Enhance Covid-19 Detection Using Inception and Attention Mechanism

Covid-19 was first discovered in Wuhan, a province in China in early 2020. This disease has a very powerful spread. Covid-19 disease will attack the lungs and cause pneumonia. Even many cases in the world cause death. Detecting Covid-19 using x-ray images is a very important method because it is inexpensive, the tools used are widely available in almost all hospitals or clinics. The use of deep learning algorithms in detecting Covid-19 is an effort that has been widely developed by researchers. Deep learning functions to automatically study data to distinguish types of lung diseases including Covid-19. However, several previous studies have addressed problems in detection that still contain detection errors. This study aims to develop an algorithm model based on Inception and Attention to produce more effective Covid-19 disease detection results. In this study, the use of a combination of the Inception and Attention algorithms succeeded in increasing accuracy by 0.96 based on testing using metric evaluations, namely accuracy tests, F1, recall, precision and confusion matrices and Area Under Curve (AUC). Compared with several studies using several types of deep learning, the combination of Inception and Attention successfully increases the effectiveness of classification results.