Human Metapneumovirus (HMPV) Classification Using Deep Learning Algorithms

Recently, following the spread of the coronavirus and the emergence of cases hmpv virus, determining the type of pneumonia is essential for taking precautionary measures. Accurate and timely diagnosis is a major challenge X-rays are used for diagnosis, and examining X-ray images and extracting results is a burden on doctors. Therefore, the use of artificial intelligence techniques is crucial to reducing effort and time. In this study, deep learning techniques were used to build a model capable of accurately diagnosing the type of infection. The model was trained on COVID-19 data and then tested on COVID-19 and hmpv data. This study designed a model for the diagnosis of pneumonia. The deep learning techniques using  based on the spareable convolution with slandered convolution, two types of data were used: covid data and hmpv data. The model aims to extract patterns from COVID-19 data and generalize them for broader  applications. This model was modified and tested on COVID data and then used to diagnose hmpv. The proposed  achieved an accuracy of 99.8% and 99.9%.