Pneumonia Detection from Chest X-ray Images Based on Convolutional Neural Network

Round 1

Reviewer 1 Report

This is a very interesting work on the application of deep learning in the diagnostic task of pneumonia detection.

Albeit the technical methods are not new they are applied with novelty, given a pneumonia model.

The paper is concise and well written. Experiments are benchmarked and comparative with competing methods.

Author Response

We gratefully thank  the reviewer for your time spend making constructive remarks and useful suggestions, which has significantly raised the quality of the manuscript and has enabled us to improve the manuscript.

Reviewer 2 Report

Authors present a method for pneumonia detection in chest x-ray images and perform extensive validation of different networks and image enhancement techniques. While the effort is laudable and authors claim to have results that outperform state of the art, the manuscript itself is badly organized and deserves a detailed revision both in terms of structure and English language/typos. Some suggestions/comments:

• page 1: 1995 was already 26 years ago; it's not an acceptable reference for the mortality of pneumonia.
• page 3: Network (CNN)
• page 3: AUC not yet defined
• page 3: pix instead of pixel (appears multiple times)
• page 4: both x-ray and X-ray appear. Make notations uniform
• page 4: 6 references for segmentation (15-20) seems too much given that the proposed work does not use segmentation at all
• page 4: References appear in both [author et al.] and [ref-number] style. It's best if the same style is used throughout the manuscript
• page 4: dateset instead of dataset (appears multiple times)
• page 4: informal language is used often such as "unbelievable" (unless you mean the literal sense of unbelievable but I guess you mean that the accuracy is outstanding?)
• page 4: Vgg or VGG?
• page 5: Researches instead of Researchers
• page 5: meat instead of meant
• page 5: outbreak is a noun and cannot be used as a verb
• page 5: informal language (plague)
• page 5: while related to pneumonia detection, COVID-19 is not targeted in this work and does not merit such a long description in my opinion
• page 6: Segmentation and generation state of the art is also not relevant for the scope of this work
• page 7: I don't understand the significance of chapter III Background
• page 8: exploded instead of explored (appears multiple times)
• page 13: migrate instead of mitigate (maybe?)

Author Response

We gratefully thank the l reviewers for your time spend making  constructive remarks and useful suggestions, which has significantly raised the quality of the manuscript and has enabled us to improve the manuscript.

Round 2

Reviewer 2 Report

The authors have addressed the main issues raised