g., according to 802.11 or...). RF telemetry of the specification set, or other communication technologies that can operate within a range greater than near-field communication technologies, can be used for communication.

[0048] External device 12 can be used to configure the operating parameters of ICM 10. External device 12 can be used to retrieve data from ICM 10, such as QT intervals. The retrieved data may include values ​​of physiological parameters measured by ICM 10, indications of arrhythmias or other disease episodes detected by ICM 10, and physiological signals recorded by ICM 10. For example, external device 12 can retrieve information related to ICM 10's detection of QT intervals, such as the mean, median, minimum, maximum, range, or mode of QT intervals over a period of time. The time period can be predetermined, such as hourly, daily, or weekly, or it can be additionally based on the timing of the last information retrieved by external device 12, or it can be determined by the user of external device 12, for example, by entering a command on external device 12 to request information from ICM 10. External device 12 can also retrieve electrocardiogram (EGM) segments recorded by ICM 10, for example, because ICM 10 determines that an arrhythmia or other disease occurred during the segment, or in response to a request to record a segment from patient 4 or another user.

[0049] For example, the processing circuitry of medical system 2, including ICM 10, external device 12, and / or one or more other computing devices, can be configured to perform the exemplary techniques of this disclosure for determining the QT interval. In some instances, the processing circuitry of medical system 2 can analyze the cardiac EGM sensed by ICM 10 to determine the QT interval within the cardiac EGM. Although described in the context of instances where the ICM 10 sensing the cardiac EGM includes an insertable cardiac monitor, exemplary systems comprising one or more implantable or external devices of any type configured to sense the cardiac EGM can be configured to implement the techniques of this disclosure.

[0050] Figure 2 This illustrates one or more technologies according to the description herein. Figure 1 A functional block diagram of an example configuration of ICM 10 is provided. In the illustrated example, ICM 10 includes electrodes 16A and 16B (collectively referred to as "electrodes 16"), antenna 26, processing circuitry 50, sensing circuitry 52, communication circuitry 54, storage device 56, switching circuitry 58, and sensor 62. Although the illustrated example includes two electrodes 16, in some examples, an IMD including or connected to two or more electrodes 16 may implement the techniques of this disclosure.

[0051] Processing circuitry 50 can include fixed function circuitry and / or programmable processing circuitry. Processing circuitry 50 can include any one or more of a microprocessor, a controller, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or equivalent discrete or analog logic circuitry. In some examples, processing circuitry 50 can include multiple components, such as any combination of one or more microprocessors, one or more controllers, one or more DSPs, one or more ASICs, or one or more FPGAs, as well as other discrete or integrated logic circuitry. The functions attributed to processing circuitry 50 herein can be embodied as software, firmware, hardware or any combination thereof.

[0052] Sensing circuitry 52 can be selectively coupled to electrodes 16 through switching circuitry 58, e.g., to select electrodes 16 and polarities for sensing a cardiac EGM, referred to as a sensing vector, as controlled by processing circuitry 50. Sensing circuitry 52 can sense signals from electrodes 16, e.g., to produce a cardiac EGM, to facilitate monitoring of electrical activity of the heart. As an example, sensing circuitry 52 can also monitor signals from sensors 62, which can include one or more accelerometers, pressure sensors, and / or optical sensors. In some examples, sensing circuitry 52 can include one or more filters and amplifiers for filtering and amplifying signals received from electrodes 16 and / or sensors 62.

[0053] Sensing circuitry 52 and / or processing circuitry 50 can be configured to detect R-waves and T-waves. In some examples, sensing circuitry 52 can include one or more rectifiers, filters, amplifiers, comparators, and / or analog-to-digital converters. In some examples, sensing circuitry 52 can output an indication to processing circuitry 50 in response to sensing an R-wave or a T-wave. In some examples, processing circuitry 50 can determine an R-wave or a T-wave in the indication from sensing circuitry 52. Processing circuitry 50 can use the indication of a detected R-wave and T-wave to determine a QT interval or a corrected QT interval (QTc).

[0054] Sensing circuitry 52 can also provide one or more digitized cardiac EGM signals to processing circuitry 50 for analysis, e.g., for cardiac rhythm discrimination, and / or for analysis to determine a QT interval or a QTc interval in accordance with the techniques of this disclosure. In some examples, processing circuitry 50 can store digitized cardiac EGMs in storage 56. Processing circuitry 50 of ICM 10 and / or processing circuitry of another device that retrieves data from ICM 10 can analyze the cardiac EGMs to determine a QT interval or a QTc interval in accordance with the techniques of this disclosure.

[0055] The communication circuitry 54 may include any suitable hardware, firmware, software, or any combination thereof for communicating with another device such as external device 12, another networked computing device, or another IMD or sensor. Under the control of the processing circuitry 50, the communication circuitry 54 may receive downlink telemetry from external device 12 or another device and transmit uplink telemetry thereto via internal or external antennas, such as antenna 26. Additionally, the processing circuitry 50 may communicate via external devices (e.g., external device 12) and, for example, Medtronic. Computer networks and networked computing devices communicate with each other. Antenna 26 and communication circuit system 54 can be configured to communicate via inductive coupling, electromagnetic coupling, near field communication (NFC), radio frequency (RF) communication, etc. WiFi or other proprietary or non-proprietary wireless communication solutions are used to transmit and / or receive signals.

[0056] In some instances, storage device 56 contains computer-readable instructions that, when executed by processing circuitry system 50, cause ICM 10 and processing circuitry system 50 to perform various functions categorized herein. Storage device 56 may include any volatile, non-volatile, magnetic, optical, or electrical medium, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), electrically erasable programmable ROM (EEPROM), flash memory, or any other digital medium. As an example, storage device 56 may store programmed values ​​of one or more operating parameters of ICM 10 and / or data collected by ICM 10 for transmission to another device using communication circuitry system 54. As an example, data stored by storage device 56 and transmitted by communication circuitry system 54 to one or more other devices may include quantification of premature ventricular contractions (PVCs) detection and / or digitization of cardiac EGM.

[0057] Figure 3 It is shown Figure 1 and 2 A conceptual side view of an ICM 10 instance configuration. Figure 3 In the example shown, the ICM10 may comprise a leadless, subcutaneously implantable monitoring device having a housing 15 and an insulating cover 76. Electrodes 16A and 16B may be formed or placed on the outer surface of the insulating cover 76. (The above refers to...) Figure 2The described circuitry 50-62 can be formed or placed on an inner surface of the insulating cover 76 or within the housing 15. In the illustrated example, the antenna 26 is formed or placed on an inner surface of the insulating cover 76, but in some examples, can be formed or placed on an outer surface. In some examples, one or more of the sensors 62 can be formed or placed on an outer surface of the insulating cover 76. In some examples, the insulating cover 76 can be positioned over the open housing 15 such that the housing 15 and the insulating cover 76 enclose the antenna 26 and the circuitry 50-62 and protect the antenna and circuitry from fluids, such as bodily fluids.

[0058] One or more of the antenna 26 or the circuitry 50-62 can be formed on the inside of the insulating cover 76, for example, by using flip-chip technology. The insulating cover 76 can be flipped onto the housing 15. When flipped and placed onto the housing 15, the components of the ICM 10 formed on the inside of the insulating cover 76 can be located in a gap 78 defined by the housing 15. The electrodes 16 can be electrically connected to the switching circuitry 58 through one or more through-holes (not shown) formed through the insulating cover 76. The insulating cover 76 can be formed of sapphire (i.e., corundum), glass, parylene, and / or any other suitable insulating material. The housing 15 can be formed of titanium or any other suitable material (e.g., a biocompatible material). The electrodes 16 can be formed of any of stainless steel, titanium, platinum, iridium, or alloys thereof. In addition, the electrodes 16 can be coated with a material such as titanium nitride or fractal titanium nitride, although other suitable materials and coatings for such electrodes can be used.

[0059] Figure 4 is a block diagram illustrating an example configuration of components of the external device 12. In Figure 4 In examples, the external device 12 includes processing circuitry 80, communication circuitry 82, storage 84, and a user interface 86.

[0060] The processing circuitry 80 can include one or more processors configured to implement functionality and / or process instructions for execution within the external device 12. For example, the processing circuitry 80 can be capable of processing instructions stored in the storage 84. The processing circuitry 80 can include, among other things, a microprocessor, a DSP, an ASIC, an FPGA, or equivalent discrete or integrated logic circuitry, or a combination of any of the foregoing devices or circuitry. Accordingly, the processing circuitry 80 can include any suitable structure, whether in hardware, software, firmware, or any combination thereof, to

[0061] Communication circuitry 82 can include any suitable hardware, firmware, software, or any combination thereof for communicating with another device, such as ICM 10. Under the control of processing circuitry 80, communication circuitry 82 can receive downlink telemetry from and send uplink telemetry to ICM 10 or another device. Communication circuitry 82 can be configured to transmit or receive signals via inductive coupling, electromagnetic coupling, near-field communication (NFC), radio frequency (RF) communication, Bluetooth® communication, WiFi, or other proprietary or non-proprietary wireless communication schemes. Communication circuitry 82 can also be configured to communicate with devices other than ICM 10 via any of a variety of forms of wired and / or wireless communication and / or network protocols.

[0062] Storage device 84 can be configured to store information within external device 12 during operation. Storage device 84 can include a computer-readable storage medium or computer-readable storage device. In some examples, storage device 84 includes one or more of a short-term memory or a long-term memory. Storage device 84 can include, for example, RAM, DRAM, SRAM, magnetic discs, optical discs, flash memory, or various forms of EPROM or EEPROM. In some examples, storage device 84 is used to store data indicative of instructions for execution by processing circuitry 80. Storage device 84 can be used by software or applications running on external device 12 to temporarily store information during program execution.

