Klasifikasi Cardio Rhythm pada Sinyal Ecg menggunakan Convolutional Neural Network

Abdullah, Muh Akram (2019) Klasifikasi Cardio Rhythm pada Sinyal Ecg menggunakan Convolutional Neural Network. Other thesis, Institut Teknlogi Sepuluh Nopember.

[thumbnail of 05111540000050-Undergraduate_Theses.pdf] Text
05111540000050-Undergraduate_Theses.pdf
Restricted to Repository staff only

Download (1MB) | Request a copy

Abstract

EKG (Elektrokardiogram) adalah suatu representasi dari pergerakan otot jantung yang didapat dengan beberapa pemeriksaan menggunakan alat elektrokardiograf. Biasanya, pemeriksaan EKG dilakukan untuk menilai apakah jantung dalam keadaan normal atau tidak. Beberapa hal yang ditunjukan oleh pemeriksaan EKG adalah laju (kecepatan denyut nadi), ritme denyut nadi dan kekuatan serta timing sinyal listrik saat melewati masing-masing bagian jantung. Pada tugas akhir ini, penulis mengusulkan sebuah sistem untuk mengklasifikasi data signal ecg yang terbagi ke dalam 4 kelas yaitu normal, AF (Atrial Fibrilation), other rhythm dan noisy. Tujuannya adalah untuk meningkatkan akurasi klasifikasi penyakit pada jantung manusia dengan metode Convolutional Neural Network. Data pelatihan dan uji coba diambil dari dataset “the PhysioNet/Computing in Cardiology Challenge 2017”. Terdapat dua uji coba yang dilakukan yaitu praproses terhadap data antara lain dilakukan pemotongan data signal dengan menyeimbangkan data tiap kelas dan menghitung spektogram satu sisi dari sinyal EKG domain dan waktu serta input dan output dan parameter dari model berupa optimizer yang digunakan. Hasil uji coba yang didapatkan dari proses pemotongan data dengan nilai akurasi sebesar 83,93% dan dari proses menghitung spektogram satu sisi dari sinyal EKG domain dan waktu dengan nilai akurasi sebesar 76.38%.
===================================================================================================================================
The ECG is a representation of the movement of the heart muscle obtained by several examinations using an electrocardiograph device. Usually, an EKG is performed to assess whether the heart is normal or not. Some of the things shown by an EKG are the rate (pulse rate), pulse rate and strength and timing of the electrical signal as it passes through each part of the heart. In this final project, the authors propose a system to classify signal ECG data which is divided into 4 classes, namely normal, AF (Atrial Fibrilation), other rhythm and noisy. The aim is to improve the accuracy of the classification of diseases of the human heart by the Convolutional Neural Network method. Training data and trials were taken from the "The PhysioNet / Computing in Cardiology Challenge 2017" dataset. There are two trials carried out, namely preprocessing on data, among others, cutting data signals by balancing each class of data and calculating a one-sided spectogram from the ECG signal domain and time and input and output and parameters of the model in the form of optimizer used. The test results obtained from the process of cutting data with an accuracy value of 83.93% and from the process of calculating a one-sided spectogram from the ECG signal domain and time with an accuracy value of 76.38%.

Item Type: Thesis (Other)
Uncontrolled Keywords: Convolutional Neural Network, Signal ECG, Dataset the PhysioNet/Computing in Cardiology Challenge 2017.
Subjects: Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science)
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5102.9 Signal processing.
Divisions: Faculty of Information Technology > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User: Muh. Akram Abdullah
Date Deposited: 23 Jul 2026 04:29
Last Modified: 23 Jul 2026 04:29
URI: http://repository.its.ac.id/id/eprint/65361

Actions (login required)

View Item View Item