University of Michigan

Virginia Commonwealth University

The University of North Carolina at Charlotte

Toyota Motor Engineering & Manufacturing North America, Inc.

Other

Denso International America, Inc.

Northern Arizona University

Kayvan Najarian

Kevin R Ward

Jonathan Gryak

Sardar Ansari

Hendrikus Derksen

Pujitha Gunaratne

Ashwin Belle

Sayedmohammadreza Soroushmehr

Zhi Li

Kenn Oldham

Heming Yao

Hasan B Alam

Wei Zhang

Nazmi Peyman

Fatemeh Afghah

Robert Mitchell Baldwin, Ii

Rodney C Daniels

Brahmajee Kartik Nallamothu

Alireza Darvish

Abolfazl Razi

Lu Wang

Ryan W Stidham

Roya Hakimzadeh

Simina Vasilache

Rebecca Smith

Joshua Drews

Kyle Gunnerson

Neriman Tokcan

Daniel Slavin

Wenan Chen

Dennis James Rivet, Ii

Harm Derksen

Jie Wu

Elyas Sabeti

Rosalyn Hobson Hargraves

Sharad Shandilya

John Frederick Reavey Cantwell

Rosalyn Stacy Hobson

Sumeyra Ummuhan Demir Kanik

Nazanin Mirshahi

Soo-Yeon Ji

A method of generating an assessment of medical condition for a patient includes obtaining a patient data tensor indicative of a plurality of tests conducted on the patient, obtaining a set of tensor factors, each tensor factor of the set of tensor factors being indicative of a decomposition of training tensor data for the plurality of tests, the decomposition amplifying low rank structure of the training tensor data, determining a patient tensor factor for the patient based on the obtained patient data tensor and the obtained set of tensor factors, applying the determined patient tensor factor to a classifier such that the determined further tensor factor establishes a feature vector for the patient, the classifier being configured to process the feature vector to generate the assessment, and providing output data indicative of the assessment.

Tensor amplification-based data processing

A system for automatically analyzing a video recording of a colonoscopy includes a processor and memory storing instructions, which when executed by the processor, cause the processor to receive the video recording of the colonoscopy performed on the colon and detect informative frames in the video recording. A frame is informative if the clarity of the frame is above a threshold or if the frame includes clinically relevant information about the colon. The instructions cause the processor to generate scores indicating severity levels of a disease for a plurality of the informative frames, estimate locations of the plurality of the informative frames in the colon, and generate an output indicating a distribution of the scores over one or more segments of the colon by combining the scores generated for the plurality of the informative frames and the estimated locations of the plurality of the informative frames in the colon.

Automated anatomic and regional location of disease features in colonoscopy videos

Computational algorithms integrate and analyze data to consider multiple interdependent, heterogeneous sources and forms of patient data, and using a classification model, provide new learning paradigms, including privileged learning and learning with uncertain clinical data, to determine patient status for conditions such as acute respiratory distress syndrome (ARDS) or non-ARDS.

Sequential minimal optimization algorithm for learning using partially available privileged information

A wearable assembly has a pulse plethysmography (PPG) sensor and a piezoelectric pressure sensor and is attachable to a patient's finger or other area corresponding to a peripheral vascular region, and further includes a signal processor configured to monitor blood flow dependent measurements and pressure measurements over time, comparing these measurements to determine properties of the vascular region, such as vascular resistance of a blood vessel, vascular radius of the blood vessel, vascular stiffness of the vascular region, blood pressure, and/or cardiac vascular power. The signal processor may apply a hysteresis comparison of the sensor outputs, e.g., using an elliptical model, and in some examples may apply an extended Kalman filter for optimizing output of the vascular region properties.

Estimation of peripheral vascular resistance using a miniature piezoelectric sensor

Systems and methods for predicting and/or detecting cardiac events based on real-time biomedical signals are discussed herein. In various embodiments, a machine learning algorithm may be utilized to predict and/or detect one or more medical conditions based on obtained biomedical signals. For example, the systems and methods described herein may utilize ECG signals to predict and detect cardiac events. In various embodiments, patterns identified within a signal may be assigned letters (i.e., encoded as distributions of letters). Based on the known morphology of a signal, states within the signal may be identified based on the distribution of letters in the signal. When applied in the in-vehicle environment, drivers or passengers within the vehicle may be alerted when an individual within the vehicle is, or is about to, experience a cardiac event.

Systems and methods for predicting and detecting a cardiac event

An automated pruning technique is proposed for reducing the size of a convolutional neural network. A large-sized network is trained and then connections between layers are explored to remove redundant parameters. Specifically, a scaling neural subnetwork is connected to the neural network and designed to infer importance of the filters in the neural network during training of the neural network. Output from the scaling neural subnetwork can then be used to remove filters from the neural network, thereby reducing the size of the convolutional neural network.