[0063] Data exchanged between external device 12 and ICM 10 can include operational parameters. External device 12 can transmit data including computer-readable instructions that, when implemented by ICM 10, can control ICM 10 to change one or more operational parameters and / or derive collected data, such as a QT interval or a QTc interval. For example, processing circuitry 80 can transmit instructions to ICM 10 requesting that ICM 10 derive collected data (e.g., QT interval data, QTc interval data, and / or digitized cardiac EGMS) to external device 12. In turn, external device 12 can receive the collected data from ICM 10 and store the collected data in storage device 84. Processing circuitry 80 can implement any of the techniques described herein to analyze a cardiac EGM received from ICM 10, for example, to determine a QT interval or a QTc interval.

[0064] A user such as a clinician or patient 4 can interact with external device 12 through user interface 86. User interface 86 includes a display (not shown), such as a liquid crystal display (LCD) or light emitting diode (LED) display or other type of screen, through which processing circuitry 80 can present information relating to ICM 10, such as an indication of a cardiac EGM, a QT interval, or a QTc interval. In addition, user interface 86 can include input mechanisms for receiving input from a user. Input mechanisms can include, for example, buttons, a keypad (e.g., an alphanumeric keypad), a peripheral pointing device, or a touch screen, or another input mechanism that permits a user to navigate through a user interface presented by processing circuitry 80 of external device 12 and provide input. In other examples, user interface 86 also includes audio circuitry for providing audible notifications, instructions, or other sounds to the user, receiving voice commands from the user, or both.

[0065] Figure 5A and 5B is a conceptual diagram illustrating example primary and secondary sensing channels for R-wave detection and T-wave detection. In Figure 5A , sensing circuitry 52 of ICM 10 can sense R-waves by using a dual channel sensing technique of Figure 5A and 5B A cardiac signal (e.g., a signal from electrode 16A and electrode 16B) can be filtered through bandpass filter 100. In some examples, bandpass filter 100 can have a passband in a range of about 10 Hz to 32 Hz. In some examples, bandpass filter 100 can have a non-linear response as shown. In other examples, bandpass filter 100 can have a generally linear response. The bandpass signal can then be rectified by rectifier 102.

[0066] The rectified signal can then be input to an automatic adjustment threshold process 104. For example, when the amplitude of the rectified signal from rectifier 102 exceeds an automatic adjustment threshold, the automatic adjustment threshold process can sense that a certain event has occurred in the cardiac signal. Automatic adjustment threshold process 104 can use an automatic adjustment sensitivity with a short blanking period (e.g., about 150 ms). During the blanking period, a sensing process such as automatic adjustment threshold process 104 or fixed threshold process 106 can not sense events in the cardiac signal to avoid a single depolarization producing multiple sensed events. Automatic adjustment threshold process 104 can form a primary sensing channel 108, which can be the main R-wave sensing mechanism in ICM 10 and can be configured to accommodate detection of both tachyarrhythmias and bradyarrhythmias.

[0067] Once the primary sensing channel 108 detects an R-wave, the threshold setting of the auto-adjusting threshold process 104 is set at 65% of the amplitude of the detected R-wave (which can be a relatively high threshold, thus not immediately detecting R-waves). The threshold is then decayed from the 65% value to 35 microvolts, such that the next R-wave can be detected. In some examples, there can be a point at which the threshold is sharply decreased, such as after an expected T-wave and P-wave, to avoid over-sensing T-waves and / or P-waves.

[0068] In some examples, the rectified signal can be input into a fixed threshold process 106. The fixed threshold process 106 can have a fixed threshold and a relatively longer blanking period (e.g., about 520 ms) compared to the auto-adjusting threshold process 104 to reduce undersensing. Similar to the auto-adjusting threshold process 104, the fixed threshold process 106 can sense an event in the cardiac signal when the amplitude of the rectified signal exceeds the fixed threshold. The output of the fixed threshold process 106 can form a secondary sensing channel 110. In other examples (not shown), the secondary sensing channel 110 can use different filtering and / or different rectification than the primary sensing channel 108.

[0069] Figure 5A And 5B The example dual-channel sensing scheme of FIGS. 1-4 can be used to avoid undersensing some R-waves, such as R-waves in PVC beats. To capture these beats, a secondary channel can be used, such as the secondary sensing channel 110 used with a lower threshold.

[0070] For example, when the primary sensing channel 108 senses an R-wave, the primary sensing channel 108 can blank the auto-adjusting threshold process 104 as well as the fixed threshold process 106 for a period of time, such as 150 ms, to avoid the secondary sensing channel 110 sensing the same beat. If the secondary sensing channel 110 senses an R-wave that was not sensed by the primary sensing channel 108, the secondary sensing channel 110 can blank the fixed threshold process 106 for 520 ms after R-wave sensing. In this example, the secondary sensing channel 110 can not blank the primary channel from sensing.

[0071] To determine the T-wave location, the ICM 10 can bandpass the EGM signal from the electrode 16A and the electrode 16B, for example, using a bandpass filter 90. In some examples, the bandpass filter can be a 6-20 Hz bandpass filter. The bandpass signal can be rectified by a rectifier 92. In Figure 5B In the example of FIG. 5, the primary sensing channel 108 and the secondary sensing channel 110 determine R-wave sensing 95. The T-wave sensor 94 can utilize the R-wave sensing 95 to determine a search window for the T-wave.

[0072] Figure 6is a conceptual diagram showing an example segment of an EGM signal 105 and a corresponding rectified waveform 107.

[0073] According to the techniques of this disclosure, the ICM 10 can determine a window following the QRS complex to search for a T-wave. The ICM 10 can determine the window based on one or more of a current RR interval (the time between two consecutive R-waves) and a RR interval of a previous beat (a previous RR interval). To accurately determine the start sample and end sample of the search window, the ICM 10 can determine an R-wave peak sample in the rectified signal (e.g., in the primary sensing channel 108 and / or the secondary sensing channel 110). To determine the R-wave peak in the rectified signal, the processing circuitry 50 of the ICM 10 can take a first predetermined number of samples, e.g., 14 samples, before a sensed R-wave and a second predetermined number of samples, e.g., 25 samples, after the sensed R-wave at a predetermined frequency, e.g., 256 Hz, and determine that the sample with the largest amplitude is the R-wave peak sample in the rectified signal. In some examples, the first predetermined number of samples and the second predetermined number of samples can be the same. In other examples, the first predetermined number of samples and the second predetermined number of samples can be different. The ICM 10 can use this technique to determine the R-wave peak samples for the current, previous, and next beats as the parameters peak current sample, peak previous sample, and peak next sample, respectively, in order to obtain the current and previous RR intervals. The ICM can determine the R-wave peak samples for the current, previous, and next beats to improve the accuracy of the window determination process in which the location of the T-wave is searched.

[0074] After determining the R-wave peak samples, the ICM 10 can determine four parameters for determining a T-wave search window. The ICM 10 can determine the parameters window start current and window end current based on the current RR interval between the current and next beats. Similarly, the ICM 10 can determine the parameters window start previous and window end previous based on the previous RR interval between the current beat and the previous beat. The following Table 1 and Table 2 show examples of determining these four parameters based on the current and previous RR intervals.

[0075]

[0076] Table 1 - Determination of window start current and window end current parameters based on current RR interval

[0077]

[0078]

[0079] Table 2 - window start previous and window end previous parameters based on previous RR interval determination

[0080] window start 1 = peak previous sample + window start previous

[0081] window end 1 = peak current sample - window end previous

[0082] previous window length = window end 1 - window start 1

[0083] Similarly, the length of the window based on the current RR interval is determined by:

[0084] window start 2 = peak current sample + window start current

[0085] window end 2 = peak next sample - window end current

[0086] current window length = window end 2 - window start 2

[0087] Figure 7 is a conceptual diagram showing an example segment of an EGM signal according to the techniques of this disclosure, depicting different parameters computed by the QT detection algorithm. In Figure 7 In the example of FIG. 4, the ICM 10 can determine the start sample and end sample of the search window for the T-wave of the current beat by:

[0088] window start = peak current sample + window start previous

[0089] window end = window start + previous window length.

[0090] Figure 8is a conceptual diagram illustrating an example segment of an EGM signal depicting determination of window_start and window_end parameters. The window_start and window_end parameters can be initially determined based on a previous RR interval as shown in Figure 8 The ICM 10 can determine the start sample and end sample of the search window based on a previous RR interval as shown in

[0091] window_end = window_end2.

[0092] The ICM 10 can set the parameter window_end = window_end2 to ensure that the P-wave of the next beat is not falsely determined as a T-wave. Thus, if the end of the window determined using the previous RR interval is close to the QRS complex of the next beat, the end of the window can be set based on the current RR interval, rather than the previous RR interval.

[0093] If the difference between the previous RR interval and the current RR interval is greater than a third predetermined number of samples, e.g., a number of samples, e.g., 128 samples at 256 Hz (500 ms), the ICM 10 can set the window_start and window_end parameters to:

[0094] window_start = window_start2

[0095] window_end = window_end2

[0096] The third predetermined number of samples can be any number of samples, e.g., 102 samples at 256 Hz (400 ms); 154 samples at 256 Hz (600 ms), or 205 samples at 256 Hz (800). In some examples, the frequency can be different than 256 Hz.

[0097] Figure 9 is a conceptual diagram illustrating an example segment of an EGM signal depicting calculation of window_start and window_end parameters if the difference between the current RR interval and the previous RR interval is greater than 500 ms (128 samples at 256 Hz). If the difference between the previous RR interval and the current RR interval is greater than a third predetermined number of samples, e.g., 128 samples (500 ms), the ICM 10 can determine the window start sample and end sample based on the current RR interval, rather than the previous RR interval, to obtain the window_start and window_end parameters asFigure 9 Similarly, if a previous beat determination was noisy, ICM 10 can determine the window start sample and end sample based on the current RR interval, rather than the previous RR interval.

