Deteksi Central Sleep Apnea Menggunakan Sensor Pyroelectric Infrared (PIR) dengan Penghapusan Motion Artifact

Hidayat, Bahari Noor (2025) Deteksi Central Sleep Apnea Menggunakan Sensor Pyroelectric Infrared (PIR) dengan Penghapusan Motion Artifact. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5023211018-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (7MB) | Request a copy

Abstract

Central sleep apnea (CSA) adalah gangguan tidur yang ditandai dengan berhentinya aktivitas pernapasan akibat kegagalan sistem saraf pusat dalam mengirimkan sinyal ke otot pernapasan. Penyakit ini umum terjadi pada penderita gagal jantung, pasien dengan kelainan neurologis, serta bayi prematur. Metode diagnosis utama, polisomnografi (PSG), memiliki keterbatasan seperti biaya tinggi dan ketidaknyamanan. Penelitian ini mengusulkan sebuah sistem deteksi CSA dengan sensor pyroelectric infrared (PIR) dilengkapi dengan algoritma penghapusan motion artifact untuk mengatasi gangguan gerakan selama tidur. Sistem dirancang untuk merekam sinyal menggunakan 3 buah sensor PIR yang mendeteksi gerakan toraks selama pernapasan dengan sampling rate 50 Hz. Sinyal diolah secara terpisah menggunakan Empirical Mode Decomposition (EMD) untuk memperoleh Intrinsic Mode Function (IMF) yang merepresentasikan sinyal pernapasan. Sinyal yang rusak direkonstruksi berdasarkan karakteristik sinyal di sekitarnya. Hasil pengolahan EMD dari masing-masing sinyal kemudian dilakukan fusion menggunakan kalman filter sehingga didapat satu sinyal akhir. Estimasi laju pernapasan dilakukan dengan analisis peak dengan hasil pengujian menunjukan MAE sebesar 1,66 BrPM, RMSE sebesar 2,02 BrPM, dan CAND sebesar 0,94 untuk posisi tidur telentang. Untuk posisi tidur menyamping, didapat MAE sebesar 2,33 BrPM, RMSE sebesar 3,20 BrPM, dan CAND sebesar 0,89. Sedangkan dalam kondisi pemberian motion artifact, didapat MAE sebesar 2,24 BrPM, RMSE sebesar 3,05 BrPM, dan CAND sebesar 0,92. Hasil deteksi CSA yang dilakukan dengan analisis amplitudo sinyal dan pemberian threshold 10 % dari baseline menunjukkan akurasi 51,8% dan sensitifitas 29%, sedangkan threshold secara manual menunjukkan akurasi 66,1%, dan sensitifitas 54,8%.
=================================================================================================================================
Central sleep apnea (CSA) is a sleep disorder characterized by the cessation of breathing activity due to the failure of the central nervous system to send signals to the respiratory muscles. This condition commonly occurs in patients with heart failure, neurological disorders, and in premature infants. The main diagnostic method, polysomnography (PSG), has limitations such as high cost and discomfort. This study proposes a CSA detection system using pyroelectric infrared (PIR) sensors equipped with a motion artifact removal algorithm to address movement disturbances during sleep. The system is designed to record signals using three PIR sensors that detect thoracic movements during breathing with a sampling rate of 50 Hz. The signals are processed separately using Empirical Mode Decomposition (EMD) to obtain Intrinsic Mode Functions (IMFs) that represent the respiratory signals. Corrupted signals are reconstructed based on the characteristics of the surrounding signals. The EMD results from each signal are then fused using a Kalman filter to obtain a single final signal. Respiratory rate estimation is carried out using peak analysis, with test results showing a MAE of 1.66 BrPM, RMSE of 2.02 BrPM, and CAND of 0.94 in the supine sleeping position. In the lateral sleeping position, a MAE of 2.33 BrPM, RMSE of 3.20 BrPM, and CAND of 0.89 were obtained. Under conditions with induced motion artifacts, an MAE of 2.24 BrPM, RMSE of 3.05 BrPM, and CAND of 0.92 were achieved. CSA detection based on signal amplitude analysis with a threshold of 10% from baseline yielded an accuracy of 51.8% and sensitivity of 29%, while manual thresholding achieved an accuracy of 66.1% and sensitivity of 54.8%.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	central sleep apnea, sensor PIR, respiratory rate, motion artifact, empirical mode decomposition, central sleep apnea, PIR sensor, respiratory rate, motion artifact, empirical mode decomposition
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5102.9 Signal processing. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7871.674 Detectors. Sensors
Divisions:	Faculty of Electrical Technology > Biomedical Engineering > 11410-(S1) Undergraduate Thesis
Depositing User:	Bahari Noor Hidayat
Date Deposited:	02 Feb 2026 05:42
Last Modified:	02 Feb 2026 05:42
URI:	http://repository.its.ac.id/id/eprint/131641

Actions (login required)

View Item