Automatic filter pruning technique for convolutional neural networks

Implementations of a system for determining the relevancy of a plurality of alarms may include: a plurality of sensors configured to be coupled to a patient, wherein the plurality of sensors is configured to gather physiological data, a medical monitoring device coupled to the plurality of sensors through a telecommunication channel, and wherein the medical monitoring device is configured to determine a physiological state of the patient using the physiological data. The medical monitoring device may be further configured to issue a plurality of alarm states, and a processing unit coupled to the medical monitoring device through a telecommunication channel, is then configured to evaluate the plurality of alarm states, determine the relevancy of each alarm state of the plurality of alarm states, and issue one or more alarms corresponding with each relevant alarm state to a computing device associated with a user.

False alarm reduction systems and related methods

The invention is a passive, wearable sensor that uses a thin piezoelectric material to produce a time history of blood pressure of the patient, with signal processing algorithms to extract physiological information. The sensor consists of a piezoelectric transducer set in a polymer laminate that can be applied to the finger or wrist of the patient. During use, a combination of compressive and bending deformation in the piezoelectric layer in response to blood pressure in the finger or wrist as a voltage output. Using signal processing techniques, the raw signal is filtered and decomposed to obtain a information to form derivative signals such as blood pressure, pulse pressure, pulse pressure variability, heart rate, heart rate variability, and respiratory rate which can be very important pre-cursors in the monitoring of the patient's physiological conditions.

Miniature piezoelectric cardiovascular monitoring system

The present application describes techniques to filter signals contaminated with blunt noise. Calculated filter coefficients may be applied to signals to generate filtered output signals without the blunt noise. Sets of filter coefficients may be calculated utilizing an &#x3b5;-tube filter process in conjunction with an autoregressive exogenous (ARX) model. Sets of filter coefficients may be calculated in accordance with a constrained optimization algorithm using data indicative of a source of the blunt noise. When the blunt noise is modeled in accordance with the ARX model, filtered output signals are generated having amplitudes constrained to a selected Epsilon value, which may be the amplitude of a primary component of the unfiltered signal. A set of filter coefficients may be calculated by determining, from the set of filter coefficients that satisfy the constrained optimization algorithm, a solution that produces a filtered output signal having the most time-invariant frequency composition.

Epsilon-tube filter for blunt noise removal

Vascular pressures and oximetry are optimized through administration of fluids and other support prior to manipulating temperature in a therapeutic hypothermia procedure. Temperature and active hemodynamic management are utilized, through measurement and adjustment of temperature and hemodynamic parameters through the use of an internal cooling device, such as endovascular catheter, and/or an external temperature-altering device, to ensure that temperature management is optimized, and hemodynamics and tissue resuscitation are optimized prior to and during temperature manipulation.

Methods and apparatus for optimizing therapeutic temperature control

A decision-support system and computer implemented method automatically measures the midline shift in a patient's brain using Computed Tomography (CT) images. The decision-support system and computer implemented method applies machine learning methods to features extracted from multiple sources, including midline shift, blood amount, texture pattern and other injury data, to provide a physician an estimate of intracranial pressure (ICP) levels. A hierarchical segmentation method, based on Gaussian Mixture Model (GMM), is used. In this approach, first an Magnetic Resonance Image (MRI) ventricle template, as prior knowledge, is used to estimate the region for each ventricle. Then, by matching the ventricle shape in CT images to the MRI ventricle template set, the corresponding MRI slice is selected. From the shape matching result, the feature points for midline estimation in CT slices, such as the center edge points of the lateral ventricles, are detected. The amount of shift, along with other information such as brain tissue texture features, volume of blood accumulated in the brain, patient demographics, injury information, and features extracted from physiological signals, are used to train a machine learning method to predict a variety of important clinical factors, such as intracranial pressure (ICP), likelihood of success a particular treatment, and the need and/or dosage of particular drugs.

Automated measurement of brain injury indices using brain CT images, injury data, and machine learning

A system for maintaining permissive hypotension in a patient includes monitoring of blood perfusion and pressure, combining blood perfusion and pressure data to determine systolic blood pressure, monitoring for detection of systolic blood pressure below a critical threshold value, and providing an alert to perfusion options to maintain permissive hypotension. The system may be adapted to adjust the critical threshold value for systolic blood pressure as treatment progresses.

System for providing continuous systolic blood pressure measurement to maintain permissive hypotension

Techniques for motion artifact (MA) reduction in impedance plethysmography (IP) and other physiological signals are provided. The techniques limit the amplitude of MA filtered signals by imposing an '&#x3b5;-tube.' The techniques may include the introduction of a regularization term to ensure that the pattern of a filtered signal is similar to the pattern of the primary component of the original, unfiltered signal by maximizing the regularity of the filtered signal. The techniques may be integrated into a portable monitoring device, such as an armband, to remove MA from various diagnostic signals and to extract primary signal components for producing enhanced device performance.

Adaptive epsilon-tube filter for blunt noise removal

Growing community of inventors

Northville, MI, United States of America

Kayvan Najarian