All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
Reviewers are satisfied with the revisions, and I concur to recommend accepting this manuscript.
[# PeerJ Staff Note - this decision was reviewed and approved by Shawn Gomez, a PeerJ Section Editor covering this Section #]
no comment
no comment
no comment
There are some remaining minor concerns that need to be addressed.
1. The manuscript is well-structured with comprehensive details for each section.
2. There are some unnecessary repetitions of some phrases, which I have indicated in the attached PDF file.
3. In my first revision of the research, I recommended making a separate section for previous studies (literature review), and it cannot be a subsection in the introduction section, as this section is important in drawing out the gaps between previous studies.
4. The tables and figures are well-organised according to the journal standards.
1. It is important to focus on the preprocessing techniques that have been utilized to prepare data for training.
2. The time needed to train the model is a very important factor in your research contribution. In addition, refer to the hardware specifications (CPU, GPU, and RAM size) used to train and test the model.
No Comment.
No comment.
The reviewers have substantial concerns about this manuscript. The authors should provide point-to-point responses to address all the concerns and provide a revised manuscript with the revised parts being marked in different color.
[# PeerJ Staff Note: The review process has identified that the English language must be improved. PeerJ can provide language editing services if you wish - please contact us at [email protected] for pricing (be sure to provide your manuscript number and title). Your revision deadline is always extended while you undergo language editing. #]
• The English language needs to be improved, and it is recommended that paragraphs be paraphrased to achieve more coherent and cohesive sentences for better understanding.
• In the introduction section, paragraphs 1 and 2 authors presented information without any references. Therefore, it is recommended to add reference citations.
• It is better to start the introduction without this sentence, “Thanks to the developing medical imaging techniques." It is not recommended in academic writing.
• The introduction should be well organised, focusing on the main keywords mentioned in the manuscript title.
• It is recommended to move the reviewed works to a separate section (2 literature review) and review more articles.
• The authors used the EfficientNet family, which consists of 8 predefined CNN-based models (B0 to B7).
• There is no modification or enhancement for the model architecture.
• The authors did not present clarification about the architecture of the EffecientNet model.
The results are measured only in terms of the AUC metric. There are other evaluation metrics such as accuracy, F1, and recall that can be used to indicate a better understanding of the results.
The manuscript, titled 'Comparison of EfficientNet CNN models for multi- label chest X-Ray disease diagnosis' presents significant technical contributions to the field. However, its clarity, accessibility, quality and impact could be further improved by addressing the following points:
1. Abstract: Clearly state the novelty (EfficientNetB7 for multi-label chest X-ray classification), quantify performance gains (e.g., AUC improvement), and briefly mention real-world applications (e.g., aiding radiologists).
2. Introduction: Define the research gap by detailing the limitations of existing methods (complexity, accuracy, real-world use). Justify the EfficientNet choice (efficiency/accuracy balance).
3. Dataset: Address class imbalance (specify the method). Justify image resizing and split ratios, explaining their impact.
4. Model Architecture: Simplify the EfficientNet description (consider a visual diagram). Justify using EfficientNetB7 (superior multi-label performance).
5. Experimental Design: Describe hyperparameter tuning and the rationale behind chosen values. State whether cross-validation was used.
6. Results: Include qualitative analysis (visual examples). Analyze why EfficientNetB7 performs better (multi-label handling, complex cases).
7. Discussion: Discuss limitations (data reliance, unseen datasets, real-time use). Expand on clinical applications (diagnosis assistance).
8. Figures/Tables: Integrate figures (ROC curves, loss charts) better into the text. Improve table visual appeal.
9. Conclusion: Suggest specific future work (other architectures, real-time testing). Emphasize broader implications (improved healthcare).
10. Limitations: Expand this section to address generalizability, data reliance, and ethical considerations.
11. Language: Simplify complex sentences. Ensure clear, professional language. Consider professional editing. (Example: "The proposed method achieved an AUC of 0.8360 for Mass disease.")
12. Reproducibility: Include a "Data and Code Availability" section with links and detailed instructions. Describe preprocessing step-by-step.
13. References: Update with recent work (EfficientNet, multi-label classification) such as Detection of tuberculosis from chest X-ray images: Boosting the performance with vision transformer and transfer learning; Automatic detection of Covid-19 from chest X-ray and lung computed tomography images using deep neural networks and transfer learning
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.