Issuu

https://doi.org/10.22214/ijraset.2022.41697

April 2022

A Review on Covid – 19 Detection Techniques using X-ray Chest Images Mohammed Rehan Javed1, Mangesh Nichat2 1

Student at Department of Computer Science and Engineering, KGIET, Darapur, Amravati, Maharashtra Professor at Department of Computer Science and Engineering, KGIET, Darapur, Amravati, Maharashtra

Abstract: Corona virus un wellness (COVID-19), is one among the foremost infectious diseases that reshaped our everyday lives globally within the twenty first century. Technology progressions have a speedy impact on each field of life, be it the medical domain or the other. more than 250 countries are suffering from COVID in in spite of time. The Indian government is creating the mandatory steps to manage the spread of virus within the society. Folks everywhere the globe are at risk of its consequences within the future. During a pandemic like this, people usually worry whether or not they show a signal of COVID-19 or not. Numerous AI strategies are applied with success in epidemic studies. Here, in this paper, we surveyed different models that can detect and predict COVID-19 from X-ray or CT of lung images. Keywords: COVID-19 detection, deep learning, chest X-ray image, CNN I. INTRODUCTION The first COVID-19 case was discovered in Wuhan, China, during December 2019 and it rapidly spread in many International Countries. The virus spread around the world in a very short period which became an epidemic and collapsed the health systems of many countries. Over the last few months, the virus has impacted severely with a continuous increase in the number of confirmed cases and deaths. According to WHO, globally 2.38M people have died out of 108M confirmed cases (WHO, 2021). India with the second largest population in the world stands in the top five affected countries in the world. The most likely symptoms of the virus are fever, sore throat, dry cough, headache, instability, muscle pain, loss of taste or smell, diarrhea, and shortness of breath. The virus is easily spread among entire human community, since the pathology is highly transmitted through contact with infected person either by touch, talking, sneezing or coughing [1]. Artificial intelligence (AI) methods are used successfully to solve many problems in healthcare. AI techniques can also be used in many epidemic studies which includes the prediction of COVID-19 epidemic studies also. The overall global economy has also been affected by this pandemic along with the health, safety and hygiene of individuals all over the world. Apart from the adverse impacts of COVID-19 there have been certain constructive influences around the world. There are no particular treatments for eliminating this disease so far, but one can minimize the COVID spread by maintaining personal sanitation and social distancing among individuals. As the world was facing losses, our nature has recovered its purity by itself from this pandemic. The harmful matters in the nature was removed from the atmosphere and the hole detected in the ozone layer was closed during this pandemic because of the complete lock down that was imposed recently. This work helps in general study of COVID-19 outbreak that utilized several machine learning classification results that visualizes the origin of disease and perform predictions and time series monitoring which helps to control the impact of disease in future. Since COVID-19 does not have very high death rate and the rate of spread of disease alone is high, when the symptoms are analyzed so far and from features, we can easily predict the important symptoms that turns disease into positive cases. Also, the healthy recovery rate implies the disease is curable yet widespread. A very important problem to be addressed is the rate of growth of infection spread by predicting the positive cases at early stage and isolating them. Presently there is considerably less number of COVID-19 analysis kits available in hospitals which are not at all enough for the increasing cases and also due to non-awareness and fear, people undergo the test for confirming whether the result is positive or negative. Hence, it is needed to realize an automatic prediction system to effectively use the analysis kits and also to stop spreading among people by giving them proper treatment at the early stage. Machine Learning (ML) and Deep Learning are actually powerful tools in the fight against the COVID-19. It can be used to manage huge data and effectively predict the spread of the disease. It helps in diagnosis and predicts COVID-19. ML/DL techniques are useful in tracing COVID cases, predicting, creating dashboards, diagnose and give proper medications, generating alerts to support social distance and also for other potential control mechanisms of the spread of virus.

©IJRASET: All Rights are Reserved | SJ Impact Factor 7.538 | ISRA Journal Impact Factor 7.894 |