[0098] ICM 10 can determine a search window for the T-wave to start at the window start sample and end at the window end sample. To determine the location of the T-wave, for each sample in the window, a fourth predetermined number of samples, such as a 9-sample median, can be taken (e.g., 4 samples before the given sample at 256 Hz and 4 samples after the given sample at 256 Hz). ICM 10 can determine the sample with the largest median to be the location of the T-wave. In some examples, ICM 10 can calculate the QT interval from the peak current sample to this largest median sample. For example, processing circuitry 50 can determine the QT interval that determines the number of samples or time between peak current sample and the largest median sample.

[0099] Normal cardiac repolarization adapts to heart rate. This phenomenon means that as heart rate increases, the myocardium remains persistently excitable, e.g., fully repolarized, until the next depolarization wave occurs. This can prevent the possibility of incomplete repolarization and subsequent development of reentrant tachycardia. In long QT syndrome, the heart’s adaptation to changes in heart rate is disrupted, promoting arrhythmias. See Postema PG, et al., “Measurement of the QT interval”, Curr Opin Cardiol. 2014; 29(3): 287-294.

[0100] The QT interval depends on the RR interval, and is longer when the heart rate is slower and shorter when the heart rate is faster. Thus, ICM 10 can calculate a corrected QT interval (QTc). Using QTc can improve detection of patients with increased risk of ventricular arrhythmias. Three methods are commonly used to calculate the QTc interval - the Bazett formula, the Fridericia formula, and the Framingham formula:

[0101] Bazett formula: QTc = QT / sqrt(RR)

[0102] Fridericia formula: QTc = QT / RR 1 / 3

[0103] Framingham formula: QTc = QT + 0.154(1-RR)

[0104] While the Bazett formula (log correction) is the most commonly used QT correction formula, this formula is not optimal outside the 60-100 heart rate range. This formula over-corrects at heart rates greater than 100 bpm and under-corrects at heart rates less than 60 bpm. Both the Friderica formula and the Framingham formula perform better for heart rates outside the 60-100 range. One study showed that the Fridericia and Framingham correction formulas exhibit better rate correction when compared to the Bazett formula and significantly improve prediction of 30-day and 1-year mortality. See Van den Berk B, et al., "Which QT Correction Formula for QT Monitoring?", Journal of the American Heart Association, Vol. 5(6), 2016. For example, ICM 10 can use the Framingham formula (linear correction formula) for calculating the QTc interval. In other examples, ICM 10 can employ the Friderica formula. In other examples, ICM 10 can employ the Bazett formula. In other examples, ICM 10 can employ some other formula or technique to determine the QTc interval. In some examples, ICM 10 can employ more than one of the Framingham, Friderica, or Bazett formulas. In such examples, ICM 10 can determine a mean, median, or mode QTc based on the formula used.

[0105] A QT detection algorithm that can be implemented in ICM 10 has been developed by using real-world clinical data from over 45 patients from the Medtronic CareLink® TM Data Warehouse. The algorithm was developed using 74 night-time transmission episodes (each 10 seconds long) and 70 patient-activated episodes (30-second EGM segments) from patients with diabetes and long QT syndrome during a 1-year follow-up of patients implanted with an ICM (such as ICM 10) for syncope indication (revealing LINQ® TM ). The development dataset had over 3,800 beats from over 45 patients analyzed. This dataset provided T-wave morphology at different locations and orientations from the ICM.

[0106] After extracting the EGM from the patient-activated episode, the R-wave is sensed by running the algorithm according to the techniques of the present disclosure. Primary and secondary event markers are used to manually annotate the location of the T-wave to obtain manual data. The algorithm results are compared to the manually annotated T-wave locations to assess the performance of the QT detection algorithm.

[0107] Figure 10is a conceptual diagram illustrating an example GUI for assisting in manually annotating a dataset. The EGM file to be annotated and the labels generated by the algorithm can be input to the GUI. The GUI can display beats for which the user should label the T-wave location. For each R-wave label 112 displayed on the GUI, the user can select the location of the T-wave 114 by clicking. In addition, the user can also assign an annotation to each beat: 1) normal beat; 2) noisy T-wave; or 3) incorrect R-wave label.

[0108] Figure 11 is a conceptual diagram illustrating an example of manual annotation. After annotating all R-waves on the GUI window (e.g., adding R-wave labels 112), the user can press the enter button to proceed to the next set of R-waves that need to be annotated. Figure 11 Annotating 1 a normal beat, 2 a noisy beat, and 3 an incorrect annotation with R-wave 116.

[0109] Figure 12 is a conceptual diagram illustrating a histogram of the difference between the manually annotated QT interval and the QT interval calculated based on the detection technique of the present disclosure for each beat in the development dataset according to the techniques of the present disclosure. To determine the performance of the QT interval detection technique of the present disclosure, the difference between the manually annotated QT interval and the QT interval calculated based on the detection technique of the present disclosure (QT(true) - QT(detected)) was calculated for each beat in the development dataset as shown in Figure 12 is a conceptual diagram illustrating a histogram of the difference between the manually annotated QT interval and the QT interval calculated based on the detection technique of the present disclosure for each beat in the development dataset according to the techniques of the present disclosure. To determine the performance of the QT interval detection technique of the present disclosure, the difference between the manually annotated QT interval and the QT interval calculated based on the detection technique of the present disclosure (QT(true) - QT(detected)) was calculated for each beat in the development dataset as shown in

[0110] Table 3 below shows the percentage of beats in the development dataset corresponding to the absolute value of the parameter (QT(true) - QT(detected)) for each beat in the development dataset in samples and ms, where the total number of beats is 3829.

[0111] Absolute value of (QT (true) - QT (detected)) samples at 256 Hz Number of beats % of beats <＝2 1778 46.4％ <＝4 2603 68％ <＝6 2999 78.3％ <＝8 3186 83.2％ <＝10 3271 85.4％ <＝12 3320 86.7％

[0112] Absolute value of (QT (true) - QT (detected)) in ms Number of beats % of beats <5 1106 29％ <10 1778 46.4％ <15 2247 58.7％ <20 2844 74.3％ <25 2999 78.3％ <30 3126 81.6％ <35 3186 83.2％ <40 3294 86％

[0113] Table 3 - Percentage of beats in the development dataset corresponding to the difference between the manually annotated QT interval and the algorithmically detected QT interval in ms and number of samples at 256 Hz.

[0114] Figure 13 is a conceptual diagram illustrating a histogram of the difference between the manually annotated QTc interval and the QTc interval calculated based on the detection technique of the present disclosure for each beat in the development dataset. The QTc interval can be calculated as the QT interval divided by the square root of the RR interval. Figure 13As observed in Table 4 below, for 46.5% of the beats in the development dataset, the difference between the manually annotated QT interval and the QT interval based on the detection technique of this disclosure was less than or equal to 2 samples (7.8 ms). For more than 78% of the beats in the development dataset, this parameter was less than 25 ms.

[0115] Absolute value of (QT (true) - QT (detected)) in ms Number of beats % of beats <5 1106 29％ <10 1778 46.4％ <15 2247 58.7％ <20 2712 70.8％ <25 2999 78.3％ <30 3126 81.6％ <35 3187 83.2％ <45 3294 86％

[0116] Table 4 - Differences between manually annotated QTc intervals and algorithm-detected QTc intervals in the development dataset.

[0117] Percentage of pulses (in milliseconds).

[0118] The QTc interval was calculated for each beat in the development dataset using the Framingham correction formula. Similar to... Figure 12 For example, for each beat in the development dataset, the difference between the QTc interval calculated based on manual annotation and the QTc interval based on the detection technique of this disclosure is calculated. Figure 13 The histogram for this parameter is shown below. The mean of the absolute values ​​of this parameter is 22.6ms and the median is 12ms. The mode is found to be 4ms.

[0119] Figure 14 This is a conceptual diagram showing a histogram of the mean difference between manually annotated QTc intervals and QTc intervals based on the detection technique of this disclosure for 46 unique devices in the development dataset. To determine the performance of the technique of this disclosure based on unique devices, a parameter (QTc interval (true) - QTc interval (detected by the algorithm)) is calculated individually for each unique ICM, and the mean of this parameter is calculated for each device. Figure 14 It was observed that 37 out of 46 devices had a mean of less than 25 ms.

[0120] Figure 15A -D is a conceptual diagram showing instance devices where the mean of (QTc interval (actual) - QTc interval (detected by the algorithm)) is greater than 25 ms. There are 9 devices with a mean greater than 25 ms. Figure 15A The -D option displays some instances from these specific ICMs. Among several of these ICMs, such as... Figure 15A As shown and in Figure 15B In some cases, the techniques of this disclosure detect T waves at different locations compared to manual annotation, but T waves are typically detected persistently at the same location. For example, in Figure 15A and 15CIn some cases, the R-wave marker 112 is followed by a T-wave 118 determined by manual annotation, followed by a T-wave 120 determined according to the detection techniques of the present disclosure. This pattern continues throughout Figure 15A and 15C In some cases, the mean of (QTc interval (true) - QTc interval (algorithm detected)) is large in these devices, but the change in QT interval can still be measured in these cases.

[0121] In some ICMs, such as shown in Figure 15D there are beat-to-beat variations in the detection of the T-wave, which in some cases is due to beat-to-beat noise or RR interval variability.

[0122] Figure 16 is a conceptual diagram showing examples of EGM strips from a development data set with T-waves having different morphologies and orientations at different RR intervals, depicting both manual annotation and detection according to the techniques of the present disclosure. These examples include night transmissions as well as patient-activated episodes from ICMs that accurately detected their QT intervals by algorithm.

