Skip to Content
MDPIMDPI

Advanced

by
  • Bushira Musa 1,
  • Ji Chen 1,* and
  • Alexander Peebles 1
  • et al.

Reviewer 1: Hamidreza Barnamehei Reviewer 2: Anonymous Reviewer 3: Anonymous

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors

Thank you for submitting you study. Your work is novel and use fine method (OpenSim Simulation). Your work is valuable, Although, I belive my comments and suggestion can improve your work. This manuscript presents a well-structured and technically sound study comparing the performance of an OpenSim musculoskeletal modeling workflow against reference datasets for estimating lower-limb joint angles and moments under various gait conditions. The study is clearly motivated by applications in astronaut health and reduced-gravity biomechanics, and the methodology is appropriately detailed. The results demonstrate generally strong correlations and acceptable errors, with insightful discussion of discrepancies. Below are my comments and suggestions to further strengthen the manuscript before publication:


Major Comments:

  1. Sample Size Justification: The study is limited to one participant per dataset. While the focus is on establishing and validating a workflow, the authors should more explicitly justify why only one subject was chosen from each publicly available dataset and clarify whether the full datasets were explored to confirm this participant was representative. A statement on the generalizability of findings given the sample size should be included in the Limitations section.
    2. Clarity on Ankle Modeling Discrepancies: The discussion of the lower PCC and higher RMSE for ankle kinematics/kinetics in Dataset 1 is informative but could be more precise. The authors mention different "ankle segment frames" and "joint center estimation methods" between models. It would be helpful to briefly specify the key differences (e.g., anatomical landmark definitions, rotation sequence) and how these inherently lead to offset biases, as the figures suggest.
    3. Force Plate Setup and Moment Accuracy (Dataset 1): The explanation that the two-force-plate setup in Dataset 1 limits the accuracy of joint moment estimation, particularly during double stance and consecutive strides, is a critical point. The authors should consider moving this from the limitations into the main results discussion (Section 4) to provide context before presenting the moment correlation values for Dataset 1. This would help readers interpret the "strong" vs. "excellent" categorization more critically.
    4. Muscle Activation and Forward Dynamics: The authors mention that RRA and CMC steps were attempted but did not yield muscle activations correlating well with EMG, and forward dynamics failed. This is a significant limitation for the stated long-term goal. The manuscript would benefit from a brief summary in the Discussion (even 1-2 sentences) of what specific challenges were encountered (e.g., which muscles, what pattern of discrepancy) and why this is a common hurdle, rather than only describing the general difficulty of configuring RRA/CMC.
    5. Future Work Specificity: The future work section mentions improving the workflow by testing more participants and conducting a new study with a body-weight offload system. Please add more specific goals for this new data collection: What specific hypotheses about reduced-gravity gait or model performance will it test? How will it address the current limitations (e.g., ankle modeling, double stance kinetics)?

 

Minor Comments:
1. Abstract: The final sentence is cut off: "...simulated". Please complete the sentence.
2. Introduction: The transition from general astronaut biomechanics to the specific gap filled by OpenSim is good. Consider adding a sentence after referencing OpenSim's five-step process to explicitly state that the present study focuses on validating the first three steps (Scaling, IK, ID) as a foundation.
3. Section 2.1 (Datasets): For Dataset 1, it states joint moments were computed using "Newton’s law of motion and principles of moments." Please specify the specific inverse dynamics approach used by the original authors if known (e.g., standard Newton-Euler inverse dynamics).
4. Section 2.2 (OpenSim Workflow): The description of marker weighting is clear. For transparency, consider adding the actual weights used for the key markers (e.g., pelvis: 500, ankle: 25) in the main text, as this is a crucial modeling choice affecting results.
5. Figures 2, 3, 4, 5: In the captions, please clarify that the shaded areas represent standard deviation across trials. Ensure all axis labels are fully defined in the captions or on the figures themselves (e.g., "% Gait Cycle," "Angle (deg)," "Moment (Nm/kg)").
6. Tables 1 & 3 (PCC): The PCC value for "Ankle dorsi/plantar flexion moment" during heel walking in Table 1 is -0.223. In the text (Page 6, Results), the average is given as 0.677 ± 0.600. The negative value for one condition drastically affects this average and standard deviation. The authors should comment briefly in the text on the implication of this negative correlation for that specific task.
7. Page 13, Discussion: "We argue that ankle flexion angles and moments calculated from our OpenSim workflow... might be more accurate..." This is a strong claim. Rephrase to be more cautious, e.g., "The pattern of ankle angles/moments estimated by our OpenSim workflow appears more physiologically plausible for heel walking, based on evidence from other gait studies..."
8. References: All references appear relevant and appropriately cited. Please ensure the formatting is consistent with the journal's style guide (e.g., use of "et al.", journal abbreviations).

 

Thank you again for your valuable work.

Comments on the Quality of English Language


Dear Authors
Your work is generally well-written in clear and professional English. The technical descriptions are precise, and the overall structure supports readability. However, there are a number of minor grammatical errors, typos, and awkward phrasings that should be corrected to meet formal publication standards. Below is a list of suggested corrections and improvements.

Detailed Suggestions:

