Oscultation device design and decision support system

Analysis of body sounds gives preliminary information about the state of the body and is a classical procedure performed as the first examination. This process is called auscultation and stethoscope is used for the procedure. From the invention of the stethoscope to the present day, it has shown a lot of development. The invention and use of the electronic stethoscope has made the procedures in the field of health faster and more practical. Thanks to the data recording feature of these stethoscopes, it has paved the way for important studies supported by artificial intelligence. The thesis work consists of two stages: First of all, it is to develop a new wearable multi-channel auscultation hardware and to provide artificial intelligence support analysis of the measured sound data. The first stage was the development of a wearable auscultation device. Thanks to the wireless communication module of the developed hardware, it offers remote treatment for healthcare professionals. In the system where we have created a multi-channel auscultation device, remote communication between the healthcare personnel and the patient will be possible thanks to the user interface. When the recorded voice recordings are sent, an artificial intelligence model has been developed that will list the region to which the voice belongs and possible diseases to the health personnel. It can also be used as a data sharing remote connection module to get opinions from different experts thanks to its user interface. The model was created by combining different data sets, creating a new data set and training this data set. The system first recognizes the Voice data regionally. After the region definition, it determines whether the voice is healthy or sick. If the voice data belongs to the patient, it is found that this disease belongs to which disease. This system will enable the development of home treatment service.