A convolutional neural network for ECG annotation as the basis for classification of cardiac rhythms

Electrocardiography is the most common tool to diagnose cardiovascular diseases. Annotation, segmentation and rhythm classification of ECGs are challenging tasks, especially in the presence of atrial fibrillation and other arrhythmias. Our aim is to increase the accuracy of heart rhythm estimation b...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Physiological measurement Ročník 39; číslo 10; s. 104005
Hlavní autori: Sodmann, Philipp, Vollmer, Marcus, Nath, Neetika, Kaderali, Lars
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: England 24.10.2018
ISSN:1361-6579, 1361-6579
On-line prístup:Zistit podrobnosti o prístupe
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Abstract Electrocardiography is the most common tool to diagnose cardiovascular diseases. Annotation, segmentation and rhythm classification of ECGs are challenging tasks, especially in the presence of atrial fibrillation and other arrhythmias. Our aim is to increase the accuracy of heart rhythm estimation by the use of extreme gradient boosting trees and the development of a deep convolutional neural network for ECG segmentation. We trained a convolutional neural network with waveforms from PhysioNet databases to annotate QRS complexes, P waves, T waves, noise and interbeat ECG segments that characterize the essences of normal and irregular heart beats. We evaluated true positive rates, positive predictive values and mean absolute differences of our annotation based on reference annotations of the QT and MIT-BIH P-wave database. Moreover, we compared the results with standard QRS detectors and Ecgpuwave. Extreme gradient boosting trees were used to determine the heart rhythm based on hand-crafted features. More precisely, a noise estimation function was used in combination with heart rate and interval data. Furthermore we defined particular features based on ECG morphology, appearance of P waves and detection of irregular beats. We examined the feature importance and identified key features for normal sinus rhythm, atrial fibrillation, alternative rhythm and noisy recordings. The classification performance was evaluated externally using F scores by applying the algorithm to the hidden test set provided by the PhysioNet/CinC Challenge 2017. The true positive rate of the convolutional neural network in detection of manually revised R peaks in the QT database was [Formula: see text] and the positive predictive value was [Formula: see text]. The detection of P and T waves reached a true positive rate of [Formula: see text] and [Formula: see text] respectively, given a 50 ms tolerance when comparing the reference to the test annotation set. The rhythm classification performance reached an overall F score of 0.82 when applying the algorithm to the hidden test set. We achieved a shared rank #9 in the post-challenge phase of the PhysioNet/CinC Challenge 2017.
AbstractList Electrocardiography is the most common tool to diagnose cardiovascular diseases. Annotation, segmentation and rhythm classification of ECGs are challenging tasks, especially in the presence of atrial fibrillation and other arrhythmias. Our aim is to increase the accuracy of heart rhythm estimation by the use of extreme gradient boosting trees and the development of a deep convolutional neural network for ECG segmentation.OBJECTIVEElectrocardiography is the most common tool to diagnose cardiovascular diseases. Annotation, segmentation and rhythm classification of ECGs are challenging tasks, especially in the presence of atrial fibrillation and other arrhythmias. Our aim is to increase the accuracy of heart rhythm estimation by the use of extreme gradient boosting trees and the development of a deep convolutional neural network for ECG segmentation.We trained a convolutional neural network with waveforms from PhysioNet databases to annotate QRS complexes, P waves, T waves, noise and interbeat ECG segments that characterize the essences of normal and irregular heart beats. We evaluated true positive rates, positive predictive values and mean absolute differences of our annotation based on reference annotations of the QT and MIT-BIH P-wave database. Moreover, we compared the results with standard QRS detectors and Ecgpuwave. Extreme gradient boosting trees were used to determine the heart rhythm based on hand-crafted features. More precisely, a noise estimation function was used in combination with heart rate and interval data. Furthermore we defined particular features based on ECG morphology, appearance of P waves and detection of irregular beats. We examined the feature importance and identified key features for normal sinus rhythm, atrial fibrillation, alternative