Disease diagnosis with sound analysis

Nowadays, lung diseases are very common. Auscultation method is primarily used in diagnosis of lung disease. Auscultation method preferred by doctors for physical examination is a low cost and efficient noninvasive method. Classic and electronic stethoscopes are used for auscultation. Although the use of stethoscope in diagnosis of lung disease is a simple and inexpensive method, accurate diagnosis of the disease requires highly experienced medical personnel. Also, the human ear to hear sounds at lower frequencies is not sensitive. For this reason, using systems which will help the doctor in making decision is important. Computerized analysis methods being used in the analysis of respiratory sounds provides a solution to these problems and contributes practically to doctors who are too busy. Respiratory sounds received from patient with stethoscope contain important information about the pathology of the lung. Properties of adventitious sounds found in respiratory sound provide guidance to physicians in the diagnosis of certain diseases. In this study, respiratory sounds are taken from patients who have different lung disease (such as asthma, bronchiectasis, chronic obstructive pulmonary disease, idiopathic pulmonary fibrosis) and respiratory sound belonging these diseases diagnosed by a doctor in advance is analysed using signal processing methods such as Wavelet Transform, Linear Prediction Coding, Short Time Fourier Transform and Mel Frequency Cepstral Coefficients and disease diagnosis is achieved by a computer.