2349

International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.538 Volume 10 Issue IV Apr 2022- Available at www.ijraset.com II. LITERATURE SURVEY L.J.Muhammad, et. al. [1] have proposed machine learning models for predicting COVID-19 positive cases. In this study the researcher applied deep learning network models for decision making, classification and regression tasks. The resulting classification models are used for detecting the onset of infection to human. This utilized the dataset that involves highly distributed clinical datasets including infectious and benign COVID-19 cases. This study had proved to be most accurate classification model when compared to naïve algorithms that showed 95% accuracy. Since it failed to analyze the high dimensional datasets the model implementation became complex and error prone. G.Monika and M.Bharathi Devi [2], developed three different Machine Learning models such as Polynomial Regression, Decision Tree Regressor, and Random Forest algorithm, for analyzing the COVID-19 cases. They found that the Polynomial Regression model produced the best accuracy of 90%. Still 90% accuracy is not sufficient for medical data analysis. Rajan Gupta, et. al. [3] used Machine Learning Models viz., SEIR and Regression model to analyze and predict the change in the spread of COVID-19 disease. The value of root mean squared log error for the SEIR model was found to be 1.52, and 1.75 for the Regression model. Celestine Iwendi, et. al. [4], proposed a Random Forest model boosted by the Ada Boost algorithm. The model predicts the severe ness of the positive cases. The model was able to give an accuracy of only 94% and an F1 score of 0.86 on the dataset used, which needs further increase in accuracy and F1 score. There exist a few research works regarding forecasting the disease spread but unfortunately, these studies were done by applying primitive statistic methods or by doing a simple survey. The implementation of advanced machine learning concepts in these studies is still in the formative stage. Qianying Lin et.al [5] propose conceptual models for the outbreak in Wuhan with the consideration of individual behavioural reaction and governmental actions (e.g., holiday extension, travel restriction, hospitalisation and quarantine). They employed the estimates of these two key components from the 1918 influenza pandemic in London, United Kingdom, incorporated zoonotic introductions and the emigration, then computed future trends and the reported ratio. The model is concise in structure, and it successfully captures the course of the COVID-19 outbreak, and thus sheds light on understanding the trends of the outbreak. Machine Learning (ML) methods have been proposed in the academic literature as alternatives to statistical ones for time series forecasting. Yet, scant evidence is available about their relative performance in terms of accuracy and computational requirements. Spyros Makridakis. et.al [7] found that the former are dominated across both accuracy measures used and for all forecasting horizons examined. Moreover, they observed that their computational requirements are considerably greater than those of statistical methods. Their paper discusses the results, explains why the accuracy of ML models is below that of statistical ones and proposes some possible ways forward. The empirical results found in our research stress the need for objective and unbiased ways to test the performance of forecasting methods that can be achieved through sizable and open competitions allowing meaningful comparisons and definite conclusions. Pneumonia is among the top diseases which cause most of the deaths all over the world. Virus, bacteria and fungi can all cause pneumonia. However, it is difficult to judge the pneumonia just by looking at chest X-rays. The aim of the study [8] is to simplify the pneumonia detection process for experts as well as for novices. Vikash Chouhan et.al suggest a novel deep learning framework for the detection of pneumonia using the concept of transfer learning. In their approach, features from images are extracted using different neural network models pretrained on ImageNet, which then are fed into a classifier for prediction. they prepared five different models and analyzed their performance. Thereafter, they proposed an ensemble model that combines outputs from all pretrained models, which outperformed individual models, reaching the state-of-the-art performance in pneumonia recognition. Our ensemble model reached an accuracy of 96.4% with a recall of 99.62% on unseen data from the Guangzhou Women and Children’s Medical Center dataset. Chronic obstructive pulmonary, pneumonia, asthma, tuberculosis, lung cancer diseases are the most important chest diseases. These chest diseases are important health problems in the world. In the study [9], a comparative chest diseases diagnosis was realized by using multilayer, probabilistic, learning vector quantization, and generalized regression neural networks. The chest diseases dataset was prepared by using patient’s epicrisis reports from a chest diseases hospital’s database. The outbreak of Corona Virus Disease 2019 (COVID-19) in Wuhan has significantly impacted the economy and society globally. Countries are in a strict state of prevention and control of this pandemic. In the study of et.al [10], the development trend analysis of the cumulative confirmed cases, cumulative deaths, and cumulative cured cases was conducted based on data from Wuhan, Hubei Province, China from January 23, 2020 to April 6, 2020 using an Elman neural network, long short-term memory (LSTM), and support vector machine (SVM). A SVM with fuzzy granulation was used to predict the growth range of confirmed new cases, new deaths, and new cured cases.

