DEEP FILTER DAN BI-LSTM UNTUK PENINGKATAN KINERJA DELINEASI SINYAL ELECTROCARDIOGRAM

PERWIRA, MUHAMMAD IKHWAN and Nurmaini, Siti (2024) DEEP FILTER DAN BI-LSTM UNTUK PENINGKATAN KINERJA DELINEASI SINYAL ELECTROCARDIOGRAM. Undergraduate thesis, Sriwijaya University.

	Text RAMA_56201_09011282025077.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (7MB) | Request a copy
	Text RAMA_56201_09011282025077_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (7MB) | Request a copy
	Text RAMA_56201_09011282025077_0002085908_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (7MB)
	Text RAMA_56201_09011282025077_0002085908_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_56201_09011282025077_0002085908_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_56201_09011282025077_0002085908_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (2MB) | Request a copy
	Text RAMA_56201_09011282025077_0002085908_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (183kB) | Request a copy
	Text RAMA_56201_09011282025077_0002085908_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (199kB) | Request a copy
	Text RAMA_56201_09011282025077_0002085908_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy

Abstract

The delineation of ECG signals is often hindered by noise, such as baseline wandering and electrode motion. This study presents a robust model for ECG signal denoising and delineation into four classes: baseline, P wave, QRS complex, and T wave, using data from multiple sources. The denoising model, based on a Multibranch LANLD architecture, was trained with noisy signals from NSTDB and clean labels from QTDB, while LUDB was used for delineation training. Fine-tuning was done by replacing the CNN output layer with a Bi-LSTM and Dense layer. The model achieved denoising up to 23 dB and delineation F1-scores of 88.2% for baseline, 84.5% for P wave, 89.7% for QRS complex, and 80.6% for T wave, with an overall accuracy of 86.4%.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Delineasi, Denoising, Multibranch LANLD, Bi-LSTM, Sinyal EKG
Subjects:	T Technology > T Technology (General) > T1-995 Technology (General)
Divisions:	09-Faculty of Computer Science > 56201-Computer Systems (S1)
Depositing User:	Muhammad Ikhwan Perwira
Date Deposited:	26 Nov 2024 06:52
Last Modified:	26 Nov 2024 06:52
URI:	http://repository.unsri.ac.id/id/eprint/159889

Actions (login required)

View Item