Article: COVID-19 multilevel severity classification using FHGSO enabled DKN EfficientNet Journal: International Journal of Advanced Mechatronic Systems (IJAMECHS) 2025 Vol.12 No.1 pp.40 - 59 Abstract: This research developed the fractional hunger game search optimisation deep Kronecker EfficientNet (FHGSO_DK EfficientNet) for classifying coronavirus disease 2019 (COVID-19). First, the image is processed using a Gaussian filter. Additionally, lung segmentation is performed using Bayesian fuzzy clustering (BFC). Then, the image segmentation process is performed. After that, feature extraction is done by feature extraction including convolutional neural network (CNN), fast robust features (SURF), fast regularisation and short band (ORB), single binary model (MBP) and statistical properties. Next, the initial classification is performed using deep Kronecker EfficientNet (DKN EfficientNet) to distinguish between normal and abnormal cases. This process is optimised with fractional hunger game search optimisation (FHGSO). If the case is abnormal, second level categorisation is executed done utilising DKN EfficientNet. For greater accuracy and efficiency, future works ought to concentrate on accurate datasets accessible in the current scenario. Inderscience Publishers - linking academia, business and industry through research

Title: COVID-19 multilevel severity classification using FHGSO enabled DKN EfficientNet

Authors: G. Venkata Rami Reddy; Abboju Niranjan

Addresses: Department of Computer Science Engineering, Jawaharlal Nehru Technological University Hyderabad (JNTUH), Kukatpally Housing Board Colony, Kukatpally, Hyderabad, Telangana 500085, India ' Department of Computer Science Engineering, Jawaharlal Nehru Technological University Hyderabad (JNTUH), Kukatpally Housing Board Colony, Kukatpally, Hyderabad, Telangana 500085, India

Abstract: This research developed the fractional hunger game search optimisation deep Kronecker EfficientNet (FHGSO_DK EfficientNet) for classifying coronavirus disease 2019 (COVID-19). First, the image is processed using a Gaussian filter. Additionally, lung segmentation is performed using Bayesian fuzzy clustering (BFC). Then, the image segmentation process is performed. After that, feature extraction is done by feature extraction including convolutional neural network (CNN), fast robust features (SURF), fast regularisation and short band (ORB), single binary model (MBP) and statistical properties. Next, the initial classification is performed using deep Kronecker EfficientNet (DKN EfficientNet) to distinguish between normal and abnormal cases. This process is optimised with fractional hunger game search optimisation (FHGSO). If the case is abnormal, second level categorisation is executed done utilising DKN EfficientNet. For greater accuracy and efficiency, future works ought to concentrate on accurate datasets accessible in the current scenario.

Keywords: COVID-19; speeded-up robust features; SURF; chest computed tomography scan; CT scan; hunger game search optimisation; HGSO; deep Kronecker network; DKN.

DOI: 10.1504/IJAMECHS.2025.144590

International Journal of Advanced Mechatronic Systems, 2025 Vol.12 No.1, pp.40 - 59

Received: 30 May 2024
Accepted: 18 Sep 2024
Published online: 23 Feb 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: COVID-19 multilevel severity classification using FHGSO enabled DKN EfficientNet

Keep up-to-date