Abstract: In situations in which an implantable medical device (e.g., a subcutaneous ICD) is co-implanted with a leadless pacing device (LPD), it may be important that the subcutaneous ICD knows when the LPD is delivering pacing, such as anti-tachycardia pacing (ATP). Techniques are described herein for detecting, with the ICD and based on the sensed electrical signal, pacing pulses and adjusting operation to account for the detected pulses, e.g., blanking the sensed electrical signal or modifying a tachyarrhythmia detection algorithm. In one example, the ICD includes a first pace pulse detector configured to obtain a sensed electrical signal and analyze the sensed electrical signal to detect a first type of pulses having a first set of characteristics and a second pace pulse detector configured to obtain the sensed electrical signal and analyze the sensed electrical signal to detect a second type of pulses having a second set of characteristics.

Abstract: A computer-implemented method includes receiving a neural network model for implementation using a processing element array, where the neural network model includes a convolution operation on a set of input feature maps and a set of filters. The method also includes determining, based on the neural network model, that the convolution operation utilizes less than a threshold number of rows in the processing element array for applying a set of filter elements to the set of input feature maps, where the set of filter elements includes one filter element in each filter of the set of filters. The method further includes generating, for the convolution operation and based on the neural network model, a first instruction and a second instruction for execution by respective rows in the processing element array, where the first instruction and the second instruction use different filter elements of a filter in the set of filters.

Abstract: Techniques are disclosed for automatically calibrating a reference orientation of an implantable medical device (IMD) within a patient. In one example, sensors of an IMD sense a plurality of orientation vectors of the IMD with respect to a gravitational field. Processing circuitry of the IMD processes the plurality of orientation vectors to identify an upright vector that corresponds to an upright posture of the patient. The processing circuitry classifies the plurality of orientation vectors with respect to the upright vector to define a sagittal plane of the patient and a transverse plane of the patient. The processing circuitry determines, based on the upright vector, the sagittal plane, and the transverse plane, a reference orientation of the IMD within the patient. As the orientation of the IMD within the patient changes over time, the processing circuitry may recalibrate its reference orientation and accurately detect a posture of the patient.

Abstract: A method for differentiating heart failure risk scores that includes receiving a current data transmission and acquiring patient metrics from a remote device, determining a daily heart failure risk score for each day occurring during a time period from a previous received data transmission to the current received data transmission based on the acquired patient metrics, determining a maximum daily heart failure risk score of the determined daily heart failure risk scores during a lookback window prior to the current received data transmission, determining a heart failure risk status alert for the received data transmission based on the temporal proximity of the determined maximum heart failure risk score and receipt of the current data transmission, selecting a type of notification based on the heart failure risk status differentiation, and indicating the transmission heart failure risk status and the heart failure risk status differentiation via the selected type of notification.

Abstract: Methods and systems for seamless adjustment of treatment are disclosed. A determination is made as to whether to intervene with a patient's treatment. Implanted device memory data is acquired over a pre-specified time period. Risk status is determined from the device memory data. Another external device memory data is acquired over a pre-specified time period. A determination is made as to whether to adjust treatment of the patient in response to the risk status, the data acquired from the implanted device memory and the external device memory data.

Abstract: An implantable medical device comprises a sensing module configured to obtain electrical signals from one or more electrodes and a control module configured to process the electrical signals from the sensing module in accordance with a tachyarrhythmia detection algorithm to monitor for a tachyarrhythmia. The control module detects initiation of a pacing train delivered by a second implantable medical device, determines a type of the detected pacing train, and modifies the tachyarrhythmia detection algorithm based on the type of the detected pacing train.

Abstract: Some aspects relate to systems, devices, and methods of delivering rate responsive pacing therapy. The method includes monitoring activity information related to an activity level of a patient and delivering rate responsive pacing (RRP) to the patient at a pacing rate corresponding to a RRP profile. The RRP profile may be used to generate the pacing rate based on the activity information and may be adjusted based on the monitored activity information.

Abstract: A medical device is configured to produce an accelerometer signal and detect a patient fall from the accelerometer signal. The device generates a body posture signal and a body acceleration signal from the accelerometer signal and detects a patient fall in response to determining that the body posture signal and the body acceleration signal meet fall detection criteria. The medical device is configured to receive a truth signal from another device that is not the medical device. The truth signal may indicate that the detected patient fall is a falsely detected patient fall and, responsive to receiving the truth signal, the medical device adjusts at least one fall detection control parameter.

Abstract: An adaptive treatment management system includes a sensor system, a treatment optimization system, and a treatment delivery system. The treatment optimization system may update a treatment regimen multiple times between physician visits to facilitate optimization of treatment. The treatment optimization system may determine a treatment regimen based on at least one of patient data, treatment compliance data, and related treatment data. Patient data may be provided by the sensor system. Treatment compliance data may be provided by the treatment delivery system. Related treatment data may be provided by a remote system.

Abstract: Examples include techniques for backing up a file to long term “cold” storage by using circuitry, and logic for execution by the circuitry, to receive a request to back up the file in a distributed file system to cold storage, to copy the file from at least one data node of the distributed file system to cold storage, to set a location of the file in cold storage in a name node of the distributed file system, and to set a length of the file to zero in the name node.

Abstract: An implantable medical device comprises a sensing module configured to obtain electrical signals from one or more electrodes and a control module configured to process the electrical signals from the sensing module in accordance with a tachyarrhythmia detection algorithm to monitor for a tachyarrhythmia. The control module detects initiation of a pacing train delivered by a second implantable medical device, determines a type of the detected pacing train, and modifies the tachyarrhythmia detection algorithm based on the type of the detected pacing train.

