ABSTRAKSI: Sistem telemonitoring ECG telah banyak dikembangkan, hingga dalam ukuran yang Portabel yaitu seperti Pocket PC sehingga pasien dapat memonitor keadaannya sendiri.Bahkan dalam beberapa tahun terakhir Elektrokardiogram (ECG) memainkan peran penting dalam diagnosa penyakit jantung, Human Computer Interface (HCI), dan penilaian kondisi emosional seseorang, dll.Secara umum, sinyal ECG dipengaruhi oleh Power line interference, Electrode contact noise, Motion artifacts/Baseline wandering, EMG noise, dan Instrumentation noise.Dalam rangka mempertahankan sinyal morfologi ECG, beberapa penelitian telah mengadopsi menggunakan metode preprocessing dengan cara yang berbeda-beda. Tetapi dalam penelitian ini, algoritma yang digunakan untuk mereduksi ganguan-gangguan yang dapat mempengaruhi sinyal ECG adalah dengan menggunakan Discrete Wavelet Transform (DWT) berbasis Waveletdetrend dan denoisingdengan menggunakan teknik thresholding, dimana teknik thresholding yang digunakan adalah soft thresholding untuk menghapus tiga sumber utama ganguan dari sinyal yang diperoleh ECG yaitu, baseline wandering, power line interference dan high frequency noises during data acquisition, dengan mengunakan fungsi Wavelet ( “Haar” , “db02”, “db03”, “db04” dan “db06”) dengan tingkat dekomposisi (“1”, “2”, dan “3”).
Ukuran kinerja untuk memilih fungsi Wavelet yang terbaik adalah dengan melihat hasil MSE dari masing-masing Wavelet yang diuji cobakan.Wavelet yang memeberikan hasil MSE yang terbaiklah yang akan diimplementasikan dalam “Perancangan dan Implementasi Telemonitoring Sistem Electrocardiograph (ECG) dengan Noise Reduction Berbasis Wavelet”.Hasil percobaan menunjukkan Wavelet"Haar" dapat menghasilkan nilai MSE yang lebih optimal.Kata Kunci : Portable telemadicine unit, Base unit, Noise reduction, Wavelet dan Sistem telemonitoringABSTRACT: ECG telemonitoring system has been developed into a small-portable size like a Pocket PC so the patients can monitor their heart condition by themselves. Today, Electrocardiogram (ECG) plays an important role in Human Computer Interface (HCI), diagnosing heart disease and assessing human emotional conditions. In general, the ECG signal is influenced by the power line interference, Electrode contact noise, motion artifacts / Baseline wandering, EMG noise, and Instrumentation noise. To maintain the morphology of the ECG signal, several ways of preprocessing methods have been adoped in several studies. This study applied algorithms to reduce the interferences affecting the ECG signal. They were Discrete Wavelet Transform (DWT)-based wavelet detrend and wavelet denoising using thresholding techniques. The thresholding technique was soft thresholding; this was to remove the three main disturbing sources of the ECG signal, namely baseline wandering, power line interference and high frequency noises during data acquisition, using wavelet functions ( " Haar " , " db02 " , " db03 " , " db04 " and " db06 " ) on the level of decomposition ( " 1 " , " 2 " , and " 3 " ).
The best performance of the wavelet was obtained by testing each wavelet and comparing the results of the MSE. The wavelet with the best MSE was then implemented in "Wavelet Based Design and Implementation of Electrocardiograph (ECG) Telemonitoring System with Noise Reduction". The results of the experiments showed that Wavelet "Haar" produced the optimum MSE score.Keyword: Telemedicine Portable Unit, Base Unit, Noise Reduction, Wavelet and Telemonitoring Systems