A Comparative Evaluation of Diverse Deep Learning Models for the COVID-19 Prediction
Bhautik Daxini1, M.K. Shah2, Rutvik K. Shukla3, Rohit Thanki4, Viral Thakar5
1Bhautik Daxini, Research Scholar, Department of Instrumentation and Control, Gujarat Technological University, Ahmedabad (Gujarat), India.
2Dr. M.K. Shah, Associate Prof. & Head, Department of Instrumentation & Control Engineering, Vishwakarma Government Engineering College, Chandkheda, (Gujarat), India.
3Rutvik K. Shukla, Assistant Prof., Department of Instrumentation & Control Engineering, Government Engineering College, Rajkot (Gujarat), India.
4Dr. Rohit Thanki, Data Scientist, KRiAN GmbH, Wolfsburg, Germany.
5Viral Thakar, Senior Machine Learning Engineer, Autodesk, Toronto, Ontario, Canada.
Manuscript received on 11 July 2023 | Revised Manuscript received on 18 July 2023 | Manuscript Accepted on 15 August 2023 | Manuscript published on 30 August 2023 | PP: 1-16 | Volume-12 Issue-9, August 2023 | Retrieval Number: 100.1/ijitee.I96960812923 | DOI: 10.35940/ijitee.I9696.0812923
Open Access | Editorial and Publishing Policies | Cite | Zenodo | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Deep learning methodologies are now feasible in practically every sphere of modern life due to technological advancements. Due to its high level of accuracy, deep learning can automatically diagnose and classify a wide range of medical conditions in the field of medicine. The coronavirus first appeared in Wuhan, China, in December 2019 and quickly spread worldwide. The COVID-19 pandemic presented significant challenges to the world’s healthcare system. PCR and medical imaging can be used to diagnose COVID-19. It hurts the health of people as well as the global economy, education, and social life. The most significant challenge in containing the rapid spread of the disease is identifying positive COVID-19 patients as promptly as possible. Because there are no automated tool kits, additional diagnostic equipment will be required. According to radiological studies, these images include essential information about the coronavirus. Accurate treatment of this virus and a solution to the problem of a lack of medical professionals in remote areas may be possible with the help of a specialized Artificial Intelligence (AI) system and radiographic pictures. We utilised pre-trained CNN models, including Xception, Inception, ResNet-50, ResNet-50V2, DenseNet-121, and MobileNetV2, to refine the COVID-19 classification analytics. In this paper, we investigate COVID-19 detection methods that utilise chest Xrays. According to the findings of our research, the pre-trained CNN Model that utilises MobileNetV2 performs better than other CNN techniques in terms of both solution size and speed. Our method may be of use to researchers fine-tuning the CNN model for efficient COVID-19 screening.
Keywords: Diabetes, Misinformation, Fake, NLP, Monitoring
Scope of the Article: Deep learning