©IJRASET: All Rights are Reserved | SJ Impact Factor 7.538 | ISRA Journal Impact Factor 7.894 |

2350

International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.538 Volume 10 Issue IV Apr 2022- Available at www.ijraset.com The experimental results showed that the Elman neural network and SVM used in this study can predict the development trend of cumulative confirmed cases, deaths, and cured cases, whereas LSTM is more suitable for the prediction of the cumulative confirmed cases. The SVM with fuzzy granulation can successfully predict the growth range of confirmed new cases and new cured cases, although the average predicted values are slightly large. Currently, the United States is the epicenter of the COVID-19 pandemic. they also used data modeling from the United States to further verify the validity of the proposed models The normalization of images applied to extract features from images, extracted features given as input to deep learning models on transfer learning scenario using pneumonia images [11]. Experiments were done on images mix of covid-19, bacterial pneumonia and control images. VGG19, MobileNetV2, Inception, Xception, and InceptionResNetV2, these 5 CNN based models were used to detect Covid 19. MobileNetV2 outperformed others. 1428 chest X-ray images comprises of COVID-19 positive, common bacterial pneumonia, and healthy i.e, no infection were considered for training a model using VGG16 pre trained network [12]. The testing shows a 96% and 92.5% accuracy in prediction COVID19 and non COVID-19 classes. III. METHODOLOGY To predict the covid-19 disease there are some phases which are shown in Figure 1.

Figure 1. Covid-19 Detection Model First of all, we need to collect covid-19 positive and negative lungs images from the dataset. Then we will extract features from available images. Once feature extraction is done the model will be train with selected training methodology. With training, we will build model which will generate result and gives us accuracy score. In Testing phase, we will test the result of model.

2351

International Journal for Research in Applied Science & Engineering Technology (IJRASET) ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.538 Volume 10 Issue IV Apr 2022- Available at www.ijraset.com IV. CONCLUSION COVID-19 pandemic now appears to be serious infected spread disease like any other wide-spread diseases. Because of the rapid rise in the number of cases during pandemic, causes struggle in healthcare sector to identify suitable and appropriate treatment. Machine Learning and Deep Learning methodologies are commonly used as alternative methods for classification and prediction. REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11]

L. J. Muhammad, Ebrahem A. Algehyne, Sani Sharif Usman, Abdulkadir Ahmad, Chinmay Chakraborty, I. A. Mohammed, “Supervised Machine Learning Models for Prediction of COVID19 Infection using Epidemiology Dataset” Computer Science, 2020. G.Monika, Dr. M.Bharathi Devi, Using Machine Learning Approach to Predict Covid-19 Progress, International Journal for Modern Trends in Science and Technology, 6(8S): 58-62, 2020. Rajan Gupta, Gaurav Pandey, Poonam Chaudhary, Saibal K. Pal, Machine Learning Models for Government to Predict COVID-19 Outbreak, Digital Government: Research and Practice, Vol. 1, No. 4, Article 26, August 2020. Celestine Iwendi, Ali Kashif Bashir, Atharva Peshkar, R. Sujatha, Jyotir Moy Chatterjee, Swetha Pasupuleti, Rishita Mishra, Sofia Pillai, Ohyun Jo, COVID19 Patient Health Prediction Using Boosted Random Forest Algorithm, Front. Public Health 2020. Q. Lin. S. Zhao, D. Gao, et al., “A conceptual model for the coronavirus disease 2019 (COVID-19) outbreak in Wuhan, China with individual reaction and governmental action,” Vol. 93, p211– 216, March 4, 2020 . Abu Al-Qumboz, M. N., & Abu-Naser, S. S, “Spinach Expert System: Diseases and Symptoms,” International Journal of Academic Information Systems Research (IJAISR), 3(3), 16-22, 2020. S. Makridakis, E. Spiliotis, and V. Assimakopoulos, "Statistical and machine learning forecasting methods: Concerns and ways forward," PloS one, vol. 13, no. 3, 2018. V. Chouhan, S.K. Singh, A. Khamparia, D. Gupta, P. Tiwari, C. Moreira, R. Damasevicius, V.H.C. De Albuquerque, A Novel transfer learning based approach for pneumonia detection in chest X-ray images, Appl. Sci. 10 (2020) 559. O. Er, N. Yumusak, F. Temurtas, Chest diseases diagnosis using artificial neural networks, Expert Syst. Appl. 37 (12) (2010) 7648e7655. L. Jia, K. Li, Y. Jiang, X. Guo, Prediction and Analysis of Coronavirus Disease 2019. arXiv Preprint, 2020, p. 05447, arXiv:2003 Apostolopoulos I.D., Mpesiana T.A. Covid-19: Automatic detection from X-ray images utilizing transfer learning with convolutional neural networks. Phys. Eng. Sci. Med. 2020;43:635–640. doi: 10.1007/s13246-020-00865-4.

2352