Systems and Methods for Generating and Applying Matrix Images to Monitor Cardiac Disease

Abstract

Systems and methods are provided for monitoring progression of a cardiac disease in a patient by providing cardio-vibrational image matrixes and/or ECG image matrices generated using sensor data supplied by a medical device. In some examples, cardio-vibrational image matrices and/or ECG image matrices are output as image files. In some implementations, systems and methods are provided for using such cardio-vibrational image matrices and/or an ECG image matrices, and/or other clinical information, using machine learning classifiers, to assess cardiac risk in a patient. In some implementations, systems and methods are provided for using cardio-vibrational image matrixes and/or ECG image matrices, and/or other clinical information for real-time analysis of cardiac risk.

What is claimed is: 
     
         1 . (canceled) 
     
     
         2 . A system for monitoring a cardiac condition of a patient using cardio-vibrational image matrix representations of cardio-vibrational signals, the system comprising:
 a wearable medical device comprising at least one vibrational sensor configured for monitoring a heart of the patient;   a non-volatile computer-readable storage medium configured to store a plurality of cardio-vibrational measurements;   at least one machine learning engine, each machine learning engine trained to identify at least one type of a set of types of cardiac conditions including at least an existence of at least one type of arrhythmia condition and a nonexistence of cardiac arrhythmia condition through analyzing cardio-vibrational image matrices; and   a plurality of operations stored as a plurality of computer-executable instructions to a non-transitory computer-readable media and/or encoded in hardware logic, wherein 
 the plurality of operations is configured to, in real-time, 
 obtain the plurality of cardio-vibrational measurements derived from cardio-vibrational signals of the patient collected by the at least one vibrational sensor, 
 store, to the non-volatile computer-readable storage medium, the plurality of cardio-vibrational measurements, 
 create a cardio-vibrational image matrix having a plurality of pixel characteristic values corresponding to the plurality of cardio-vibrational measurements, the cardio-vibrational image matrix graphically representing a time progression of a plurality of adjacent cardiac portions, a same at least one deflection feature being captured in each cardiac portion, and 
 using one or more machine learning engines of the at least one machine learning engine, classify the contents of the cardio-vibrational image matrix as a determined type of the set of types of cardiac conditions; 
   wherein, upon identifying that the determined type corresponds to one of the at least one type of arrhythmia condition, the plurality of operations is configured to initiate an electrical therapy via the wearable medical device.   
     
     
         3 . The system of  claim 2 , wherein the plurality of operations is configured to divide the plurality of cardio-vibrational measurements into each cardiac portion of the plurality of adjacent cardiac portions such that each cardiac portion comprises the same at least one deflection feature. 
     
     
         4 . The system of  claim 2 , wherein the plurality of operations is configured to map parameter values of the plurality of cardio-vibrational measurements to the plurality of pixel characteristic values. 
     
     
         5 . The system of  claim 2 , wherein the plurality of pixel characteristic values comprises between around at least 3 and 16 different colors. 
     
     
         6 . The system of  claim 2 , wherein the same at least one deflection feature is selected from an S1 peak and an S2 peak. 
     
     
         7 . The system of  claim 2 , wherein the wearable medical device comprises a wearable cardioverter defibrillator. 
     
     
         8 . The system of  claim 2 , wherein the at least one type of arrhythmia condition comprises an arrhythmia classification including at least one of a duration, a rate, or a mechanism of arrhythmia. 
     
     
         9 . The system of  claim 2 , wherein the at least one type of arrhythmia condition comprises at least one of a supraventricular tachycardia (SVT), a ventricular tachycardia, ventricular fibrillation, tachycardia, bradycardia, asystole, a heart pause condition, pulseless electrical activity, or atrial fibrillation. 
     
     
         10 . The system of  claim 2 , wherein the wearable medical device comprises the non-transitory computer readable media and/or hardware logic. 
     
     
         11 . The system of  claim 2 , wherein the electrical therapy comprises at least one of a defibrillating shock or a pacing pulse. 
     