rhythm and noisy recordings. The classification performance was evaluated externally using F 1 scores by applying the algorithm to the hidden test set provided by the PhysioNet/CinC Challenge 2017.APPROACHWe trained a convolutional neural network with waveforms from PhysioNet databases to annotate QRS complexes, P waves, T waves, noise and interbeat ECG segments that characterize the essences of normal and irregular heart beats. We evaluated true positive rates, positive predictive values and mean absolute differences of our annotation based on reference annotations of the QT and MIT-BIH P-wave database. Moreover, we compared the results with standard QRS detectors and Ecgpuwave. Extreme gradient boosting trees were used to determine the heart rhythm based on hand-crafted features. More precisely, a noise estimation function was used in combination with heart rate and interval data. Furthermore we defined particular features based on ECG morphology, appearance of P waves and detection of irregular beats. We examined the feature importance and identified key features for normal sinus rhythm, atrial fibrillation, alternative rhythm and noisy recordings. The classification performance was evaluated externally using F 1 scores by applying the algorithm to the hidden test set provided by the PhysioNet/CinC Challenge 2017.The true positive rate of the convolutional neural network in detection of manually revised R peaks in the QT database was [Formula: see text] and the positive predictive value was [Formula: see text]. The detection of P and T waves reached a true positive rate of [Formula: see text] and [Formula: see text] respectively, given a 50 ms tolerance when comparing the reference to the test annotation set. The rhythm classification performance reached an overall F 1 score of 0.82 when applying the algorithm to the hidden test set.MAIN RESULTSThe true positive rate of the convolutional neural network in detection of manually revised R peaks in the QT database was [Formula: see text] and the positive predictive value was [Formula: see text]. The detection of P and T waves reached a true positive rate of [Formula: see text] and [Formula: see text] respectively, given a 50 ms tolerance when comparing the reference to the test annotation set. The rhythm classification performance reached an overall F 1 score of 0.82 when applying the algorithm to the hidden test set.We achieved a shared rank #9 in the post-challenge phase of the PhysioNet/CinC Challenge 2017.SIGNIFICANCEWe achieved a shared rank #9 in the post-challenge phase of the PhysioNet/CinC Challenge 2017.
Electrocardiography is the most common tool to diagnose cardiovascular diseases. Annotation, segmentation and rhythm classification of ECGs are challenging tasks, especially in the presence of atrial fibrillation and other arrhythmias. Our aim is to increase the accuracy of heart rhythm estimation by the use of extreme gradient boosting trees and the development of a deep convolutional neural network for ECG segmentation. We trained a convolutional neural network with waveforms from PhysioNet databases to annotate QRS complexes, P waves, T waves, noise and interbeat ECG segments that characterize the essences of normal and irregular heart beats. We evaluated true positive rates, positive predictive values and mean absolute differences of our annotation based on reference annotations of the QT and MIT-BIH P-wave database. Moreover, we compared the results with standard QRS detectors and Ecgpuwave. Extreme gradient boosting trees were used to determine the heart rhythm based on hand-crafted features. More precisely, a noise estimation function was used in combination with heart rate and interval data. Furthermore we defined particular features based on ECG morphology, appearance of P waves and detection of irregular beats. We examined the feature importance and identified key features for normal sinus rhythm, atrial fibrillation, alternative rhythm and noisy recordings. The classification performance was evaluated externally using F scores by applying the algorithm to the hidden test set provided by the PhysioNet/CinC Challenge 2017. The true positive rate of the convolutional neural network in detection of manually revised R peaks in the QT database was [Formula: see text] and the positive predictive value was [Formula: see text]. The detection of P and T waves reached a true positive rate of [Formula: see text] and [Formula: see text] respectively, given a 50 ms tolerance when comparing the reference to the test annotation set. The rhythm classification performance reached an overall F score of 0.82 when applying the algorithm to the hidden test set. We achieved a shared rank #9 in the post-challenge phase of the PhysioNet/CinC Challenge 2017.
