Detection of heart diseases using artificial intelligence algorithms and smart phones / tablets

Within the scope of this thesis, an artificial intelligence supported mobile application has been developed to analyze heart sounds in terms of disease using mobile devices and digital stethoscope. The application in question uses the heart sound obtained with a digital stethoscope. It produces disease diagnosis results every five seconds by analyzing heart sounds in the artificial intelligence module. The forecast is presented on the screen of the mobile device, and the graphic of the heart sound is also displayed on the screen. All calculations are performed instantly on the hardware of the mobile device. Therefore, it is not necessary to transfer the heart sound to the cloud using the internet and to perform the calculations on the cloud hardware. The developed mobile application has been tested on 162 patients in the Department of Cardiology Surgery of Ankara City Hospital, and it has been determined that the mobile application can predict the disease with an accuracy rate of 92%. This prediction accuracy is a very good result, as such a high disease prediction rate can only be achieved by well-trained cardiologists. As a result, it is considered that auscultation, which is one of the most common and oldest methods used in the diagnosis of heart diseases, will undergo a radical change with the application of artificial intelligence algorithms on mobile devices and will enable pre-diagnosis to be made independently of the specialist doctor.