Diagnosis using pulmonary sounds and design of an electronic auscultation device

The aim of this thesis is to find out relations between sounds and pulmonarydiseases. Every pulmonary disease has distinguishing sound. Physicians diagnosepulmonary disease by using a stethoscope. This is a qualitative approach, but thisshould be measurable and recordable for a quantitative diagnosis. Mathematicaltransformations are applied to signals to obtain further information from that signalthat is not readily available in the raw signal. In this study, a literature survey isperformed on analyzing lung sounds and diagnosis as in computer based approach.There are many articles about electronic recording and analysis of sounds. They arenot completely distinguishable due to the mixing of heart, muscle and other sounds.Some of them can be filtered easily. It is possible to distinguish Chronic ObstructivePulmonary Disease (COPD) and other diseases using sounds. Unfortunately, there isno clear distinction among diseases. Nevertheless, digital signal processing and itsfiltering methods can be used to analyze pulmonary sounds. Many parameters (age,sex, smoking, pulmonary drugs, weight, muscle and heart sounds etc.) can affect thegroupings in the signal analyzing systems and their meanings. These can beeliminated by means of separation technique and filtering methods. New techniquesmay make to pre-diagnosis and pulmonary disease relations possible. Thus, soundrecordings of the all patients can be easily stored by using computer as sound library.In the study the patients are diagnosed and the sounds are recorded at UniversityHospital in Gaziantep. FFT and STFT techniques are used to analyze the soundrecordings. Pneumonia and wheezing characteristics are observed after the analysis.It is aimed to have a database about the patients; the sound recordings are stored in acomputer and will be used for revisiting and comparisons. Title includes design of anauscultation device. Unfortunately, thoughts are improved related to design of deviceand signal processing methods were indicated from computer based systems.Software is improved to be used for diagnosis by the clinicians.Key Words: Respiratory sound analysis, Respiratory disease, FFT, STFT