Next Article in Journal
Thermal Hysteresis in Melting–Solidification of Nanoparticles
Next Article in Special Issue
Node2Node: Self-Supervised Cardiac Diffusion Tensor Image Denoising Method
Previous Article in Journal
An Approach Based on Simulation and Optimisation for the Intermodal Dispatching of Public Transport and Ride-Pooling Services
Previous Article in Special Issue
Explainable-AI in Automated Medical Report Generation Using Chest X-ray Images
 
 
Article
Peer-Review Record

Artificial Intelligence Approach for Early Detection of Brain Tumors Using MRI Images

Appl. Sci. 2023, 13(6), 3808; https://doi.org/10.3390/app13063808
by Adham Aleid *, Khalid Alhussaini, Reem Alanazi, Meaad Altwaimi, Omar Altwijri and Ali S. Saad *
Reviewer 1: Anonymous
Reviewer 2:
Appl. Sci. 2023, 13(6), 3808; https://doi.org/10.3390/app13063808
Submission received: 9 February 2023 / Revised: 9 March 2023 / Accepted: 13 March 2023 / Published: 16 March 2023
(This article belongs to the Special Issue Advances in Medical Image Analysis and Computer-Aided Diagnosis)

Round 1

Reviewer 1 Report

The authors have presented a model entitled "Artificial Intelligence for Early MRI Brain Tumor Recognition." The subject is relevant to medical image processing. Yet, the paper's primary contribution is rather modest, which hinders its effectiveness.

1. The title of the manuscript is ambiguous and can be improved

2. The abstract of a manuscript typically includes a problem statement, followed by a discussion of the methodology, the results, and concluding remarks.

3. It is unclear what the main contribution of this paper is; it appears that the authors have rehashed known technologies.

4. What is the purpose of this investigation? Yet, other approaches have already achieved excellent accuracy in detecting cancers and are documented in the literature.

5. Authors should strengthen their literature work with recent technologies, 

http://www.aimspress.com/article/doi/10.3934/mbe.2021292

https://www.sciencedirect.com/science/article/abs/pii/S0045790622003603

6. What problems do existing approaches present? How the recommended solution to resolving these issues should be addressed in depth is required.

7. The methods proposed should be explored in detail; the functioning mechanism is unclear.

8. The results are insufficient to justify publication; authors should compare their work to the most relevant and established procedures for the same purpose.

9. How fast is the procedure? Can you explain the trade-off between speed and accuracy?

10. Should the presentation, coherence, and English of the manuscript be improved?

Author Response

Response to reviewer 1 comments

 

First we would like to thank the reviewer for the valuable comments and suggestions which certainly improves the quality of the manuscript.

The following is the answer to the reviewer comments and suggestions.

Best regards.

*1. The title of the manuscript is ambiguous and can be improved

Ans :  The title was slightly modified second reviewer is happy with the title

*2. The abstract of a manuscript typically includes a problem statement, followed by a discussion of the methodology, the results, and concluding remarks.

Ans: Abstract was updated

  1. It is unclear what the main contribution of this paper is; it appears that the authors have rehashed known technologies.

Ans: It is clear that we developed a classic harmonic search algorithm to suit MRI brain segmentation and compared to CNN and deep learning methods

  1. What is the purpose of this investigation? Yet, other approaches have already achieved excellent accuracy in detecting cancers and are documented in the literature.

Ans: Accuracy is not the only factor used for brain segmentation the majority of researchers focus on Dice Index which provides a more reliable segmentation context. I our proposed method it almost real-time diagnosis while other methods need hours.

*5. Authors should strengthen their literature work with recent technologies, 

Ans: literature work is strengthened with recent publications (introduction section page2 )

  1. What problems do existing approaches present? How the recommended solution to resolving these issues should be addressed in depth is required?

Ans: It was added according to reviewer's comments

  1. The methods proposed should be explored in detail; the functioning mechanism is unclear.

Ans: the schematic diagram  of the algorithm added to provide more details about the method and the functioning mechanism. (figure 2)

  1. The results are insufficient to justify publication; authors should compare their work to the most relevant and established procedures for the same purpose.

Ans: Comparison is done with other new most relevant methods for the same purpose and on the same database BraTS dataset (see table 1)

  1. How fast is the procedure? Can you explain the trade-off between speed and accuracy?

Ans: The average execution time was added in the table 1 of comparison

  1. Should the presentation, coherence, and English of the manuscript be improved?

Done.

 

 

Author Response File: Author Response.docx

Reviewer 2 Report

Authors have taken an interesting topic of early detection of brain tumors using MRI images.

From the point of innovation and framework of the paper, this paper may be considered if the following comments are taken care of:

1. The presentation of the work is clear with regard to the flow of content.

2. The topic is current. The paper integrates methodological solutions which already exist, which is my great concern and remark.

3. The title and abstract of the paper clearly reflect its content. However, in the abstract, the author needs to specify the novelty of the proposed work.

4. . The authors pay less attention to the related progress. The paper should be updated to include more recent references, preferably from the last 2 or 3 years.

4. The advantages and disadvantages of the previous work are not clearly expounded, in other words, the motivation for writing the paper is not explained. Better highlight the novelty of the study

5. Authors mentioned, "In this study, we translated data from 3D volumes into 2D images to reduce the computational cost and ease the segmentation problem.". The translation process should be explained in a detailed manner.

6. It would be nice to make a picture of the algorithm of the realized research.

7. The proposed method is compared against time-consuming DL  methods, it would be nice to include the computational complexity of the proposed model along with the other benchmarked methods.

