Systems and methods for transitioning patient care from signal-based monitoring to risk-based monitoring
Inventors
Baronov, Dimitar V. • Butler, Evan J. • Lock, Jesse M. • McManus, Michael F.
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-10796804-B2
Publication Date
2020-10-06
Expiration Date
Abstract
A risk-based patient monitoring system for critical care patients combines data from multiple sources to assess the current and the future risks to the patient, thereby enabling providers to review a current patient risk profile and to continuously track a clinical trajectory. A physiology observer module in the system utilizes multiple measurements to estimate Probability Density Functions (PDF) of a number of Internal State Variables (ISVs) that describe a components of the physiology relevant to the patient treatment and condition. A clinical trajectory interpreter module in the system utilizes the estimated PDFs of ISVs to identify under which probable patient states the patient can be currently categorized and assign a probability value that the patient will be in each of the identified states. The combination of patient states and their probabilities is defined as the clinical risk to the patient.
Core Innovation
The disclosed system and method perform risk-based monitoring of a critical care patient by substantially continuously acquiring physiological data from a plurality of sensors, including at least a heart rate sensor and an SpO2 sensor, attached to the patient. The computer generates a clinical trajectory described by probabilities of possible patient states over a range of time steps, where the possible patient states cannot be directly measured.
A clinical trajectory interpreter module uses a physiology observer module that estimates probability density functions for internal state variables through a dynamic/observation model. Predicted probability density functions are generated from posterior estimated probability density functions using Bayes theorem and conditional probability density functions of acquired data given internal state variables, and a set of possible states is determined for hidden internal state variables.
The monitoring computes and displays risk-related outputs and interpretive structures derived from the internal-state uncertainty. Hazard levels are determined for possible patient states, and probabilities are displayed as graphical indicators that graphically identify the probability that the patient is in a corresponding patient state at a given point in a range of time, with a timeline controller for dynamic selection of time points to display the evolution of patient-state probabilities.
Claims Coverage
The partial content provides two independent claims. The inventive concepts are centered on estimating and propagating probability density functions for internal state variables with Bayes theorem, deriving probabilities for patient states that cannot be directly measured, and displaying evolving risk-related graphical indicators over time.
Risk-based monitoring with attachable sensors and Bayesian probability density functions
A method of risk-based monitoring of a critical care patient using a plurality of sensors including at least a heart rate sensor and an SpO2 sensor attachable to the patient; substantially continuously acquiring physiological data; substantially continuously estimating a clinical trajectory described by probabilities of possible patient states using predicted probability density functions and posterior predicted probability density functions for internal state variables generated using Bayes theorem and conditional probability density functions; determining possible states of a hidden internal state variable; generating probabilities of possible patient states that cannot be directly measured; and substantially continuously displaying the clinical trajectory on a graphical user interface as probabilities across time steps.
Graphical hazard-level indicators with backward-evolved probability density functions
A computer-implemented method for risk-based monitoring of a patient with a computer, display device, and memory; providing a plurality of physiological sensors including at least a heart rate sensor and an SpO2 sensor physically attached to a critical care patient; acquiring data associated with a plurality of internal state variables and storing it; generating predicted probability density functions at time step t(k) and at a previous time step t(k-1) by evolving backwards; generating posterior probability density functions for internal state variables at the previous time step using Bayes theorem; determining a set of possible states of a hidden internal state variable; generating probabilities of possible patient states that cannot be directly measured; causing display of graphical indicators corresponding to possible patient states with hazard level and probability over a range of time; and causing display of a timeline controller that allows dynamic selection of time points to show evolution over the range of time.
The independent claims collectively cover Bayesian estimation and propagation of probability density functions for internal state variables, derivation of probabilities for possible patient states from hidden internal state variables, and presentation of the resulting risk trajectory via graphical hazard-level indicators that change over time with a timeline controller.
Stated Advantages
Displays probabilities of possible patient states that cannot be directly measured as a function of a plurality of time steps on a graphical user interface.
Associates graphical indicators for possible patient states with hazard levels to support risk-based monitoring.
Supports dynamic exploration of patient-state evolution over time using a timeline controller.
Enables risk-based estimation by generating predicted and posterior predicted probability density functions using Bayes theorem and conditional probability density functions.
Computes measurement utility using Kullback-Leibler divergence and integrates external computations or attributes to expand state space.
Documented Applications
Critical-care modeling for HLHS stage 1 physiology using a dynamic Bayesian network and particle filtering.
Blood transfusion decision support using risk/benefit tradeoffs and patient-state modeling.
Nitric oxide treatment risk stratification using standardized clinical plans.
Outpatient chronic-condition monitoring using dynamic Bayesian network-based state tracking under uncertainty and missing measurements, including ADHD state distributions and measurement modalities under uncertainty.
Interested in licensing this patent?