Pemrosesan Mula untuk Deteksi Puncak Gelombang R

Irwanto, Naufal Athallah (2026) Pemrosesan Mula untuk Deteksi Puncak Gelombang R. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 07211840000061-Undergraduate-Thesis.pdf - Accepted Version Restricted to Repository staff only Download (5MB) | Request a copy

Abstract

Elektrokardiogram (ECG) merupakan rekaman aktivitas listrik jantung yang digunakan untuk mengamati irama dan denyut jantung. Salah satu bagian penting pada sinyal ECG adalah puncak gelombang R atau R-Peak yang terdapat di dalam kompleks QRS. Ketepatan deteksi R-Peak dapat dipengaruhi oleh gangguan seperti baseline wander, interferensi jaringan listrik, artefak gerakan, dan gangguan frekuensi tinggi. Penelitian ini bertujuan menganalisis pengaruh pemrosesan mula terhadap performa deteksi R-Peak menggunakan algoritma Pan–Tompkins yang dimodifikasi. Metode pemrosesan mula yang dibandingkan meliputi Bandpass Filter, Wavelet Denoising, dan kombinasi Bandpass Filter dengan Wavelet Denoising atau Hybrid, sedangkan sinyal Raw digunakan sebagai kondisi pembanding tanpa pemrosesan mula. Pengujian dilakukan menggunakan 48 rekaman MIT-BIH Arrhythmia Database dengan total 109.150 anotasi denyut. Seluruh skenario diproses menggunakan tahapan dan parameter deteksi yang sama. Performa deteksi dievaluasi berdasarkan nilai Sensitivity (Se), Positive Predictivity (+P), dan Detection Error Rate (DER). Hasil pengujian menunjukkan bahwa sinyal Raw menghasilkan performa agregat terbaik dengan Se sebesar 99,08%, +P sebesar 98,25%, dan DER sebesar 2,68%. Di antara metode pemrosesan mula, Bandpass Filter memberikan hasil terbaik dengan Se sebesar 97,16%, +P sebesar 99,51%, dan DER sebesar 3,32%. Wavelet Denoising memperoleh Se sebesar 97,44%, +P sebesar 97,80%, dan DER sebesar 4,75%, sedangkan Hybrid menghasilkan Se sebesar 95,52%, +P sebesar 99,15%, dan DER sebesar 5,30%. Hasil tersebut menunjukkan bahwa pemrosesan mula tidak selalu meningkatkan performa deteksi R-Peak. Pemilihan metode dan parameter pemrosesan mula perlu disesuaikan dengan karakteristik sinyal ECG serta algoritma deteksi yang digunakan.
====================================================================================================================================
An electrocardiogram (ECG) is a recording of the electrical activity of the heart that can be used to observe heart rhythm and heartbeat. One of the important components of an ECG signal is the R-wave peak, or R-peak, located within the QRS complex. The accuracy of R-peak detection can be affected by disturbances such as baseline wander, power-line interference, motion artifacts, and high-frequency noise. This study aims to analyze the effects of preprocessing on R-peak detection performance using a modified Pan–Tompkins algorithm. The preprocessing methods evaluated in this study include Bandpass Filtering, Wavelet Denoising, and a combination of Bandpass Filtering and Wavelet Denoising referred to as the Hybrid method, while the Raw signal is used as a baseline condition without additional preprocessing. The evaluation was conducted using 48 records from the MIT-BIH Arrhythmia Database, containing a total of 109,150 annotated heartbeats. All scenarios were processed using the same detection stages and parameters. Detection performance was evaluated based on Sensitivity (Se), Positive Predictivity (+P), and Detection Error Rate (DER). The results show that the Raw signal achieved the best aggregate performance, with a Se of 99.08%, a +P of 98.25%, and a DER of 2.68%. Among the preprocessing methods, the Bandpass Filter provided the best performance, with a Se of 97.16%, a +P of 99.51%, and a DER of 3.32%. Wavelet Denoising achieved a Se of 97.44%, a +P of 97.80%, and a DER of 4.75%, while the Hybrid method achieved a Se of 95.52%, a +P of 99.15%, and a DER of 5.30%. These results indicate that preprocessing does not necessarily improve R-peak detection performance. Therefore, the preprocessing method and its parameters should be selected according to the characteristics of the ECG signal and the detection algorithm being used.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Elektrokardiogram (ECG), Pemrosesan Mula, Bandpass Filter, Wavelet Denoising, Pan–Tompkins, Deteksi Puncak Gelombang R.Electrocardiogram (ECG), Preprocessing, Bandpass Filter, Wavelet Denoising, Pan–Tompkins, R-Wave Peak Detection.
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5102.9 Signal processing.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	Naufal Athallah Irwanto
Date Deposited:	28 Jul 2026 07:41
Last Modified:	28 Jul 2026 07:41
URI:	http://repository.its.ac.id/id/eprint/137451

Actions (login required)

View Item