Convolutional Neural Networks for patient-specific ECG classification

S. Kiranyaz; T. Ince; R. Hamila; M. Gabbouj

doi:10.1109/EMBC.2015.7318926

Convolutional Neural Networks for patient-specific ECG classification

S. Kiranyaz, T. Ince, R. Hamila, M. Gabbouj

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

69 Citations (Scopus)

Abstract

We propose a fast and accurate patient-specific electrocardiogram (ECG) classification and monitoring system using an adaptive implementation of 1D Convolutional Neural Networks (CNNs) that can fuse feature extraction and classification into a unified learner. In this way, a dedicated CNN will be trained for each patient by using relatively small common and patient-specific training data and thus it can also be used to classify long ECG records such as Holter registers in a fast and accurate manner. Alternatively, such a solution can conveniently be used for real-time ECG monitoring and early alert system on a light-weight wearable device. The experimental results demonstrate that the proposed system achieves a superior classification performance for the detection of ventricular ectopic beats (VEB) and supraventricular ectopic beats (SVEB).

Original language	English
Title of host publication	2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)
Pages	2608-2611
Number of pages	4
DOIs	https://doi.org/10.1109/EMBC.2015.7318926
Publication status	Published - 1 Aug 2015
Publication type	A4 Article in conference proceedings
Event	Annual International Conference of the IEEE Engineering in Medicine and Biology Society - Duration: 1 Jan 1900 → …

Conference

Conference	Annual International Conference of the IEEE Engineering in Medicine and Biology Society
Period	1/01/00 → …

Keywords

electrocardiography
feature extraction
medical signal processing
neural nets
signal classification
1D convolutional neural network
ECG monitoring
ECG record classification
light-weight wearable device
patient-specific ECG classification
patient-specific electrocardiogram classification
supraventricular ectopic beat detection
Convolution
Databases
Electrocardiography
Feature extraction
Neural networks
Neurons
Training

Publication forum classification

Publication forum level 1

Access to Document

10.1109/EMBC.2015.7318926

Cite this

@inproceedings{cabebbfb4e7d442fa7a838b06c6de7ed,

title = "Convolutional Neural Networks for patient-specific ECG classification",

abstract = "We propose a fast and accurate patient-specific electrocardiogram (ECG) classification and monitoring system using an adaptive implementation of 1D Convolutional Neural Networks (CNNs) that can fuse feature extraction and classification into a unified learner. In this way, a dedicated CNN will be trained for each patient by using relatively small common and patient-specific training data and thus it can also be used to classify long ECG records such as Holter registers in a fast and accurate manner. Alternatively, such a solution can conveniently be used for real-time ECG monitoring and early alert system on a light-weight wearable device. The experimental results demonstrate that the proposed system achieves a superior classification performance for the detection of ventricular ectopic beats (VEB) and supraventricular ectopic beats (SVEB).",

keywords = "electrocardiography, feature extraction, medical signal processing, neural nets, signal classification, 1D convolutional neural network, ECG monitoring, ECG record classification, light-weight wearable device, patient-specific ECG classification, patient-specific electrocardiogram classification, supraventricular ectopic beat detection, Convolution, Databases, Electrocardiography, Feature extraction, Neural networks, Neurons, Training",

author = "S. Kiranyaz and T. Ince and R. Hamila and M. Gabbouj",

note = "EXT={"}Hamila, R.{"} EXT={"}Kiranyaz, S.{"} EXT={"}Ince, T.{"}; Annual International Conference of the IEEE Engineering in Medicine and Biology Society ; Conference date: 01-01-1900",

year = "2015",

month = aug,

day = "1",

doi = "10.1109/EMBC.2015.7318926",

language = "English",

isbn = "978-1-4244-9270-1",

pages = "2608--2611",

booktitle = "2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)",

}

Kiranyaz, S, Ince, T, Hamila, R & Gabbouj, M 2015, Convolutional Neural Networks for patient-specific ECG classification. in 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). pp. 2608-2611, Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 1/01/00. https://doi.org/10.1109/EMBC.2015.7318926

TY - GEN

T1 - Convolutional Neural Networks for patient-specific ECG classification

AU - Kiranyaz, S.

AU - Ince, T.

AU - Hamila, R.

AU - Gabbouj, M.

N1 - EXT="Hamila, R." EXT="Kiranyaz, S." EXT="Ince, T."

PY - 2015/8/1

Y1 - 2015/8/1

N2 - We propose a fast and accurate patient-specific electrocardiogram (ECG) classification and monitoring system using an adaptive implementation of 1D Convolutional Neural Networks (CNNs) that can fuse feature extraction and classification into a unified learner. In this way, a dedicated CNN will be trained for each patient by using relatively small common and patient-specific training data and thus it can also be used to classify long ECG records such as Holter registers in a fast and accurate manner. Alternatively, such a solution can conveniently be used for real-time ECG monitoring and early alert system on a light-weight wearable device. The experimental results demonstrate that the proposed system achieves a superior classification performance for the detection of ventricular ectopic beats (VEB) and supraventricular ectopic beats (SVEB).

AB - We propose a fast and accurate patient-specific electrocardiogram (ECG) classification and monitoring system using an adaptive implementation of 1D Convolutional Neural Networks (CNNs) that can fuse feature extraction and classification into a unified learner. In this way, a dedicated CNN will be trained for each patient by using relatively small common and patient-specific training data and thus it can also be used to classify long ECG records such as Holter registers in a fast and accurate manner. Alternatively, such a solution can conveniently be used for real-time ECG monitoring and early alert system on a light-weight wearable device. The experimental results demonstrate that the proposed system achieves a superior classification performance for the detection of ventricular ectopic beats (VEB) and supraventricular ectopic beats (SVEB).

KW - electrocardiography

KW - feature extraction

KW - medical signal processing

KW - neural nets

KW - signal classification

KW - 1D convolutional neural network

KW - ECG monitoring

KW - ECG record classification

KW - light-weight wearable device

KW - patient-specific ECG classification

KW - patient-specific electrocardiogram classification

KW - supraventricular ectopic beat detection

KW - Convolution

KW - Databases

KW - Electrocardiography

KW - Feature extraction

KW - Neural networks

KW - Neurons

KW - Training

U2 - 10.1109/EMBC.2015.7318926

DO - 10.1109/EMBC.2015.7318926

M3 - Conference contribution

SN - 978-1-4244-9270-1

SP - 2608

EP - 2611

BT - 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

T2 - Annual International Conference of the IEEE Engineering in Medicine and Biology Society

Y2 - 1 January 1900

ER -

Convolutional Neural Networks for patient-specific ECG classification

Abstract

Conference

Keywords

Publication forum classification

Access to Document

Fingerprint

Cite this