Medical device systems include processing circuitry configured to acquire sensed cardiac signals associated with cardiac activity of a heart of a patient, and to analyze the sensed cardiac signals to determine if a noise signal is present within the cardiac signals.

Systems and methods for detecting arrhythmias in cardiac activity are provided and include memory to store specific executable instructions. One or more processors are configured to execute the specific executable instructions for obtaining first and second far field cardiac activity (CA) data sets over primary and secondary sensing channels, respectively, in connection with a series of beats. The system detects candidate atrial features from the second CA data set, identifies ventricular features from the first CA data set and utilizes the ventricular features to separate beat segments within the second CA data set. The system automatically iteratively analyzes the beat segments by overlaying an atrial activity search window with the second CA data set and determines whether one or more of the candidate atrial features occur within the atrial activity search window. The system adjusts an atrial sensitivity profile based on whether the atrial activity search window includes the one or more of the candidate atrial features and detects atrial events based on the atrial sensitivity profile.

Systems and methods for detecting arrhythmias in cardiac activity are provided and include memory to store specific executable instructions. One or more processors are configured to execute the specific executable instructions for obtaining first and second far field cardiac activity (CA) data sets over primary and secondary sensing channels, respectively, in connection with a series of beats. The system detects candidate atrial features from the second CA data set, identifies ventricular features from the first CA data set and utilizes the ventricular features to separate beat segments within the second CA data set. The system automatically iteratively analyzes the beat segments by overlaying an atrial activity search window with the second CA data set and determines whether one or more of the candidate atrial features occur within the atrial activity search window. The system adjusts an atrial sensitivity profile based on whether the atrial activity search window includes the one or more of the candidate atrial features and detects atrial events based on the atrial sensitivity profile.

Various methods and systems are provided for analyzing an electrocardiogram (ECG) in real-time using machine learning to identify heartbeats, calculate a cycle length for each heartbeat, and display the cycle length for each heartbeat at a user interface. Waveform morphology of ECG data is continuously learned to identify recurrent signals and generate templates based on recurrent signals, to which ECG data is compared to identify and display heartbeats. Generated templates are continuously updated to reflect changing waveform morphologies.

Various methods and systems are provided for analyzing an electrocardiogram (ECG) in real-time using machine learning to identify heartbeats, calculate a cycle length for each heartbeat, and display the cycle length for each heartbeat at a user interface. Waveform morphology of ECG data is continuously learned to identify recurrent signals and generate templates based on recurrent signals, to which ECG data is compared to identify and display heartbeats. Generated templates are continuously updated to reflect changing waveform morphologies.

A smart hardware security engine using biometric features and hardware-specific features is provided. The smart security engine can combine one or more entropy sources, including individually distinguishable biometric features, and hardware-specific features to perform secret key generation for user registration and authentication. Such hybrid signatures may be distinct from person-to-person (e.g., due to the biometric features) and from device-to-device (e.g., due to the hardware-specific features) while varying over time. Thus, embodiments described herein can be used for personal device authentication as well as secret random key generation, significantly reducing the scope of an attack.

A method of analysing changes in the electrical activity of a patient's heart between a reference state and a test state, the method using a reference data set of electrophysiological data captured from the patient in the reference state and at least one test data set of electrophysiological data captured from the patient in the test state, each data set defining a plurality of electrograms for a respective plurality of spatial locations relative to the heart, the method comprising processing the electrophysiological data by, matching each electrogram in the reference data set to a corresponding electrogram in the at least one test data set to create a pair of electrograms for each of the plurality of spatial locations, and deriving a time delay for each spatial location by calculating the time delay between the electrograms of the pair of matched electrograms for that spatial location.

A method of analysing changes in the electrical activity of a patient's heart between a reference state and a test state, the method using a reference data set of electrophysiological data captured from the patient in the reference state and at least one test data set of electrophysiological data captured from the patient in the test state, each data set defining a plurality of electrograms for a respective plurality of spatial locations relative to the heart, the method comprising processing the electrophysiological data by, matching each electrogram in the reference data set to a corresponding electrogram in the at least one test data set to create a pair of electrograms for each of the plurality of spatial locations, and deriving a time delay for each spatial location by calculating the time delay between the electrograms of the pair of matched electrograms for that spatial location.

Techniques are disclosed for using feature delineation to reduce the impact of machine learning cardiac arrhythmia detection on power consumption of medical devices. In one example, a medical device performs feature-based delineation of cardiac electrogram data sensed from a patient to obtain cardiac features indicative of an episode of arrhythmia in the patient. The medical device determines whether the cardiac features satisfy threshold criteria for application of a machine learning model for verifying the feature-based delineation of the cardiac electrogram data. In response to determining that the cardiac features satisfy the threshold criteria, the medical device applies the machine learning model to the sensed cardiac electrogram data to verify that the episode of arrhythmia has occurred or determine a classification of the episode of arrhythmia.

Techniques are disclosed for using feature delineation to reduce the impact of machine learning cardiac arrhythmia detection on power consumption of medical devices. In one example, a medical device performs feature-based delineation of cardiac electrogram data sensed from a patient to obtain cardiac features indicative of an episode of arrhythmia in the patient. The medical device determines whether the cardiac features satisfy threshold criteria for application of a machine learning model for verifying the feature-based delineation of the cardiac electrogram data. In response to determining that the cardiac features satisfy the threshold criteria, the medical device applies the machine learning model to the sensed cardiac electrogram data to verify that the episode of arrhythmia has occurred or determine a classification of the episode of arrhythmia.