Pengolahan Data Isyarat EKG pada Emosi Manusia berbasis Transformasi Wavelet

JAENAL ARIFIN, Ir. Oyas Wahyunggoro, M.T., Ph.D ; Ir. Rudy Hartanto, M.T

2014 | Tesis | S2 Teknik Elektro

Sistem pengolahan data isyarat jantung dapat dikaji dan diterapkan dalam beberapa penelitian, baik psikologis atau monitoring perawatan keadaan jantung manusia. Dengan menggunakan electrokardiografi aktivitas detak jantung manusia dapat direkam atau dimonitoring. Aspek emosi seseorang dapat dimungkinkan diamati pada saat orang tersebut telah dilakukan perekaman aktivitas detak jantungnya. Data emosi dikumpulkan dan dikelompokkan berdasarkan emosi yang nantinya diamati. Keadaan emosi yang diamati kondisi normal, senang dan sedih. Khusus pada kondisi sedih penelitian ini diberikan rangsangan berupa mendengarkan musik dan menonton film. Tahapan penelitian ini meliputi pengukuran, perekaman, print out data ekg dan selanjutnya discanning. Tahap berikutnya pre-processing data citra ekg, meliputi median filter, penentuan nilai ambang dan morphology. Tahap selanjutnya menggunakan methode thinning dan spasial to time dalam pengolahan citra ekgnya. Tahap terakhir ekstraksi ciri menggunakan transformasi wavelet. Dari ekstraksi ciri didapatkan energi rerata sinyal. Energi rerata ini yang dapat mengidentifikasi emosi seseorang dalam keadaan normal, senang dan sedih.

Heart signal data processing system can be studied and applied in several research works, psychological studies or in the monitoring and care of human heart state. Through the use of electrocardiogram (ECG), the activity of human heartbeat can be recorded and monitored. The emotional aspect of a human can likely be observed when the activity of his/her heartbeats had been recorded previously. The data of emotions were then collected and grouped based on emotions that will be observed. The emotional states being observed were normal, happy and sad. Particularly on sad condition, subjects were given stimulations by listening to the music and watching a movie. Stages of research included measurement, recording, ECG data printout and scanning. The next step was pre-processing of ECG image data. This involved median filter, threshold and morphology. Subsequently, thinning method was employed in the processing of ECG images. The last stage was feature extraction by using wavelet transformation. From this stage, the average energy of signals was obtained. The experiment result shows that the average energy from the last stage can be used to identify someone's emotion whether it is normal, happy or sad.

Kata Kunci : Electrocardiography (ECG), wavelet transformation, physiological signal processing.