Author Vollmer, Marcus
Sodmann, Philipp
Nath, Neetika
Kaderali, Lars
Author_xml – sequence: 1
  givenname: Philipp
  surname: Sodmann
  fullname: Sodmann, Philipp
  organization: Institute of Bioinformatics, University Medicine Greifswald, Greifswald, Germany. DZHK (German Center for Cardiological Research), partner site Greifswald, Greifswald, Germany
– sequence: 2
  givenname: Marcus
  surname: Vollmer
  fullname: Vollmer, Marcus
– sequence: 3
  givenname: Neetika
  surname: Nath
  fullname: Nath, Neetika
– sequence: 4
  givenname: Lars
  surname: Kaderali
  fullname: Kaderali, Lars
BackLink https://www.ncbi.nlm.nih.gov/pubmed/30235165$$D View this record in MEDLINE/PubMed
BookMark eNpNkL1PwzAUxC1URGlhZ0IeWUL9kbjOWFWlIFVigTl6tZ_VQGIXOwH1v6e0BXHLnd799IYbkYEPHgm54eyeM60nXCqeqWJaTgBQsvyMXP6dBv_ykIxSemOMcy2KCzKUTMiCq-KS2Bk1wX-Gpu_q4KGhHvt4sO4rxHfqQqSL-ZKC96GDH4ZCot0G6RpSnQ69aSCl2tXm2AdHDURbg6Fxs-s2bboi5w6ahNcnH5PXh8XL_DFbPS-f5rNVZqRUXWaskqUrnGKIUq9BGwalcaDyqdMCtJASp8pK5WRe6txiYdFZoUBbFMiYGJO7499tDB89pq5q62SwacBj6FMl-F5FPi3VHr09of26RVttY91C3FW_y4hvR7xpFw
CitedBy_id crossref_primary_10_1109_TIM_2022_3224535
crossref_primary_10_3389_fcvm_2024_1341786
crossref_primary_10_1007_s41870_023_01241_7
crossref_primary_10_1038_s41598_022_18664_0
crossref_primary_10_1016_j_bspc_2020_101874
crossref_primary_10_3390_s22166071
crossref_primary_10_1109_ACCESS_2022_3181727
crossref_primary_10_1142_S0218126625300053
crossref_primary_10_1016_j_compbiomed_2020_103801
crossref_primary_10_1016_j_compbiomed_2022_106511
crossref_primary_10_1016_j_bspc_2022_104106
crossref_primary_10_1093_cvr_cvab169
crossref_primary_10_1161_JAHA_119_013924
crossref_primary_10_1016_j_ins_2022_01_030
crossref_primary_10_3390_s23167204
crossref_primary_10_1109_JSEN_2022_3162691
crossref_primary_10_1016_j_jelectrocard_2019_12_009
crossref_primary_10_3390_electronics11234021
crossref_primary_10_1080_17434440_2022_2115887
crossref_primary_10_1016_j_compbiomed_2022_105331
crossref_primary_10_1038_s41598_020_79512_7
crossref_primary_10_3390_s21206848
crossref_primary_10_1088_1361_6579_ac72f5
crossref_primary_10_1109_ACCESS_2023_3288700
crossref_primary_10_1016_j_eswa_2025_127955
crossref_primary_10_1088_2057_1976_ac2b4e
crossref_primary_10_1088_1361_6579_ac08e6
crossref_primary_10_1136_jim_2021_002278
crossref_primary_10_1016_j_cmpb_2020_105740
crossref_primary_10_3390_bioengineering10050552
crossref_primary_10_1016_j_imu_2020_100507
crossref_primary_10_1016_j_compbiomed_2022_105445
crossref_primary_10_1109_ACCESS_2025_3567093
crossref_primary_10_3390_s22062329
crossref_primary_10_1155_2021_9926769
crossref_primary_10_3389_fneur_2022_943816
crossref_primary_10_1016_j_bspc_2020_102194
crossref_primary_10_4018_IJACI_300796
crossref_primary_10_1007_s00034_021_01889_1
crossref_primary_10_1016_j_cmpb_2019_02_005
crossref_primary_10_1109_JBHI_2023_3308856
crossref_primary_10_3390_s20247307
crossref_primary_10_1016_j_bbe_2021_04_015
crossref_primary_10_1007_s00521_021_06005_7
crossref_primary_10_1007_s10916_019_1511_2
crossref_primary_10_1155_2022_9592970
crossref_primary_10_1007_s42979_023_01970_3
crossref_primary_10_1016_j_measurement_2023_113205
crossref_primary_10_1088_1361_6579_ab15a2
crossref_primary_10_3390_metabo13121204
crossref_primary_10_1007_s11042_022_13821_z
crossref_primary_10_1038_s41746_022_00709_3
crossref_primary_10_1109_ACCESS_2021_3092631
ContentType Journal Article
DBID NPM
7X8
DOI 10.1088/1361-6579/aae304
DatabaseName PubMed
MEDLINE - Academic
DatabaseTitle PubMed
MEDLINE - Academic
DatabaseTitleList MEDLINE - Academic
PubMed
Database_xml – sequence: 1
  dbid: NPM
  name: PubMed
  url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed
  sourceTypes: Index Database
– sequence: 2
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod no_fulltext_linktorsrc
Discipline Medicine
Engineering
Physics
EISSN 1361-6579
ExternalDocumentID 30235165
Genre Journal Article
GroupedDBID ---
-~X
123
1JI
4.4
53G
5B3
5VS
5ZH
7.M
7.Q
AAGCD
AAJIO
AAJKP
AALHV
AATNI
ABCXL
ABHWH
ABJNI
ABQJV
ABVAM
ACAFW
ACGFS
ACHIP
AEFHF
AENEX
AFYNE
AKPSB
ALMA_UNASSIGNED_HOLDINGS
AOAED
ASPBG
ATQHT
AVWKF
AZFZN
CBCFC
CEBXE
CJUJL
CRLBU
CS3
DU5
EBS
EDWGO
EJD
EMSAF
EPQRW
EQZZN
F5P
HAK
IHE
IJHAN
IOP
IZVLO
KOT
LAP
M45
N5L
N9A
NPM
NT-
NT.
