Document Type

Dissertation

Degree

Doctor of Philosophy (PhD)

Major/Program

Electrical and Computer Engineering

First Advisor's Name

Malek Adjouadi

First Advisor's Committee Title

committee chair

Second Advisor's Name

Mercedes Cabrerizo

Second Advisor's Committee Title

committee member

Third Advisor's Name

Jean Andrian

Third Advisor's Committee Title

committee member

Fourth Advisor's Name

Armando Barreto

Fourth Advisor's Committee Title

committee member

Fifth Advisor's Name

Naphtali Rishe

Fifth Advisor's Committee Title

committee member

Sixth Advisor's Name

Sharan Ramaswamy

Sixth Advisor's Committee Title

committee member

Keywords

myocardial infarction, machine learning, real-time, frequency independence

Date of Defense

3-26-2021

Abstract

The central aim of this research is the development and deployment of a novel multilayer machine learning design with unique application for the diagnosis of myocardial infarctions (MIs) from individual heartbeats of single-lead electrocardiograms (EKGs) irrespective of their sampling frequencies over a given range. To the best of our knowledge, this design is the first to attempt inter-patient myocardial infarction detection from individual heartbeats of single-lead (lead II) electrocardiograms that achieves high accuracy and near real-time diagnosis. The processing time of 300 milliseconds to a diagnosis is just at the time range in between extremely fast heartbeats of around 300 milliseconds, or 200 beats per minute. The design achieves stable performance metrics over the frequency range of 202Hz to 2.8kHz with an accuracy of 77.12%, positive predictive value (PPV) of 75.85%, and a negative predictive value (NPV) of 83.02% over the entire PTB database; 85.07%, 81.54%, 87.31% over the PTB-XL (the largest EKG database available for research) validation set, and 84.17%, 78.37%, 87.55% over the PTB-XL test set. Major design contributions and findings of this work reveal (1) a method for the realtime detection of ventricular depolarization events in the PQRST complex from 12-lead electrocardiograms using Independent Component Analysis (ICA), with a slightly different use of ICA proposed for electrocardiogram analysis and R-peak detection/localization; (2) a multilayer Long-Short Term Memory (LSTM) neural network design that identifies

infarcted patients from a single heartbeat of a single-lead (lead II) electrocardiogram; (3) and integrated LSTM neural network with an algorithm that detects the R-peaks in real time for instantaneous detection of myocardial infarctions and for effective monitoring of patients under cardiac stress and/or at risk of myocardial infarction; (4) a fully integrated 12-lead real-time classifier with even higher detection metrics and a deeper neural architecture, which could serve as a near real-time monitoring tool that could gauge disease progression and evaluate benefits gained from early intervention and treatment planning; (5) a real-time frequency-independent design based on a single-lead single-beat MI detector, which is of pivotal importance to deployment as there is no standard sampling frequency for EKGs, making them span a wider frequency spectrum. vii

Identifier

FIDC009685

ORCID

0000-0001-5282-4480

Previously Published In

H. Martin, W. Izquierdo, M. Cabrerizo, M. Adjouadi, "Real-time R-spike detection in the cardiac waveform through independent component analysis", 2017 IEEE Signal Processing in Medicine and Biology Symposium (SPMB), Dec. 2017 Philadelphia, PA, USA

H. Martin, W. Izquierdo, U. Morar, M. Cabrerizo, A. Cabrera, and M. Adjouadi, "A Fast and Accurate Myocardial Infarction Detector", The 2020 International Conference on Computational Science and Computational Intelligence (CSCI’20), Las Vegas, Nevada, USA, Dec. 2020

H. Martin, W. Izquierdo, M. Cabrerizo, A. Cabrera, and M. Adjouadi, "Near Real-Time Single-Beat Myocardial Infarction Detection from Single-Lead Electrocardiogram using Long Short-Term Memory Neural Network", Biomedical Signal Processing and Control, (under review), Feb. 2021

H. Martin,U. Morar,W. Izquierdo, M. Cabrerizo, A. Cabrera, and M.Adjouadi, "Real-time Frequency-Independent Single-Lead and Single-Beat Myocardial Infarction Detection", Artificail Intelligence in Medicine, (under review), Feb. 2021

Share

COinS
 

Rights Statement

Rights Statement

In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/
This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).