Pemodelan Sinyal Elektrokardiogram Menggunakan Markov Switching Autoregressive (MSAR) dalam Monitoring Kondisi Jantung Aritmia, Gagal Jantung Kongestif, dan Ritme Sinus Normal

Sianturi, Elizabeth (2025) Pemodelan Sinyal Elektrokardiogram Menggunakan Markov Switching Autoregressive (MSAR) dalam Monitoring Kondisi Jantung Aritmia, Gagal Jantung Kongestif, dan Ritme Sinus Normal. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003211119_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (12MB) | Request a copy

Abstract

Gangguan jantung seperti aritmia dan gagal jantung kongestif adalah penyebab kematian utama global, sehingga pemantauan akurat melalui sinyal elektrokardiogram (EKG) sangat krusial. EKG adalah alat diagnostik utama dalam mendeteksi gangguan jantung, namun karakteristik sinyal EKG bersifat fluktuatif, non-linier, dan non-stasioner, sehingga memerlukan pendekatan permodelan yang adaptif terhadap perubahan data. Penelitian ini bertujuan untuk memodelkan sinyal EKG menggunakan metode Markov Switching Autoregressive (MSAR) guna menganalisis pola transisi antar kondisi jantung pada tiga kategori utama, yaitu Arrhythmia (ARR), Congestive Heart Failure (CHF), dan Normal Sinus Rhythm (NSR). Data yang digunakan merupakan sinyal EKG teranotasi dengan fokus pada analisis probabilitas transisi antar state untuk memahami dinamika perubahan kondisi jantung. Dua pendekatan digunakan dalam pemodelan MSAR, yaitu EM-Gaussian dan Bayesian-Exponential Power, dengan pemilihan model terbaik berdasarkan nilai Akaike’s Information Criterion (AIC) terkecil. Hasil pemodelan dengan pendekatan EM-Gaussian menunjukkan bahwa residual pada masing-masing regime tidak mengikuti distribusi normal, sementara model terbaik yang dihasilkan pada ketiga kondisi jantung adalah MS(3)AR(1). Adapun pendekatan Bayesian-Exponential Power mampu mengelompokkan kondisi jantung ke dalam tiga regime, yaitu kondisi tenang, kondisi transisi bergejolak, dan kondisi gejolak ekstrem. Berdasarkan hasil analisis, nilai Average Run Length (ARL) tertinggi untuk regime gejolak ekstrem tercatat pada kondisi ARR sebesar 8,16 milidetik, CHF sebesar 6,99 milidetik , dan NSR sebesar 2,96 milidetik. Temuan ini memberikan gambaran kuantitatif terhadap durasi dominasi masing-masing regime pada berbagai kondisi jantung.
=========================================================================================================================================
Cardiac disorders such as arrhythmia and congestive heart failure are among the leading causes of death worldwide, making accurate monitoring through electrocardiogram (ECG) signals critically important. ECG is a primary diagnostic tool for detecting heart conditions; however, its signal characteristics are highly fluctuating, nonlinear, and non-stationary. This necessitates the use of modeling approaches that can adapt to changes in the data. This study aims to model ECG signals using the Markov Switching Autoregressive (MSAR) method to analyze transition patterns between cardiac states across three major categories: Arrhythmia (ARR), Congestive Heart Failure (CHF), and Normal Sinus Rhythm (NSR). The dataset consists of annotated ECG signals, with a focus on analyzing transition probabilities between states to understand the dynamics of cardiac condition changes. Two approaches were employed in the MSAR modeling: EM-Gaussian and Bayesian-Exponential Power, with model selection based on the smallest Akaike’s Information Criterion (AIC) value. The EM-Gaussian approach resulted in residuals that did not follow a normal distribution, while the best-fitting model for all three cardiac conditions was MS(3)AR(1). Meanwhile, the Bayesian-Exponential Power approach was able to classify cardiac conditions into three regimes resting state, transitional state, and extreme fluctuation state. Based on the analysis, the highest Average Run Length (ARL) for the extreme fluctuation regime was observed in ARR at 8.16 milliseconds, followed by CHF at 6.99 milliseconds, and NSR at 2.96 milliseconds. These findings provide a quantitative representation of the dominant duration of each regime across different heart conditions.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Jantung, EKG, Markov Switching Autoregressive, Aritmia, Gagal Jantung Kongestif, Ritme Sinus Normal, Gaussian, Exponential Power. Heart, EKG, Markov Switching Autoregressive, Arrhythmia, Congestive Heart Failure, Normal Sinus Rhythm, Gaussian, Exponential Power.
Subjects:	Q Science > QA Mathematics > QA274.2 Stochastic analysis Q Science > QA Mathematics > QA279.5 Bayesian statistical decision theory.
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Elizabeth Sianturi
Date Deposited:	06 Aug 2025 05:54
Last Modified:	06 Aug 2025 05:56
URI:	http://repository.its.ac.id/id/eprint/126340

Actions (login required)

View Item