-
An Effective Detection of Covid-19 Using Deep Feature Extraction with Enhanced Quantum Neural Support Vector Machine
- Karthick V
Assistant Professor, Department of Computer Science and Engineering Saveetha School of Engineering,Saveetha Institute of Medical and Technical Sciences, Chennai, India.
vkarthiksse@gmail.com
- Gayathri A
Professor, Department of Computer Science and Engineering, Saveetha School of Engineering, Saveetha Institute of Medical and Technical Sciences, Chennai, India.
saranyaresearchin@gmail.com
Keywords: Test
Abstract
Early detection of COVID-19 disease can help devise a suitable treatment plan and disease containment decisions. In the medical field, Computed Tomography (CT) scan provides better utility for diagnosis and severity information about covid-19. However, the detection of covid-19 is not possible due to automatically delineating the exact regions in chest CT scan images, and the severity of the disease is also not performed based on quantitative analysis. The newly proposed method has introduced an enhanced quantum neural support vector machine (EQN-SVM) model to classify the CT scan images whether the image is contracted with covid-19 (or) not. The operation is mainly classified into four parts: pre-processing, extraction, classification, and hyperparameter selection. Initially, the image pre-processing removes the redundant images and identifies the exact lung region. Secondly, the image extraction process is involved for the CT scan image using the Densely Connected Convolutional Network (DenseNet) approach. The EQN-SVM model is introduced based on the image classification process to detect the exact results. Finally, the hyperparameter selection uses the quantum-inspired differential evolution (QIDE) algorithm. The performance of the new method is evaluated using the MATLAB tool. The new proposed results are obtained in terms of Accuracy (99.15%), Precision (98.32%), Recall (99.13%), F-Score (98.11%) and Receiver operating characteristics (ROC) (0.98) curve.