PONE-D-25-14929

Trustworthy AI for medical decisions: Adversarial robust and fair machine learning prediction for Parkinson’s disease

PLOS ONE

Dear Dr. Rahim,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

ACADEMIC EDITOR:

Required:

• The abstract should more clearly state the study’s objectives, dataset characteristics, and the rationale for including adversarial attack analysis.
• In Section 3.2.2, several claims—particularly regarding model robustness, federated learning, and explainability—require proper justification and supporting references.
• The discussion on bias related to gender, age, and race should be more clearly substantiated with examples and explanations.
• The mention of "research focused on algorithmic fairness" in Section 3.3 lacks citation, and Section 3.4.2 appears misaligned with its intended focus on group fairness.
• A clearer distinction between sections 4.2 and 4.4 is needed as they currently overlap in content.
• Furthermore, the limitations of the experimental setup should be explicitly addressed.
• Key related works and comparisons, particularly those using similar datasets and standard models like SVMs or neural networks, should be included.
• Feature selection choices—especially the exclusion of age and race—should be justified.
• Metric inconsistencies should be addressed, with valid ranges and statistical significance testing included.
• Imbalances in class distribution should be noted and mitigated appropriately.
• Data availability should align with reproducibility standards, and the ethical treatment of sensitive attributes—such as discretization of age or race—requires clear explanation.
• It is important to explicitly note limitations, such as the exclusion of deep learning models, and discuss strategies to manage accuracy-fairness trade-offs in future work.
• Language should be polished for grammatical accuracy and reduced complexity.
• Formatting and consistency in references, notation, and tables should also be reviewed

Recommendations:

• The paper would benefit from increased originality and refinement in both content and presentation.
• The authors are encouraged to compare their findings with current state-of-the-art methodologies to better contextualize their contribution.
• The introduction contains an extensive list of contributions, which could be narrowed to emphasize the main innovations.
• The concept of fairness is frequently mentioned but not always supported by concrete evidence.
• A comparative table summarizing models, datasets, pros, cons, and performance would be a valuable addition.
• Certain sections could be restructured for clarity.
• Also, the methodology and results sections would benefit from added detail on experimental setup, parameters, and statistical significance.
• A dedicated discussion section analyzing the results would strengthen the manuscript.
• Methodological transparency can be improved by detailing fairness mitigation algorithms, feature selection techniques, and adversarial attack parameters.

==============================

Please submit your revised manuscript by Jun 05 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Rizik M. H. Al-Sayyed, Ph.D.

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please note that PLOS ONE has specific guidelines on code sharing for submissions in which author-generated code underpins the findings in the manuscript. In these cases, we expect all author-generated code to be made available without restrictions upon publication of the work. Please review our guidelines at https://journals.plos.org/plosone/s/materials-and-software-sharing#loc-sharing-code and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse.

3. Thank you for stating the following in the Acknowledgments Section of your manuscript:

“This work was supported by the National Research Foundation of Korea (NRF) grant 716

funded by the Korea government (MSIT) (No. RS-2023-00221186) and also supported 717

by the National Research Foundation of Korea (NRF) grant funded by the Korea 718

government (MSIT) (No. RS-2022-00166733).”

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

“The author(s) received no specific funding for this work.”

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

4. When completing the data availability statement of the submission form, you indicated that you will make your data available on acceptance. We strongly recommend all authors decide on a data sharing plan before acceptance, as the process can be lengthy and hold up publication timelines. Please note that, though access restrictions are acceptable now, your entire data will need to be made freely accessible if your manuscript is accepted for publication. This policy applies to all data except where public deposition would breach compliance with the protocol approved by your research ethics board. If you are unable to adhere to our open data policy, please kindly revise your statement to explain your reasoning and we will seek the editor's input on an exemption. Please be assured that, once you have provided your new statement, the assessment of your exemption will not hold up the peer review process.

Additional Editor Comments:

Thank you for your submission. While your work raising one of the important issues in healthcare AI, especially pertaining to fairness and adversarial robustness, requires substantial work. The reviewers determined that the manuscript fails to be coherent, methodologically sophisticated, and scientifically thorough. Primary issues include lack of reasonable justification for certain arguments (especially in Sections 3.2.2, 3.3), not addressing contemporary relevant literature, insufficient detailing of bias, redundancy in data pre-processing explanations, and no benchmarking against contemporary literature. Add the problem, dataset, and technologies used in the abstract. Furthermore, inconsistencies such as having FPR greater than one, vague defining of terms such as mitigation of fairness, adversarial attacks, or mistreatment as well as fuzzy explanations should be fixed. Revise the document so that all the statements made are supported by relevant academic literature. Make sure claims are backed by references, redefine the results and discussions as well as limit the conclusions to frameworks set out in the objectives. More focusing on aspect of providing rationale for why concepts such as mitigation requires more elaboration. The manuscript requires, at minimum, this level of overhaul in conjunction with enhancing the scientific as well as editorial quality of the manuscript.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: 1. Your work lacks sufficient originality and does not meet the minimum quality standards of PLOS ONE Journal.

