Performance Result for Detection of COVID-19 using Deep Learning
Harit Ahuja1, Shubharthi Dey2, Bagish Choudhury3, Aishwarya D4, Sanvinoth P S S5
1Harit Ahuja, Department of Computer Science and Engineering Department, SRM Institute of Science and Technology, Kattankulathur, Chennai, India.
2Shubharthi Dey, Department of Computer Science and Engineering Department, SRM Institute of Science and Technology, Kattankulathur, Chennai, India.
3Bagish Choudhury, Department of Computer Science and Engineering De-partment, SRM Institute of Science and Technology, Kattankulathur, Chennai, India.
4Aishwarya D, Department of Computer Science and Engineering Depart-ment, SRM Institute of Science and Technology, Kattankulathur, Chennai, India.
5Sanvinoth P S S, Department of Computer Science and Engineering Department, SRM Institute of Science and Technology, Kattankulathur, Chennai, India.
Manuscript received on April 20, 2020. | Revised Manuscript received on April 30, 2020. | Manuscript published on May 10, 2020. | PP: 699-703 | Volume-9 Issue-7, May 2020. | Retrieval Number: G5684059720/2020©BEIESP | DOI: 10.35940/ijitee.G5684.059720
Open Access | Ethics and Policies | Cite | Mendeley
© 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: The 2019 novel coronavirus (COVID-19), which has sprawled fleetly among masses residing in distant nations, had a prefatory juncture in China. From both a safeness and a lucrative outlook, it has staggered the world with its hasty diffusion with conjectural vicious generic repercussions for the masses. Consequent to the escalating cases daily, there is a constricted fraction of COVID-19 inspection kits acquirable in healthcare institutions. Ergo, to obviate COVID-19 propagating betwixt masses, it is imperative to enforce an instinctive unveiling network as a prompt jack legging diagnosis appendage. The contemplated method embroils a convolutional neural network- based model, namely ResNet50, concerted with a Fully Connected Layer (FCL), reinforced by Rectified Linear Unit (ReLU) for the unearthing of coronavirus pneumonia imparted sufferer by harnessing chest X-ray radiographs. The endorsed classification model, i.e. resnet50 affirmed by FCL and ReLU, compassed accuracy of 94% for unearthing COVID-19. When equated to diverse classification models, the purported model is preeminent. The aftereffect is premised on the attested X-ray images from the data appropriable in the arsenal of Kaggle.
Keywords: ResNet50, Deep learning, COVID-19.
Scope of the Article: Deep Learning