Noninvasive infarct size determination

Abstract

The present application relates to using noninvasive techniques to determine whether a patient has experienced a cardiac event. The present application also relates to using noninvasive techniques to determine a size of a myocardial infarction experienced by a patient. In some embodiments, arterial pressure waveforms may be obtained, and from the arterial pressure waveform, a set of cardiac parameters may be extracted. The extracted cardiac parameters may be provided, as input, to the trained machine learning model, which may output a result indicating whether the patient experienced a cardiac event, a size of a myocardial infarction experience by a patient, or other information about the patient's cardiac health.

What is claimed is: 
     
         1 . A system for detecting a size of a myocardial infarction using non-invasive tests, the system comprising:
 a wearable device worn by a patient comprising at least one sensor configured to capture a set of arterial pressure measurements of the patient that experienced a myocardial infarction, the wearable device being further configured to generate an arterial pressure waveform representing the set of arterial pressure measurements;   a computing system comprising one or more processors executing computer program instructions to effectuate operations comprising:
 obtaining, from the wearable device, first data comprising the arterial pressure waveform; 
 computing, based on the first data comprising the arterial pressure waveform, a first intrinsic frequency describing dynamics of a systolic phase of a cardiac cycle of the patient, a second intrinsic frequency describing a diastolic phase of the cardiac cycle of the patient, a relative height of a dicrotic notch (RHDN) of the cardiac cycle of the patient, and an envelope ratio (ER) of the cardiac cycle of the patient; 
 determining, based on the first intrinsic frequency and the second intrinsic frequency, respectively, a systolic intrinsic phase angle describing the systolic phase of the cardiac cycle of the patient and a diastolic intrinsic phase angle describing the diastolic phase of the cardiac cycle of the patient; 
 providing (i) at least one of the first intrinsic frequency or the second intrinsic frequency, (ii) at least one of the systolic intrinsic phase angle or the diastolic intrinsic phase angle, (iii) the RHDN of the cardiac cycle of the patient, and (iv) the ER of the cardiac cycle of the patient to a trained artificial neural network (ANN) configured to determine a size of a myocardial infarction experienced by the patient; 
 obtaining, from the trained ANN, the size of the myocardial infarction experienced by the patient; 
 providing second data comprising the size of the myocardial infarction experienced by the patient to one or more client devices of the patient, a medical provider of the patient, or the patient and the medical provider of the patient. 
   
     
     
         2 . The system of  claim 1 , wherein the wearable device is further configured to capture a set of pulse-oxygen level measurements of the patient, and generate a pulse-oxygen level waveform representing the set of pulse-oxygen level measurements, the operations further comprise:
 obtaining, from the wearable device, third data comprising the pulse-oxygen level waveform;   computing, based on the third data representing the pulse-oxygen level waveform, the first intrinsic frequency, the second intrinsic frequency, the systolic intrinsic phase angle, the diastolic intrinsic phase angle, the RHDN of the cardiac cycle of the patient, and the ER of the cardiac cycle of the patient.   
     
     
         3 . The system of  claim 1 , wherein the operations further comprise:
 retrieving, from a database, a plurality of datasets comprising arterial vessel wall displacement waveforms, arterial pressure waveforms, or arterial vessel wall displacement waveforms and arterial pressure waveforms, wherein each dataset of the plurality of datasets is associated with a given patient of a plurality of patients that have experienced a myocardial infarction and includes an indication of a size of the myocardial infarction experienced by the given patient, wherein the size of the myocardial infarction indicates an amount of cardiac tissue that has been damaged by the myocardial infarction;   for each of the arterial vessel wall displacement waveforms, the arterial pressure waveforms, or the arterial vessel wall displacement waveforms and the arterial pressure waveforms, computing a first intrinsic frequency, a second intrinsic frequency, a systolic intrinsic phase angle, a diastolic intrinsic phase angle, a relative height of a dicrotic node (RDHN) of a cardiac cycle of a respective patient, and an envelope ratio (ER) of the cardiac cycle of the respective patient to obtain a plurality of first intrinsic frequencies, a plurality of second intrinsic frequencies, a plurality of systolic intrinsic phase angles, a plurality of diastolic intrinsic phase angles, a plurality of RDHNs, and a plurality of ERs;   generating and storing training data comprising respective tuples of one of the plurality of first intrinsic frequencies, one of the plurality of second intrinsic frequencies, one of the plurality of systolic intrinsic phase angles, one of the plurality of diastolic intrinsic phase angles, one of the plurality of RDHNs, and one of the plurality of ERs; and   training and testing an artificial neural network (ANN) based on the training data to obtain the trained ANN.   
     
     
         4 . The system of  claim 1 , wherein:
 the envelope ratio is a ratio of an energy carried by the arterial pressure waveform during systole to an energy carried by the arterial pressure waveform during diastole;   the myocardial infarction experienced by the patient occurred within twenty-four (24) hours of the first data being captured;   the trained ANN comprises a plurality of layers including an input layer, one or more hidden layers, and an output layer;   the input layer comprises four nodes respectively corresponding to one of:
 (i) a first intrinsic frequency of a cardiac cycle of a given patient, a systolic intrinsic phase angle of the cardiac cycle of the given patient, an RDHN of the cardiac cycle of the given patient, and an ER of the cardiac cycle of the given patient, or 
 (ii) a second intrinsic frequency of a cardiac cycle of a given patient, a diastolic intrinsic phase angle of the cardiac cycle of the given patient, an RDHN of the cardiac cycle of the given patient, and an ER of the cardiac cycle of the given patient; and 
   nodes of the one or more hidden layers are fully connected to the four nodes of the input layer, the one or more hidden layers being configured to estimate a size of a myocardial infraction of the given patient.   
     
