Machine Learning Improves the Detection of Misplaced V1 and V2 Electrodes During 12-Lead Electrocardiogram Acquisition

Khaled Rjoob, RR Bond, D Finlay, V. E. McGilligan, Stephen Leslie, Aleeha Iftikhar, Daniel Gueldenring, Ali Rababah, Charles Knoery, Aaron Peace

Research output: Contribution to conferencePaper

Abstract

Electrode misplacement during 12-lead
Electrocardiogram (ECG) acquisition can adversely cause
false ECG interpretation, diagnosis and subsequent
incorrect clinical treatment or lack thereof. A common
misplacement errors are the. superior placement of V1
and V2 electrodes. The analysis of ECG signals that were
recorded from ECGs with vertically misplaced leads V1
and V2 can yield a false diagnosis of Brugada syndrome,
myocardial infarction (MI) or left ventricular hypertrophy
(LVH). The aim of the current research was to detect lead
V1 and V2 misplacement using feature engineered
machine learning algorithms to enhance ECG data quality
to improve clinical decision making in cardiac care. In
this particular study, we reasonably assume that V1 and
V2 are concurrently superiorly misplaced together. ECGs
for 450 patients, (normal n=150, LVH n=150, MI n=150)
were extracted from body surface potential maps. ECG
signals were extracted using correct and incorrectly
placed V1 and V2 electrodes, i.e. leads derived from the
fourth intercostal space (ICS) as well as the first ICS,
second ICS, and third ICS. The prevalence for correct and
incorrect leads were 50%. Sixteen features were extracted
including: morphological, statistical and time-frequency
features. Two feature selection approaches (filter method
and wrapper method) were applied to find an optimal set
of features that provide a high accuracy when used with a
machine learning model. To ensure accuracy, six
classifiers were applied including: fine tree, coarse tree,
bagged tree, Linear Support Vector Machine (LSVM),
Quadratic Support Vector Machine (QSVM) and logistic
regression. The accuracy of V1 and V2 misplacement
detection was 94.3% in the first ICS, 92.7% in the second
ICS and 70% in third ICS respectively. Based on
accuracy results, bagged tree was the best classifier in the
first, second and third ICS to detect V1 and V2
misplacement.

Conference

ConferenceComputing in Cardiology
Abbreviated titleCinC 2019
CountrySingapore
CityBiopolis
Period8/09/1911/09/19
Internet address

Fingerprint

Electrocardiography
Electrodes
Left Ventricular Hypertrophy
Myocardial Infarction
Learning
Brugada Syndrome
Lead
Machine Learning
Research
Support Vector Machine
Therapeutics

Keywords

  • machine Learning
  • ECG
  • Lead misplacement

Cite this

Rjoob, Khaled ; Bond, RR ; Finlay, D ; McGilligan, V. E. ; Leslie, Stephen ; Iftikhar, Aleeha ; Gueldenring, Daniel ; Rababah, Ali ; Knoery, Charles ; Peace, Aaron. / Machine Learning Improves the Detection of Misplaced V1 and V2 Electrodes During 12-Lead Electrocardiogram Acquisition. Paper presented at Computing in Cardiology, Biopolis, Singapore.
@conference{3d6f67527420487989743869719286a2,
title = "Machine Learning Improves the Detection of Misplaced V1 and V2 Electrodes During 12-Lead Electrocardiogram Acquisition",
abstract = "Electrode misplacement during 12-leadElectrocardiogram (ECG) acquisition can adversely causefalse ECG interpretation, diagnosis and subsequentincorrect clinical treatment or lack thereof. A commonmisplacement errors are the. superior placement of V1and V2 electrodes. The analysis of ECG signals that wererecorded from ECGs with vertically misplaced leads V1and V2 can yield a false diagnosis of Brugada syndrome,myocardial infarction (MI) or left ventricular hypertrophy(LVH). The aim of the current research was to detect leadV1 and V2 misplacement using feature engineeredmachine learning algorithms to enhance ECG data qualityto improve clinical decision making in cardiac care. Inthis particular study, we reasonably assume that V1 andV2 are concurrently superiorly misplaced together. ECGsfor 450 patients, (normal n=150, LVH n=150, MI n=150)were extracted from body surface potential maps. ECGsignals were extracted using correct and incorrectlyplaced V1 and V2 electrodes, i.e. leads derived from thefourth intercostal space (ICS) as well as the first ICS,second ICS, and third ICS. The prevalence for correct andincorrect leads were 50{\%}. Sixteen features were extractedincluding: morphological, statistical and time-frequencyfeatures. Two feature selection approaches (filter methodand wrapper method) were applied to find an optimal setof features that provide a high accuracy when used with amachine learning model. To ensure accuracy, sixclassifiers were applied including: fine tree, coarse tree,bagged tree, Linear Support Vector Machine (LSVM),Quadratic Support Vector Machine (QSVM) and logisticregression. The accuracy of V1 and V2 misplacementdetection was 94.3{\%} in the first ICS, 92.7{\%} in the secondICS and 70{\%} in third ICS respectively. Based onaccuracy results, bagged tree was the best classifier in thefirst, second and third ICS to detect V1 and V2misplacement.",
keywords = "machine Learning, ECG, Lead misplacement",
author = "Khaled Rjoob and RR Bond and D Finlay and McGilligan, {V. E.} and Stephen Leslie and Aleeha Iftikhar and Daniel Gueldenring and Ali Rababah and Charles Knoery and Aaron Peace",
year = "2019",
month = "9",
day = "24",
language = "English",
note = "Computing in Cardiology, CinC 2019 ; Conference date: 08-09-2019 Through 11-09-2019",
url = "http://www.cinc.org",

}

Rjoob, K, Bond, RR, Finlay, D, McGilligan, VE, Leslie, S, Iftikhar, A, Gueldenring, D, Rababah, A, Knoery, C & Peace, A 2019, 'Machine Learning Improves the Detection of Misplaced V1 and V2 Electrodes During 12-Lead Electrocardiogram Acquisition' Paper presented at Computing in Cardiology, Biopolis, Singapore, 8/09/19 - 11/09/19, .