[0123] Another observation from the analysis of the development data set is that the ICMs are able to capture beat-to-beat changes in QT interval well, as shown in Figure 17A -B. In Figure 17A , the QTc changes by 43 ms between beat 122 and beat 124, which is captured well by the ICM. Similarly, in Figure 17B , the QTc interval changes by 54 ms between beat 126 and beat 128, and by 15 ms between beat 128 and the next beat, which is also captured by the ICM 10.

[0124] In some examples, the amplitudes of the samples in the search window can be weighted by giving more weight (weighting more heavily) to the amplitudes of the samples located at the most likely detection of the T-wave based on the previous QT or QTc intervals (e.g., 12 QT or QTc intervals) and less weight to the samples at the end of the window. For example, the processing circuitry 50 of the ICM 10 can take the last 12 QT or QTc intervals and form a weighted window based on the location of the T-wave in the fastest QT or QTc interval and the location of the T-wave in the slowest QT or QTc interval and apply a weight to the amplitudes of the samples within the window such that the amplitudes of the samples are greater than they would be otherwise. For example, the processing circuitry 50 of the ICM 10 can apply a weight to the amplitudes of the samples outside of the window such that the amplitudes of the samples are less than they would be otherwise. In some examples, the weighted window can be equal to the location of the T-wave in the fastest QT or QTc interval and the location of the T-wave in the slowest QT or QTc interval. In other examples, the weighted window can be greater than or less than the location of the T-wave in the fastest QT or QTc interval and the location of the T-wave in the slowest QT or QTc interval.

[0125] In some examples, the ICM 10 can use the noise detection techniques, amplitude screening techniques, and / or confidence techniques according to the present disclosure. The ICM 10 can use such techniques individually or in any combination. These techniques can be used to improve the accuracy of the determination of the QT or QTc intervals.

[0126] The ICM 10 can incorporate noise detection techniques to detect noisy beats by detecting noisy QRS complexes and determining whether the search window for the T-wave after the QRS is noisy. In some examples, the ICM 10 can use the noise detection techniques of the present disclosure to determine whether an R-wave is noisy and whether the ECG around a detected T-wave is noisy. In some examples, the ICM 10 can analyze the zero crossings and rate of change (e.g., slope) of the cardiac signal to determine whether the cardiac signal is noisy. For example, if the zero crossings have a high rate of change, this can indicate a noisy cardiac signal.

[0127] Figure 18 is a flowchart illustrating example noise determination techniques of the present disclosure. The sensing circuitry 52 can determine an R-wave of the cardiac signal (202). For example, the auto-adjusting threshold process 104 and / or the fixed threshold process 106 can determine that the rectified signal from the rectifier 102 has crossed the auto-adjusting threshold and / or the fixed threshold.

[0128] Processing circuitry 50 can determine whether the R-wave is noisy (204). For example, processing circuitry 50 can determine whether a rate of change of amplitudes of a first plurality of samples of the cardiac signal around an R-wave peak indicates that the R-wave is noisy. For example, processing circuitry 50 can determine an R-wave peak. In some examples, as part of determining the R-wave peak, processing circuitry 50 can determine a mean, median, or mode of amplitudes of a second plurality of samples of the cardiac signal, a first sample of the second plurality of samples being before the determined R-wave and a last sample of the second plurality of samples being after the determined R-wave. For example, processing circuitry 50 can store a predetermined number of samples before the determined R-wave and a predetermined number of samples after the determined R-wave in storage 84. In some examples, the predetermined number of samples before the determined R-wave and the predetermined number of samples after the determined R-wave are the same. In other examples, the predetermined number of samples before the determined R-wave and the predetermined number of samples after the determined R-wave are different. In some examples, the predetermined number of samples before the determined R-wave and the predetermined number of samples after the determined R-wave are 25 samples. However, the predetermined number of samples before the determined R-wave and the predetermined number of samples after the determined R-wave can be any number of samples. Processing circuitry 50 can determine a mean, median, or mode of the amplitudes of the second plurality of samples of the cardiac signal. A first sample of the second plurality of samples is before the determined R-wave and a last sample of the second plurality of samples is after the determined R-wave. For example, processing circuitry 50 can determine a mean of the second plurality of samples.

[0129] Processing circuitry 50 can determine a respective mean, median, or mode difference between the mean, median, or mode and each of the second plurality of samples. In other words, processing circuitry 50 can determine a difference between each of the samples of the second plurality of samples and the mean, median, or mode of the second plurality of samples. Processing circuitry 50 can determine an absolute value of each of the respective mean, median, or mode differences and determine the R-wave peak to be a highest absolute value of each of the respective mean, median, or mode differences. For example, processing circuitry 50 can determine a magnitude of a result of a subtraction, regardless of whether the result of the subtraction is a positive number or a negative number. For example, if the result of the subtraction is -3, processing circuitry 50 can determine an absolute value of the result of the subtraction to be 3. If the result of the subtraction is 3, processing circuitry 50 can determine an absolute value of the result of the subtraction to be 3. For example, processing circuitry 50 can compare the absolute values of each sample and determine the sample with the highest absolute value of the group of samples to be the R-wave peak.

[0130] Processing circuitry 50 can store a predetermined number of the first plurality of samples preceding the determined R-wave peak and a predetermined number of the first plurality of samples following the determined R-wave peak in storage 84. In some examples, the predetermined number of samples preceding the determined R-wave peak and the predetermined number of samples following the determined R-wave peak are the same. In some examples, the predetermined number of samples preceding the determined R-wave peak and the predetermined number of samples following the determined R-wave peak are different. In some examples, the predetermined number of samples preceding the determined R-wave peak is 15 and the predetermined number of samples following the determined R-wave peak is 16, although the predetermined number of samples preceding the determined R-wave peak and the predetermined number of samples following the determined R-wave peak can be any number.

[0131] Processing circuitry 50 can determine a plurality of first differences between the first plurality of samples of the cardiac signal. A first sample of the first plurality of samples precedes the R-wave peak and a last sample of the first plurality of samples follows the R-wave peak. Each of the plurality of first differences can be an amplitude difference of a respective sample of the first plurality of samples and a next sample of the first plurality of samples. For example, the first differences can include an amplitude difference between a first sample and a second sample of the first plurality of samples, an amplitude difference between the second sample and a third sample of the first plurality of samples, and so on. Thus, in an example in which there are 32 samples, processing circuitry 50 can determine 31 first difference values.

[0132] Processing circuitry 50 can determine a plurality of second differences between each of the plurality of first differences and a respective next difference of the plurality of first differences. For example, the plurality of second differences can include a difference between a first first difference and a second first difference, a difference between the second first difference and a third first difference, and so on. Thus, in an example in which there are 31 first difference values, IMD 10 can determine 30 second difference values.

[0133] Processing circuitry 50 can determine whether there is a sign change (e.g., from positive to negative, or from negative to positive) between each of the plurality of first differences and a respective next difference in the plurality of first differences. For example, processing circuitry 50 can determine that a first first difference is positive and a second first difference is negative. This would be a sign change. If the first first difference is negative and the second first difference is positive, this would also be a sign change. For each determined sign change, processing circuitry 50 can determine whether a magnitude of a respective second difference value is greater than a first predetermined noise threshold. For example, the first predetermined noise threshold can be 40. In some examples, the magnitude is an absolute value of the respective determined sign change. In some examples, there can be a positive first predetermined noise threshold and a negative first predetermined noise threshold, and processing circuitry 50 can determine whether the magnitude of the corresponding respective second difference value is greater than the positive first predetermined noise threshold or below the negative first predetermined noise threshold.

[0134] Processing circuitry 50 can count a total number of sign changes of the first difference values in which the respective second difference value is greater than the positive first predetermined noise threshold or below the negative first predetermined noise threshold. For example, processing circuitry 50 can count the number of sign changes of the first difference values in which the respective second difference value is greater than the positive first predetermined noise threshold or below the negative first predetermined noise threshold. Processing circuitry 50 can determine whether the counted total number is greater than a second predetermined noise threshold (e.g., 8). For example, if the result of the count is higher than the second predetermined noise threshold, processing circuitry 50 can consider the R wave to be noisy. In this case, the beat associated with the noisy R wave can not be used to determine the QT or QTc interval. If the result of the count is equal to or lower than the second predetermined noise threshold, processing circuitry 50 can consider the R wave to be not noisy.

[0135] In some examples, processing circuitry 50 can use other techniques to determine whether the R wave is noisy. For example, processing circuitry 50 can utilize a morphology comparison to one or more templates. The one or more templates can be derived from recent data acquired from the patient or from long-term historical data from the patient. In some examples, processing circuitry 50 can perform a root mean square error analysis to determine whether the R wave is noisy.

[0136] Based on the R wave being not noisy, processing circuitry 50 can determine whether the cardiac signal around the determined T wave is noisy (206). For example, to check for noise around the determined T wave, processing circuitry 50 can first map the T wave location of the sample from the rectified signal that can be determined to have the highest slope to the actual ECG. Because the rectified signal is shifted to the right compared to the actual signal, the T wave location in the actual signal can be, for example, the 18th sample before the T wave location sample as determined by the algorithm in the rectified signal.

[0137] For example, processing circuitry 50 can store a predetermined number of third plurality of samples before the determined T-wave actual location (e.g., the mapped T-wave) and a predetermined number of the third plurality of samples after the determined T-wave actual location in storage 84. In some examples, the predetermined number of samples before the determined T-wave actual location and the predetermined number of samples after the determined T-wave actual location are the same. In some examples, the predetermined number of samples before the determined T-wave actual location and the predetermined number of samples after the determined T-wave actual location are different. In some examples, the predetermined number of samples before the determined T-wave actual location is 15 and the predetermined number of samples after the determined T-wave actual location is 16, although the predetermined number of samples before the determined T-wave actual location and the predetermined number of samples after the determined T-wave actual location can be any number.