     
         5 . A non-transitory computer-readable medium storing computer program instructions that, when executed by one or more processors of a computing system, effectuate operations comprising:
 obtaining first data representing a hemodynamic waveform of a patient;   determining, based on the first data, a set of cardiovascular parameters of the patient, the set of cardiovascular parameters comprising a first intrinsic frequency of a cardiac cycle of the patient, a first intrinsic phase angle of the cardiac cycle, a relative height of a dicrotic notch (RHDN) of the cardiac cycle, and an envelope ratio (ER) of the cardiac cycle;   providing the set of cardiovascular parameters to a trained neural network configured to determine a size of a myocardial infarction experienced by the patient;   obtaining, from the trained neural network, the size of the myocardial infarction experienced by the patient; and   providing second data comprising the size of the myocardial infarction experienced by the patient to one or more client devices of the patient, a medical provider of the patient, or the patient and the medical provider of the patient.   
     
     
         6 . The non-transitory computer-readable medium of  claim 5 , wherein the set of cardiovascular parameters further comprises a second intrinsic frequency of the cardiac cycle, and a second intrinsic phase angle of the cardiac cycle. 
     
     
         7 . The non-transitory computer-readable medium of  claim 6 , wherein:
 the first intrinsic frequency describes a systolic phase of the cardiac cycle of the patient;   the second intrinsic frequency describes a diastolic phase of the cardiac cycle of the patient;   the first intrinsic phase angle comprises a systolic intrinsic phase angle describing the systolic phase of the cardiac cycle of the patient; and   the second intrinsic phase angle comprises a diastolic intrinsic phase angle describing the diastolic phase of the cardiac cycle of the patient.   
     
     
         8 . The non-transitory computer-readable medium of  claim 5 , wherein the hemodynamic waveform of the patient comprises at least one of an arterial pressure waveform or a vessel wall displacement waveform. 
     
     
         9 . The non-transitory computer-readable medium of  claim 5 , wherein the operations further comprise:
 retrieving a plurality of hemodynamic waveforms, each of the plurality of hemodynamic waveforms respectively corresponding to one of a plurality of patients;   determining, for each of the plurality of hemodynamic waveforms, a set of cardiovascular parameters of a cardiac cycle of a respective one of the plurality of patients to obtain a plurality of sets of cardiovascular parameters, wherein each set of cardiovascular parameters of the plurality of sets of cardiovascular parameters comprises: (i) one or more intrinsic frequencies of the cardiac cycle of the respective one of the plurality of patients, (ii) one or more intrinsic phase angles of the cardiac cycle of the respective one of the plurality of patients, (iii) a relative height of a dicrotic notch (RHDN) of the cardiac cycle of the respective one of the plurality of patients, and (iv) an envelope ratio (ER) of the cardiac cycle of the respective one of the plurality of patients;   generating and storing training data comprising tuples of the one or more intrinsic frequencies, the one or more intrinsic phase angles, the RHDN, and the ER for each of the plurality of patients, wherein each tuple includes an indication of a size of a myocardial infarction experienced by the respective one of the plurality of patients; and   training and testing a neural network based on the training data to obtain the trained neural network.   
     
     
         10 . The non-transitory computer-readable medium of  claim 5 , wherein the first data representing the hemodynamic waveform of a patient are obtained from a client device of the patient, wherein the client device is operatively coupled to at least one sensor configured to capture hemodynamic measurements of the cardiac cycle of the patient, generate the first data representing the hemodynamic waveform of the cardiac cycle of the patient, and at least one of output or store, in memory, the first data representing the hemodynamic waveform of the cardiac cycle of the patient. 
     
     
         11 . The non-transitory computer-readable medium of  claim 10 , wherein the client device is a wearable device. 
     
     
         12 . The non-transitory computer-readable medium of  claim 5 , wherein the operations further comprise:
 updating training data used to train the trained neural network based on the set of cardiovascular parameters and the size of myocardial infarction of the patient to obtain updated training data, wherein the updated training data comprises a tuple including the set of cardiovascular parameters and the size of the myocardial infarction of the patient.   
     
     
         13 . The non-transitory computer-readable medium of  claim 5 , wherein the operations further comprise:
 steps for training a trained neural network to obtain the trained neural network.   
     
     
         14 . The non-transitory computer-readable medium of  claim 5 , wherein the first data representing the hemodynamic waveform of the patient are captured using cardiovascular measurement means. 
     
     
         15 . A method implemented by a computing system executing computer program instructions, the method comprising:
 obtaining first data representing a hemodynamic waveform of a patient;   determining, based on the first data, a set of cardiovascular parameters of the patient, the set of cardiovascular parameters comprising a first intrinsic frequency of a cardiac cycle of the patient, a first intrinsic phase angle of the cardiac cycle, a relative height of a dicrotic notch (RHDN) of the cardiac cycle, and an envelope ratio (ER) of the cardiac cycle;   providing the set of cardiovascular parameters to a trained neural network configured to determine a size of a myocardial infarction experienced by the patient;   obtaining, from the trained neural network, the size of the myocardial infarction experienced by the patient; and   providing second data comprising the size of the myocardial infarction experienced by the patient to one or more client devices of the patient, a medical provider of the patient, or the patient and the medical provider of the patient.