Page 1, Abstract:
   · "Correspondance" → Correspondence
   · "We speculate that lower PCC values and higher RMSE present with ankle dorsi/plantar flexion angle..." → We speculate that the lower PCC values and higher RMSE observed for ankle dorsi/plantar flexion angle...
   · The final sentence is cut off: "...in simulated" → ...in simulated reduced-gravity environments.
2. Page 1, Keywords:
   · "Keywords:musculoskeletal modelling," → Keywords: musculoskeletal modelling, (add space after colon)
3. Page 2, Introduction:
   · "DeWitt et. al." → DeWitt et al. (standard format for "et al.")
   · "McCrory et. al." → McCrory et al.
   · "...during in-flight 0G treadmill running exercise, in which the mechanical load to joints can be increased..." → ...during in-flight 0G treadmill running exercise, where mechanical load on the joints can be increased...
   · "As decades long effort," → As part of a decades-long effort,
4. Page 3, Introduction:
   · "OpenSim overcomes these limitations by combining inverse dynamics with optimization methods to distribute joint torques across individual muscles based on their anatomical properties." → This sentence is clear but slightly long. It is acceptable as is.
5. Page 4, Materials and Methods:
   · "LAMB prototocol" → LAMB protocol
   · "...and the LAMB prototocol based full-body marker set..." → ...and the LAMB protocol-based full-body marker set...
   · "Joint moments were therefore computed on one leg..." → Joint moments were therefore computed for one leg...
6. Page 4, Section 2.2:
   · "The OpenSim workflow begins with the scaling step, where we customized a generic OpenSim model..." → The OpenSim workflow begins with a scaling step, in which we customize a generic OpenSim model... (Use present tense for describing methodology).
   · "Markers like the pelvis and knee are assigned with heavier weights" → Markers such as those on the pelvis and knee are assigned heavier weights
   · "...markers are placed further away from their corresponding anatomical landmarks in foot segment of the OpenSim model for dataset2, compared to those for dataset1, as the dataset2 participant walked with shoes on whereas the dataset1 participant walked with bare foot." → This is awkward. Suggest: For Dataset 2, markers were placed further from their corresponding anatomical landmarks on the foot segment of the OpenSim model compared to Dataset 1, as the participant in Dataset 2 wore shoes, while the participant in Dataset 1 was barefoot.
7. Page 5, Figure 1 Caption:
   · "Pink markers represent the anatomical landmarks. The positive X, Y and Z of global coordinates are showed as the lines of red, green and blue color respectively..." → Pink markers represent anatomical landmarks. The positive X, Y, and Z axes of the global coordinate system are shown as red, green, and blue lines, respectively...
   · "a) the scale model" → a) the scaled model
   · "b) the scaled model using lower limb marker set for dataset 2." → b) the scaled model using the lower-limb marker set for Dataset 2.
8. Page 5, Section 2.3:
   · "Joint moment data from OpenSim ID output was normalized to participant mass and was also time-normalized..." → Joint moment data from the OpenSim ID output was normalized to participant body mass and time-normalized...
   · "Custom MATLAB scripts were developed to compare normalized joint angles and moments between the reference data and OpenSim data across different walking conditions." → This is clear.
9. Page 6, Section 2.3:
   · "It is categorized into four levels of similarity: excellent when PCC is greater than 0.9, strong similarity when it is between 0.7 and 0.9, moderate if it is between 0.4 and 0.7, low if it is less than 0.4 [18], [19]." → Similarity is categorized into four levels: excellent (PCC > 0.9), strong (0.7 ≤ PCC ≤ 0.9), moderate (0.4 ≤ PCC < 0.7), and low (PCC < 0.4) [18,19].
10. Page 8, Table 1 Title:
    · "Correlation coefficients (PCC) between OpenSim results and dataset1 values across four walking conditions." → Correlation coefficients (PCC) between OpenSim results and Dataset 1 reference values across four walking conditions.
11. Page 9, Text above Table 2:
    · "...for the hip and knee flexion/extension angles, as well as ankle dorsiflexion/plantarflexion angles respectively..." → ...for hip flexion/extension, knee flexion/extension, and ankle dorsiflexion/plantarflexion angles, respectively... (Apply same correction to the moment list).

12. Page 12, Table 4:
    · "Knee flexion/extension moment (Nm/kg)" → The values in this row (e.g., 0.113, 0.094) seem inconsistent with the high PCCs (~0.97). Please verify the units and values. The note in the Mean column says (0.031) but the text on Page 12 says 0.078 ± 0.0131. Ensure consistency.
13. Page 12, Discussion:
    · "Most of our simulated results show either excellent or strong correlation with reference data for hip, knee and ankle mechanics across various conditions, except for ankle moment during heel walking condition in dataset1." → Most of our simulated results show excellent or strong correlation with reference data for hip, knee, and ankle mechanics across various conditions, except for the ankle moment during the heel-walking condition in Dataset 1.
    · "Smaller RMSE values (less than 5 deg) are present with hip and knee flexion/extension angles between two approaches for both datasets." → Smaller RMSE values (< 5 deg) were found for hip and knee flexion/extension angles between the two approaches for both datasets.
    · "This discrepancy might be due to lower weight we placed for ankle and toe markers in OpenSim IK setup file and as well as difference in ankle and toe joint center estimation methods..." → This discrepancy might be due to the lower weights assigned to ankle and toe markers in the OpenSim IK setup file, as well as differences in ankle and toe joint center estimation methods...
14. Page 13, Discussion:
    · "We argue that ankle flexion angles and moments calculated from our OpenSim workflow (Fig.2) might be more accurate..." → As suggested earlier, rephrase for caution: The ankle flexion angles and moments calculated by our OpenSim workflow (Fig. 2) appear more physiologically plausible...
    · "This is because ankle flexion angle and moment from reference data remains negative through gait cycle, suggesting that the participant's ankle stays in a plantarflexed position across all conditions." → This is because the ankle flexion angle and moment from the reference data remain negative throughout the gait cycle, suggesting the participant's ankle was maintained in a plantarflexed position across all conditions.
    · "...the study of dataset1 didn't perform matrix transformation..." → ...the study associated with Dataset 1 may not have performed a matrix transformation...
15. Page 13, Discussion (towards end):
    · "There might be due to some kinematic inconsistency..." → This might be due to kinematic inconsistency...
    · "High PCC correlation and low RMSE can still be observed despite the absence of more filtering, supporting the functionality of the workflow (Figures 4 and 5)." → High PCCs and low RMSEs were still observed despite the absence of additional filtering, supporting the functionality of the workflow (Figures 4 and 5).
16. Page 14, Future Work:
    · "Regarding to RRA and CMC steps..." → Regarding the RRA and CMC steps...
    · "...with decreased optimal force set for residual actuators to reduce tracking error and residual actuator forces..." → ...with a decreased optimal force set for residual actuators to reduce tracking errors and residual actuator forces...
17. Throughout: Standardize the formatting of "Dataset 1" and "Dataset 2" (with a space and capital 'D').

 

Thank you again for your well-written manuscript.