[0138] For example, processing circuitry 50 can determine a plurality of third differences between a third plurality of samples of the cardiac signal, a first sample of the third plurality of samples being before the mapped T-wave and a last sample of the third plurality of samples being after the mapped T-wave. Each of the plurality of third differences can be an amplitude difference of a respective sample of the third plurality of samples and a next sample of the third plurality of samples. For example, the plurality of third differences can include an amplitude difference between a first sample and a second sample of the third plurality of samples, an amplitude difference between a second sample and a third sample of the third plurality of samples, and so on.

[0139] Processing circuitry 50 can determine a plurality of fourth differences between each of the plurality of third differences and a respective next difference of the plurality of third differences. For example, a fourth difference can include a difference between a first third difference and a second third difference, a difference between a second third difference and a third third difference, and so on.

[0140] For example, by determining the plurality of third differences and the plurality of fourth differences, processing circuitry 50 can effectively high-pass filter the mapped T-wave in the time domain with a lag to determine a frequency cutoff. The mapped T-wave can be slower than the R-wave, so the lag can be longer.

[0141] Processing circuitry 50 can determine whether there is a sign change (e.g., from positive to negative, or from negative to positive) between each of the plurality of third differences and a respective next difference in the plurality of third differences. For example, processing circuitry 50 can determine that a first third difference is positive and a second third difference is negative. This would be a sign change. If the first third difference is negative and the second third difference is positive, this would also be a sign change. For each determined sign change, processing circuitry 50 can determine whether a magnitude of a corresponding respective fourth difference value is greater than a positive third predetermined noise threshold or lower than a negative third predetermined noise threshold. For example, the positive third predetermined noise threshold can be 20 and the negative third predetermined threshold can be -20. In some examples, the magnitude is an absolute value of a respective magnitude of a respective fourth difference, and processing circuitry 50 can determine whether a magnitude of a corresponding respective fourth difference value is greater than a third predetermined noise threshold (e.g., 20).

[0142] For example, processing circuitry 50 can count a total number of respective fourth difference values having a magnitude greater than a positive third predetermined noise threshold or lower than a negative third predetermined noise threshold. In examples where the magnitude is an absolute value, processing circuitry 50 can count a number of respective fourth difference values having a magnitude greater than a positive third predetermined noise threshold. Processing circuitry 50 can determine whether the counted total number is greater than a fourth predetermined noise threshold (e.g., 5). For example, if the result of the count is higher than the fourth predetermined noise threshold, processing circuitry 50 can consider the cardiac signal around the T-wave to be noisy. In this case, beats associated with the noisy ECG around the T-wave can not be used to determine QT or QTc intervals. If the result of the count is equal to or lower than the predetermined number, IMD 10 can consider the ECG around the T-wave to be not noisy.

[0143] In some examples, processing circuitry 50 can use other techniques to determine whether the ECG around the T-wave is noisy. For example, processing circuitry 50 can utilize a morphology comparison to one or more templates. The one or more templates can be derived from recent data acquired from the patient or from long-term historical data from the patient. In some examples, processing circuitry 50 can perform a root mean square error analysis to determine whether the ECG around the T-wave is noisy.

[0144] Based on the determined lack of noise in the cardiac signal around the T-wave, processing circuitry 50 can determine a QT interval or a corrected QT interval based on the determined T-wave and the determined R-wave (208). For example, processing circuitry 50 can determine the time between the R-wave peak and the determined T-wave as the QT interval. In other examples, processing circuitry 50 can determine the QT interval as the number of samples between the R-wave peak and the determined T-wave. Processing circuitry 50 can also determine a QTc interval (150). For example, processing circuitry 50 can apply at least one of the Framingham formula, the Friderica formula, or the Bazett formula, or another formula or technique, to the QT interval to determine the QTc interval. In some examples, if the R-wave of a beat is noisy or the cardiac signal around the T-wave location of a beat is noisy, processing circuitry 50 can consider the beat to be noisy and exclude the beat from the QT interval (or QTc) calculation. In some examples, if processing circuitry 50 determines that a beat is not noisy, processing circuitry 50 can employ the features (e.g., amplitude) screening techniques of the present disclosure. Because processing circuitry 50 only looks for noise around the detected T-wave location, the processing circuitry is not affected by noise in other portions of the cardiac signal in the current R-R segment.

[0145] In some examples, processing circuitry 50 can attempt to detect a beat that is significantly different from other beats. In some examples, processing circuitry 50 can analyze the morphology of the beats and compare one or more features of the beats to determine that a certain beat is significantly different from other beats. In some examples, processing circuitry 50 can map the determined T-wave, as discussed above. In some examples, processing circuitry 50 can not map the determined T-wave. In some examples, processing circuitry 50 can determine whether at least one feature of the determined T-wave (whether mapped or not) exceeds a predetermined difference threshold (e.g., 200) from a mean, median, or mode of the at least one feature of a plurality of other T-waves. This can be performed to ensure that a P-wave or noise is not detected as a T-wave. In some examples, the at least one feature includes amplitude. In some examples, the determination of the QT interval is further based on the determined T-wave not exceeding a predetermined difference threshold from a mean, median, or mode of the at least one feature of the plurality of other T-waves (e.g., 8 other T-waves). For example, processing circuitry 50 can store a predetermined number of detected T-wave location amplitudes before the T-wave and a predetermined number of detected T-wave location amplitudes after the detected T-wave. In some examples, the predetermined number of detected T-wave location amplitudes before the T-wave and the predetermined number of detected T-wave location amplitudes after the detected T-wave are the same. In other examples, the predetermined number of detected T-wave location amplitudes before the T-wave and the predetermined number of detected T-wave location amplitudes after the detected T-wave are different. In some examples, the predetermined number of detected T-wave location amplitudes before the T-wave and the predetermined number of detected T-wave location amplitudes after the detected T-wave are each 4, but can be any number.

[0146] In some examples, as part of determining whether the determined T-wave exceeds a predetermined difference threshold from a plurality of other T-waves, processing circuitry 50 can determine whether a number of noisy beats in a predetermined number of beats is greater than a predetermined noisy beat threshold, a first beat of the predetermined number of beats being before the determined T-wave and a last beat of the predetermined number of beats being after the determined T-wave. In some examples, the predetermined noisy beat threshold can be a percentage of the predetermined number of beats, such as 50%. For example, processing circuitry 50 can determine whether a beat is noisy by employing the noise detection techniques discussed above with respect to each of the beats. If the number of noisy beats exceeds the predetermined noisy beat threshold, processing circuitry 50 can exclude the current beat from the QT interval analysis.

[0147] In some examples, if the number of noisy beats does not exceed the predetermined noisy beat threshold, processing circuitry 50 can determine a mean, median, or mode of the amplitudes of the predetermined number of beats that are not noisy beats. For example, IMD 10 can determine a median T-wave amplitude of the predetermined number of detected T-wave location amplitudes preceding the T-wave and the predetermined number of detected T-wave location amplitudes following the T-wave that are not noisy beats. Processing circuitry 50 can also determine an absolute value of the difference between the mean, median, or mode of the amplitudes of the beats that are not noisy beats and the determined T-wave. Processing circuitry 50 can also determine whether the absolute value of the difference is greater than a predetermined difference threshold, e.g., 200. If the absolute value of the difference is greater than the predetermined difference threshold, IMD 10 can exclude the current beat from the QT interval (or QTc) calculation.

[0148] An IMD 10 implementing the amplitude screening techniques of the present disclosure can exclude beats whose R-waves are incorrectly detected as T-waves. In some examples, an IMD 10 implementing the amplitude screening techniques of the present disclosure can exclude PVC beats. In some examples, an IMD 10 implementing the amplitude screening techniques of the present disclosure can exclude beats having a T-wave location that is significantly different from other beats.

[0149] In some examples, processing circuitry 50 can determine a confidence of the determined T-wave. In some examples, the confidence is based on a flatness of the cardiac signal around the determined T-wave. In some examples, the confidence metric can include, e.g., a high, medium, or low confidence. In some examples, processing circuitry 50 can determine the flatness based on an area under the curve between selected points on either side of the determined T-wave, an amplitude of samples around the T-wave, a slope analysis, or other techniques that would indicate the flatness of the cardiac signal around the determined T-wave.

[0150] In some examples, as part of determining the flatness, processing circuitry 50 can determine a mapped T-wave location as discussed above. In some examples, processing circuitry 50 can determine a maximum amplitude and a minimum amplitude of the amplitudes of a fourth plurality of samples of the cardiac signal, a first sample of the fourth plurality of samples being before the mapped T-wave location and a last sample of the fourth plurality of samples being after the mapped T-wave location. For example, processing circuitry 50 can store in storage 84 a predetermined number of samples preceding the mapped T-wave and a predetermined number of samples following the mapped T-wave. The T-wave can be mapped to an actual location of the T-wave, as mentioned above. In some examples, the predetermined number of samples preceding the T-wave and the predetermined number of samples following the detected T-wave are the same. In other examples, the predetermined number of samples preceding the T-wave and the predetermined number of samples following the detected T-wave are different. In some examples, the predetermined number of samples preceding the T-wave and the predetermined number of samples following the detected T-wave are each 20, but can be any number.

[0151] In other examples, processing circuitry 50 can determine the mapped T-wave location by comparing different hysteresis differences or by morphological comparison to one or more templates. The one or more templates can be derived from recent data acquired from the patient or from long-term historical data from the patient.

