Litcius/Paper detail

COVID-19 Detection in X-ray Images using CNN Algorithm

Areej A.wahab Ahmed Musleh, Ashraf Y. A. Maghari

202048 citationsDOI

Abstract

Based on the best published research from Stanford University, the CheXNet algorithm was developed to diagnose and detect pneumonia from chest X-rays. To achieve better performance than experienced radiologists from the same university, simple changes were made to the algorithm to diagnose 14 pathological condition in the chest X-ray with a performance that exceeds all Previously developed deep learning [1]. In this paper, we experimented with applying a convolutional neural networks (CNN) algorithm in a similar way to the mechanism of work in CheXNet algorithm by using a dataset of 550 Chest X-ray images collected from Kaggle website, some of them are infected with Covid-19 virus. We had an acceptable prediction accuracy of 89.7% which is closed to the results of CheXNet algorithm.

Topics & Concepts

Convolutional neural networkComputer scienceCoronavirus disease 2019 (COVID-19)AlgorithmDeep learningArtificial intelligencePneumoniaSevere acute respiratory syndrome coronavirus 2 (SARS-CoV-2)Pattern recognition (psychology)MedicinePathologyInfectious disease (medical specialty)DiseaseInternal medicineCOVID-19 diagnosis using AIRadiomics and Machine Learning in Medical ImagingAI in cancer detection