Method and system for characterizing cardiovascular systems from single channel data

• US 10,441,216 B2
• Filed: 03/20/2018
• Issued: 10/15/2019
• Est. Priority Date: 08/17/2012
• Status: Active Grant

• Current Assignee: Analytics For Life
• Original Assignee: Analytics For Life, Inc.
• Inventors: Burton, Timothy, Ramchandani, Shyamlal, Howe-Patterson, Matthew, Yazdi, Mohsen, Gupta, Sunny
• Primary Examiner(s): Patel, Niketa I
• Assistant Examiner(s): Levicky, William J
• Application Number:
  US15/926,290
• Publication Number:
  US 20180206787A1
• Time in Patent Office: 574 Days
• CPC Class Codes: A61B 5/02028 Determining haemodynamic parameters not otherwise provided for, e.g. cardiac contractility or left ventricular ejection fraction
• CPC Class Codes: A61B 5/316 Modalities, i.e. specific diagnostic methods
• CPC Class Codes: A61B 5/318 Heart-related electrical modalities, e.g. electrocardiography [ECG]
• CPC Class Codes: A61B 5/341 Vectorcardiography [VCG]
• CPC Class Codes: A61B 5/364 Detecting abnormal ECG interval, e.g. extrasystoles, ectopic heartbeats
• CPC Class Codes: A61B 5/4848 Monitoring or testing the effects of treatment, e.g. of medication
• CPC Class Codes: A61B 5/6804 Garments; Clothes
• CPC Class Codes: A61B 5/681 Wristwatch-type devices
• CPC Class Codes: A61B 5/7203 for noise prevention, reduction or removal
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: A61B 5/7267 involving training the classification device
• CPC Class Codes: A61B 5/7275 Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders
• CPC Class Codes: G16Z 99/00 Subject matter not provided for in other main groups of this subclass

Abstract

Methods to identify and risk stratify disease states, cardiac structural defects, functional cardiac deficiencies induced by teratogens and other toxic agents, pathological substrates, conduction delays and defects, and ejection fraction using single channel biological data obtained from the subject. A modified Matching Pursuit (MP) algorithm may be used to find a noiseless model of the data that is sparse and does not assume periodicity of the signal. After the model is derived, various metrics and subspaces are extracted to characterize the cardiac system. In another method, space-time domain is divided into a number of regions (which is largely determined by the signal length), the density of the signal is computed in each region and input to a learning algorithm to associate them to the desired cardiac dysfunction indicator target.

Claims

1. A method for detecting a cardiac dysfunction using a computing device, comprising:

receiving only one-dimensional electrophysiological data;

constructing multi-dimensional electrophysiological data from the only one-dimensional electrophysiological data;

defining a phase space system from the multi-dimensional electrophysiological data;

detecting the cardiac dysfunction from the phase space system; and

providing an indicator of the cardiac dysfunction.

2. The method of claim 1, wherein the step of constructing multi-dimensional electrophysiological data comprises:

denoising the one-dimensional electrophysiological data to generate denoised one-dimensional data;

removing a baseline of the denoised one-dimensional electrophysiological data; and

taking fractional derivatives of the denoised one-dimensional electrophysiological data to construct the multi-dimensional electrophysiological data.

3. The method of claim 1, wherein the one-dimensional electrophysiological data comprises at least one cardiac cycle.

4. The method of claim 3, wherein the cardiac cycle corresponds to a vector sum of electrical activation of the heart.

5. The method of claim 4, wherein the step of detecting the cardiac dysfunction from the phase space system further comprises:

using variables from an N dimensional space-time information to determine feature vectors; and

correlating the feature vectors with attributes that are associated with the cardiac dysfunction.

6. The method of claim 5, wherein the step of correlating is performed using one or more machine learning algorithms.

7. The method of claim 6, wherein the one or more machine learning algorithms includes a genetic algorithm mediated map generation or artificial neural network.

8. The method of claim 5, further comprising:

dividing the N dimensional space-time information into a predetermined number of regions;

determining a density of a time series of each region; and

using the density of each region in the step of correlating.

9. The method of claim 1, wherein the cardiac dysfunction is at least one of disease states, cardiac structural defects, functional cardiac deficiencies induced by teratogens and toxic agents, pathological substrates, conduction delays and defects, and ejection fraction.

10. The method of claim 1, wherein the step of providing an indicator of the cardiac dysfunction further comprises generating a 3-D phase space plot containing the indicator.

11. The method of claim 10, wherein the indicator is a complex subharmonic frequency (CSF) trajectory that is associated with the cardiac dysfunction.

12. The method of claim 1, further comprising the step of receiving the one-dimensional electrophysiological data from one of a one-dimensional recorder, an implantable telemeter, a smartphone, a smart handheld consumer device, a smart watch, a perfusion sensor, and clothing embedded with biometrics sensors.

13. A method for detecting a cardiac dysfunction using a computing device, comprising:

receiving only one-dimensional electrophysiological data;

constructing multi-dimensional electrophysiological data from the only one-dimensional electrophysiological data;

defining a phase space system from the multi-dimensional electrophysiological data;

determining by a baseline measurement of a subject'"'"'s physiological condition from the phase space system;

determining a first risk score associated with the baseline measurement of the subject'"'"'s physiological condition;

applying a therapy to the subject and subsequently receiving only second one-dimensional electrophysiological data;

constructing second multi-dimensional electrophysiological data from the only second one-dimensional electrophysiological data;

defining a second phase space system from the second multi-dimensional electrophysiological data;

determining a second measurement of the subject'"'"'s physiological condition from the second phase space system after the therapy is applied;

determining a second risk score associated with the second measurement of the subject'"'"'s physiological condition; and

comparing the first risk score and the second risk score to determine a utility of patient therapies.

14. A non-transitory computer readable medium having computer readable instructions stored thereon that when executed by a computing device causes the computing device to perform a method for detecting a cardiac dysfunction, the method comprising:

receiving only one-dimensional electrophysiological data;

constructing multi-dimensional electrophysiological data from the only one-dimensional electrophysiological data;

defining a phase space system from the multi-dimensional electrophysiological data;

detecting the cardiac dysfunction from the phase space system; and

providing an indicator of the cardiac dysfunction.

15. The non-transitory computer readable medium of claim 14, further comprising instructions for:

denoising the one-dimensional electrophysiological data to generate denoised one-dimensional data;

removing a baseline of the denoised one-dimensional electrophysiological data; and

constructing the multi-dimensional electrophysiological data by taking fractional derivatives of the denoised one-dimensional electrophysiological data.

16. The non-transitory computer readable medium of claim 14, further comprising instructions for:

using variables from an N dimensional space-time information to determine feature vectors; and

correlating the feature vectors with attributes that are associated with the cardiac dysfunction.

17. The non-transitory computer readable medium of claim 16, further comprising instructions for:

dividing the N dimensional space-time information into a predetermined number of regions;

determining a density of a time series of each region; and

using the density of each region in the step of correlating.

18. The non-transitory computer readable medium of claim 14, further comprising instructions for generating a 3-D phase space plot containing the indicator.

19. The non-transitory computer readable medium of claim 14, wherein the cardiac dysfunction is at least one of disease states, cardiac structural defects, functional cardiac deficiencies induced by teratogens and toxic agents, pathological substrates, conduction delays and defects, and ejection fraction.

20. The non-transitory computer readable medium of claim 14, wherein the step of detecting of the cardiac dysfunction is performed using at least one machine learning algorithm.