Klasifikasi Aritmia Berbasis Morfologi Gelombang Komplek QRS Menggunakan Deep Learning

Kurniawan, Arief (2023) Klasifikasi Aritmia Berbasis Morfologi Gelombang Komplek QRS Menggunakan Deep Learning. Doctoral thesis, Institut Teknologi Sepuluh Nopember.

Text 07111960010020-Disertation.pdf - Accepted Version Restricted to Repository staff only until 1 September 2025. Download (1MB)

Abstract

Jantung berperan dalam sistem peredaran darah, aktivitas jantung menyebabkan beda potensial pada permukaan tubuh atau sinyal ECG. Dalam 1 periode, gelombang sinyal ECG terdiri dari gelombang P, komplek QRS dan T. Saat kondisi normal dan tanpa beraktifitas berat terdapat 60 sampai 100 periode gelombang sinyal ECG atau detak jantung. Jika kurang dari 60 atau lebih 100 detak jantung permenit seseorang mengalami aritmia. Klasifikasi aritmia untuk menetukan apakah detak jantung tersebut normal atau bukan normal. Salah satu klasifikasi aritmia yaitu klasifikasi berbasis morfologi gelombang komplek QRS pada sinyal ECG. Klasifikasi aritmia berbasis morfologi gelombang komplek QRS diawali dengan deteksi gelombang komplek QRS. Deteksi komplek QRS menggunakan algorima heuristik yang terdiri dari implementasi dari metoda QVAT. Klasifikasi aritmia berdasarkan bentuk atau morfologi gelombang QRS yang diusulkan menggunakan deep learning Multi layer perceptron (MLP). Input dari deep learning adalah 1 segmen gelombang komplek QRS. Klasifikasi aritmia dengan menggunakan metoda MLP mempunyai akurasi, sensitivitas, spesifisitas, dan prediksi positif sebesar 99,77%, 99,55%, 99,55%, dan 99,85%. Hasil ini menunjukkan klasifikasi aritmia yang diusulkan mempunyai kualitas yang lebih baik jika dibandingkan dengan metode lain.
=====================================================================================================================================
In the circulatory system, heart activity causes a potential difference on the body surface or ECG signal. Within one period, the ECG signal wave consists of P, QRS complexes and T waves. During normal conditions, there are 60 to 100 periods of ECG signal waves or heartbeats. If less than 60 or more than 100 heartbeats per minute, a person has an arrhythmia. Classification of arrhythmias determines whether the heartbeat is normal or not. One arrhythmia classification is based on the morphology of the QRS complex wave on the ECG signal. Classification of arrhythmias based on the morphology of the QRS complex wave begins with detecting the QRS complex wave. The detection of the QRS complex uses a heuristic algorithm consisting of implementing the QVAT method. Classification of arrhythmias based on the shape or morphology of the QRS wave is proposed using deep learning Multi-layer perceptron (MLP). The input from \textit{deep learning} is one segment of the QRS complex wave. Arrhythmia classification using the MLP method has an accuracy, sensitivity, specificity, and positive prediction of 99,77%, 99,55%, 99,55%, and 99,85%. These results indicate that the proposed arrhythmia classification has better quality when compared to other methods.

Item Type:	Thesis (Doctoral)
Uncontrolled Keywords:	Arrhythmia Detection, Arrhythmia Classification, Deep Learning, Electrocardiogram
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105.546 Computer algorithms
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20001-(S3) PhD Thesis
Depositing User:	Arief Kurniawan
Date Deposited:	07 Aug 2023 07:47
Last Modified:	15 May 2024 09:52
URI:	http://repository.its.ac.id/id/eprint/104215

Actions (login required)

View Item