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 Dr. Yoo,
Thank you for submitting your revised manuscript titled "A retrospective evaluation of individual thigh muscle volume disparities based on hip fracture types in Followed-up Patients: An AI-Based Segmentation Approach Using UNETR". After reading your response letter and the revised manuscript I’m happy to let you know that decision is “Accept”.
[# PeerJ Staff Note - this decision was reviewed and approved by Jeremy Loenneke, a PeerJ Section Editor covering this Section #]
Dear Dr. Yoo,
Thank you for resubmitting your manuscript titled "A retrospective evaluation of individual thigh muscle volume disparities based on hip fracture types in followed-up patients: An AI-Based Segmentation Approach Using UNETR". The manuscript is improved, yet it needs some additional work, thus the decision is “Minor revision”.
Some of your responses were not incorporated into the manuscript:
(2) Based on figure 2 – only part of the abdominal oblique mm. and rectus abdominis m. were segmented. How did that affect their analysis?
Response:
The starting slice location for the CT scans varied across subjects (L3 or L4), resulting in an inability to fully capture the complete extent of the abdominal oblique mm and rectus abdominis muscle. As a result, comparing or analyzing these muscles individually was impractical and could potentially lead to inaccurate conclusions. Therefore, these muscles were excluded from the analysis and categorized under the “Else” group. In this study, we primary focus on evaluating muscle disparities in the thigh region. Figure 2 illustrates the expected outcome of the segmentation model rather than the specific data utilized for the analysis.
- Please add a short statement to that effect in the materials and methods section.
(4) A comment from a previous reviewer was not fully addressed yet - The authors should add some evaluations of muscle quality in their study, or at least give a detailed discussion about this topic. The assessment of muscle fat infiltration is an obvious advantage of CT or MR imaging, so when the study just focuses on muscle size it is of less of interests.
Response: We acknowledge the significance of evaluating the intermuscular or intramuscular adipose tissue in aging population and their physical capabilities. However, our model was developed utilizing a labeled dataset, which was derived through applying a thresholding on fat tissues to segregate them as background. Consequently, this results in a model output that exclusively identifies ‘muscle’ tissue, omitting any fat content within muscle. This constraint challenges our ability to precisely evaluate the percentage of muscle fat infiltration in the outputs from our segmentation model. We are planning to conduct further research to revise both the training dataset and the model comprehensively. Presently, our study is centered on the disparities in muscle volume observed before and after surgical intervention.
- Please add a short statement to that effect in the materials and methods section.
In addition, please address these items:
(1) Nicholson & Kotowicz, 2012 was added to the main text (L301), but it is missing from the references.
(2) L245-8: “However, in the male group, some major muscles like the Sartorius (2.4%), Vastus intermedius (2.8%), Vastus medialis (0.8%) and Gluteus maximus (3.6%) demonstrated particularly pronounced disparities for intertrochanteric fractures when compared to femoral neck fractures”. This text was added in reference to table 3. Can you please explain the % values as I can’t correlate the text to the table.
(3) Figures 7 and 8 – please explain the change of values in the y-axis. This seems to be a major change in data analysis, yet the authors did not address it in their letter of response.
(4) Table 1 (new) – please make sure that all values have the same significant decimal numbers (i.e., 1 should be 1.0 etc.). Furthermore, please note that the "Femoral Neck Fracture group" and "Intertrochanteric Fracture group" columns sum up to more than 100% (100.2% and 100.1% respectively).
Please ensure that all editorial, and staff comments are addressed in a response letter and any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate.
Please note that submitting a revision of your manuscript does not guarantee eventual acceptance, and that your revision may be subject to re-review by the reviewer(s) before a decision is rendered.
**PeerJ Staff Note:** Please ensure that all review and editorial 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. It is a common mistake to address reviewer questions in the response letter but not in the revised manuscript. If a reviewer raised a question then your readers will probably have the same question so you should ensure that the manuscript can stand alone without the response letter. Directions on how to prepare a response letter can be found at: https://peerj.com/benefits/academic-rebuttal-letters/.