[0152] Processing circuitry 50 can determine whether the absolute value of the difference between the maximum amplitude and the minimum amplitude is greater than or equal to a first predetermined confidence threshold, whether the absolute value of the difference is less than or equal to a second predetermined confidence threshold, or whether the absolute value of the difference is between the first predetermined confidence threshold and the second predetermined confidence threshold. For example, processing circuitry 50 can determine a maximum amplitude between stored samples and a minimum amplitude between the stored fourth plurality of samples. For example, processing circuitry 50 can determine that the amplitude of the sample with the highest amplitude is the maximum amplitude and that the amplitude of the sample with the lowest amplitude is the minimum amplitude.

[0153] Processing circuitry 50 can determine a difference between the maximum amplitude and the minimum amplitude and can determine an absolute value of the difference.

[0154] Processing circuitry 50 can compare the absolute value of the difference to an upper threshold and / or a lower threshold. In some examples, the upper threshold is 35 and the lower threshold is 25. If the absolute value of the difference is greater than or equal to the upper threshold, processing circuitry can assign a high confidence to the beat. If the absolute value of the difference is less than or equal to the lower threshold, processing circuitry 50 can assign a low confidence to the beat. If the absolute value of the difference is between the upper threshold and the lower threshold, processing circuitry 50 can assign a medium confidence to the beat. In some examples, processing circuitry 50 can exclude beats with a low confidence measure from a QT interval (or QTc) analysis.

[0155] In some examples, processing circuitry 50 of ICM 10 can determine a mean, median, mode, standard deviation, or any other trend of the determined QT intervals or QTc intervals over time. ICM 10 can communicate the mean, median, mode, standard deviation, or any other trend of the determined QTc intervals to external device 12. In some examples, ICM 10 can determine a time or count of QT intervals or QTc intervals that are longer than a predetermined threshold. For example, this predetermined threshold can be about 500 ms, as QT intervals greater than 500 ms can be associated with a higher risk of torsades de pointes. In some examples, ICM 10 can determine a time or count of QT intervals or QTc intervals that change by more than a threshold. For example, ICM 10 can determine a time or count of QT intervals or QTc intervals that change by more than 30 ms or 40 ms or some other threshold, which can even be patient-specific, over a particular time period.

[0156] Figure 19A is a flowchart illustrating example techniques of the present disclosure. Sensing circuitry 52 of ICM 10 can sense a cardiac signal (130). In some examples, sensing circuitry 52 can apply one or more bandpass filters (e.g., bandpass filter 100) or rectifiers (e.g., rectifier 102) to the cardiac signal. In some examples, sensing circuitry 52 can sense the cardiac signal using primary sensing channel 108 and secondary sensing channel 110.

[0157] Sensing circuitry 52 can determine an R-wave of the cardiac signal (132). For example, auto-adjust threshold process 104 and / or fixed threshold process 106 can determine that the rectified signal from rectifier 102 has crossed an auto-adjust threshold and / or a fixed threshold. Processing circuitry 50 of ICM 10 can determine a previous RR interval (134). For example, to accurately determine the previous RR interval, processing circuitry 50 can determine peak R values (e.g., peak current sample and peak previous sample) in two consecutive sensed R-waves. Processing circuitry 50 can determine the previous RR interval as the time between the two consecutive peak R values, peak current sample and peak previous sample. In some examples, processing circuitry 50 can take a first predetermined number of samples before a sensed R-wave and a second predetermined number of samples after the sensed R-wave at a predetermined frequency, and determine the sample with the largest amplitude as the R-wave peak sample.

[0158] Processing circuitry 50 of ICM 10 can also determine a current RR interval (136). For example, to accurately determine the current RR interval, processing circuitry 50 can determine a peak R value (e.g., peak next sample) of a next sensed R-wave. Processing circuitry 50 can determine peak next sample in the same or similar manner as used to determine peak current sample and peak previous sample. Processing circuitry 50 can determine the current RR interval as the time between the two consecutive peak R values, peak next sample and peak current sample.

[0159] Processing circuitry 50 can then determine a search window for searching for a T-wave based on one or more of the current RR interval or previous RR intervals. Processing circuitry 50 can determine the search window to start a number of samples after the R-wave peak and end a different number of samples after the R-wave peak. The number of samples can be based on the length of the current RR interval or previous RR intervals. In some examples, the number of samples can be stored in storage 56, such as in a lookup table. In some examples, the number of samples can be those set forth above in Tables 1 and 2.

[0160] In some examples, processing circuitry 50 can utilize a threshold to assist in determining a T-wave. In other examples, processing circuitry can not utilize a threshold. For example, processing circuitry 50 can determine whether the amplitude of a sample is less than a threshold (or in some cases less than or equal to a threshold) (140). If the amplitude of the sample is less than the threshold (or in some cases, less than or equal to the threshold) (Yes path of A), processing circuitry 50 can determine that the sample is not a T-wave (142). Processing circuitry 50 can then examine the next sample. If the amplitude of the sample is equal to or greater than the threshold (or in some cases greater than the threshold), processing circuitry 50 can keep the sample as a candidate for a T-wave (No path of A). Figure 18 Figure 18 A’s No path).

[0161] Processing circuitry 50 can determine a T-wave of the cardiac signal in the search window (144). For example, processing circuitry 50 can determine the highest amplitude sample in the search window as the T-wave. In other examples, processing circuitry 50 can take a predetermined number of samples around a given sample and determine a median, mean, or mode, and determine the T-wave as the largest amplitude median, mean, or mode in the search window.

[0162] In some examples, processing circuitry 50 can determine a confidence of the T-wave (146). In some examples, processing circuitry 50 can not be able to determine a confidence of the T-wave. For example, processing circuitry 50 can determine the confidence based on one or more of a predetermined number of previous T-wave amplitudes or QT intervals. For example, if the amplitude of the detected T-wave is too low compared to previous T-waves, or if the QT interval is very different from the previous 12 QT intervals, ICM 10 can provide a low confidence to the detected T-wave.

[0163] Figure 19B is Figure 19A ​The processing circuit system 50 can determine the QT interval (148) based on the determined T wave and the determined R wave. For example, the processing circuit system 50 can determine the QT interval as the time between the peak value of the R wave and the determined T wave. In other instances, the processing circuit system 50 can determine the QT interval as the number of samples between the peak value of the R wave and the determined T wave. The processing circuit system 50 can also determine the QTc interval (150). For example, the processing circuit system 50 can apply at least one of the Framingham formula, the Friderica formula, or the Bazett formula, or another formula or technique, to the QT interval to determine the QTc interval.

[0164] The processing circuit system 50 can determine the trend of the QT interval or QTc interval over time (152). For example, the processing circuit system 50 can determine the mean, median, mode, standard deviation, or any other trend of the determined QT interval or QTc interval over time.

[0165] Figure 20 It is depicted, for example, as a defined search window ( Figure 19A As part of section 138), a flowchart illustrates how the processing circuitry 50 can determine whether to base the search window on an instance of a previous RR interval or the current RR interval. Figure 20 In one example, the processing circuit system 50 can determine whether the difference between the current RR interval and the previous RR interval is greater than a predetermined amount (160). For example, the processing circuit system 50 can determine whether the difference between the current RR interval and the previous RR interval is greater than a predetermined time period. In another example, the processing circuit system 50 can determine whether the difference between the current RR interval and the previous RR interval is greater than a predetermined number of samples. In some examples, if the difference between the current RR interval and the previous RR interval is not greater than a predetermined amount (160), the processing circuit system 50 can determine whether the difference between the current RR interval and the previous RR interval is greater than a predetermined number of samples. Figure 20 If the "No" path in the previous RR interval is selected, the processing circuit system 50 can determine the search window (162) based on the previous RR interval. If the difference between the current RR interval and the previous RR interval is greater than a predetermined amount ( Figure 20 If the path is "yes" in the code, then the processing circuit system 50 can determine the search window (164) based on the current RR interval. In some instances, the predetermined amount can be 500 ms. In some instances, the predetermined amount can be 128 samples at 256 Hz.

[0166] Figure 21 It is depicted, for example, as a defined search window ( Figure 19A As part of section 138), a flowchart describes how the processing circuitry system 50 can determine whether to base the search window on a previous RR interval or another instance of the current RR interval. Figure 21In the example of FIG. 17, processing circuitry can determine whether the previous beat was noisy (170). For example, ICM 10 can be configured to detect noisy beats. For instance, processing circuitry 50 can determine whether a beat is noisy by determining whether a current R-wave is noisy or whether a segment of ECG between a current R-wave and a next R-wave is noisy, or both. For instance, processing circuitry can determine whether a current R-wave is noisy by defining a window of a number of samples before the R-wave peak and a number of samples after the R-wave peak. Processing circuitry 50 can determine a noise count by counting a number of samples in the window that have a sign change greater than a first threshold (e.g., 50) or a sign change less than a second threshold (e.g., -50). If this noise count is greater than or equal to a threshold (e.g., 5), processing circuitry 50 can determine that the beat is noisy. Processing circuitry 50 can determine whether a segment between a current R-wave and a next R-wave is noisy in a similar manner. In this case, processing circuitry 50 can define a window that starts a number of samples after the current R-wave and ends a number of samples before the next R-wave. Processing circuitry 50 can determine a noise count in a similar manner, but the threshold can be different. If the noise count is greater than or equal to another threshold (e.g., 5 or some other count), processing circuitry can determine that the beat is noisy.

[0167] If processing circuitry 50 determines that the previous beat was not noisy (the“NO” path in FIG. 17), Figure 21 If processing circuitry 50 determines that the previous beat was not noisy (the“NO” path in FIG. 17), Figure 21 If processing circuitry 50 determines that the previous beat was not noisy (the“NO” path in FIG. 17),

[0168] While techniques herein are described as being performed by various elements such as sensing circuitry 52 and processing circuitry 50, in some examples other elements or combinations of elements can perform the techniques. For example, sensing circuitry 52 can perform techniques described as being performed by processing circuitry 50, processing circuitry 50 can perform techniques described as being performed by sensing circuitry 52, or a combination of sensing circuitry 52 and processing circuitry 50 can perform techniques described as being performed by either.

