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.
Dear Authors,
Your paper has been revised. It has been accepted for publication in PEERJ Computer Science. Thank you for your fine contribution.
[# PeerJ Staff Note - this decision was reviewed and approved by Mehmet Cunkas, a PeerJ Section Editor covering this Section #]
Good
Good
Good
All the reviewer comments are considered and excuted
1. Highlight your contributions more precisely.
2. Add one paragraph to discuss a more recent non-deep learning-based works including the following work and discuss their limitations in the introduction section to present the importance of the deep learning methods.
3. Add one paragraph to discuss more recent advancements in deep learning for other computer vision applications, specifically for medical applications
no comment
no comment
Highlight the limitations of existing works discussed in the literature (related work).
Dear Authors,
Your paper has been revised. It needs minor revisions before being accepted for publication in PEERJ Computer Science. More precisely;
1) The considered dataset shows a data imbalance between normal and malignancy images. In their analysis, the authors should consider an equal amount of images for both.
2) The ROC curve slope shown in Figure 9 has a strange behavior. The authors should check it.
It is well and good in the revised manuscript.
The reviewer's comments are brilliantly incorporated in the revised manuscript.
The dataset is collected from 118 patients, and it has two classes, including normal cells and leukemia blasts. The normal class contains 3389 microscopic images, and the leukemia blast consists of 7272 microscopic images. It shows data imbalance between normal and malignancy images. The authors should consider an equal amount of images for both.
The ROC curve slope shown in Figure 9 is abnormal. AUC is less in the given figure 9. It should come in a reverse manner (refer to attachment)
The manuscript has been improved.
The experimental section has been improved.
The validity and findings have been improved.
The ROC curve should be above the baseline. Please check
Please address all of the reviewers' points.
**PeerJ Staff Note:** Please ensure that all review, editorial, and staff comments are addressed in a response letter and that any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate.
-
1. There is no preprocessing involved, which is very much required for accurate results. The size of the images is not focused.
2. The evaluation method is done by using usual CNN models and Grey Wolf optimization. Many research works are already available in the same methodology. Authors should highlight the novelty of their work.
1. Novelty is missing.
2. The modified Tiny ViT performance is less than InceptionV3. Being a medical image processing work, accuracy is more important than the complexity of the network.
3. There is no specification of colors used in Figure 1.
4. So many optimization algorithms have evolved nowadays than a grey wolf. Why didn't the authors compare the Grey Wolf algorithm with others?
1. The input image split ratio is given as a table and also as a figure. Authors can avoid repetition of information.
2. There is no clarity of explanation in Figure 8.
3. All the results are obtained nicely. But there is no inference of the results throughout the paper.
4. The Grey Wolf algorithm is most common. There is no need for an explanation of the step-by-step procedure.
5. Authors can use some interesting graphs, like Glyph, spider plots, etc., to show their results instead of the usual bar chart.
The Introduction should be revised and highlight the research gap.
Highlight the role of deep learning and ViT-based models in the the manuscript to underline the technological significance of the work “Enhancing Blood Cell Diagnosis Using Hybrid Residual and Dual Block Transformer Network”, “Blood Cancer Detection Leveraging Deep Learning Fusion Techniques: An Optimized Model Based on VGG16 and AlexNet”, “Morphological diagnosis of hematologic malignancy using feature fusion-based deep convolutional neural network”, “Feature Fusion based Deep Learning method for Leukemia cell classification”
The CNN model description should be revised.
Please provide the complexity of the model.
How proposed model react to the noises
Define each variable value used in the study.
How Gray Wolf Optimization improved results. An ablation study can be added.
Visualize the learning process of the model, including training and validation curves, and analyze any observed trends to provide deeper insights into the optimization behavior.
The results would benefit from a more rigorous statistical analysis to support the claims of performance improvements. Including confidence intervals or significance testing would lend credibility to the findings.
Compared the results of the proposed model with deep learning and ViT-based methods like ConvNeXt, DaViT and CrossViT.
The ROC Plot-based comparison of the methods discussed in Table 6 can be added.
The method should be validated on another dataset for the generalization of the model.
The conclusion should be revised.
How many heads did you select in the proposed study?
Please specify the feature dimension and patch size used in the transformer.
The discussion section should be improved.
Lime XAI need further explanation.
. Highlight your contributions more precisely.
2. Add one paragraph to discuss a more recent non-deep learning-based work, including the following work, and discuss its limitations in the introduction section to present the importance of deep learning methods.
-
Compare the proposed method with existing at least deep learning-based methods published in 2024 or 2025 (at least compare with the presented performance in these works with giving proper citation).
Highlight the future scopes in the conclusion.
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.