Abdomen Fat and Liver Segmentation of CT Scan Images for Determining Obesity and Fatty Liver Correlation
Round 1
Reviewer 1 Report
This manuscript introduces a method to determine the correlation between fatty liver and obesity from analysis of liver segmentation and fat liver from CT images. The manuscript has provided some good findings. However, the author needs to clarify more about the research methods and the process of analyzing the research results. Some comments and suggestions for the author are as follows:
#1: The introduction to datasets should be moved to section 2 "materials and methods".
#2: The Results and Discussion section focuses on providing the empirical results of the study. The results also cause much confusion for readers. The author has presented models as well as flowcharts of the implementation processes, but the results section has not shown the detailed effects of the processes. Authors should add illustrative results for liver and fat segmentation processes from CT images. The results should be presented in tables.
#3: Table 3 shows the correlation coefficients. Could the author explain in more detail about these parameters and how to calculate the extraction to get these results?
#4: There should be more information in the caption of figure 5. What do the colors in the figure mean?
#5: Same goes for Figures 8, 9, and 10. Explanations should be added for the objects in the figure.
#6: Please correct this sentence: “Table 2 shows the interpretation of the correlation coefficient used in this thesis”- Line 285-286. The word "thesis" should be replaced.
Author Response
We would like to thank you for reviewing our manuscript. Your feedback has definitely improved the quality of our manuscript. We have addressed each comment individually and have updated the manuscript accordingly.
The response to each comment is mentioned as follows:
Reviewer 1:
Comment 1: The introduction to datasets should be moved to section 2 "materials and methods".
Response – Thanks for the suggestion. The introduction on the data sets has been moved to section 2 as advised.
Comment 2: The Results and Discussion section focuses on providing the empirical results of the study. The results also cause much confusion for readers. The author has presented models as well as flowcharts of the implementation processes, but the results section has not shown the detailed effects of the processes. Authors should add illustrative results for liver and fat segmentation processes from CT images. The results should be presented in tables.
Response- Thanks for the comment. We appreciate your valuable concern. The results have been demonstrated in tables as recommended.
Comment 3: Table 3 shows the correlation coefficients. Could the author explain in more detail about these parameters and how to calculate the extraction to get these results?
Response- Thanks for the comment. Further details about the parameters on how to calculate the extraction have been given in the revised version of the manuscript.
Comment 4: There should be more information in the caption of figure 5. What do the colors in the figure mean?
Response- Thanks for the suggestion. More details relevant to the figure have been written.
Comment 5: Same goes for Figures 8, 9, and 10. Explanations should be added for the objects in the figure.
Response- Thanks for the comment. A further explanation has been given for the above-mentioned figures.
Comment 6: Please correct this sentence: “Table 2 shows the interpretation of the correlation coefficient used in this thesis”- Line 285-286. The word "thesis" should be replaced.
Response- Thanks for the concern with the text relevant to Table 2. The term thesis has been replaced as advised.
Reviewer 2 Report
This is a study to perform correlation between abdominal fat vs. liver fat using CT of 125 subjects. Segmentation will be performed using computer algorithms to determine regions corresponding to subcutaneous fat and visceral fat, as well as the whole liver. Fat content of the liver will be determined indirectly from Hounsfield Units (HU) where lower value indicates greater fat content.
The significance / purpose of the study is unclear. I understand that the authors are trying to show correlation between how much fatty tissue a person has in different compartments vs. how fatty the liver is, but scientific or medical significance of this information needs to be better presented.
Overall, the methodology to perform segmentation and correlation appear sound, with appropriate figures. However, mathematical / algorithmic details regarding automated segmentation could be presented better, using equations, etc. rather than in words (for example line 177 states: “Gray level, Gaussian gradients, region growing algorithm, ….” but these cannot be used by a reader to replicate your method).
Equation 4, 5, 6 and Table 2 could be removed, as these are standard statistics of linear correlation.
There are so many typographical error throughout the paper, starting with the very first sentence of the Abstract. Please perform a spelling check, a grammar check, and read thoroughly before submitting.
line 10: “T”his research
line 17: what is “abdomen size ration” do you mean ratio? Why would you get a ratio of visceral fat to abdomen size?
line 93: “Tow” should be “Two”
Author Response
We would like to thank you for reviewing our manuscript. Your feedback has definitely improved the quality of our manuscript. We have addressed each comment individually and have updated the manuscript accordingly.
