Epidemics like the COVID-19 epidemic are major global problems, some of which spread rapidly, causing high mortality rates. The rapid spread of such diseases leaves behind the production and global use of medical testing tools, and during this period, the undetected disease spreads quickly and becomes hard to control. Therefore, non-medical digital technologies with faster disease prediction are needed. The studies on deep learning, a subfield of artificial intelligence, reveal that it is possible to predict disease more quickly. It will be possible to stop the pandemic progress with minimal damage if the best predictive approach is used. Lives are saved when the disease is detected early. A dataset containing low-dose CT scan images, gender, age, weight, COVID-19 PCR test result, and symptoms including cough, fever, shortness of breath, chest pain, and fatigue were used in this study. The primary goal is developing an assistive system for diagnosing COVID-19 disease based on deep learning using this dataset. The deep learning technique used in this study has a multi-branch architecture consisting of both classical artificial neural networks and convolutional neural networks. As a result of this study, it was concluded that the developed technique can be used not only for COVID-19 but also for different diseases and is open to development with data from more patients.