COVINet: A hybrid model for classification of COVID and Non-COVID pneumonia in CT and X-Ray imagery
The COVID-19 pandemic has resulted in a significant increase in the number of pneumonia cases, including those caused by the Coronavirus. To detect COVID pneumonia, RT-PCR is used as the primary detection tool for COVID-19 pneumonia but chest imaging, including CT scans and X-Ray imagery, can also b...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
KeAi Communications Co., Ltd.
2023-06-01
|
| Series: | International Journal of Cognitive Computing in Engineering |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2666307423000153 |
| _version_ | 1827400267640864768 |
|---|---|
| author | Vasu Mittal Akhil Kumar |
| author_facet | Vasu Mittal Akhil Kumar |
| author_sort | Vasu Mittal |
| collection | DOAJ |
| description | The COVID-19 pandemic has resulted in a significant increase in the number of pneumonia cases, including those caused by the Coronavirus. To detect COVID pneumonia, RT-PCR is used as the primary detection tool for COVID-19 pneumonia but chest imaging, including CT scans and X-Ray imagery, can also be used as a secondary important tool for the diagnosis of pneumonia, including COVID pneumonia. However, the interpretation of chest imaging in COVID-19 pneumonia can be challenging, as the signs of the disease on imaging may be subtle and may overlap with normal pneumonia. In this paper, we propose a hybrid model with the name COVINet which uses ResNet-101 as the feature extractor and classical K-Nearest Neighbors as the classifier that led us to give automated results for detecting COVID pneumonia in X-Rays and CT imagery. The proposed hybrid model achieved a classification accuracy of 98.6%. The model's precision, recall, and F1-Score values were also impressive, ranging from 98-99%. To back and support the proposed model, several CNN-based feature extractors and classical machine learning classifiers have been exploited. The outcome with exploited combinations suggests that our model can significantly enhance the accuracy and precision of detecting COVID-19 pneumonia on chest imaging, and this holds the potential of being a valuable resource for early identification and diagnosis of the illness by radiologists and medical practitioners. |
| first_indexed | 2024-03-08T19:58:36Z |
| format | Article |
| id | doaj.art-d2a7c6dc8ae74bc1a98f817d43d51ffb |
| institution | Directory Open Access Journal |
| issn | 2666-3074 |
| language | English |
| last_indexed | 2024-03-08T19:58:36Z |
| publishDate | 2023-06-01 |
| publisher | KeAi Communications Co., Ltd. |
| record_format | Article |
| series | International Journal of Cognitive Computing in Engineering |
| spelling | doaj.art-d2a7c6dc8ae74bc1a98f817d43d51ffb2023-12-24T04:46:39ZengKeAi Communications Co., Ltd.International Journal of Cognitive Computing in Engineering2666-30742023-06-014149159COVINet: A hybrid model for classification of COVID and Non-COVID pneumonia in CT and X-Ray imageryVasu Mittal0Akhil Kumar1School of Computer Science & Engineering, Vellore Institute of Technology, Chennai, IndiaCorresponding author.; School of Computer Science & Engineering, Vellore Institute of Technology, Chennai, IndiaThe COVID-19 pandemic has resulted in a significant increase in the number of pneumonia cases, including those caused by the Coronavirus. To detect COVID pneumonia, RT-PCR is used as the primary detection tool for COVID-19 pneumonia but chest imaging, including CT scans and X-Ray imagery, can also be used as a secondary important tool for the diagnosis of pneumonia, including COVID pneumonia. However, the interpretation of chest imaging in COVID-19 pneumonia can be challenging, as the signs of the disease on imaging may be subtle and may overlap with normal pneumonia. In this paper, we propose a hybrid model with the name COVINet which uses ResNet-101 as the feature extractor and classical K-Nearest Neighbors as the classifier that led us to give automated results for detecting COVID pneumonia in X-Rays and CT imagery. The proposed hybrid model achieved a classification accuracy of 98.6%. The model's precision, recall, and F1-Score values were also impressive, ranging from 98-99%. To back and support the proposed model, several CNN-based feature extractors and classical machine learning classifiers have been exploited. The outcome with exploited combinations suggests that our model can significantly enhance the accuracy and precision of detecting COVID-19 pneumonia on chest imaging, and this holds the potential of being a valuable resource for early identification and diagnosis of the illness by radiologists and medical practitioners.http://www.sciencedirect.com/science/article/pii/S2666307423000153COVID-19COVID pneumoniaResNet-101Machine learningDeep learning |
| spellingShingle | Vasu Mittal Akhil Kumar COVINet: A hybrid model for classification of COVID and Non-COVID pneumonia in CT and X-Ray imagery International Journal of Cognitive Computing in Engineering COVID-19 COVID pneumonia ResNet-101 Machine learning Deep learning |
| title | COVINet: A hybrid model for classification of COVID and Non-COVID pneumonia in CT and X-Ray imagery |
| title_full | COVINet: A hybrid model for classification of COVID and Non-COVID pneumonia in CT and X-Ray imagery |
| title_fullStr | COVINet: A hybrid model for classification of COVID and Non-COVID pneumonia in CT and X-Ray imagery |
| title_full_unstemmed | COVINet: A hybrid model for classification of COVID and Non-COVID pneumonia in CT and X-Ray imagery |
| title_short | COVINet: A hybrid model for classification of COVID and Non-COVID pneumonia in CT and X-Ray imagery |
| title_sort | covinet a hybrid model for classification of covid and non covid pneumonia in ct and x ray imagery |
| topic | COVID-19 COVID pneumonia ResNet-101 Machine learning Deep learning |
| url | http://www.sciencedirect.com/science/article/pii/S2666307423000153 |
| work_keys_str_mv | AT vasumittal covinetahybridmodelforclassificationofcovidandnoncovidpneumoniainctandxrayimagery AT akhilkumar covinetahybridmodelforclassificationofcovidandnoncovidpneumoniainctandxrayimagery |