System and method for improved cardiac rhythm classification from the time between heart beat intervals using non-linear dynamics

Abstract

A system for classifying cardiac rhythms is disclosed. The system includes one or more processors, one or more computer-readable tangible storage devices, and program instructions stored on at least one of the one or more storage devices for execution by at least one of the one or more processors. The program instructions include first program instructions times between heartbeats. The program instructions further comprise second program instructions to segment the time series into a plurality of segments. The program instructions further comprise third program instructions to calculate a plurality of parameters corresponding to each of the 30-second segments. The program instructions further comprise fourth program instructions to analyze the obtained data and the calculated parameters using a plurality of multivariable algorithms for rhythm classification. The program instructions further comprise fifth program instructions to synthesize the results of the plurality of multivariable algorithms to formulate a single rhythm classification.

1 - 10 . (canceled) 
     
     
         11 . A system for configured to classify cardiac rhythms, the system comprising a processor, a computer-readable tangible storage device, and program instructions stored on the computer-readable tangible storage device for execution by the processor, the program instructions comprising:
 first program instructions to obtain data in form of an RR time interval series of the times between heartbeats of a patient;   second program instructions to segment the time series into a plurality of segments;   third program instructions to calculate a parameter corresponding to a segment of the plurality of segments, wherein the parameter includes a coefficient of sample entropy, a detrended fluctuation analysis, and/or a local dynamics score;   fourth program instructions to analyze the obtained data and the parameter using a trained multivariable algorithm for rhythm classification so as to obtain a probability estimate of a candidate rhythm in form of normal sinus rhythm, atrial fibrillation and/or sinus rhythm with ectopy; and   fifth program instructions to obtain a rhythm classification based on the probability estimate.   
     
     
         12 . The system of  claim 11 , wherein a time series of the time series is 10 minutes in duration. 
     
     
         13 . The system of  claim 11 , wherein a segment of the plurality of segments includes a 30 second interval. 
     
     
         14 . The system of  claim 11 , wherein the trained multivariable algorithm includes a neural network, a Bayesian network, a decision tree algorithm, a rough set theory algorithm, a logistic regression algorithm, an ordinal regression algorithm, a nearest neighbor analysis, or data mining analysis. 
     
     
         15 . The system of  claim 11 , wherein the third program instructions calculate a parameter corresponding to any one or combination of segments of the plurality of segments. 
     
     
         16 . The system of  claim 11 , wherein the third program instructions calculate a plurality parameters, wherein at least one parameter corresponds to any one or combination of segments of the plurality of segments. 
     
     
         17 . The system of  claim 11 , wherein the third program instructions calculate a plurality parameters, wherein any one or combination of parameters corresponds to each segment of the plurality of segments. 
     
     
         18 . A computer program product to be executed on a computer for classifying cardiac rhythms, the computer program product including a computer-readable tangible storage device and program instructions stored on the computer-readable tangible storage device, the program instructions comprising:
 first program instructions to obtain data representative of an RR time interval series of the times between heartbeats of a patient;   second program instructions to segment the time series into a plurality of segments;   third program instructions to calculate a parameter corresponding to a segment of the plurality of segments, wherein the parameter includes a coefficient of sample entropy, a detrended fluctuation analysis, and/or a local dynamics score;   fourth program instructions to analyze the obtained data and the parameter using a trained multivariable algorithm for rhythm classification so as to obtain a probability estimate of a candidate rhythm in form of normal sinus rhythm, atrial fibrillation, and/or sinus rhythm with ectopy; and   fifth program instructions to obtain a rhythm classification based on the probability estimate.   
     
     
         19 . The computer program product of  claim 18 , wherein the time series is 10 minutes in duration. 
     
     
         20 . The computer program product of  claim 18 , wherein a segment of the plurality of segments includes a 30 second interval. 
     
     
         21 . The computer program product of  claim 18 , wherein the trained multivariable algorithm includes a neural network, a Bayesian network, a decision tree algorithm, a rough set theory algorithm, a logistic regression algorithm, an ordinal regression algorithm, a nearest neighbor analysis, or data mining analysis. 
     
     
         22 . The computer program product of  claim 18 , wherein the third program instructions calculate a parameter corresponding to any one or combination of segments of the plurality of segments. 
     
     
         23 . The computer program product of  claim 18 , wherein the third program instructions calculate a plurality parameters, wherein at least one parameter corresponds to any one or combination of segments of the plurality of segments. 
     
     
         24 . The computer program product of  claim 18 , wherein the third program instructions calculate a plurality parameters, wherein any one or combination of parameters corresponds to each segment of the plurality of segments. 
     
     
         25 . A system for detecting arrhythmia, the system comprising:
 a heart rhythm recording device including electrocardiogram (ECG) leads and/or a digital recording device;   a classifier module comprising a processor, a computer-readable tangible storage device, and program instructions stored on the computer-readable tangible storage device for execution by the processor, the program instructions comprising:
 first program instructions to cause the processor to obtain data related to a time series of times between heartbeats from the heart rhythm recording device; 
 second program instructions to cause the processor to segment the time series into a plurality of segments; 
 third program instructions to cause the processor to calculate a parameter corresponding to a segment, the parameter including any one or combination of:
 an entropy measure that detects atrial fibrillation in interval time series data by counting how many template patterns repeat themselves; 
 a quantitative measure of fractal-like scaling properties of interval time series data; or 
 a local dynamics measure that determines how often individual templates in interval time series data match each other; 
 
 fourth program instructions to cause the processor to execute a trained multivariable algorithm for determining a probability that one or more of sinus rhythm, premature atrial contractions, premature ventricular contractions, atrial fibrillation, or normal sinus rhythm is a predictor for rhythm classification; and 
 fifth program instructions to cause the processor to formulate a rhythm classification based on predictor probability; and 
   a processing device to receive the rhythm classification and detect arrhythmia based on the rhythm classification.   
     
     
         26 . The system of  claim 25 , wherein a time series of the time series is 10 minutes in duration. 
     
     
         27 . The system of  claim 25 , wherein a segment of the plurality of segments includes a 30 second interval. 
     
     
         28 . The system of  claim 25 , wherein the trained multivariable algorithm includes a neural network, a Bayesian network, a decision tree algorithm, a rough set theory algorithm, a logistic regression algorithm, an ordinal regression algorithm, a nearest neighbor analysis, or data mining analysis. 
     
     
         29 . The system of  claim 25 , wherein the third program instructions calculate a parameter corresponding to any one or combination of segments of the plurality of segments. 
     
     
         30 . The system of  claim 25 , wherein the third program instructions calculate a plurality parameters, wherein at least one parameter corresponds to any one or combination of segments of the plurality of segments. 
     
     
         31 . The system of  claim 25 , wherein the third program instructions calculate a plurality parameters, wherein any one or combination of parameters corresponds to each segment of the plurality of segments.