Abstract: An implantable monitoring device is disclosed for monitoring a patient's heart rate variability over time. The device includes a cardiac electrogram amplifier, a sensing electrode coupled to an input of the amplifier, timing circuitry, processing circuitry and a memory. The timing circuitry defines successive shorter time periods during each monitoring period. The processing circuitry relies upon electrogram activity that occurs during rest periods that extend as long as T1, all of which is stored into memory. Active periods are not considered as part of the heart rate variability calculation. The processing circuitry calculates median intervals between depolarizations of the patient's heart sensed by the amplifier during the shorter time periods and calculates a standard deviation of the median intervals during T2, a longer monitoring period.

Abstract: A method and device for differentiating heart failure risk scores that includes determining receipt of a current data transmission and acquiring patient metrics from a remote monitoring device, determining a daily heart failure risk score for each day occurring during a time period from a previous received data transmission to the current received data transmission based on the acquired patient metrics, and determining a maximum daily heart failure risk score of the determined daily heart failure risk scores during a lookback window prior to the current received data transmission. A heart failure risk score is determined for the received data transmission based on the determined maximum daily heart failure risk score, and a heart failure risk score alert is determined for the received data transmission based on the proximity of the determined maximum heart failure risk score and the current received data transmission.

Abstract: A quaternion deep neural network (QTDNN) includes a plurality of modular hidden layers, each comprising a set of QT computation sublayers, including a quaternion (QT) general matrix multiplication sublayer, a QT non-linear activations sublayer, and a QT sampling sublayer arranged along a forward signal propagation path. Each QT computation sublayer of the set has a plurality of QT computation engines. In each modular hidden layer, a steering sublayer precedes each of the QT computation sublayers along the forward signal propagation path. The steering sublayer directs a forward-propagating quaternion-valued signal to a selected at least one QT computation engine of a next QT computation subsequent sublayer.

Abstract: Systems and methods include differential diagnosis for acute heart failure to provide treatment to a patient including determining whether the patient has cardiac volume overload, determining whether the patient has decreased abdominal venous system volume, and providing the appropriate treatment in response to the determinations. A multi-sensor system may be used to determine cardiac volume and abdominal venous system volume. Fluid redistribution treatment may be provided when cardiac volume overload is accompanied by a decrease in abdominal venous system volume. Fluid accumulation treatment may be provided when cardiac volume overload is not accompanied by a decrease in abdominal venous system volume.

Abstract: In some examples, a system comprises one or more medical devices configured to, for each of a plurality of periods, determine a respective value for each of a plurality of parameters of a patient and processing circuitry. The processing circuitry can for each of the plurality of periods, execute an algorithm to determine at least one of a heart failure risk score or status for the patient based on the determined values for the period, determine a number of the determined heart failure risk scores or statuses that were false determinations, determine that the number of false determinations for the patient satisfies a false determination threshold, and modify the algorithm for the patient in response to the determination that the number of false determinations for the patient satisfies the false determination threshold.

Abstract: A deep neural network (DNN) includes hidden layers arranged along a forward propagation path between an input layer and an output layer. The input layer accepts training data comprising quaternion values, outputs a quaternion-valued signal along the forward path to at least one of the hidden layers. At least some of the hidden layers include quaternion layers to execute consistent quaternion (QT) forward operations based on one or more variable parameters. A loss function engine produces a loss function representing an error between the DNN result and an expected result. QT backpropagation-based training operations include computing layer-wise QT partial derivatives, consistent with an orthogonal basis of quaternion space, of the loss function with respect to a QT conjugate of the one or more variable parameters and of respective inputs to the quaternion layers.

Abstract: Provided is a method, system and/or apparatus for determining prospective heart failure event risk. Acquired from a device memory are a heart failure patient's current and preceding risk assessment periods. Counting detected data observations in the current risk assessment period for a current risk assessment total amount and counting detected data observations in the preceding risk assessment period for a preceding risk assessment period total amount. Associating the current risk assessment and preceding risk assessment total amounts with a lookup table to acquire prospective risk of heart failure (HF) event for the preceding risk assessment period and the current risk assessment period. Employing weighted sums of the prospective risk of the HF event for the preceding risk assessment period and the current risk assessment period to calculate a weighted prospective risk of the HF event for a patient. Displaying on a graphical user interface the weighted prospective risk of the HF event for the patient.

Abstract: A machine-learning system includes a quaternion (QT) computation engine. Input data to the QT computation engine includes quaternion values, each comprising a real component and three imaginary components, represented as a set of real-valued tensors. A single quaternion value is represented as a 1-dimensional real-valued tensor having four real-valued components, wherein a first real-valued component represents the real component of the single quaternion value, and wherein a second, a third, and a fourth real-valued component each respectively represents one of the imaginary components. A quaternion-valued vector having a size N is represented as a 2-dimensional real-valued tensor comprising N 1-dimensional real-valued tensors. A quaternion-valued matrix having N×M dimensions is represented as a 3-dimensional real-valued tensor comprising M 2-dimensional real-valued tensors comprising N 1-dimensional real-valued tensors.

Abstract: Various systems and methods for implementing context-based digital signal processing are described herein. An object detection system includes a processor to: access sensor data from a first sensor and a second sensor integrated in a vehicle; access an operating context of the vehicle; assign a first weight to a first object detection result from sensor data of the first sensor, the first weight adjusted based on the operating context; assign a second weight to a second object detection result from sensor data of the second sensor, the second weight adjusted based on the operating context; and perform a combined object detection technique by combining the first object detection result weighted by the first weight and the second object detection result weighted by the second weight.