Improved Qubit Neural Network Based Computer Aided Detection Model for COVID-19 on Chest Radiographs

The significance of the advanced computation model for medical diagnoses has been gradually recognized with the growth in e-health. Specifically, in the pandemic COVID19, the advanced computational modelling role is crucial plays in understanding complex clinical data in future years. Computed tomography (CT) and Chest X-ray (CXR) images were presented for the diagnosis of COVID19. The deep learning (DL) method has proven better and more efficient in several medical imaging and computer vision applications. With the increase in the COVID outbreak, researcher workers are making use of the DL method to identify coronavirus infection. This study develops a hybrid squirrel search algorithm with improved deep learning-based detection (HSSA-IDLD) system for the diagnoses of COVID-19. The major intention of the HSSA-IDLD technique lies in the recognition and discrimination against COVID-19 in healthcare images. Initially, the medical images are processed using a contrast enhancement approach. For feature extraction, an improved EfficietNet model is utilized in this study. In the presented HSSA-IDLD technique, the HSSA is derived as a hyperparameter optimizer. Moreover, Qubit neural network (QBNN) approach was exploited for the recognition and classification of COVID-19. A widespread simulation process was carried out to demonstrate the enhanced performance of the HSSA-IDLD algorithm. An extensive comparison study reported improvements in the HSSA-IDLD technique compared to recent DL models.