The response to each comment is mentioned as follows:
Reviewer 2
Comment 1: The significance / purpose of the study is unclear. I understand that the authors are trying to show correlation between how much fatty tissue a person has in different compartments vs. how fatty the liver is, but scientific or medical significance of this information needs to be better presented.
Response- Thanks for the comment. The significance of the study has been elaborated throughout the paper as advised.
Comment 2: There are so many typographical error throughout the paper, starting with the very first sentence of the Abstract. Please perform a spelling check, a grammar check, and read thoroughly before submitting.
line 10: “T”his research
line 17: what is “abdomen size ration” do you mean ratio? Why would you get a ratio of visceral fat to abdomen size?
line 93: “Tow” should be “Two”
Response- Thank you for your comment. The paper has been checked and edited by all authors and all grammar and typos mistakes have been corrected.
Round 2
Reviewer 1 Report
Thanks for the authors' responses and revisions. I am satisfied with your replies.
Author Response
Thank you so much for reviewing our manuscript.
Reviewer 2 Report
It is difficult to review this paper since there is no line-by-line response to indicate which part of the critique has been addressed. All I see is a manuscript with bunch of yellow highlights.
Having said that, and reading through the paper quickly, I still feel that the rational for the study is not sufficiently provided.
The entire study is justified by one sentence, "There is a hypothesis that visceral fat produces free fatty acids which exposes the liver to fat accumulation [18]" That is insufficient, as correlation does not infer causality, and the practical significance of the study is also unclear.
Equations regarding t-test still remain, even though they were asked to be removed.
Author Response
We would like to thank you for reviewing our manuscript, it has helped us to improve the quality of the manuscript.
We have answered all the comments and have provided line-by-line responses as follows:
Comment 1: It is difficult to review this paper since there is no line-by-line response to indicate which part of the critique has been addressed. All I see is a manuscript with bunch of yellow highlights.
Response- Thank you for your feedback and please accept our sincere apologies if we may not address your valuable comments properly. All comments have been properly addressed and made our best effort to provide the most adequate answers to your questions/comments. For easy tracking with the comments and as advised by the journal editor, we have highlighted the text in the manuscript to indicate the changes that have been made throughout the paper. Besides, a line-by-line response to indicate which part of the critique has been addressed is also given in our responses to the reviewers’ report (See the below responses).
Comment 2: Having said that, and reading through the paper quickly, I still feel that the rational for the study is not sufficiently provided.
Response- Thank you for your comment. The rationale of the study has been discussed in the paper. We explained why this study is needed. We also described how the research community can benefit from our study. (Please see lines 99-106).
Comment 3: The significance / purpose of the study is unclear. I understand that the authors are trying to show correlation between how much fatty tissue a person has in different compartments vs. how fatty the liver is, but scientific or medical significance of this information needs to be better presented.
Response – We appreciate your valuable comment. The scientific and medical significance of this study has been elaborated (Please see lines 99 – 106).
Comment 4: Overall, the methodology to perform segmentation and correlation appear sound, with appropriate figures. However, mathematical/algorithmic details regarding automated segmentation could be presented better, using equations, etc. rather than in words (for example line 177 states: “Gray level, Gaussian gradients, region growing algorithm, ….” but these cannot be used by a reader to replicate your method).
Response- Thank you for the comment. The mathematical/algorithmic detail steps regarding automated segmentation have been discussed deliberately. There are five main steps in the process and each step has been explained in detail (Please see 230 – 439).
Comment 5: The entire study is justified by one sentence, "There is a hypothesis that visceral fat produces free fatty acids which exposes the liver to fat accumulation [18]" That is insufficient, as correlation does not infer causality, and the practical significance of the study is also unclear.
Response- Thanks for the suggestion. The practical significance of the study has been explained in the abstract (Please see lines 23 – 28). We explain the real-world applications that could benefit from our proposed work. We also explain how our proposed solution can improve in prevention and treatment of fatty liver disease. Most importantly, the main contribution of the work has been discussed in the revised version of the manuscript (Please see lines 124 – 135).
Comment 6: Equation 4, 5, 6 and Table 2 could be removed, as these are standard statistics of linear correlation.
Response- Thank you for the suggestion. Equations 4, 5, and 6 have been removed as suggested and equations have been re-numbered accordingly (Please see lines 582 – 591). We also removed Table 2 from the revised version of the manuscript as advised (Please see lines 540 – 543).
Comment 7: Why would you get a ratio of visceral fat to abdomen size?
Response- Thank you for your concern. The reason why we get a ratio of visceral fat to abdomen size has been described in both the abstract and in the results and discussions sections (please see lines 18-21) and (lines 565- 568).