 

 

 

Author Response

We thank the reviewers for all these comments. We have provided our reply (highlighted in red) to each comment shown below.

Review 1 Comments:

Major Comments:

  1. Sample Size Justification: The study is limited to one participant per dataset. While the focus is on establishing and validating a workflow, the authors should more explicitly justify why only one subject was chosen from each publicly available dataset and clarify whether the full datasets were explored to confirm this participant was representative. A statement on the generalizability of findings given the sample size should be included in the Limitations section.

Reply:   We agree that we only analyzed data of one participant per dataset. Dataset1 has a total of 50 participants. Dataset2 currently comprises of two participants. As stated in the last part of the manuscript introduction section, this paper focuses on establishing an OpenSim workflow, not on population based clinical analysis. Both participants we have analyzed in this paper are representative of young and able-bodied participants who completed all gait conditions described in each study with good data quality. We have added the relevant context in the section of Materials and Methods. (page 4, lines 144 – 148).

We have also added a statement on the generalizability of findings given the sample size in the limitation part of discussion section. (page 15, lines 377 – 381)

  1. Clarity on Ankle Modeling Discrepancies: The discussion of the lower PCC and higher RMSE for ankle kinematics/kinetics in Dataset 1 is informative but could be more precise. The authors mention different "ankle segment frames" and "joint center estimation methods" between models. It would be helpful to briefly specify the key differences (e.g., anatomical landmark definitions, rotation sequence) and how these inherently lead to offset biases, as the figures suggest.

Reply: We thank the reviewer for this comment regarding ankle modeling discrepancies. We suspect that the offset stems from the difference of shank segment definition between models. For the OpenSim model, the shank segment frame originates at the knee joint center, and the foot segment frame originates at the ankle joint center as shown in Figure 8A. The physical model used in the study of dataset1 follows the International Society of Biomechanics (ISB) definition of joint coordinate system, in which origins of both shank and foot segment frames are coincident with ankle joint center. The foot segment frame from OpenSim model aligns with the foot segment frame of dataset1 physical model. However, X coordinate of shank segment frame in physical model is defined as the line perpendicular to the torsional plane of the tibia/fibula and pointing anteriorly. Torsional planes are defined as the planes contain three anatomical landmarks: the inter-condylar point located midway between medial tibial condyle and lateral tibial condyle (IC), medial malleolus (MM), and lateral malleolus (LM). Its Z coordinate is defined as the line connecting MM and LM, which is pointing to the right. Y coordinate is the common line perpendicular to X and Z axis of foot segment. During quiet stance for calibration, the shank segment coordinates are considered being not aligned with the foot coordinates in the physical model. Especially in sagittal plane, the Y axis of shank points up and backward away from tibia and fabular bones whereas the Y axis of shank in OpenSim model points upward along the longitudinal axis of tibia/fibula. Ankle dorsi/plantar flexion angle is the angle by which Y axis of foot segment is rotated to align with that axis of shank segment. The angle becomes positive in OpenSim model when the rotation of alignment is performed clockwise, which means ankle performs dorsiflexion. However, the angle becomes negative in the physical model as the rotation of alignment was performed counterclockwise, which means ankle performs plantarflexion. This difference can be seen when the joint angle data from quiet stance of physical model was imported to scaled OpenSim model as shown in Figure 8B. (pages 13-14, lines 322-347)

Figure 8: Leg model difference during quiet stance calibration. A) OpenSim frame definition. X, Y and Z coordinates of each segment frame in OpenSim model are represented by thin red, green and blue lines of right leg. The thick red, green and blue lines represent X, Y and Z coordinates of global frames. Ankle angle in the sagittal plane is 12 degrees meaning dorsiflexion B) OpenSim representation of physical leg model. Ankle angle in the sagittal plane is -20 degrees meaning plantarflexion

Shank Segment frame

Foot Segment frame

  1. Force Plate Setup and Moment Accuracy (Dataset 1): The explanation that the two-force-plate setup in Dataset 1 limits the accuracy of joint moment estimation, particularly during double stance and consecutive strides, is a critical point. The authors should consider moving this from the limitations into the main results discussion (Section 4) to provide context before presenting the moment correlation values for Dataset 1. This would help readers interpret the "strong" vs. "excellent" categorization more critically.

Reply: We agree with the reviewer’s comment. We have moved the explanation of two-force-plate setup to the beginning part of the discussion section to provide reader context before discussing the moment correlation values for Dataset1. (page 12 lines 295-309)


  1. Muscle Activation and Forward Dynamics: The authors mention that RRA and CMC steps were attempted but did not yield muscle activations correlating well with EMG, and forward dynamics failed. This is a significant limitation for the stated long-term goal. The manuscript would benefit from a brief summary in the Discussion (even 1-2 sentences) of what specific challenges were encountered (e.g., which muscles, what pattern of discrepancy) and why this is a common hurdle, rather than only describing the general difficulty of configuring RRA/CMC.

Reply: I agree with the reviewer’s comment. We have described specific challenges and our next step to address them in the limitation part of discussion.

Furthermore, muscle activations comparable to reference data have not been produced in our OpenSim workflow. We have developed setup files of both RRA and CMC steps for a sample trial in Dataset 2 to calculate the muscle activation of targeted muscles while walking at 1G level. The muscle activation estimated by our OpenSim workflow currently does not have a high correlation with direct EMG measurements in the reference data, and OpenSim’s forward dynamics thus didn’t reconstruct the similar joint kinematics of the sample trial. Challenges in our RRA configuration include weights selected for tracking a joint coordinate and optimal forces selected for residual and reserve actuators. Challenge in configuring CMC setup files is to balance muscle and reserve actuators. There are 16 degrees of freedom (DOFs), 13 rotational DOFs at pelvis, hip, knee and ankle joints, and 3 translational DOFs at pelvis. We have created two level of weights: high level of weights (500/1000) are used for pelvis tilt and list, ankle dorsi and plantar flexion, low level of weight (20) to all other DOFs. We utilized magnitude of optimal forces and excitation similar to those in OpenSim Gait2354 and Gait2392 RRA and CMC configurations for residual and reserve actuators of our OpenSim model. Minimizing residual errors is a difficult, time-consuming, and iterative process that involves making small adjustments to the model’s trajectory and the torso segment’s center of mass, while appropriate tracking weights are selected for each coordinate [30]. After Pelvis COM and kinematics adjustment of several iterations, average residual forces at pelvis joint in all directions are reduced to a good level (below 5N for force, below 30Nm for torque), except for its residual force in vertical (Y axis) direction (greater than 10N). The current CMC output doesn’t produce muscle activations matching directed EMG measurements reported in dataset 2, although it has average position errors for 3 translational DOFs below 2mm and 13 rotational DOFs below 2 degrees.  To generate matching muscle activations, we will follow best practices and troubleshooting tips in OpenSim documentation by feeding the improved CMC task coordinate weights back to the RRA setup file in iterations while reducing the optimal force set for residual actuators to further reduce tracking errors and residual actuator forces. (page 15, lines 385-411)