[0169] The techniques described in this disclosure can be implemented, at least in part, in hardware, software, firmware or any combination thereof. For example, various aspects of the techniques can be implemented within one or more processors, DSPs, ASICs, FPGAs or any other equivalent integrated or discrete logic QRS circuitry as well as any combinations of such components, embodied in external devices, such as physician or patient programmers, simulators, or other devices. The terms "processor" "processing circuitry, "controller" or "control module" shall mean any of the foregoing integrated or discrete logic QRS circuitry as well as any combinations of such components, alone or in combination with other logic circuitry or any other equivalent circuitry.

[0170] For aspects implemented in software, at least some of the functionality attributed to the systems and devices described in this disclosure can be embodied as instructions on a non-transitory computer-readable storage medium such as a RAM, a ROM, a NVRAM, an EEPROM, a flash memory, a magnetic or optical media, and the like. The instructions can be executed to support one or more aspects of the functionality described in this disclosure.

[0171] This disclosure includes the following non-limiting examples.

[0172] Example 1. A device comprising: one or more electrodes; sensing circuitry configured to sense a cardiac signal via the one or more electrodes; and processing circuitry configured to: determine an R-wave of the cardiac signal; determine whether the R-wave is noisy; based on the R-wave not being noisy, determine whether the cardiac signal around a determined T-wave is noisy; and based on the cardiac signal around the determined T-wave not being noisy, determine a QT interval or a corrected QT interval based on the determined T-wave and a determined R-wave.

[0173] Example 2. The device of example 1, wherein as part of determining whether the R-wave is noisy, the processing circuitry is configured to: determine whether a rate of change of amplitudes of a plurality of samples of the cardiac signal around an R-wave peak indicates that the R-wave is noisy.

[0174] Example 3. The device of Example 2, wherein as part of determining whether the rate of change of the amplitudes of the plurality of samples of the cardiac signal around the R-wave peak indicates that the R-wave is noisy, the processing circuitry is configured to: determine a peak value of the R-wave; determine a plurality of first differences between a first plurality of samples of the cardiac signal, a first sample of the first plurality of samples being before the R-wave peak and a last sample of the first plurality of samples being after the R-wave peak, each of the plurality of first differences being a difference in amplitude of a respective sample of the first plurality of samples and a next sample of the first plurality of samples; determine a plurality of second differences between each of the plurality of first differences and a respective next difference of the plurality of first differences; determine whether there is a sign change between each of the plurality of first differences and the respective next difference of the plurality of first differences; for each determined sign change, determine whether a magnitude of a corresponding respective second difference value is greater than a first positive predetermined noise threshold or below a first negative predetermined noise threshold; calculate a total number of the respective second difference values having the magnitude greater than the positive first predetermined noise threshold or below the negative first predetermined noise threshold; and determine whether the calculated total number is greater than a second predetermined noise threshold.

[0175] Example 4. The device of Example 3, wherein as part of determining the R-wave peak, the processing circuitry is configured to: determine a mean, median, or mode of amplitudes of a plurality of second samples of the cardiac signal, a first sample of the second plurality of samples being before the determined R-wave and a last sample of the second plurality of samples being after the determined R-wave; determine a respective mean, median, or mode difference between the mean, median, or mode and each of the plurality of second samples; determine an absolute value of each of the respective mean, median, or mode differences; and determine the R-wave peak to be a highest absolute value of each of the respective mean, median, or mode differences.

[0176] Example 5. The device of any combination of examples 1-4, wherein as part of determining whether the cardiac signal around the determined T-wave is noisy, the processing circuitry is configured to: determine a plurality of third differences between a third plurality of samples of the cardiac signal, a first sample of the third plurality of samples being before the determined T-wave and a last sample of the third plurality of samples being after the determined T-wave, each of the plurality of third differences being a difference in amplitude of a respective sample of the third plurality of samples and a next sample of the third plurality of samples; determine a plurality of fourth differences between each of the plurality of third differences and a respective next difference of the plurality of third differences; determine whether there is a sign change between each of the plurality of third differences and the respective next difference of the plurality of third differences; for each determined sign change, determine whether a magnitude of a corresponding respective fourth difference value is greater than a positive third predetermined noise threshold or lower than a negative third predetermined noise threshold; calculate a total number of the respective second difference values having the magnitude that is greater than the positive third predetermined noise threshold or lower than the negative third predetermined noise threshold; and determine whether the calculated total number is greater than a fourth predetermined noise threshold.

[0177] Example 6. The device of any combination of examples 1-5, wherein the processing circuitry is further configured to: determine whether at least one feature of the determined T-wave exceeds a predetermined difference threshold from a mean, median, or mode of the at least one feature of a plurality of other T-waves, wherein the determination of the QT interval is further based on the determined T-wave not exceeding the predetermined difference threshold from the mean, median, or mode of the at least one feature of the plurality of other T-waves.

[0178] Example 7. The device of example 6, wherein as part of determining whether the determined T-wave exceeds the predetermined difference threshold from a plurality of other T-waves, the processing circuitry is configured to: determine whether a number of noisy beats of a predetermined number of beats is greater than a predetermined noisy beat threshold, a first beat of the predetermined number of beats being before the determined T-wave and a last beat of the predetermined number of beats being after the determined T-wave.

[0179] Example 8. The device of example 7, wherein as part of determining whether the determined T-wave exceeds the predetermined difference threshold from a plurality of other T-waves, the processing circuitry is configured to: determine a mean, median, or mode of amplitudes of non-noisy beats of the predetermined number of beats; determine an absolute value of a difference between the mean, median, or mode of the amplitudes of the non-noisy beats and the determined T-wave; and determine whether the absolute value of the difference is greater than the predetermined difference threshold.

[0180] Example 9. The apparatus of any combination of examples 1-8, wherein the processing circuitry is further configured to: determine a confidence of the determined T-wave.

[0181] Example 10. The apparatus of example 9, wherein the confidence is based on a flatness of the cardiac signal around the determined T-wave.

[0182] Example 11. The apparatus of example 10, wherein the processing circuitry is configured to determine the flatness, wherein as part of determining the flatness, the processing circuitry is configured to: determine a mapped T-wave location; determine a maximum amplitude and a minimum amplitude of amplitudes of a plurality of samples of the cardiac signal, a first sample of the plurality of samples being before the mapped T-wave location and a last sample of the plurality of samples being after the mapped T-wave location; and determine whether an absolute value of a difference between the maximum amplitude and the minimum amplitude is greater than or equal to a first predetermined confidence threshold, the absolute value of the difference is less than or equal to a second predetermined confidence threshold, or the absolute value of the difference is between the first predetermined confidence threshold and the second predetermined confidence threshold.

[0183] Example 12. A method comprising: determining, by processing circuitry, an R-wave of a cardiac signal; determining, by the processing circuitry, whether the R-wave is noisy; determining, by the processing circuitry, whether the cardiac signal around a determined T-wave is noisy based on the R-wave not being noisy; and determining, by the processing circuitry, a QT interval or a corrected QT interval based on the determined T-wave and a determined R-wave based on the cardiac signal around the determined T-wave not being noisy.

[0184] Example 13. The method of example 12, wherein determining whether the R-wave is noisy comprises: determining, by the processing circuitry, whether a rate of change of amplitudes of a plurality of samples of the cardiac signal around an R-wave peak indicates that the R-wave is noisy.

[0185] Example 14. The method of Example 13, wherein determining whether the rate of change of the amplitudes of the plurality of samples of the cardiac signal around the R-wave peak indicates that the R-wave is noisy comprises: determining, by the processing circuitry, a peak value of the R-wave; determining, by the processing circuitry, a plurality of first differences between a first plurality of samples of the cardiac signal, a first sample of the first plurality of samples being before the R-wave peak and a last sample of the first plurality of samples being after the R-wave peak, each of the plurality of first differences being a difference in amplitude of a respective sample of the first plurality of samples and a next sample of the first plurality of samples; determining, by the processing circuitry, a plurality of second differences between each of the plurality of first differences and a respective next difference of the plurality of first differences; determining, by the processing circuitry, whether there is a sign change between each of the plurality of first differences and the respective next difference of the plurality of first differences; determining, by the processing circuitry for each determined sign change, whether a magnitude of a corresponding respective second difference value is greater than a first positive predetermined noise threshold or below a first negative predetermined noise threshold; calculating, by the processing circuitry, a total number of the respective second difference values having the magnitude greater than the positive first predetermined noise threshold or below the negative first predetermined noise threshold; and determining, by the processing circuitry, whether the calculated total number is greater than a second predetermined noise threshold.

[0186] Example 15. The method of Example 14, wherein determining the R-wave peak comprises: determining, by the processing circuitry, a mean, median, or mode of amplitudes of a plurality of second samples of the cardiac signal, a first sample of the second plurality of samples being before the determined R-wave and a last sample of the second plurality of samples being after the determined R-wave; determining, by the processing circuitry, a respective mean, median, or mode difference between the mean, median, or mode and each of the plurality of second samples; determining, by the processing circuitry, an absolute value of each of the respective mean, median, or mode differences; and determining, by the processing circuitry, the R-wave peak to be a highest absolute value of each of the respective mean, median, or mode differences.

