Pre-trained deep learning models in automatic COVID-19 diagnosis

Reza, Ahmed Wasif; Hasan, Mahamudul; Nowrin, Nazla; Shibly, Mir Moynuddin Ahmed

doi:10.11591/ijeecs.v22.i3.pp1540-1547

Cited by 8 publications

(6 citation statements)

References 21 publications

(17 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The empirical findings were likewise outstanding. This type of computerized system can assist us in combating the deadly viral spread [32].…”

Section: Methodsmentioning

confidence: 99%

An approach to categorize chest X-ray images using sparse categorical cross entropy

Ruby

et al. 2021

IJEECS

View full text Add to dashboard Cite

The Coronavirus disease (COVID-19) pandemic is the most recent threat to global health. Reverse transcription-polymerase chain reaction (RT-PCR) testing, computed tomography (CT) scans, and chest X-ray (CXR) images are being used to identify Coronavirus, one of the most serious community viruses of the twenty-first century. Because CT scans and RT-PCR analyses are not available in most health divisions, CXR images are typically the most time-saving and cost-effective tool for physicians in making decisions. Artificial intelligence and machine learning have become increasingly popular because of recent technical advancements. The goal of this project is to combine machine learning, deep learning, and the health-care sector to create a categorization technique for detecting the Coronavirus and other respiratory disorders. The three conditions evaluated in this study were COVID-19, viral Pneumonia, and normal lungs. Using X-ray pictures, this research developed a sparse categorical cross-entropy technique for recognizing all three categories. The proposed model had a training accuracy of 91% and a training loss of 0.63, as well as a validation accuracy of 81% and a validation loss of 0.7108.

show abstract

“…The empirical findings were likewise outstanding. This type of computerized system can assist us in combating the deadly viral spread [32].…”

Section: Methodsmentioning

confidence: 99%

An approach to categorize chest X-ray images using sparse categorical cross entropy

Ruby

et al. 2021

IJEECS

View full text Add to dashboard Cite

show abstract

“…Similarly, Apostolopoulos and Mpesiana (2020) [6] explored the use of transfer learning on various pre-trained CNNs, including VGG19 and MobileNet, to detect COVID-19 in chest X-rays, reporting over 90% accuracy. Khan et al (2020) [2] developed CoroNet, an approach that leverages Xception architecture, pre-trained on the ImageNet dataset. This method showed significant promise, especially in correctly identifying COVID-19 from other types of pneumonia and normal cases.…”

Section: Ivliterature Reviewmentioning

confidence: 99%

Efficient COVID-19 Detection From Chest X-Ray Images Using Deep Transfer Learning

Singh,

Srivastava

2024

eatp

View full text Add to dashboard Cite

The new coronavirus, COVID-19, is causing a global pandemic that is now underway, making quick and precise diagnostic techniques necessary. The effectiveness of applying transfer learning techniques for the identification of COVID-19 in chest Xray pictures is thoroughly examined in this research work. The research uses pretrained convolutional neural networks (CNNs) to improve COVID-19 diagnostic performance in terms of both speed and accuracy. Using a dataset of chest X-ray pictures from patients with and without COVID-19 infection, we investigate several approaches for preprocessing, selecting models, and implementing them to have the best detection results. Our findings show that, when compared to conventional techniques, transfer learning may greatly increase diagnosis accuracy, making it a potentially useful tool for medical practitioners. The suggested model's performance in comparison to current approaches is also covered in the report, along with its therapeutic implications, any drawbacks, and suggestions for further study. Our goal in conducting this study is to add a dependable, scalable, and effective diagnostic method to the expanding body of knowledge in the fight against COVID-19. I.INTRODUCTIONThe new coronavirus SARS-CoV-2 is the cause of the COVID-19 pandemic, which poses an unprecedented challenge to international health systems. For the virus to be contained and its effects reduced early and precise diagnosis is essential. The gold standard for COVID-19 diagnosis is reverse transcription-polymerase chain reaction (RT-PCR) testing, yet these procedures are frequently limited by time, sensitivity, and availability concerns. Consequently, there is a strong demand for complementary diagnostic techniques to RT-PCR, especially in environments with restricted resources. Because COVID-19 largely affects the respiratory system, medical imaging, particularly chest X-rays, has become an important diagnostics approach. Whenever it relates to testing, chest X-rays are quicker and easier to get than RT-PCRs. Radiologists, on the other hand, must manually examine chest X-rays, which takes time and is vulnerable to inter-observer variability. This highlights the necessity of automated, dependable, and effective diagnostic methods. Numerous uses of artificial intelligence (AI), primarily deep learning, have showed promise in the discipline of medical imaging. A effective method for COVID-19 identification is provided by transfer learning, a branch of deep learning that uses pre-trained models customised for a certain job. This strategy makes use of characteristics that have previously been learnt from big datasets. This work explores the use of transfer learning to create a reliable and effective diagnostic model for COVID-19 detection using chest X-ray images. The objective in this study is to assess how well distinct transfer learning strategies work for correctly identifying COVID-19 from chest Xray images. To choose the best course of action, we will carefully examine the dataset, preprocessing techniques,...

show abstract

“…To perform the segmentation of the moving object, we employ the auto-encoder as supervised learning for both approaches. For the first approach, we construct the network by fine-tuning the VGG-16 network [29], [30] that was pre-trained trained on the famous ImageNet dataset [29], [31] as the encoder part. Then, we changed the fully connected layers with a latent space.…”

Section: Evaluation Protocolmentioning

confidence: 99%

End-to-end deep auto-encoder for segmenting a moving object with limited training data

Kebir¹,

Taibi²

2022

IJECE

View full text Add to dashboard Cite

<span lang="EN-US">Deep learning-based approaches have been widely used in various applications, including segmentation and classification. However, a large amount of data is required to train such techniques. Indeed, in the surveillance video domain, there are few accessible data due to acquisition and experiment complexity. In this paper, we propose an end-to-end deep auto-encoder system for object segmenting from surveillance videos. Our main purpose is to enhance the process of distinguishing the foreground object when only limited data are available. To this end, we propose two approaches based on transfer learning and multi-depth auto-encoders to avoid over-fitting by combining classical data augmentation and principal component analysis (PCA) techniques to improve the quality of training data. Our approach achieves good results outperforming other popular models, which used the same principle of training with limited data. In addition, a detailed explanation of these techniques and some recommendations are provided. Our methodology constitutes a useful strategy for increasing samples in the deep learning domain and can be applied to improve segmentation accuracy. We believe that our strategy has a considerable interest in various applications such as medical and biological fields, especially in the early stages of experiments where there are few samples.</span>

show abstract

Pre-trained deep learning models in automatic COVID-19 diagnosis

Cited by 8 publications

References 21 publications

An approach to categorize chest X-ray images using sparse categorical cross entropy

An approach to categorize chest X-ray images using sparse categorical cross entropy

Efficient COVID-19 Detection From Chest X-Ray Images Using Deep Transfer Learning

End-to-end deep auto-encoder for segmenting a moving object with limited training data

Contact Info

Product

Resources

About