  1. Future Work Specificity: The future work section mentions improving the workflow by testing more participants and conducting a new study with a body-weight offload system. Please add more specific goals for this new data collection: What specific hypotheses about reduced-gravity gait or model performance will it test? How will it address the current limitations (e.g., ankle modeling, double stance kinetics)?

Reply: We have added specific activities for our future work and description on how to address the current limitations using the following paragraph.

In addition, we will test more participants in existing datasets, the goal of which is to further validate the angle output from IK step and moment output from ID step by taking ankle modelling difference into account, and to implement the RRA and CMC steps to produce and compare estimation of muscle activations with direct EMG measurement. We also plan to conduct a reduced gravity simulation study with a body-weight offload system ZeroG (Aretech LLC, Sterling, VA) in a newly established gait and balance lab. To capture double stance kinetics, ten force plates will be used in two rows along the walking path and each row includes 5 plates in series in the lab. The goal of this experimental study is to examine the effect of reduced gravities on lower limb mechanics and muscle activation patterns during walking and weighted squat tasks by using our OpenSim workflow. Simulated gravity levels will include Lunar, Martian, micro gravities and as well as gravity values used in dataset 2. We hypothesize that altered muscle activation patterns with respect to amplitude and timing will depend not only on decreased level of reduced gravity but also on specific foot contact strategies participants reply on when performing these tasks. Results of this upcoming study will provide more evidence on identifying muscles with higher rate of atrophy and bones with higher rate of density loss so that customized workout program aboard ISS can be designed by clinicians to strengthen these muscles and bones of astronauts. Pages 15-16 lines 411- 429

 

Minor Comments:
1. Abstract: The final sentence is cut off: "...simulated". Please complete the sentence.

Reply: We have addressed this comment in the final sentence of abstract.


  1. Introduction: The transition from general astronaut biomechanics to the specific gap filled by OpenSim is good. Consider adding a sentence after referencing OpenSim's five-step process to explicitly state that the present study focuses on validating the first three steps (Scaling, IK, ID) as a foundation.

Reply: We thank you for your comment. This sentence, “In this study, we focus on validating the first three steps of the OpenSim workflow (scaling, inverse kinematics, and inverse dynamics) as a foundational phase toward comprehensive joint mechanics estimation.” has been added after the description of OpenSim’s five-step process to state that the present study focuses on validating the first three steps, scaling, inverse kinematics, inverse dynamics, as a foundational step.


  1. Section 2.1 (Datasets): For Dataset 1, it states joint moments were computed using "Newton’s law of motion and principles of moments." Please specify the specific inverse dynamics approach used by the original authors if known (e.g., standard Newton-Euler inverse dynamics).

Reply: The original authors used the standard Newton-Euler inverse dynamics approach, with recorded GRF and COP, calculated joint motion and properties such as joint length and center of mass. In this approach, Newton’s and Euler’s equations were applied recursively for each lower limb segment to solve for moments according to conservation of force and moment, starting at foot and ending at torso. (Page 4, lines 162-167)


  1. Section 2.2 (OpenSim Workflow): The description of marker weighting is clear. For transparency, consider adding the actual weights used for the key markers (e.g., pelvis: 500, ankle: 25) in the main text, as this is a crucial modeling choice affecting results.

Reply: We have added the actual weights used for the anatomical and tracking markers on the pelvis, left and right legs in the main text.

With respect to the lower limb, key anatomical markers on the pelvis, thigh, and shank are assigned heavier weights , as they strongly influence body alignment and gait, and they experience less impact from soft tissue movement artifacts, while tracking markers are assigned lower weights . In detail: anterior Superior Illiac Spines (LxAsis and RxAsis): 10. Mid point of Posterior Superior Illiac Spine (Psis): 10; Lateral Condyle markers (LxLatCon and RxLatCon) : 10; Fibular Head markers (LxFH and RxFH):10; Lateral Malleous Markers (LxLatMal and RxLatMal): 1; Heel markers (LxHeel and RxHeel): 10.; Toe markers (LxToe and RxToe): 10;  5th Metatarsal markers (LxMeta5 and RxMeta5): 1; Thigh markers (LxThigh and RxThigh): 1; Shank markers (LxShank and RxShank): 1. (page 5 lines 208-217)

For example, shoulder marker’s weights from both sides (LxShoulder and RxShoulder) are 5.This weighting system directs the IK solver to prioritize minimizing errors at significant anatomical landmarks. (page 5, lines 220-222)

  1. Figures 2, 3, 4, 5: In the captions, please clarify that the shaded areas represent standard deviation across trials. Ensure all axis labels are fully defined in the captions or on the figures themselves (e.g., "% Gait Cycle," "Angle (deg)," "Moment (Nm/kg)").

Reply: We have updated captions of Figures 2,3, 4 and 5 to include the definition of axis labels and the definition of shared areas in figures as standard deviations across trials. For Figures 2 and 4, X axis represents the progress of gait cycle of right leg as % Gait Cyle, and Y axis represents the joint angle in degree. For Figures 3 and 5, X axis represents the progress of gait cycle of right leg as % Gait Cyle, and Y axis represents the normalized joint moment in Nm/kg.


