Perancangan Alat Deteksi Ritme Shockable Dan Non-Shockable Dari Henti Jantung Menggunakan Machine Learning

Shafira, Nadirra Shifa Zuhra (2022) Perancangan Alat Deteksi Ritme Shockable Dan Non-Shockable Dari Henti Jantung Menggunakan Machine Learning. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 07311840000063-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (3MB)

Abstract

Cardiac Arrest atau henti jantung adalah salah satu penyakit yang menyumbang kematian terbesar di dunia mencapai 7 juta per tahun. Henti jantung memiliki dua ritme yang secara fisiologis mirip, yaitu ritme shockable dan non- shockable. Pemberian shock kepada penderita yang non-shockable dapat mengakibatkan kematian. Pendeteksian yang tertanam pada Automated External Defibrillator (AED) memiliki akurasi yang berbeda-beda bergantung pada merknya, tetapi rata-rata AED dinilai hanya baik dalam menilai ritme VF (Ventrikular Fibrilasi) yang termasuk ritme shockable, sedangkan masih ada ritme shockable lain sehingga perlu dikembangkan dari segi pendeteksinya. Oleh karena itu, pada penelitian ini diusulkan perangkat pendeteksi ritme henti jantung menggunakan perangkat ECG (Electrocardiography) yang dilengkapi klasifikasi ritme shockable dan non-shockable dengan gabungan ekstraksi fitur Count1, Count2 dan Lk dan untuk pengambilan keputusannya digunakan Multilayer Perceptron (MLP). MLP merupakan salah satu arsitektur deep learning dengan model ANN (Artificial Neural Network). Deep learning termasuk bagian dari machine learning. MLP terdiri dari layer dengan neuron yang saling terkoneksi atau dapat disebut Fully- Connected Layer. MLP menggunakan activation function yang bersifat non-linear pada seluruh neuron di hidden layer. MLP ini membutuhkan setidaknya tiga layer neural network. Hasil performa algoritma yang didapatkan menggunakan data test didapatkan akurasi 94,15%, sensitivitas 89,61%, spesifisitas 98,70%, dan presisi 98,57%. Untuk data train didapatkan akurasi 95,61%, sensitivitas 93,86%, spesifisitas 97,36%, dan presisi 97,27%.
==============================================================================================================================
Cardiac Arrest or cardiac arrest is one of the diseases that accounts for the largest deaths in the world reaching 7 million per year. Cardiac arrest has two physiologically similar rhythms, namely shockable and non-shockable rhythms. Giving shock to a patient who is non-shockable can result in death. The detection that is embedded in the Automated External Defibrillator (AED) has different accuracy depending on the brand, but on average the AED is rated only good in assessing VF (ventricular fibrillation) rhythms which are shockable rhythms, while there are other shockable rhythms that need to be developed in terms of detection. Therefore, in this study, a device for detecting cardiac arrest rhythm using an ECG (Electrocardiography) device is proposed which is equipped with a classification of shockable and non-shockable rhythms with a combination of Count1, Count2 and Lk feature extraction and Multilayer Perceptron (MLP) is used for decision making. MLP is a deep learning architecture with an ANN (Artificial Neural Network) model. Deep learning is part of machine learning. MLP consists of layers with interconnected neurons or can be called Fully-Connected Layer. MLP uses non-linear activation functions on all neurons in the hidden layer. This MLP requires at least three neural network layers. The results of the algorithm performance obtained using test data obtained 94.15% accuracy, 89.61% sensitivity, 98.70% specificity, and 98.57% precision. For train data, the accuracy is 95.61%, sensitivity is 93.86%, specificity is 97.36%, and precision is 97.27%.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Henti jantung, non-shockable, shockable, AED, ECG, MLP. Cardiac arrest, non-shockable, shockable, AED, ECG, MLP.
Subjects:	R Medicine > R Medicine (General) > R857.M3 Biomedical materials. Biomedical materials--Testing.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Biomedical Engineering > 11410-(S1) Undergraduate Thesis
Depositing User:	Mr. Marsudiyana -
Date Deposited:	18 Jun 2026 06:56
Last Modified:	18 Jun 2026 06:56
URI:	http://repository.its.ac.id/id/eprint/133897

Actions (login required)

View Item