Prediction of COVID-19 & Pneumonia using Machine Learning & Deep Learning Model
2nd International Conference on Signal and Information Processing, IConSIP 2022
; 2022.
Article
in English
| Scopus | ID: covidwho-2233270
ABSTRACT
As a result of the COVID-19 pandemic, medical examinations (RTPCR, X-ray, CT-Scan, etc.) may be required to make a medical decision. COVID-19's SARS-CoV-2 virus infects and spreads in the lungs, which can be easily recognized by chest X-rays or CT scans. However, along with COVID-19 instances, cases of another respiratory ailment known as Pneumonia began to rise. As a result, clinicians are having difficulty distinguishing between COVID-19 and Pneumonia. So, more tests were required to identify the condition. After a few days, the COVID-19 SARS-CoV-2 virus multiplied in the lungs, causing pneumonia and COVID-19 named Novel Corona virus infected Pneumonia. We employ Machine Learning and Deep Learning models to predict diseases such as COVID-19 Positive, COVID-19 Negative, and Viral Pneumonia in this research. A dataset of data is used in a Machine Learning model. A dataset of 120 images was used in the Machine Learning model. By extracting eight statistical elements from an image texture, we calculated accuracy. Adaboost, Decision Tree & Naive Bayes have overall accuracy of 88.46%, 86.4% and 80%, respectively. When we compared the algorithms, Adaboost algorithm performs the best, with overall accuracy of 88.46%, sensitivity of 84.62%, specificity of 92.31%, F1-score of 88% and Kappa of 0.8277. VGG16 Architecture is used in CNN model for 838 images in Deep Learning model. The model's total accuracy is 99.17 %. © 2022 IEEE.
Adaboost; CNN; Deep Learning; Machine learning; VGG16; Adaptive boosting; Computerized tomography; Decision trees; Diagnosis; Image texture; Learning systems; Textures; Viruses; Condition; CT-scan; Learning models; Machine learning models; Machine-learning; Medical decision making; Overall accuracies; X-ray CT; COVID-19
Full text:
Available
Collection:
Databases of international organizations
Database:
Scopus
Type of study:
Prognostic study
Language:
English
Journal:
2nd International Conference on Signal and Information Processing, IConSIP 2022
Year:
2022
Document Type:
Article
Similar
MEDLINE
...
LILACS
LIS