Welch Allyn, Inc.

Intel Corporation

Microsoft Technology Licensing, LLC

Protocol Systems, Inc.

Farzin Guilak

Steven D Baker

Daniel L Williams

James P Welch

Anand Sampath

Daniel Lulich

Vivek K De

Narayan Srinivasa

David I Poisner

Sriram R Vangal

Miguel Bautista Gabriel

Sangeeta Ghangam

Doye Chukwuyem Emelue

Roy L Doyal

Hyochan An

Pratik Dasharathkumar Patel

An integrated circuit (IC) is provided for in-situ anomaly detection. Sensors in the IC generates sensor datasets including information indicating conditions in the IC. A processing unit in the IC uses a sensor dataset and a model to detect and classify the anomaly. The processing unit may filter the sensor dataset, extract features from the filtered sensor dataset, and input the features into the model. The model outputs one or more classifications of the anomaly. A feature may be a distance vector that represents a difference between a data value in the filtered sensor dataset from a reference data value. The model may be a network of bit-cells in the IC. The model may be continuously trained in-situ, i.e., on the IC. The processing unit may provide the classifications to another processing unit in the IC. The other processing unit may mitigate the anomaly based on the classifications.

In-situ detection of anomalies in integrated circuits using machine learning models

An edge device monitors a mechanical system for changes in vibration state. When an overall vibration state of the mechanical system deviates by at least a threshold from an expected vibration state, the edge device can send the data to a server for diagnostic processing. The edge device can thus monitor for a pre-failure condition that suggests an upcoming failure or a need for service. With selective sending of sensor data, the bandwidth and processing associated with diagnostics can be reserved for situations more likely indicative of a condition needing diagnosing, rather than performing constant diagnostic processing.

Machine diagnostics based on overall system energy state

A method for communicating data using at least one network linking at least one respondent device with at least one monitoring device, said method comprising the steps of: continuously measuring at least one physiologic parameter for purposes of detecting a predetermined event using said at least one monitoring device; operating said network in an off state in which said network is off except for the periodic transmission of patient status data while said network is in a first operative state wherein said predetermined event has not occurred and transmitting said measured data along said at least one said network in a second state when said predetermined event has occurred.

Personal status physiologic monitor system and architecture and related monitoring methods

A method for performing context management, said method comprising the steps of: producing a continuous physiologic signal, as detected by a monitoring device; associating at least one unique hardware identifier to said continuous physiologic signal and binding a unique patient identifier to said continuous signal wherein a change in said physiologic signal in which said signal is no longer continuous will cause the unique patient identifier to unbind from said signal.

A system for actively monitoring a patient includes at least one body-worn monitoring device that has at least one sensor capable of measuring at least one physiologic parameter and detecting at least one predetermined event. At least one intermediary device is, linked to the body-worn monitoring device by means of a first wireless network and at least one respondent device is linked to said at least one intermediary device by a second wireless network wherein the respondent device is programmed to perform a specified function automatically when the at least one predetermined event is realized. The monitoring device operates to periodically transmit patient status data to the intermediary device but the system predominantly operates in a quiet state, providing very low power consumption.

A method and apparatus for automatically determining salient features for object classification is provided. In accordance with one embodiment, one or more unique features are extracted from a first content group of objects to form a first feature list, and one or more unique features are extracted from a second anti-content group of objects to form a second feature list. A ranked list of features is then created by applying statistical differentiation between unique features of the first feature list and unique features of the second feature list. A set of salient features is then identified from the resulting ranked list of features.

Method and apparatus for automatically determining salient features for object classification

A method and apparatus for efficiently classifying and categorizing data objects such as electronic text, graphics, and audio based documents within very-large-scale hierarchical classification trees is provided. In accordance with one embodiment of the invention, a first node of a plurality of nodes of a subject hierarchy is selected. Previously classified data objects corresponding to a selected first node of a subject hierarchy as well as any associated sub-nodes of the selected node are aggregated to form a content class of data objects. Similarly, data objects corresponding to sibling nodes of the selected node and any associated sub-nodes of the sibling nodes are then aggregated to form an anti-content class of data objects. Features are then extracted from each of the content class of data objects and the anti-content class of data objects to facilitate characterization of said previously classified data objects.

Very-large-scale automatic categorizer for web content

A high-quality, low-bit-rate method of compressing physiological-waveform data obtained from a signal sent from a diagnostic medical device is disclosed. Associated with the signal are time and voltage characteristics. The method includes the steps of (1) examining such signal during a periodic event in which such data occurs, and which data includes a first region dominated by high-frequency information and a second region dominated by low-frequency information, and (2) removing the first region from the event based upon preselected criteria, which criteria are independent of indicia related to the first region. Also included is the step of compressing the data by processing the removed first region according to a first preselected plan which emphasizes low-bit-rate compression, and by processing the data of the entire event according to a second preselected plan which emphasizes high-quality compression. The removing step may further include the substeps of identifying a set of potential plural peaks associated with the first region, and confirming that the potential peaks are the desired ones by using a preselected frequency analysis.

Growing community of inventors

Beaverton, OR, United States of America

Farzin Guilak