Automated detection of COVID-19 through convolutional neural network using chest x-ray images

  1. Rubina Sarki
  2. Khandakar Ahmed
  3. Hua Wang
  4. Yanchun Zhang
  5. Kate Wang

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

The COVID-19 epidemic has a catastrophic impact on global well-being and public health. More than 27 million confirmed cases have been reported worldwide until now. Due to the growing number of confirmed cases, and challenges to the variations of the COVID-19, timely and accurate classification of healthy and infected patients is essential to control and treat COVID-19. We aim to develop a deep learning-based system for the persuasive classification and reliable detection of COVID-19 using chest radiography. Firstly, we evaluate the performance of various state-of-the-art convolutional neural networks (CNNs) proposed over recent years for medical image classification. Secondly, we develop and train CNN from scratch. In both cases, we use a public X-Ray dataset for training and validation purposes. For transfer learning, we obtain 100% accuracy for binary classification (i.e., Normal/COVID-19) and 87.50% accuracy for tertiary classification (Normal/COVID-19/Pneumonia). With the CNN trained from scratch, we achieve 93.75% accuracy for tertiary classification. In the case of transfer learning, the classification accuracy drops with the increased number of classes. The results are demonstrated by comprehensive receiver operating characteristics (ROC) and confusion metric analysis with 10-fold cross-validation.

Article activity feed

  1. Version published to 10.1371/journal.pone.0262052
  2. ScreenIT

    SciScore for 10.1101/2021.02.06.21251271: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    NIH rigor criteria are not applicable to paper type.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    This study used the Keras DL Framework by F. Chollet. [43] and TensorFlow as backend, which includes pre-trained, DL models made available in Keras Applications alongside weights.
    TensorFlow
    suggested: (tensorflow, RRID:SCR_016345)

    Results from OddPub: Thank you for sharing your code and data.

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    Approach Limitation: Limitations of the study can be summarized as follows; Initially, due to the limited availability of high quality COVID-19 public image, only a small-to-moderate data set size used in the study. The literature review suggested for this research is at its initial stage, because most of the existing AI systems introduced in the research literature have been closed-source and inaccessible for the scientific community to develop a better understanding and expansion of these systems. DL models pre-trained on an ImageNet dataset used as a compensation procedure due to the limited number of datasets. Several experiments performed with different parameters to achieve high accuracy. However, the history of DL architectures is not well known in many situations and is regarded as a black box. Thus, the proper structure and optimum values for the number of layers and nodes in different layers are still challenging. The selection of values for the learning rate, number of epochs, and regularizer intensity often involve unique domain knowledge. In this study, a pre-trained VGG16 network with defined hyperparameters yielded the highest performance accuracy than other tested systems in binary classification and achieved. Similarly, a newly developed model with defined layers and hyperparameter yielded the highest accuracy in multi-class classification. Although with certain limitations, we believe our system can be advantageous to assist experts in identifying COVID-19 i...

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: Please consider improving the rainbow (“jet”) colormap(s) used on page 11. At least one figure is not accessible to readers with colorblindness and/or is not true to the data, i.e. not perceptually uniform.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  3. Version published to 10.1101/2021.02.06.21251271v1 on medRxiv