P2P
PJBAE
R4D
RIN
RNS
RO9
ROL
RPA
SY9
UCJ
W28
XPP
ZMT
7X8
ADEQX
AEINN
ID FETCH-LOGICAL-c336t-cd639f5f60ee38ba8c0a9cfa647f82a8233e76d36f34984de5defd26a8de2e002
IEDL.DBID 7X8
ISICitedReferencesCount 61
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000448419800001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1361-6579
IngestDate Thu Oct 02 05:04:40 EDT 2025
Wed Feb 19 02:36:21 EST 2025
IsPeerReviewed true
IsScholarly true
Issue 10
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c336t-cd639f5f60ee38ba8c0a9cfa647f82a8233e76d36f34984de5defd26a8de2e002
Notes ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Undefined-1
ObjectType-Feature-3
content type line 23
PMID 30235165
PQID 2111154796
PQPubID 23479
ParticipantIDs proquest_miscellaneous_2111154796
pubmed_primary_30235165
PublicationCentury 2000
PublicationDate 2018-10-24
PublicationDateYYYYMMDD 2018-10-24
PublicationDate_xml – month: 10
  year: 2018
  text: 2018-10-24
  day: 24
PublicationDecade 2010
PublicationPlace England
PublicationPlace_xml – name: England
PublicationTitle Physiological measurement
PublicationTitleAlternate Physiol Meas
PublicationYear 2018
SSID ssj0011825
Score 2.455672
Snippet Electrocardiography is the most common tool to diagnose cardiovascular diseases. Annotation, segmentation and rhythm classification of ECGs are challenging...
SourceID proquest
pubmed
SourceType Aggregation Database
Index Database
StartPage 104005
Title A convolutional neural network for ECG annotation as the basis for classification of cardiac rhythms
URI https://www.ncbi.nlm.nih.gov/pubmed/30235165
https://www.proquest.com/docview/2111154796
Volume 39
WOSCitedRecordID wos000448419800001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV3PS8MwFH6oU9GDP-av-YsIXsu6pE3Tk4yx6WVjB4XdRvaSsB1s5zoF_3uTtHNeBMFLCw0tJf3y8jXfl_cA7hPEFueog9hM7A8KRRFInIQBdSl-UIacKuGLTSSDgRiN0mG14FZUtspVTPSBWuXo1sib1I3tOEpS_jB_C1zVKKeuViU0NqHGLJVxqE5GaxXBcue43HfVchaPtJIp7cBqfl9rSqmZL9P2C8H0E03v8L-veAQHFcUk7RITx7Chszrs_0g8WIfdfiWp12HHe0CxOAHVJs6DXmHRPsHluvQn7xQnlt6SbueRyCzLSwGfyIJYAknsVDgrfDs6Mu7cR2V7bgh6CCJZTD-X09fiFF563efOU1AVYQiQMb4MUFkOY2LDQ62ZmEiBoUzRSB4lRlApKGM64Ypxw6JURErHShtFuRRKU23j7RlsZXmmL4BoSjGSJknThEUTbVIZc2YDHCoUaELdgLtVv44tyJ1yITOdvxfjdc824Lz8OON5mY1j7KoexS0eX_7h7ivYo277QsgDGl1Dzdghrm9gGz-Ws2Jx69Fjj4Nh_wsF6dA3
linkProvider ProQuest
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=A+convolutional+neural+network+for+ECG+annotation+as+the+basis+for+classification+of+cardiac+rhythms&rft.jtitle=Physiological+measurement&rft.au=Sodmann%2C+Philipp&rft.au=Vollmer%2C+Marcus&rft.au=Nath%2C+Neetika&rft.au=Kaderali%2C+Lars&rft.date=2018-10-24&rft.issn=1361-6579&rft.eissn=1361-6579&rft.volume=39&rft.issue=10&rft.spage=104005&rft_id=info:doi/10.1088%2F1361-6579%2Faae304&rft.externalDBID=NO_FULL_TEXT
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1361-6579&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1361-6579&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1361-6579&client=summon