  1. Tables 1 & 3 (PCC): The PCC value for "Ankle dorsi/plantar flexion moment" during heel walking in Table 1 is -0.223. In the text (Page 6, Results), the average is given as 0.677 ± 0.600. The negative value for one condition drastically affects this average and standard deviation. The authors should comment briefly in the text on the implication of this negative correlation for that specific task.

Reply: We agree with the reviewer’s comment. Ankle dorsi/plantar flexion moment PCC in heel walking condition skews drastically the averaged PCCs from excellent level ( greater than 0.96 in other three conditions) to moderate level across all conditions. (Page 15, Lines 355-357)


  1. Page 13, Discussion: "We argue that ankle flexion angles and moments calculated from our OpenSim workflow... might be more accurate..." This is a strong claim. Rephrase to be more cautious, e.g., "The pattern of ankle angles/moments estimated by our OpenSim workflow appears more physiologically plausible for heel walking, based on evidence from other gait studies..."

Reply: We agree with the reviewer’s concern. The statement has been revised to a more cautious, emphasizing physiological plausibility rather than greater accuracy. (Page 14, lines 357 – 360)


  1. References: All references appear relevant and appropriately cited. Please ensure the formatting is consistent with the journal's style guide (e.g., use of "et al.", journal abbreviations).

Reply: I have checked the formatting of context in manuscript is consistent with the journal’s style.

 

 Thank you again for your valuable work.

 

Comments on the Quality of English Language


Dear Authors
Your work is generally well-written in clear and professional English. The technical descriptions are precise, and the overall structure supports readability. However, there are a number of minor grammatical errors, typos, and awkward phrasings that should be corrected to meet formal publication standards. Below is a list of suggested corrections and improvements.

Page 1, Abstract: 
   · "Correspondance" → Correspondence 
   · "We speculate that lower PCC values and higher RMSE present with ankle dorsi/plantar flexion angle..." → We speculate that the lower PCC values and higher RMSE observed for ankle dorsi/plantar flexion angle... 
   · The final sentence is cut off: "...in simulated" → ...in simulated reduced-gravity environments. 

Reply: We thank the reviewer for these suggestions. The spelling of “correspondence” has been corrected. The sentence has been revised to read "We speculate that lower PCC values and higher RMSE present with ankle dorsi/plantar flexion angle...". The abstract sentence has been completed to read “...in simulated reduced-gravity environments.”  


  1. Page 1, Keywords:
       · "Keywords:musculoskeletal modelling," → Keywords: musculoskeletal modelling, (add space after colon) 

Reply: We thank the reviewer for this comment. A space has been added after the colon so that the keywords now read “Keywords: musculoskeletal modelling,”. 


  1. Page 2, Introduction: 
       · "DeWitt et. al." → DeWitt et al. (standard format for "et al.") 
       · "McCrory et. al." → McCrory et al. 
       · "...during in-flight 0G treadmill running exercise, in which the mechanical load to joints can be increased..." → ...during in-flight 0G treadmill running exercise, where mechanical load on the joints can be increased... 
       · "As decades long effort," → As part of a decades-long effort, 

Reply: We thank the reviewer for these suggestions. The references have been updated to the standard format for et al. The sentences "...during in-flight 0G treadmill running exercise, in which the mechanical load to joints can be increased..." and “As decades long effort,” have been revised as recommended. 


  1. Page 3, Introduction:
       · "OpenSim overcomes these limitations by combining inverse dynamics with optimization methods to distribute joint torques across individual muscles based on their anatomical properties." → This sentence is clear but slightly long. It is acceptable as is. 

Reply: We thank the reviewer for this comment. The sentence has been retained without modification. 


  1. Page 4, Materials and Methods:
       · "LAMB prototocol" → LAMB protocol 
       · "...and the LAMB prototocol based full-body marker set..." → ...and the LAMB protocol-based full-body marker set... 
       · "Joint moments were therefore computed on one leg..." → Joint moments were therefore computed for one leg... 

Reply: We thank the reviewer for this comment. “LAMB protocol” has been corrected in the revised English-edited version of the manuscript. The sentence "Joint moments were therefore computed on one leg..." has been updated to the recommended edit. 


  1. Page 4, Section 2.2:
       · "The OpenSim workflow begins with the scaling step, where we customized a generic OpenSim model..." → The OpenSim workflow begins with a scaling step, in which we customize a generic OpenSim model... (Use present tense for describing methodology). 
       · "Markers like the pelvis and knee are assigned with heavier weights" → Markers such as those on the pelvis and knee are assigned heavier weights 
       · "...markers are placed further away from their corresponding anatomical landmarks in foot segment of the OpenSim model for dataset2, compared to those for dataset1, as the dataset2 participant walked with shoes on whereas the dataset1 participant walked with bare foot." → This is awkward. Suggest: For Dataset 2, markers were placed further from their corresponding anatomical landmarks on the foot segment of the OpenSim model compared to Dataset 1, as the participant in Dataset 2 wore shoes, while the participant in Dataset 1 was barefoot. 

Reply: We thank the reviewer for this comment. The sentence “The OpenSim workflow begins with the scaling step, where we customized a generic OpenSim model..." has been revised to use present tense, and wording related to marker weighting and foot marker placement has been clarified as suggested. All changes have been incorporated into the English-edited version of the manuscript. 


  1. Page 5, Figure 1 Caption:
       · "Pink markers represent the anatomical landmarks. The positive X, Y and Z of global coordinates are showed as the lines of red, green and blue color respectively..." → Pink markers represent anatomical landmarks. The positive X, Y, and Z axes of the global coordinate system are shown as red, green, and blue lines, respectively... 
       · "a) the scale model" → a) the scaled model 
       · "b) the scaled model using lower limb marker set for dataset 2." → b) the scaled model using the lower-limb marker set for Dataset 2. 

Reply: We thank the reviewer for this comment. The figure 1 caption has been revised to improve grammar and clarity as suggested.  


  1. Page 5, Section 2.3:
       · "Joint moment data from OpenSim ID output was normalized to participant mass and was also time-normalized..." → Joint moment data from the OpenSim ID output was normalized to participant body mass and time-normalized... 
       · "Custom MATLAB scripts were developed to compare normalized joint angles and moments between the reference data and OpenSim data across different walking conditions." → This is clear. 