Machine Learning Improves the Detection of Misplaced V1 and V2 Electrodes During 12-Lead Electrocardiogram Acquisition. / Rjoob, Khaled; Bond, RR; Finlay, D; McGilligan, V. E.; Leslie, Stephen; Iftikhar, Aleeha; Gueldenring, Daniel; Rababah, Ali; Knoery, Charles; Peace, Aaron.

2019. Paper presented at Computing in Cardiology, Biopolis, Singapore.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Machine Learning Improves the Detection of Misplaced V1 and V2 Electrodes During 12-Lead Electrocardiogram Acquisition

AU - Rjoob, Khaled

AU - Bond, RR

AU - Finlay, D

AU - McGilligan, V. E.

AU - Leslie, Stephen

AU - Iftikhar, Aleeha

AU - Gueldenring, Daniel

AU - Rababah, Ali

AU - Knoery, Charles

AU - Peace, Aaron

PY - 2019/9/24

Y1 - 2019/9/24

N2 - Electrode misplacement during 12-leadElectrocardiogram (ECG) acquisition can adversely causefalse ECG interpretation, diagnosis and subsequentincorrect clinical treatment or lack thereof. A commonmisplacement errors are the. superior placement of V1and V2 electrodes. The analysis of ECG signals that wererecorded from ECGs with vertically misplaced leads V1and V2 can yield a false diagnosis of Brugada syndrome,myocardial infarction (MI) or left ventricular hypertrophy(LVH). The aim of the current research was to detect leadV1 and V2 misplacement using feature engineeredmachine learning algorithms to enhance ECG data qualityto improve clinical decision making in cardiac care. Inthis particular study, we reasonably assume that V1 andV2 are concurrently superiorly misplaced together. ECGsfor 450 patients, (normal n=150, LVH n=150, MI n=150)were extracted from body surface potential maps. ECGsignals were extracted using correct and incorrectlyplaced V1 and V2 electrodes, i.e. leads derived from thefourth intercostal space (ICS) as well as the first ICS,second ICS, and third ICS. The prevalence for correct andincorrect leads were 50%. Sixteen features were extractedincluding: morphological, statistical and time-frequencyfeatures. Two feature selection approaches (filter methodand wrapper method) were applied to find an optimal setof features that provide a high accuracy when used with amachine learning model. To ensure accuracy, sixclassifiers were applied including: fine tree, coarse tree,bagged tree, Linear Support Vector Machine (LSVM),Quadratic Support Vector Machine (QSVM) and logisticregression. The accuracy of V1 and V2 misplacementdetection was 94.3% in the first ICS, 92.7% in the secondICS and 70% in third ICS respectively. Based onaccuracy results, bagged tree was the best classifier in thefirst, second and third ICS to detect V1 and V2misplacement.

AB - Electrode misplacement during 12-leadElectrocardiogram (ECG) acquisition can adversely causefalse ECG interpretation, diagnosis and subsequentincorrect clinical treatment or lack thereof. A commonmisplacement errors are the. superior placement of V1and V2 electrodes. The analysis of ECG signals that wererecorded from ECGs with vertically misplaced leads V1and V2 can yield a false diagnosis of Brugada syndrome,myocardial infarction (MI) or left ventricular hypertrophy(LVH). The aim of the current research was to detect leadV1 and V2 misplacement using feature engineeredmachine learning algorithms to enhance ECG data qualityto improve clinical decision making in cardiac care. Inthis particular study, we reasonably assume that V1 andV2 are concurrently superiorly misplaced together. ECGsfor 450 patients, (normal n=150, LVH n=150, MI n=150)were extracted from body surface potential maps. ECGsignals were extracted using correct and incorrectlyplaced V1 and V2 electrodes, i.e. leads derived from thefourth intercostal space (ICS) as well as the first ICS,second ICS, and third ICS. The prevalence for correct andincorrect leads were 50%. Sixteen features were extractedincluding: morphological, statistical and time-frequencyfeatures. Two feature selection approaches (filter methodand wrapper method) were applied to find an optimal setof features that provide a high accuracy when used with amachine learning model. To ensure accuracy, sixclassifiers were applied including: fine tree, coarse tree,bagged tree, Linear Support Vector Machine (LSVM),Quadratic Support Vector Machine (QSVM) and logisticregression. The accuracy of V1 and V2 misplacementdetection was 94.3% in the first ICS, 92.7% in the secondICS and 70% in third ICS respectively. Based onaccuracy results, bagged tree was the best classifier in thefirst, second and third ICS to detect V1 and V2misplacement.

KW - machine Learning

KW - ECG

KW - Lead misplacement

UR - http://www.cinc.org/2019/Program/accepted/35_CinCFinalPDF.pdf

M3 - Paper

ER -