Forecasting uterine activity
AssigneeCerner Innovation, Inc.
InventorsAndrew Roberts, Sasanka Are, Jeffrey W. Wall
Life of this patent
Estimating the term…
Abstract
A decision support tool is provided for predicting uterine contractions. The predicted contractions are determined from measurements of uterine activity (UA). The contraction forecast may be made using a plurality of trained predictive models. The forecasts are formed with linear regressive models based, at least in part, on UA data from a reference population. The predicted contractions may be predictions of contraction duration, contraction peak, and time between contractions. In this way, these predicted contractions may be used for decision support for care plans during labor, such as increased monitoring and/or modifying the care plan.
Claims22
Non-transitory computer-readable media having computer-executable instructions embodied thereon that when executed, provide a method for a decision support system for patients in labor using uterine activity forecasts, the method comprising:
- receiving a plurality of uterine activity (UA) measurements for a patient in labor, the plurality of UA measurements being acquired over a time span and corresponding to a set of contractions experienced by the patient;
- constructing a time series from the plurality of UA measurements;
- based on the time series of UA measurements corresponding to the set of contractions, forecasting one or more future contractions to occur within a future time interval for the patient using a plurality of models, wherein each model within the plurality of models is trained to predict a UA feature value for each future contraction within the one or more future contractions, wherein the plurality of models comprises at least:
- a first reference model trained with UA data from a reference population to predict at least a first value for a first UA feature of future contractions using the time series of UA measurements;
- and
- a first patient-specific model trained with UA measurements of the patient to predict at least a second value for the first UA feature of future contractions using a time series of UA measurements of the patient including the at least first value of the first UA feature;
- and
- based on the one or more future contractions that are forecasted, initiating a response action for the patient.
A system forecasting uterine activity within a future time interval, the system comprising:
- one or more processors;
- memory storing computer-usable instructions that, when executed by the one or more processors, implement a method comprising:
- receiving a plurality of uterine activity (UA) measurements for a patient in labor, the plurality of UA measurements being acquired over a time span and corresponding to a set of contractions experienced by the patient;
- constructing a time series from the plurality of UA measurements;
- based on the time series of UA measurements corresponding to the set of contractions, forecasting one or more future contractions to occur within a future time interval for the patient using a plurality of trained models, wherein the plurality of trained models comprises at least:
- a first reference model trained with UA data from a reference population to predict at least a first value for a first UA feature of future contractions using the time series of UA measurements;
- and,
- a first patient-specific model trained with UA measurements of the patient to predict at least a second value for the first UA feature of future contractions using a time series of UA measurements of the patient including the at least first value of the first UA feature;
- and
- based on the one or more future contractions that are forecasted, initiating a response action.
A computerized method for forecasting uterine activity within a future time interval, the method comprising:
- receiving a plurality of uterine activity (UA) measurements for a patient in labor, the plurality of UA measurements being acquired over a time span and corresponding to a set of contractions experienced by the patient;
- constructing a time series from the plurality of UA measurements;
- based on the time series of UA measurements corresponding to the set of contractions, forecasting one or more future contractions to occur within a future time interval for the patient using a plurality of models, wherein each model within the plurality of models is trained to predict a UA feature for the future time interval using an adapted peak-over-threshold approach for identifying contractions in the time series, wherein the plurality of models comprises at least:
- a first reference model trained with UA data from a reference population to predict at least a first value for a first UA feature duration of a future contraction using the time series of UA measurements, and
- a first patient-specific model trained with UA measurements of the patient to predict at least a second value for the first UA feature of magnitude of the future contractions using a time series of UA measurements of the patient including the at least first value of the first UA feature;
- and
- based on the one or more future contractions that are forecasted, initiating a response action.
Dependencies are inferred from the claim text.