Reply: We thank the reviewer for this comment. The sentence has been revised to read “Joint moment data from the OpenSim ID output was normalized to participant body mass and time-normalized…”. The sentence "Custom MATLAB scripts…" has been retained without modification. 


  1. Page 6, Section 2.3:
       · "It is categorized into four levels of similarity: excellent when PCC is greater than 0.9, strong similarity when it is between 0.7 and 0.9, moderate if it is between 0.4 and 0.7, low if it is less than 0.4 [18], [19]." → Similarity is categorized into four levels: excellent (PCC > 0.9), strong (0.7 ≤ PCC ≤ 0.9), moderate (0.4 ≤ PCC < 0.7), and low (PCC < 0.4) [18,19]. 

Reply: We thank you for this comment. The sentence has been revised as suggested to clarify the categorization of similarity levels based on Pearson’s correlation coefficient. 


  1. Page 8, Table 1 Title:
        · "Correlation coefficients (PCC) between OpenSim results and dataset1 values across four walking conditions." → Correlation coefficients (PCC) between OpenSim results and Dataset 1 reference values across four walking conditions. 

Reply: We thank the reviewer for this comment. The table title has been revised to include the term “reference”, and the formatting for “Dataset 1” has been standardized throughout the English-edited version of the manuscript.  


  1. Page 9, Text above Table 2:
        · "...for the hip and knee flexion/extension angles, as well as ankle dorsiflexion/plantarflexion angles respectively..." → ...for hip flexion/extension, knee flexion/extension, and ankle dorsiflexion/plantarflexion angles, respectively... (Apply same correction to the moment list). 

Reply: Thank you for your comment. The text has been revised to clarify the listing of hip, knee, and ankle variables, and the same correction has been applied to the joint moments in the English-edited version. 

 

  1. Page 12, Table 4:
        · "Knee flexion/extension moment (Nm/kg)" → The values in this row (e.g., 0.113, 0.094) seem inconsistent with the high PCCs (~0.97). Please verify the units and values. The note in the Mean column says (0.031) but the text on Page 12 says 0.078 ± 0.0131. Ensure consistency.

Reply:  Thanks for this comment. We have corrected the PCCs values of not only knee moment, but also of hip and ankle moments, and of ankle angles in the original Table 3, and reflect them in the Boxplot 3. The mean and standard deviation of knee flexion/extension moment are  0.078 and 0.031 Nm/kg respectively. We have updated the standard deviation from 0.0131 Nm/kg to 0.031 Nm/kg.


  1. Page 12, Discussion:
        · "Most of our simulated results show either excellent or strong correlation with reference data for hip, knee and ankle mechanics across various conditions, except for ankle moment during heel walking condition in dataset1." → Most of our simulated results show excellent or strong correlation with reference data for hip, knee, and ankle mechanics across various conditions, except for the ankle moment during the heel-walking condition in Dataset 1. 
        · "Smaller RMSE values (less than 5 deg) are present with hip and knee flexion/extension angles between two approaches for both datasets." → Smaller RMSE values (< 5 deg) were found for hip and knee flexion/extension angles between the two approaches for both datasets. 
        · "This discrepancy might be due to lower weight we placed for ankle and toe markers in OpenSim IK setup file and as well as difference in ankle and toe joint center estimation methods..." → This discrepancy might be due to the lower weights assigned to ankle and toe markers in the OpenSim IK setup file, as well as differences in ankle and toe joint center estimation methods... 

Reply: Thank you for these comments. The manuscript has been revised for grammar, clarity, and consistency. We refined the correlation descriptions, RMSE reporting, and expanded the discussion on potential sources of discrepancy. Formatting for Dataset 1 has been standardized throughout. 


  1. Page 13, Discussion:
        · "We argue that ankle flexion angles and moments calculated from our OpenSim workflow (Fig.2) might be more accurate..." → As suggested earlier, rephrase for caution: The ankle flexion angles and moments calculated by our OpenSim workflow (Fig. 2) appear more physiologically plausible... 
        · "This is because ankle flexion angle and moment from reference data remains negative through gait cycle, suggesting that the participant's ankle stays in a plantarflexed position across all conditions." → This is because the ankle flexion angle and moment from the reference data remain negative throughout the gait cycle, suggesting the participant's ankle was maintained in a plantarflexed position across all conditions. 
        · "...the study of dataset1 didn't perform matrix transformation..." → ...the study associated with Dataset 1 may not have performed a matrix transformation... 

Reply: We thank the reviewer for these comments. The manuscript has been revised to adopt a more cautious and objective tone, focusing on the physiological plausibility of the results. Additionally, we have addressed grammar errors and speculative statements have been qualified for a more balanced discussion. 


  1. Page 13, Discussion (towards end):
        · "There might be due to some kinematic inconsistency..." → This might be due to kinematic inconsistency... 
        · "High PCC correlation and low RMSE can still be observed despite the absence of more filtering, supporting the functionality of the workflow (Figures 4 and 5)." → High PCCs and low RMSEs were still observed despite the absence of additional filtering, supporting the functionality of the workflow (Figures 4 and 5). 

Reply: We thank the reviewer for these comments. The sentences have been revised to correct grammar errors and clarify the discussion regarding kinematic inconsistencies and better reflect the interpretation of the PCC and RMSE results.  


  1. Page 14, Future Work:
        · "Regarding to RRA and CMC steps..." → Regarding the RRA and CMC steps... 
        · "...with decreased optimal force set for residual actuators to reduce tracking error and residual actuator forces..." → ...with a decreased optimal force set for residual actuators to reduce tracking errors and residual actuator forces... 

Reply: Thank you for your comment. The phrasing has been revised for grammatical accuracy, and we have included the sentence “…while reducing the optimal force set for residual actuators…” to clarify the RRA and CMC implementation. 


  1. Throughout: Standardize the formatting of "Dataset 1" and "Dataset 2" (with a space and capital 'D').

Reply: We thank the reviewer for this comment. The formatting of “Dataset 1” and “Dataset 2” has been standardized throughout the manuscript. 

