COMPARISON OF SOUND ACOUSTIC PARAMETERS IN VOWEL PRONUNCIATION BETWEEN HEALTHY SUBJECTS AND CORONARY HEART PATIENTS

• المؤلف الرئيسي: Khairina, Jasmine
• التنسيق: Theses
• اللغة: Indonesia
• الوصول للمادة أونلاين: https://digilib.itb.ac.id/gdl/view/79352

The acoustic characteristics of voice have been associated with a number of health problems. Research on voice acoustic continues to develop and provides links to various organs and mechanisms of the human body. One of them is the circulatory process involving the heart. This study aims to investigate the correlation between human voice and coronary heart disease through voice analysis based on acoustic parameters which is done by comparing the value of sound parameters from healthy subjects and coronary heart patient subjects. The number of samples in this study were 36 healthy subjects with an age range of 26-78 years, and 33 samples of coronary heart patient subjects with an age range of 24-71 years. We perform quantitative analysis the sounds of five vowels namely /a/, /i/, /u/, /e/, and /o/ using PRAAT software, to obtain five values of acoustic parameters, namely formant frequency, Harmonic to Noise Ratio (HNR), jitter, shimmer, and intensity. We analyse the oscillogram, spectrogram, and wave spectrum over a specific time period. By using the Mann Whitney U test, we identify several parameters which can distinguish healthy subjects from subjects with coronary heart disease. A significant difference is shown by the Shimmer parameter with an average p-value of 0,0007, while the parameter that does not differ is F4 (p-value 0.704). If we look at each vowel, the vowel /i/ (p-value 0,186) can differentiate between healthy subjects and subjects with coronary heart disease well, while the vowel /a/ (p-value 0,417) cannot differentiate between the two subjects.