Deep Learning Model for COVID-19-Infected Pneumonia Diagnosis Using Chest Radiography Images

Round 1

Reviewer 1 Report

Authors proposed a deep learning (DL)-based computer-aided diagnosis system for 15 rapid and easy detection of pneumonia using X-ray images. To improve classification accuracy and 16 faster conversion of the models, we employ transfer learning techniques using well-known DL mod- 17 els such as VGG19 and ResNet50. Experiment results conducted on the COVID-QU-Ex dataset 18 demonstrate that the proposed method is effective in diagnosing X-ray images onto three classes, 19 such as COVID-19-infected pneumonia, non-COVID-19 infections (other viral and bacterial pneu- 20 monia), and normal (uninfected) images, with an average classification accuracy of 96.6%. 

Following improvements are suggested:

1. The abstract should be a single paragraph and should follow the style of structured abstracts, but without headings: 1) Background: Place the question addressed in a broad context and highlight the purpose of the study; 2) Methods: Describe briefly the main methods or treatments applied. Include any relevant preregistration numbers, and species and strains of any animals used. 3) Results: Summarize the article's main findings; and 4) Conclusion: Indicate the main conclusions or interpretations. The abstract should be an objective representation of the article: it must not contain results which are not presented and substantiated in the main text and should not exaggerate the main conclusions. The conclusion is missing from your abstract.

2. In the introduction, what key theoretical perspectives and empirical findings in the main literature have already informed the problem formulation? What major, unaddressed puzzle, controversy, or paradox does this research address?

3. Why does it need to be addressed?  Why it should be now - not in the past?

2. Please add some section of eXplanaable DL section with respect to healthcare.

3. Currently deepRadiomics is going to cover a lot of attention. How does AI with Radiomics helps need to be included.

4. Further, in the introduction, what is the recent knowledge gap of the main literature that the author needs to write this research? What we have known and what we have not known? What is missing from current works? Please explain and give examples!

5. In terms of the knowledge gap, it will be best if the research challenge/knowledge gap could be stated in one article or more articles in the main literature . Assure that you have included all key articles  in the main literature.  https://doi.org/10.3390/diagnostics12102549 in the overall prospective, https://doi.org/10.3390/diagnostics12102420 is bone prospective https://doi.org/10.1016/j.compbiomed.2022.106083   https://doi.org/10.1007/s12553-022-00700-8 is in lungs related optimization  https://doi.org/10.1007/s13721-022-00394-y    https://doi.org/10.1016/j.compeleceng.2022.108259 pattern based,  https://doi.org/10.1016/j.procs.2016.05.244  https://doi.org/10.1007/978-3-030-27272-2_14 https://doi.org/10.1080/03772063.2017.1331757  6. Research question must be explicitly stated in the introduction. Show how the main literature informs the formulation of your research question(s).

7.  Besides, it feels to me you just put the tables and figures in result section rather than their explanation. Try to explain each table and figure in detail in the result section. Discussion is a separate section.

8. Abalation study is needed to show the roubous ness of the model.

9. Class imbalance is not addressed well, need to be taken carefully.

Author Response

Thank you for your suggestions. Please, refer to the revised manuscript for changes and find Author's response in the uploaded file below where we responded to each comment by the reviewer.

Author Response File: Author Response.docx

Reviewer 2 Report

The subject is somewhat clear, and it has been explored much more than the current introduction gives credit. The article presents a good idea.  Although the initial question is interesting, I have a few issues with the study.

This article fits into the framework of translational medicine: how to fill the gap between basic sciences and clinical sciences. How to understand  COVID-19 using Artificial Intelligence and medical image. The author can add a paragraph in the introduction to explain the idea in the context of translational medicine. Example of an article should be cited:

doi: 10.1080/03007995.2017.1385450

doi: 10.1097/BCO.0000000000000846

1. Title: good

2. Abstract: it captures the appropriate essence of the manuscript. Excellent.

3. Introduction: The introduction identifies the problem that is being addressed in the manuscript and develops and states the purpose of the manuscript.

4. Tables and figures: Quality of figures is so important too. Please provide some high-resolution figures. Some figures have a poor resolution.

5. References: I have verified all references and all key references are correct. Please update the reference according to the article proposed above.

6. Methods: The Data used  was appropriate to validate this kind of study.

7. Discussion:

* The authors  don't discuss the  limitations of the study correctly. Please add it

8. Conclusion: The conclusion is justified by the methods and results.

9. There are still some mistakes in grammar and misprints, the authors should carefully check this manuscript.

* I have enjoyed reading, and I am in favor of publication after suitable.

Author Response

Thank you for your suggestions. Please, refer to the revised manuscript for changes and find Author's response in the uploaded file below where we responded to each comment by the reviewer.

Author Response File: Author Response.docx