{"title":"基于计算机断层肺图像深度学习的COVID-19分割分类框架","authors":"Wessam M. Salama , Moustafa H. Aly","doi":"10.1016/j.jnlest.2022.100161","DOIUrl":null,"url":null,"abstract":"<div><p>Corona Virus Disease 2019 (COVID-19) has affected millions of people worldwide and caused more than 6.3 million deaths (World Health Organization, June 2022). Increased attempts have been made to develop deep learning methods to diagnose COVID-19 based on computed tomography (CT) lung images. It is a challenge to reproduce and obtain the CT lung data, because it is not publicly available. This paper introduces a new generalized framework to segment and classify CT images and determine whether a patient is tested positive or negative for COVID-19 based on lung CT images. In this work, many different strategies are explored for the classification task. ResNet50 and VGG16 models are applied to classify CT lung images into COVID-19 positive or negative. Also, VGG16 and ReNet50 combined with U-Net, which is one of the most used architectures in deep learning for image segmentation, are employed to segment CT lung images before the classifying process to increase system performance. Moreover, the image size dependent normalization technique (ISDNT) and Wiener filter are utilized as the preprocessing techniques to enhance images and noise suppression. Additionally, transfer learning and data augmentation techniques are performed to solve the problem of COVID-19 CT lung images deficiency, therefore the over-fitting of deep models can be avoided. The proposed frameworks, which comprise of, end-to-end, VGG16, ResNet50, and U-Net with VGG16 or ResNet50, are applied on the dataset which is sourced from COVID-19 lung CT images in Kaggle. The classification results show that using the preprocessed CT lung images as the input for U-Net hybrid with ResNet50 achieves the best performance. The proposed classification model achieves 98.98% accuracy (ACC), 98.87% area under the ROC curve (AUC), 98.89% sensitivity (Se), 97.99% precision (Pr), 97.88% <em>F</em><sub>1</sub>- score, and 1.8974-second computational time.</p></div>","PeriodicalId":53467,"journal":{"name":"Journal of Electronic Science and Technology","volume":"20 3","pages":"Article 100161"},"PeriodicalIF":0.0000,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1674862X22000143/pdfft?md5=ab3b5702d4ce126d6b35c5ff609a0562&pid=1-s2.0-S1674862X22000143-main.pdf","citationCount":"12","resultStr":"{\"title\":\"Framework for COVID-19 segmentation and classification based on deep learning of computed tomography lung images\",\"authors\":\"Wessam M. Salama , Moustafa H. Aly\",\"doi\":\"10.1016/j.jnlest.2022.100161\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><p>Corona Virus Disease 2019 (COVID-19) has affected millions of people worldwide and caused more than 6.3 million deaths (World Health Organization, June 2022). Increased attempts have been made to develop deep learning methods to diagnose COVID-19 based on computed tomography (CT) lung images. It is a challenge to reproduce and obtain the CT lung data, because it is not publicly available. This paper introduces a new generalized framework to segment and classify CT images and determine whether a patient is tested positive or negative for COVID-19 based on lung CT images. In this work, many different strategies are explored for the classification task. ResNet50 and VGG16 models are applied to classify CT lung images into COVID-19 positive or negative. Also, VGG16 and ReNet50 combined with U-Net, which is one of the most used architectures in deep learning for image segmentation, are employed to segment CT lung images before the classifying process to increase system performance. Moreover, the image size dependent normalization technique (ISDNT) and Wiener filter are utilized as the preprocessing techniques to enhance images and noise suppression. Additionally, transfer learning and data augmentation techniques are performed to solve the problem of COVID-19 CT lung images deficiency, therefore the over-fitting of deep models can be avoided. The proposed frameworks, which comprise of, end-to-end, VGG16, ResNet50, and U-Net with VGG16 or ResNet50, are applied on the dataset which is sourced from COVID-19 lung CT images in Kaggle. The classification results show that using the preprocessed CT lung images as the input for U-Net hybrid with ResNet50 achieves the best performance. The proposed classification model achieves 98.98% accuracy (ACC), 98.87% area under the ROC curve (AUC), 98.89% sensitivity (Se), 97.99% precision (Pr), 97.88% <em>F</em><sub>1</sub>- score, and 1.8974-second computational time.</p></div>\",\"PeriodicalId\":53467,\"journal\":{\"name\":\"Journal of Electronic Science and Technology\",\"volume\":\"20 3\",\"pages\":\"Article 100161\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2022-09-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://www.sciencedirect.com/science/article/pii/S1674862X22000143/pdfft?md5=ab3b5702d4ce126d6b35c5ff609a0562&pid=1-s2.0-S1674862X22000143-main.pdf\",\"citationCount\":\"12\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Electronic Science and Technology\",\"FirstCategoryId\":\"95\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S1674862X22000143\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"Engineering\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Electronic Science and Technology","FirstCategoryId":"95","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1674862X22000143","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"Engineering","Score":null,"Total":0}
Framework for COVID-19 segmentation and classification based on deep learning of computed tomography lung images
Corona Virus Disease 2019 (COVID-19) has affected millions of people worldwide and caused more than 6.3 million deaths (World Health Organization, June 2022). Increased attempts have been made to develop deep learning methods to diagnose COVID-19 based on computed tomography (CT) lung images. It is a challenge to reproduce and obtain the CT lung data, because it is not publicly available. This paper introduces a new generalized framework to segment and classify CT images and determine whether a patient is tested positive or negative for COVID-19 based on lung CT images. In this work, many different strategies are explored for the classification task. ResNet50 and VGG16 models are applied to classify CT lung images into COVID-19 positive or negative. Also, VGG16 and ReNet50 combined with U-Net, which is one of the most used architectures in deep learning for image segmentation, are employed to segment CT lung images before the classifying process to increase system performance. Moreover, the image size dependent normalization technique (ISDNT) and Wiener filter are utilized as the preprocessing techniques to enhance images and noise suppression. Additionally, transfer learning and data augmentation techniques are performed to solve the problem of COVID-19 CT lung images deficiency, therefore the over-fitting of deep models can be avoided. The proposed frameworks, which comprise of, end-to-end, VGG16, ResNet50, and U-Net with VGG16 or ResNet50, are applied on the dataset which is sourced from COVID-19 lung CT images in Kaggle. The classification results show that using the preprocessed CT lung images as the input for U-Net hybrid with ResNet50 achieves the best performance. The proposed classification model achieves 98.98% accuracy (ACC), 98.87% area under the ROC curve (AUC), 98.89% sensitivity (Se), 97.99% precision (Pr), 97.88% F1- score, and 1.8974-second computational time.
期刊介绍:
JEST (International) covers the state-of-the-art achievements in electronic science and technology, including the most highlight areas: ¨ Communication Technology ¨ Computer Science and Information Technology ¨ Information and Network Security ¨ Bioelectronics and Biomedicine ¨ Neural Networks and Intelligent Systems ¨ Electronic Systems and Array Processing ¨ Optoelectronic and Photonic Technologies ¨ Electronic Materials and Devices ¨ Sensing and Measurement ¨ Signal Processing and Image Processing JEST (International) is dedicated to building an open, high-level academic journal supported by researchers, professionals, and academicians. The Journal has been fully indexed by Ei INSPEC and has published, with great honor, the contributions from more than 20 countries and regions in the world.