     
         12 . The system of  claim 2 , wherein the plurality of operations is configured to, upon identifying the determined type corresponds to one of the at least one type of arrhythmia condition, issue a warning to a clinician, a caretaker, and/or a wearer of the wearable medical device. 
     
     
         13 . The system of  claim 2 , wherein the plurality of operations is configured to, upon identifying the determined type corresponds to one of the at least one type of arrhythmia condition, select a pacing routine. 
     
     
         14 . The system of  claim 2 , wherein obtaining the plurality of cardio-vibrational measurements comprises obtaining at least a predetermined duration of cardio-vibrational measurements. 
     
     
         15 . The system of  claim 14 , wherein the plurality of operations is configured to divide the cardio-vibrational measurements into the plurality of adjacent cardiac portions each having a duration smaller than the predetermined duration. 
     
     
         16 . The system of  claim 2 , wherein creating the cardio-vibrational image matrix comprises:
 plotting the plurality of adjacent cardiac portions along a first axis; and   plotting the plurality of pixel characteristic values of each cardiac portion of the plurality of adjacent cardiac portions on a second axis.   
     
     
         17 . The system of  claim 2 , wherein the plurality of operations is configured to train at least a first machine learning engine of the one or more machine learning engines at least in part using historic cardio-vibrational image matrices generated from signals produced from monitoring the patient, thereby recognizing a unique cardiac signature of the patient. 
     
     
         18 . A system for monitoring a cardiac condition of a patient using cardio-vibrational image matrix representations of cardio-vibrational signals, the system comprising:
 a wearable medical device comprising at least one vibrational sensor configured for monitoring a heart of the patient;   a non-volatile computer-readable storage medium configured to store a plurality of cardio-vibrational measurements;   at least one machine learning engine configured to apply one or more of a plurality of cardiac risk biomarker classifiers, each cardiac risk biomarker classifier trained, using a pre-existing plurality of cardio-vibrational image matrices corresponding to each of a progression of heart failure classifications, to identify at least one heart failure biomarker of a set of heart failure biomarkers through analyzing cardio-vibrational image matrices, wherein 
 the set of heart failure biomarkers correspond to the progression of heart failure classifications; and 
   a plurality of operations stored as a plurality of computer executable instructions to a non-transitory computer readable media and/or encoded in hardware logic, wherein 
 the plurality of operations is configured to, in real-time, 
 obtain the plurality of cardio-vibrational measurements derived from cardio-vibrational signals of the patient collected by the at least one vibrational sensor, 
 store, to the non-volatile computer-readable storage medium, the plurality of cardio-vibrational measurements, 
 divide the plurality of cardio-vibrational measurements into a plurality of adjacent cardiac portions, a same at least one deflection feature being captured in each cardiac portion of the plurality of adjacent cardiac portions, wherein the at least one deflection feature is selected from an S1 peak and an S2 peak, 
 map parameter values of the plurality of cardio-vibrational measurements to a plurality of pixel characteristic values, 
 create a cardio-vibrational image matrix of the mapped parameter values, the cardio-vibrational image matrix graphically representing a time progression of the plurality of adjacent cardiac portions, 
 using one or more machine learning engines of the at least one machine learning engine, screen the contents of the cardio-vibrational image matrix for at least one identified biomarker of the set of heart failure biomarkers, and 
 determine, based at least in part on the at least one identified biomarker, a present classification of the progression of heart failure classifications. 
   
     
     
         19 . The system of  claim 18 , wherein the plurality of operations further comprises comparing the present classification to at least one historic classification of the patient to determine patient progression related to the set of heart failure biomarkers. 
     
     
         20 . The system of  claim 18 , wherein the set of heart failure biomarkers relates to at least one of sudden cardiac arrest (SCA) or low ejection fraction (EF). 
     
     
         21 . The system of  claim 18 , wherein the set of heart failure biomarkers comprises one or more of electromechanical activation over time (EMAT), left ventricular systolic time (LVST), S3 intensity, or S3 width.