[0187] Example 16. The method of any combination of examples 12-15, wherein determining whether the ECG around the determined T-wave is noisy comprises: determining, by the processing circuitry, a plurality of third differences between a third plurality of samples of the cardiac signal, a first sample of the third plurality of samples being before the determined T-wave and a last sample of the third plurality of samples being after the determined T-wave, each of the plurality of third differences being a difference in amplitude of a respective sample of the third plurality of samples and a next sample of the third plurality of samples; determining, by the processing circuitry, a plurality of fourth differences between each of the plurality of third differences and a respective next difference of the plurality of third differences; determining, by the processing circuitry, whether a sign change exists between each of the plurality of third differences and the respective next difference of the plurality of third differences; determining, by the processing circuitry for each determined sign change, whether a magnitude of a corresponding respective fourth difference value is greater than a positive third predetermined noise threshold or below a negative third predetermined noise threshold; calculating, by the processing circuitry, a total number of the respective second difference values having the magnitude greater than the positive third predetermined noise threshold or below the negative third predetermined noise threshold; and determining, by the processing circuitry, whether the calculated total number is greater than a fourth predetermined noise threshold.

[0188] Example 17. The method of any combination of examples 12-16, further comprising: determining, by the processing circuitry, whether at least one feature of the determined T-wave exceeds a predetermined difference threshold from a mean, median, or mode of the at least one feature of a plurality of other T-waves, wherein the determination of the QT interval is further based on the determined T-wave not exceeding the predetermined difference threshold from the mean, median, or mode of the at least one feature of the plurality of other T-waves.

[0189] Example 18. The method of example 17, wherein determining whether the determined T-wave exceeds the predetermined difference threshold from a plurality of other T-waves comprises: determining, by the processing circuitry, whether a number of noisy beats in a predetermined number of beats is greater than a predetermined noisy beat threshold, a first beat of the predetermined number of beats being before the determined T-wave and a last beat of the predetermined number of beats being after the determined T-wave.

[0190] Example 19. The method of example 18, wherein determining whether the determined T-wave exceeds the predetermined difference threshold from a plurality of other T-waves comprises: determining, by the processing circuitry, a mean, median, or mode of amplitudes of beats that are not noisy in the predetermined number of beats; and determining, by the processing circuitry, an absolute value of a difference between the mean, median, or mode of the amplitudes of the beats that are not noisy and the determined T-wave; and determining, by the processing circuitry, whether the absolute value of the difference is greater than the predetermined difference threshold.

[0191] Example 20. The method of any combination of examples 12-19, further comprising determining, by the processing circuitry, a confidence of the determined T-wave.

[0192] Example 21. The method of example 20, wherein the confidence is based on a flatness of the cardiac signal around the determined T-wave.

[0193] Example 22. The method of example 21, further comprising: determining, by the processing circuitry, the flatness, wherein determining the flatness comprises: determining, by the processing circuitry, a mapped T-wave location; determining, by the processing circuitry, a maximum amplitude and a minimum amplitude of amplitudes of a fourth plurality of samples of the cardiac signal, a first sample of the fourth plurality of samples being before the mapped T-wave location and a last sample of the fourth plurality of samples being after the mapped T-wave location; and determining, by the processing circuitry, whether an absolute value of a difference between the maximum amplitude and the minimum amplitude is greater than or equal to a first predetermined confidence threshold, whether the absolute value of the difference is less than or equal to a second predetermined confidence threshold, or whether the absolute value of the difference is between the first predetermined confidence threshold and the second predetermined confidence threshold.

[0194] Example 23. A non-transitory computer-readable storage medium storing a set of instructions, which when executed, cause a system to: determine an R-wave of the cardiac signal; determine whether the R-wave is noisy; based on the R-wave not being noisy, determine whether the cardiac signal around the determined T-wave is noisy; and based on the cardiac signal around the determined T-wave not being noisy, determine a QT interval or a corrected QT interval based on the determined T-wave and a determined R-wave.

[0195] Various examples have been described. These and other examples are within the scope of the following claims.

Claims

1. An apparatus for detection of a QT interval or a corrected QT interval in a cardiac signal, comprising: one or more electrodes; sensing circuitry configured to sense the cardiac signal via the one or more electrodes; and processing circuitry configured to: determine an R wave of the cardiac signal; determine whether the R wave is noisy, wherein as part of determining whether the R wave is noisy, the processing circuitry is configured to determine whether a rate of change of amplitudes of a plurality of samples of the cardiac signal around an R wave peak indicates that the R wave is noisy; based on the R wave not being noisy, determine whether the cardiac signal around a determined T wave is noisy; and based on the cardiac signal around the determined T wave not being noisy, determine the QT interval or the corrected QT interval based on the determined T wave and the determined R wave.

2. The apparatus of claim 1, wherein as part of determining whether a rate of change of amplitudes of a plurality of samples of the cardiac signal around an R wave peak indicates that the R wave is noisy, the processing circuitry is configured to: determine an R wave peak; determine a plurality of first differences between a first plurality of samples of the cardiac signal, a first sample of the first plurality of samples being before the R wave peak and a last sample of the first plurality of samples being after the R wave peak, each of the plurality of first differences being a difference in amplitude of a respective sample of the first plurality of samples and a next sample of the first plurality of samples; determine a plurality of second differences between each of the plurality of first differences and a respective next difference of the plurality of first differences; determine whether a sign change exists between each of the plurality of first differences and the respective next difference of the plurality of first differences; for each determined sign change, determine whether a magnitude of a corresponding respective second difference value is greater than a first positive predetermined noise threshold or lower than a first negative predetermined noise threshold; calculate a total number of the respective second difference values having the magnitude greater than the positive first predetermined noise threshold or lower than the negative first predetermined noise threshold; and determine whether the calculated total number is greater than a second predetermined noise threshold.

3. The apparatus of claim 2, wherein as part of determining the R wave peak, the processing circuitry is configured to: determine a mean, median, or mode of amplitudes of a plurality of second samples of the cardiac signal, a first sample of the plurality of second samples being before the determined R wave and a last sample of the plurality of second samples being after the determined R wave; determine a respective mean, median, or mode difference between the mean, median, or mode and each of the plurality of second samples; determine an absolute value of each of the respective mean, median, or mode differences; and determine the R wave peak to be a highest absolute value of each of the respective mean, median, or mode differences.

4. The apparatus of claim 1, wherein as part of determining whether the cardiac signal around the determined T wave is noisy, the processing circuitry is configured to: determining a third plurality of differences between samples of the cardiac signal, a first sample of the third plurality of samples being before a mapped T-wave and a last sample of the third plurality of samples being after the mapped T-wave, each of the third plurality of differences being a difference in amplitude of a respective sample of the third plurality of samples and a next sample of the third plurality of samples; determining a fourth plurality of differences between each of the third plurality of differences and a respective next difference of the third plurality of differences; determining whether there is a sign change between each of the third plurality of differences and the respective next difference of the third plurality of differences; for each determined sign change, determining whether a magnitude of a corresponding respective fourth difference value is greater than a positive third predetermined noise threshold or lower than a negative third predetermined noise threshold; calculating a total number of respective second difference values having the magnitude greater than the positive third predetermined noise threshold or lower than the negative third predetermined noise threshold; and determining whether the calculated total number is greater than a fourth predetermined noise threshold.

5. The device of claim 1, wherein the processing circuitry is further configured to: determine whether at least one feature of the determined T-wave exceeds a predetermined difference threshold from a mean, median, or mode of the at least one feature of a plurality of other T-waves, wherein the determination of the QT interval is further based on the determined T-wave not exceeding the predetermined difference threshold from the mean, median, or mode of the at least one feature of the plurality of other T-waves.

6. The device of claim 5, wherein as part of determining whether the determined T-wave exceeds the predetermined difference threshold from a plurality of other T-waves, the processing circuitry is configured to: determine whether a number of noisy beats of a predetermined number of beats is greater than a predetermined noisy beat threshold, a first beat of the predetermined number of beats being before the determined T-wave and a last beat of the predetermined number of beats being after the determined T-wave.

7. The device of claim 6, wherein as part of determining whether the determined T-wave exceeds the predetermined difference threshold from a plurality of other T-waves, the processing circuitry is configured to: determine a mean, median, or mode of amplitudes of beats of the predetermined number of beats that are not noisy; determine an absolute value of a difference between the mean, median, or mode of the amplitudes of the beats that are not noisy and the determined T-wave; and determine whether the absolute value of the difference is greater than the predetermined difference threshold.

8. The device of claim 1, wherein the processing circuitry is further configured to: determine a confidence of the determined T-wave.

9. The device of claim 8, wherein the confidence is based on a flatness of the cardiac signal around the determined T-wave.

10. The device of claim 9, wherein the processing circuitry is configured to determine the flatness, wherein as part of determining the flatness, the processing circuitry is configured to: determine a mapped T-wave location; determining a maximum amplitude and a minimum amplitude of amplitudes of a fourth plurality of samples of the cardiac signal, a first sample of the fourth plurality of samples being before the mapped T-wave location and a last sample of the fourth plurality of samples being after the mapped T-wave location; and determining whether an absolute value of a difference between the maximum amplitude and the minimum amplitude is greater than or equal to a first predetermined confidence threshold, the absolute value of the difference is less than or equal to a second predetermined confidence threshold, or the absolute value of the difference is between the first predetermined confidence threshold and the second predetermined confidence threshold.

11. A non-transitory computer-readable storage medium storing a set of instructions, the set of instructions, when executed, cause a system to: determine an R-wave of a cardiac signal; determine whether the R-wave is noisy, wherein determining whether the R-wave is noisy comprises determining a rate of change of amplitudes of a plurality of samples of the cardiac signal around an R-wave peak whether the R-wave is noisy; based on the R-wave not being noisy, determine whether the cardiac signal around a determined T-wave is noisy; and based on the cardiac signal around the determined T-wave not being noisy, determine a QT interval or a corrected QT interval based on the determined T-wave and a determined R-wave.

Images