Dear Dr. Yoo,
Your manuscript titled "A retrospective evaluation of individual thigh muscle volume disparities based on hip fracture types in followed-up patients: An AI-Based Segmentation Approach Using UNETR" was considered by an expert reviewer and based on their opinion and my review, the decision is “Major revision”.
Please carefully read the reviewers’ comments and address them fully in your revised manuscript. Specifically, please address the following points:
General comments
(1) Abstract results are still given in mm^3
(2) Based on figure 2 – only part of the abdominal oblique mm. and rectus abdominis m. were segmented. How did that affect their analysis?
(3) Discussion – The discussion section is still very limited and needs revision. Currently it does not include much literature and previous relevant papers. How are current findings corresponding to previous ones? Are they in agreement or not. If not – what are some possible reasons for the differences?
(4) A comment from a previous reviewer was not fully addressed yet - The authors should add some evaluations of muscle quality in their study, or at least give a detailed discussion about this topic. The assessment of muscle fat infiltration is an obvious advantage of CT or MR imaging, so when the study just focuses on muscle size it is of less of interests.
(5) Figures and tables legends are still very short and are missing a lot of information (e.g., Figure 1: Dataflow, Figure 2: Ground truth image etc.). Each figure and table should be a “standalone”, meaning the reader should not need to search the text to understand the figure/table but rather be able to read the legend and understand what they see.
(6) Table 2: (i) please explain in the legend all the missing loss ratios (e.g., Male FNF loss ratio raw is missing many values). (ii) Many of the loss values are in the hundreds (cm^3). For example, loss averages for gluteus maximus are 300-740 cm^3. These are huge losses. Did the authors check the deep learning output to see if no errors were introduced?
Specific comments
L35: “relative absolute volume difference 0.032 ± 0.095”. Are there any units for this value? (same thing in L81).
L46: “presented a unique pattern”. Vague. Please be specific.
L56: “A femoral neck fractures” should be “fracture”.
L57: “occurs just below the ball of the hip joint, in atrophy in the hip flexors”. This sentence seems to miss text. It is incomprehensible.
L70: “Automated Muscle Segmentation from Clinical CT Using Bayesian U-Net for Personalized Musculoskeletal Modeling”. Is that a reference? Not sure what it means.
L80: Please explain in text what “disc core” means.
L130: In materials and methods section please specify the average age for males and females.
L196: should be measured in cm^3.
L218: “5.44” should be “5.4”
Figure 1: in excluded box – what is “No available CT cut (n=17)”?
Figure 3: please add the general location of each presented slice (e.g., 60th slice, mid thigh, etc.).
Figures 6/7/8: please remove superimposed text (A/B/C/D) from boxplots and bars. It does not add any information.
Figures 7/8: text of x-axis seems to be truncated.
Figures 7/8 and Table 2: ITF and FNF abbreviations were never explained before in the text.
Please ensure that all review, editorial, and staff comments are addressed in a response letter and any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate.
Please note that submitting a revision of your manuscript does not guarantee eventual acceptance, and that your revision may be subject to re-review by the reviewer(s) before a decision is rendered.
The manuscript is well written and with the changes easier to read. The authors have mostly addressed all previous concerns.
no comment
no comment
The results of the abstract should match the results section, specifically the units and number of significant digits used.
The discussion is still very light on comparisons with other previous studies, e.g. Hardy et al. It would strengthen the conclusions if you related your results with those previously published. That is, are your results similar or dissimilar to other studies? What does that mean?
Dear Dr. Yoo,
Your manuscript titled "A retrospective evaluation of individual thigh muscle volume disparities based on hip fracture types in followed-up patients: An AI-Based Segmentation Approach Using UNETR" was considered by two expert reviewers and based on their opinions and my review, the decision is “Major revision”.