8. The authors need to clearly provide several solid future research directions. Clearly state your unique research contributions in the conclusion section. Add limitations of the model.

9. Lastly, some spelling, punctuation and grammatical mistakes can be taken care of with the help of a native speaker.

ex. 

Line 133. 

threshold based segmentation techniques could be local or global one. - Line should start with capital letter.

10. References should be throughly checked as per the format of the journal. Ref. [8a] to be rechecked - Line no. 167

Good luck

Regards

 

Author Response

Response to reviewer 2

 

First we would like to thank the reviewer for the valuable comments and suggestions which certainly improves the quality of the manuscript.

The following is the answer to the reviewer comments and suggestions.

Best regards.

 

Comments and Suggestions for Authors

Authors have taken an interesting topic of early detection of brain tumors using MRI images.

From the point of innovation and framework of the paper, this paper may be considered if the following comments are taken care of:

  1. The presentation of the work is clear with regard to the flow of content.

No action

  1. The topic is current. The paper integrates methodological solutions which already exist, which is my great concern and remark.

ANS: The methods are already exist but never applied to MRI brain images, the algorithm was modified to fit to the MRI brain segmentation.

  1. The title and abstract of the paper clearly reflect its content. However, in the abstract, the author needs to specify the novelty of the proposed work.

Ans: The novelty was introduced in the abstract.  And at the end of the introduction.

  1. The authors pay less attention to the related progress. The paper should be updated to include more recent references, preferably from the last 2 or 3 years.

Ans: The recent publication in the field was add in the introduction.

  1. The advantages and disadvantages of the previous work are not clearly expounded, in other words, the motivation for writing the paper is not explained. Better highlight the novelty of the study

Ans: It was added in the abstract and last section of the introduction

  1. Authors mentioned, "In this study, we translated data from 3D volumes into 2D images to reduce the computational cost and ease the segmentation problem.". The translation process should be explained in a detailed manner.

Ans: It was explained in the method page 4

  1. It would be nice to make a picture of the algorithm of the realized research.
  2. The proposed method is compared against time-consuming DL  methods, it would be nice to include the computational complexity of the proposed model along with the other benchmarked methods.

Execution time and computational complexity were added in table 1.

  1. The authors need to clearly provide several solid future research directions. Clearly state your unique research contributions in the conclusion section. Add limitations of the model.

Ans: conclusion section was updated according to the reviewer comment.

  1. Lastly, some spelling, punctuation and grammatical mistakes can be taken care of with the help of a native speaker.
  2. Line 133. threshold based segmentation techniques could be local or global one. - Line should start with capital letter.

English improvement was Done for the whole manuscript.

  1. References should be thoroughly checked as per the format of the journal. Ref. [8a] to be rechecked - Line no. 167done

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors modified the article and highlighted the text in the revised manuscript. However, the authors did provide the response to the reviewer letter, I did not fine how the authors response to the reviewers' comments. The authors should provide a detailed response to reviewers comment doc. The following modifications are required:

1. According to the revised article, the title of the article is  is ambiguous. The order of the words is wrong, i.e., Artificial Intelligence Approach for Early Detection of Tumors  Brain using MRI images, it should be Artificial Intelligence Approach for Early Detection of Brain Tumors using MRI images,

2. Furthermore, the presented study compared few CNN and DLA segmentation techniques; however the authors should compare the proposed model with latest start -of-art-approaches. 

3. There is no flowchart of the proposed framework. How the model perform the brain tumors task. 

4. The algorithmic steps are just in simple language. It should in pseudo code form. 

 

 

Author Response

Response to Reviewer 1

we thank the reviewer for the valuable comments which certainly will improve the manuscript.

Reviewer 1 comments and response to them

The following modifications are required:

  1. According to the revised article, the title of the article is  is ambiguous. The order of the words is wrong, i.e., Artificial Intelligence Approach for Early Detection of Tumors  Brain using MRI images, it should be Artificial Intelligence Approach for Early Detection of Brain Tumors using MRI images,

Ans : The title was adjusted according to the reviewer comment.

  1. Furthermore, the presented study compared few CNN and DLA segmentation techniques; however the authors should compare the proposed model with latest start -of-art-approaches. 

Ans: The reviewer wants us to compare with the latest “start of art approaches”. We suppose he meant “state of the art approaches”. We did the comparison with a chosen set of methods which is participated in an international competition “BraTS-2017-2021” and used the same set of data “BraTS 2017-2021”, and these data were segmented manually by experts in the field. We can not compare with methods using different set of data and does not calculate the Dice Index. The justification is added to the manuscript.

  1. There is no flowchart of the proposed framework. How the model perform the brain tumors task.

Ans: Figure 1. Represent the general flowchart of the proposed method, and figure 2 represent the segmentation process based on multilevel threshold using HSO algorithm.

  1. The algorithmic steps are just in simple language. It should in pseudo code form. 

Ans: We did it in simple language for clarity and simplicity for the readers.

Regards

 

Author Response File: Author Response.docx

Reviewer 2 Report

The Authors have responded to the comments to some extent. I recommend the paper can be published after minor grammatical editions. 

Author Response

Response to reviewer 2

We thank the reviewer for the valuable comments which will improve the manuscript

The Authors have responded to the comments to some extent. I recommend the paper can be published after minor grammatical editions. 

Ans: The English was revised and grammatical editions were performed.

Regards

 

Author Response File: Author Response.docx

Back to TopTop