 

 

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Brief summary: This study aimed to develop an OpenSim workflow using publicly available gait data and to evaluate its performance by comparing joint angles and moments in terms of PCC and RMSE values.

Major comments:

The topic of this work is very interesting, and the paper is well written. However, substantial changes are required to improve the content, clarity of presentation, and overall comprehensibility of the study.

  1. The abstract is too long and overly detailed. It should provide a clear overview of the paper highlighting the research gap, aims, methods, and main results, without focusing on excessive specific details. I recommend shortening and refining it to improve clarity and readability.
  2. In the paragraph corresponding to lines 95–110, the authors describe the limitations of traditional motion capture methods and state that OpenSim can address this gap; however, no reference is made to previous studies that have used OpenSim for this purpose. Without a comparison with existing literature, it is unclear what specific novelty the authors’ contribution provides. It would therefore be appropriate to add one or two references to works that employ OpenSim to bridge this gap, even if not specifically focused on the topic of the present paper.
  3. The Results section should only present the findings, without any interpretation, which should instead be moved to the Discussion section. The table of PCC values is not very readable. I suggest creating boxplots showing the mean and standard deviation, along with horizontal lines indicating the four thresholds of similarity mentioned in the text. The same comment also applies to RMSE values and the thresholds reported in the literature.
  4. The Discussions section is too descriptive and mixes results with technical details, which makes it difficult to follow. It would benefit from a clearer structure that separates the main findings, their interpretation, and the study novelty. Many technical details on marker setup and RRA/CMC configuration could be moved to the Materials and Methods section, while the Discussion should focus on the impact of these choices on the results. Additionally, the section would be stronger if the authors explicitly highlighted what is novel in their OpenSim workflow compared to previous studies, summarized limitations in concise points, and clearly outlined the implications and future perspectives. Finally, adding a short Conclusion section would help to clearly summarize the main findings of the study.

Minor comments:

Please, follow the template considering margins and interline. Lines 57-74 and Lines 134-152. The text within the paragraph is not justified (i.e., the margins are not aligned on both sides).

Author Response

We thank the reviewers for all these comments. We have provided our reply (highlighted in red) to each comment shown below.

Reviewer 2 Comments

Major comments:

The topic of this work is very interesting, and the paper is well written. However, substantial changes are required to improve the content, clarity of presentation, and overall comprehensibility of the study.

  1. The abstract is too long and overly detailed. It should provide a clear overview of the paper highlighting the research gap, aims, methods, and main results, without focusing on excessive specific details. I recommend shortening and refining it to improve clarity and readability.

Reply:   We agree with the reviewer’s comment. The abstract has been shorted to improve clarity and readability, by only highlighting the most relevant aspects of the work. Page 1 (lines 11-39)

  1. In the paragraph corresponding to lines 95–110, the authors describe the limitations of traditional motion capture methods and state that OpenSim can address this gap; however, no reference is made to previous studies that have used OpenSim for this purpose. Without a comparison with existing literature, it is unclear what specific novelty the authors’ contribution provides. It would therefore be appropriate to add one or two references to works that employ OpenSim to bridge this gap, even if not specifically focused on the topic of the present paper.

Reply:  We thank the reviewer for pointing out this concern. A paragraph has been added to describe the research works that employ OpenSim to address the limitations of conventional motion capture paradigm, as well as to summarize the gaps in current OpenSim validation studies in the areas of reduced gravity. Pages 2-3 (lines 88-97 and lines 100 - 116)

  1. The Results section should only present the findings, without any interpretation, which should instead be moved to the Discussion section. The table of PCC values is not very readable. I suggest creating boxplots showing the mean and standard deviation, along with horizontal lines indicating the four thresholds of similarity mentioned in the text. The same comment also applies to RMSE values and the thresholds reported in the literature.

Reply:    We agree with reviewer comments. The results section has been updated to present the findings. Moreover, we have created boxplots to show PCC and RMSE mean and standard deviation values of each dataset (Figure 4 and Figure 7), along with horizontal lines indicating the four thresholds of similarity for PCC and agreeable levels for RMSE. 

  1. The Discussions section is too descriptive and mixes results with technical details, which makes it difficult to follow. It would benefit from a clearer structure that separates the main findings, their interpretation, and the study novelty. Many technical details on marker setup and RRA/CMC configuration could be moved to the Materials and Methods section, while the Discussion should focus on the impact of these choices on the results. Additionally, the section would be stronger if the authors explicitly highlighted what is novel in their OpenSim workflow compared to previous studies, summarized limitations in concise points, and clearly outlined the implications and future perspectives. Finally, adding a short Conclusion section would help to clearly summarize the main findings of the study.

Reply: We agree with reviewer’s suggestions. First, a brief paragraph was created to describe the section of discussion which is structured to include main findings, results interpretation, study novelty and limitations, and future research plan. Page 12  (lines 292-293)

Secondly, technical details on marker setup have been moved to the subsection of OpenSim Workflow in Materials and Methods section. Page 5 (lines 193-200)

General description of RRA and CMC configurations has been shorted. We have added description of specific challenges in tailoring these configurations to produce muscle activations for this study. Page 15 (lines 386-413)

Thirdly, we have described novelty in our OpenSim workflow, and have revised limitations (page 15, lines 377-386) as well as outlined the implications and future perspectives. Pages 15-16 (lines 414-431)

Lastly, we have added a conclusion section to summarize the main findings of the study. Page 16 (lines 433-446)

 

Minor comments:

Please, follow the template considering margins and interline. Lines 57-74 and Lines 134-152. The text within the paragraph is not justified (i.e., the margins are not aligned on both sides).

Reply:    We have addressed this comment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

ABSTRACT

Comment 1. The abstract is very informative but particularly long. Shorten it by trying to highlight the most relevant aspects of the work carried out. 

INTRODUCTION

Comment 2. The introduction provides a broad overview of biomechanics for bodyweight offload applications, but a concise paragraph could be added summarizing the gaps in current OpenSim validation studies in the areas of reduced gravity. This could emphasize the relevance of the study.

Comment 3. The description of OpenSim's five-step workflow is clear but too specific for the Introduction. I recommend moving the technical details to the Materials and Methods section.

