Artificial Intelligence Applied to Chest X-Ray Images for the Automatic Detection of COVID-19. A Thoughtful Evaluation Approach

Arias-Londoño, Julián D.; García, Jorge Andrés Gómez; Moro-Velázquez, Laureano; Godino-Llorente, Juan Ignacio

doi:10.1109/access.2020.3044858

Cited by 83 publications

(73 citation statements)

References 40 publications

(48 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The pandemic event called COVID19 has significantly affected the performance of centers such as small clinics, doctors' offices, emergency care centers and large hospitals with emergency rooms. [6,7]. The use of intelligent systems plays a considerable role in the process of treating the patient.…”

Section: Introductionmentioning

confidence: 99%

Densely Connected Convolutional Networks (DenseNet) for Diagnosing Coronavirus Disease (COVID-19) from Chest X-ray Imaging

Tabrizchi

Mosavi

Vamossy

et al. 2021

2021 IEEE International Symposium on Medical Measurements and Applications (MeMeA)

View full text Add to dashboard Cite

Since the beginning of the coronavirus disease (COVID-19) pandemic several machine learning and deep learning methods had been introduced to detect the infected patients using the X-Ray or CT scan images. Numerous sophisticated data-driven methods had been introduced to improve the performance and the accuracy of the diagnosis models. This paper proposes an improved densely connected convolutional networks (DenseNet) method based on transfer learning (TL) to enhance the model performance. The results show promising model accuracy.

show abstract

Section: Introductionmentioning

confidence: 99%

Densely Connected Convolutional Networks (DenseNet) for Diagnosing Coronavirus Disease (COVID-19) from Chest X-ray Imaging

Tabrizchi

Mosavi

Vamossy

et al. 2021

2021 IEEE International Symposium on Medical Measurements and Applications (MeMeA)

View full text Add to dashboard Cite

show abstract

“…This is of vital importance, since as will be shown below, when not only the lung region is used, the models tend to focus on regions where its association with disease in question is unclear. Figure 2 (extracted from [ 52 ]) shows an example of how when using the whole image, CNNs may use as most important regions for classification areas that are not within the lungs. This means that there are regions that provide enough information to adequately separate the classes with features not related to the disease that they are trying to classify.…”

Section: Explanatory Ai Methods In the Identification Of Covid-19 Using Cxrmentioning

confidence: 99%

“…In the same study, the absence of an ood set to assess generalizability is strongly criticized. On the other hand, another report [ 52 ] recognized that most of published work has not performed any analysis to demonstrate the reliability of network predictions. In the context of medical tasks, this is particularly relevant.…”

Section: Evidence Of Shortcuts Learning In Cxr Classificationmentioning

confidence: 99%

Current limitations to identify covid-19 using artificial intelligence with chest x-ray imaging (part ii). The shortcut learning problem

et al. 2021

View full text Add to dashboard Cite

Since the outbreak of the COVID-19 pandemic, computer vision researchers have been working on automatic identification of this disease using radiological images. The results achieved by automatic classification methods far exceed those of human specialists, with sensitivity as high as 100% being reported. However, prestigious radiology societies have stated that the use of this type of imaging alone is not recommended as a diagnostic method. According to some experts the patterns presented in these images are unspecific and subtle, overlapping with other viral pneumonias. This report seeks to evaluate the analysis the robustness and generalizability of different approaches using artificial intelligence, deep learning and computer vision to identify COVID-19 using chest X-rays images. We also seek to alert researchers and reviewers to the issue of “shortcut learning”. Recommendations are presented to identify whether COVID-19 automatic classification models are being affected by shortcut learning. Firstly, papers using explainable artificial intelligence methods are reviewed. The results of applying external validation sets are evaluated to determine the generalizability of these methods. Finally, studies that apply traditional computer vision methods to perform the same task are considered. It is evident that using the whole chest X-Ray image or the bounding box of the lungs, the image regions that contribute most to the classification appear outside of the lung region, something that is not likely possible. In addition, although the investigations that evaluated their models on data sets external to the training set, the effectiveness of these models decreased significantly, it may provide a more realistic representation as how the model will perform in the clinic. The results indicate that, so far, the existing models often involve shortcut learning, which makes their use less appropriate in the clinical setting.

show abstract

“…J. D. Arias-Londoño, et al [10] presented an evaluation of different methods based on a deep CNN automatic COVID-19 diagnosis tool using chest X-Ray images. The aim of the authors was to evaluate how preprocessing the data affects the results and improves it's explain ability, and the critical analysis was also analyzed in this system, the accuracy of proposed method had achieved 91.5% and an 87.4% average recall for the worst.…”

Section: Introductionmentioning

confidence: 99%

COVID-19 Detection Based on Image Regrouping and Resnet-SVM Using Chest X-Ray Images

Zhou

Zhang

et al. 2021

IEEE Access

View full text Add to dashboard Cite

As the COVID-19 spread worldwide, countries around the world are actively taking measures to fight against the epidemic. To prevent the spread of it, a high sensitivity and efficient method for COVID-19 detection is necessary. By analyzing the COVID-19 chest X-ray images, a combination method of image regrouping and ResNet-SVM was proposed in this study. The lung region was segmented from the original chest X-ray images and divided into small pieces, and then the small pieces of lung region were regrouped into a regular image randomly. Furthermore the regrouped images were fed into the deep residual encoder block for feature extraction. Finally the extracted features were as input into support vector machine for recognition. The visual attention was introduced in the novel method, which paid more attention to the features of COVID-19 without the interference of shapes, rib and other related noises. The experimental results showed that the proposed method achieved 93% accuracy without large number of training data, outperformed the existing COVID-19 detection models.

show abstract

Artificial Intelligence Applied to Chest X-Ray Images for the Automatic Detection of COVID-19. A Thoughtful Evaluation Approach

Cited by 83 publications

References 40 publications

Densely Connected Convolutional Networks (DenseNet) for Diagnosing Coronavirus Disease (COVID-19) from Chest X-ray Imaging

Densely Connected Convolutional Networks (DenseNet) for Diagnosing Coronavirus Disease (COVID-19) from Chest X-ray Imaging

Current limitations to identify covid-19 using artificial intelligence with chest x-ray imaging (part ii). The shortcut learning problem

COVID-19 Detection Based on Image Regrouping and Resnet-SVM Using Chest X-Ray Images

Contact Info

Product

Resources

About