Paroxysmal atrial fibrillation classification from single-lead ECG using combined CNN and LSTM

Name: Nopporn Subsermsong

LCSH: Cardiovascular system -- Diseases -- Diagnosis

LCSH: Neural networks (Computer science)

LCSH: Convolutions (Mathematics)

Abstract: Atrial Fibrillation (AF) is a type of cardiac arrhythmia which has a close connection to many cardiovascular diseases. Detection of atrial fibrillation is difficult to detect since many cases are usually undiagnosed. เท particular, paroxysmal atrial fibrillation (PAF) is a form of AF that occurs occasionally and has a lower probability of being detected. เท this study, we used the data from Dynamic ECG Recordings: The 4th China Physiological Signal Challenge 2021. The single lead of ECG signals from a 30 second recording window, over a 15-minute data segment, are fed into a deep learning (DL) model with a bandpass filtering technique for denoising ECG signals and SMOTE for solving the imbalanced datasets. The DL model consists of Convolution Neural Network (CNN) and Long-Short Term Memory (LSTM) for classification of PAF which give the accuracy of lead I of the ECG signal is 94.15% and the accuracy of lead II of the ECG signal is 95.18%

King Mongkut's Institute of Technology Ladkrabang. Central Library

Address: BANGKOK

Email: Lifelong@kmitl.ac.th

Name: Laor Boongasame

Role: Thesis Advisor

Created: 2022

Modified: 2566-05-23

Issued: 2023-05-23

วิทยานิพนธ์/Thesis

application/pdf

eng

DegreeName: Master of Science

Level: Master's Degree

Descipline: Data Science and Analytics

Grantor: King Mongkut's Institute of Technology Ladkrabang

©copyrights King Mongkut's Institute of Technology Ladkrabang

ลำดับที่.	ชื่อแฟ้มข้อมูล	ขนาดแฟ้มข้อมูล	จำนวนเข้าถึง	วัน-เวลาเข้าถึงล่าสุด
1	Nopporn Subsermsong.pdf	1.87 MB	4	2024-07-28 01:03:13

ใช้เวลา

0.024429 วินาที