MATERIALS AND METHODS

Comment 4. In section 2.1, why only one able-bodied participant was selected for each dataset? Please, justify appropriately this choice. For example, clarify whether this choice was dictated by the completeness of the data or by computational constraints. Then, discuss the implications in terms of generalizability as appropriate.

Comment 5. In section 2.2, the authors state (lines 170-172): “Markers like the pelvis and knee are assigned with heavier weights to prioritize matching these points more closely, while other markers are given lower weights.” At lines 180-189 more details are provided about the weights selection. But was a sensitivity analysis performed? The choice of the weights should be more clearly described.

Comment 6. In order to improve the reproducibility and clarity of the marker sets, add a representation of the retroreflective markers to Figure 1, along with a representation of the correspondence between the experimental marker sets and the OpenSim model markers.

RESULTS

Comment 7. The results are clearly presented, have a good structure, and are in line with the aims of the study.

DISCUSSION

Comment 8.

In the Discussion section, the interpretation and argumentation of the results are comprehensive, and appropriately place the main findings in the context of the existing literature.

Author Response

We thank the reviewers for all these comments. We have provided our reply (highlighted in red) to each comment shown below.

Reviewer 3 Comments:

ABSTRACT

Comment 1. The abstract is very informative but particularly long. Shorten it by trying to highlight the most relevant aspects of the work carried out. 

Reply:       We agree with the reviewer’s comment. The abstract has been shorted by only highlighting the most relevant aspects of the work. Page 1 (lines 11-39)

 

INTRODUCTION

Comment 2. The introduction provides a broad overview of biomechanics for bodyweight offload applications, but a concise paragraph could be added summarizing the gaps in current OpenSim validation studies in the areas of reduced gravity. This could emphasize the relevance of the study.

Reply:       A concise paragraph has been added to summarize the gaps in current OpenSim validation studies in the areas of reduced gravity, along with reliability evidence provided by OpenSim validation studies conducted in sports and rehabilitation medicine. Pages 2-3 (lines 88-97 and lines 100 - 116)

 

Comment 3. The description of OpenSim's five-step workflow is clear but too specific for the Introduction. I recommend moving the technical details to the Materials and Methods section.

Reply:      We agree with the comment. The description of five-step workflow has been shortened and moved to Subsection 2.2 OpenSim Workflow under Section 2 Materials and Methods.  Page 4 (lines 180-182)

 

MATERIALS AND METHODS

Comment 4. In section 2.1, why was only one able-bodied participant selected for each dataset? Please, justify appropriately this choice. For example, clarify whether this choice was dictated by the completeness of the data or by computational constraints. Then, discuss the implications in terms of generalizability as appropriate.

Reply:   We thank the reviewer for this comment. We agree that we only analyzed data of one participant per dataset. Dataset1 has a total of 50 participants. Dataset2 currently comprises of two participants. As stated in the last part of the manuscript introduction section, this paper focuses on establishing an OpenSim workflow, not on population based clinical analysis. Both participants we have analyzed in this paper are representative of young and able-bodied participants who completed all gait conditions described in each study with good data quality. We have added the relevant context in the section of Materials and Methods. (page 4, lines 144 – 148).

We have also added a statement on the generalizability of findings given the sample size in the limitation part of discussion section. (page 15, lines 377 – 381)

 

Comment 5. In section 2.2, the authors state (lines 170-172): “Markers like the pelvis and knee are assigned with heavier weights to prioritize matching these points more closely, while other markers are given lower weights.” At lines 180-189 more details are provided about the weights selection. But was a sensitivity analysis performed? The choice of the weights should be more clearly described.

Reply: We thank the reviewer for this comment. The goal of weight selection is to reconstruct the core kinematic components of gait which are pelvis, hip, knee and ankle joints. We didn’t conduct a systematic analysis on how different set of weights have on marker’s root mean square of error. However, the choice of the weights on markers is based on heuristic analysis we conducted on assigning marker weights to different markers, in which we have tried different weights for key markers and utilized weights assigned to markers in other OpenSim gait models. The actual weights used for the anatomical and tracking markers on the pelvis, left and right legs have been in the main text.

With respect to the lower limb, key anatomical markers on the pelvis, thigh, and shank are assigned heavier weights , as they strongly influence body alignment and gait, and they experience less impact from soft tissue movement artifacts, while tracking markers are assigned lower weights . In detail: anterior Superior Illiac Spines (LxAsis and RxAsis): 10. Mid point of Posterior Superior Illiac Spine (Psis): 10; Lateral Condyle markers (LxLatCon and RxLatCon) : 10; Fibular Head markers (LxFH and RxFH):10; Lateral Malleous Markers (LxLatMal and RxLatMal): 1; Heel markers (LxHeel and RxHeel): 10.; Toe markers (LxToe and RxToe): 10;  5th Metatarsal markers (LxMeta5 and RxMeta5): 1; Thigh markers (LxThigh and RxThigh): 1; Shank markers (LxShank and RxShank): 1. (page 5 lines 208-217)

For example, shoulder marker’s weights from both sides (LxShoulder and RxShoulder) are 5.This weighting system directs the IK solver to prioritize minimizing errors at significant anatomical landmarks. (page 5, lines 220-222)

 

Comment 6. In order to improve the reproducibility and clarity of the marker sets, add a representation of the retroreflective markers to Figure 1, along with a representation of the correspondence between the experimental marker sets and the OpenSim model markers.

Reply:     We agree with the reviewer’s comment. Representation of the retroreflective markers have been added to Figure 1, along with corresponding OpenSim model markers to give an example of our OpenSim model performance in IK analysis for both datasets. The caption has been updated accordingly as well. Page 6 (line 230)

 

RESULTS

Comment 7. The results are clearly presented, have a good structure, and are in line with the aims of the study.

Reply:    We thank this comment from the reviewer.

 

DISCUSSION

Comment 8.

In the Discussion section, the interpretation and argumentation of the results are comprehensive, and appropriately place the main findings in the context of the existing literature.

Reply:       We thank this comment from the reviewer.

 

 

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Authors have made all the requested changes, the paper can now be accepted.