Please carefully read the reviewers’ comments and address them fully in your revised manuscript. Specifically, please address the following points:
(1) Both reviewers commented on the small sample size (n =18), which gets even smaller as the authors divide their cohort into several sub-groups (first - femoral neck and intertrochanteric fractures, and then into female and male within these groups). Reviewer #1 suggested the possibility of merging some subgroups to avoid extremely small sample sizes (e.g., male/female; n = 2). Please address this point in your response and explain if this is possible or not.
(2) Revise discussion. The current discussion is too short and has limited benefit to the reader. Please add a literature review and specify how your results correspond and relate to previous studies.
(3) Figures and tables legends are very short and are missing a lot of information. Each figure and table should be a “standalone”, meaning the reader should not need to search the text to understand the figure/table but rather be able to read the legend and understand what they see.
(4) Table 1: Significant decimal digits – 1 significant decimal digit, rather than 3, is more than enough. Switching to 1 significant decimal digit would not change the information but would make the table much easier to read.
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.
Please note that submitting a revision of your manuscript does not guarantee eventual acceptance and that your revision may be subject to re-review by the reviewer(s) before a decision is rendered.
The objectives and aims of the study are quite generally described. The manuscript would benefit from making them more specific.
The Conclusions should be shortened and made more to the point. Specifically, it should clearly relate to the objectives and aims of the study.
The enclosed consent form does not seem to be the one actually used, since it is in English and still contains the instructions for writing it. The manuscript does not contain any specific mention of the informed consent. Otherwise a high ethical standard is apparent.
One limitation, which is correctly described, is that segmentations could have been affected by screws or other implants. It is not described whether any visual validation of the post-operative segmentations was done. Since the sample size is so small, it would be trivial to manually review all segmentations, which would add a lot of strength to the study.
The sample size is quite small, especially given that subgroups are analysed. One subgroup consists of only two patients.
The paragraph in the results with percentages related to intertrochanteric and femoral neck fractures is a little difficult to follow. It would probably be easier if it was described and justify in the Methods section, to make the Results more concise.
The last paragraph of the Results is argumentative and doesn't really add any results. Would be better placed in the discussion.
The Discussions should be expanded with a discussion of how these results relate to other studies and previous knowledge.
The muscle volumes are given as cubic millimetres with 3 decimals. This measurement uncertainties makes this unreasonable. Would suggest that the volumes are described as cubic centimetres (ml) with one or two decimals instead. This would also make the numbers easier to read.
A minor point in paragraph 2 of the Results section is that negative losses are described, which implies an increase in volume. The negative signs should be removed.
The tables and figures would benefit from a little more detailed legends and defining the abbreviations. For example, it is not obvious what units the tables use. Would suggest that you try to make the tables and figures self-contained, so they can be understood on their own without a detailed reading of the text of the manuscript.
The MS is well written and good organized. The specific muscles segmentation is very time-consuming so I appreciate the authors’ efforts on this study. However, there are several concerns need to be addressed:
1. A nice review (Edward J. O. Hardy et al. DOI: 10.1002/jcsm.13067) found the rates of atrophy appeared to vary between muscle groups (greatest in triceps surae (11.2% day 28), followed by quadriceps (9.2% day 28), then hamstrings 6.5% day 28). The data seem largely different from the loss rates in your study, how to explain this discrepancy? I also did not see the time intervals between pre- and post-frx CT scans.
2. This study only focused on muscle volume or size change, however, several recent studies (DOI: 10.1002/jcsm.12996; DOI: 10.1002/jcsm.13261) high value the muscle radiodensity of hip muscles associated with clinical outcomes in hip fracture patients as well as the physical performance (https://doi.org/10.1016/j.jamda.2020.06.052). Therefore, why not add muscle fat infiltration or density variables in the study?
3. Did all the hip frx patients have surgery? This will affect the mobility after hip frx.
4. The big limitation of this study is the pretty small sample size, in particular the authors separate the groups by frx type and gender. Pls give the sample size analysis. I would suggest the authors combine the fracture type groups or the gender groups, but make frx type or gender as a covariate.
.
.
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.