2. You need to enhance the English writing in all paper sections. Several statements and paragraphs are not understood.

3. Rewrite the abstract section to specify the problem directly and describe your solution in brief.

4. Compare your results with state-of-the-art approaches.

5. In Adversarial Robustness section 3.2.2, you draw conclusions without justifications. (There is a lot of bias in the information when it comes to gender, age, and race, which shows how these traits are uneven across classes. Compared to gender features, age and race features are more skewed). Explain how bias exists?

6. In section 3.2.2 there are conclusions without justification, such as:

• Adversarial training is one of the strong training methods that boost model generalization and manipulation resistance.

• Resilience added to federated learning systems increases data privacy and model efficiency.

• Explainable artificial intelligence (XAI) may highlight the evolution and upkeep of robustness.

• Furthermore, standardizing model robustness assessment will be a new criterion for healthcare artificial intelligence.

7. In section 3.3, you wrote (In research focused on algorithmic fairness, which involves the removal of biases as illustrated in Figure 1) which research do you mean? No reference is given.

8. The content of section 3.4.2 Group fairness does not relate to Group fairness, it relates to the model’s assessment.

9. Section 4.2 Data prepossessing and section 4.4 Data preparation have the same contents. What are the differences?

10. Experiments limitations should be discussed.

Reviewer #2: This study presented a CAD for Parkinson’s disease that studied the impact on health-related quality of life and revealed disparities based on gender, age, and race. The authors applied Random Forest (RF) and Decision Tree (DT) on the Parkinson’s disease dataset.

There are some comments that need to be considered in the article:

The abstract lacks clarity regarding the objective of using machine learning techniques. Additionally, a description of the dataset is also not given in the abstract. Adversarial attacks are not clear, and why to add it to the data.

In the introduction,

The number of primary contributions mentioned is huge; it is better to be reduced to your main contribution only.

The authors utilize the word fairness many times in many sections without proving if they fairly use it. For example, in the literature review they make, however, there are many articles not included in the study.

in subsection 2.1, "such as those covered here," unclear” pronoun those”

In Related Works, it is better to add the most related models that have already been applied to the same dataset of Parkinson. Did any authors utilize other standard ML techniques such as SVM or neural networks?

It is better to add a table with the ML techniques, dataset, pros and cons of each study, and performance.

Why add this "section 3.3 Mitigate ML bias"?

In methodology, just mention the main methods that you are utilizing in your study or that you wish to compare with only.

3.2.1 Model robustness: In lines 191–192, it is better to add a reference for your claim.

Add reference for lines 378 and 379: XGBoost and SULOV. Which one did you utilize in your study?

In data splitting, there is no reason to make 5-fold cross-validation after the initial separation of the data.

In the data collection, the number of features is around 15. Could you please explain the reasoning behind using feature selection?

And if you used FS, why did you exclude age and race? You can let FS method selects the most valuable features.

In the results section,

First, it is better to add in the experiment setup.

The parameters for the decision tree and random forest

Comparing your results with the most recent studies already discussed in your related works is preferable.

and it is better to compute the significance of your obtained results.

In addition, it is better to add a discussion section, which discusses the finding behind the obtained results.

Reviewer #3: This manuscript significantly contributes to the discourse on trustworthy AI in healthcare, and by addressing the following outlined points, the authors can enhance the rigor, clarity, and impact of their work. The study’s focus on fairness and adversarial robustness is commendable, and the suggested revisions will help it meet the high standards expected by PLOS ONE. Therefore, it is recommended for acceptance with minor revisions and a medium confidence level.

Minor Revisions

1. Statistical Analysis and Metric Validity:

o Address Metric Anomalies: Several tables report implausible values (e.g., FPR >1). Revise calculations and ensure all metrics (FPR, FNR) are within valid ranges (0–1).

o Statistical Significance Testing: To confirm significance, include p-values or confidence intervals for key results (e.g., accuracy drops post-mitigation/adversarial attacks).

o Class Imbalance: Acknowledge and mitigate the impact of the PD: HC imbalance (≈3:2) using techniques like SMOTE, stratified sampling, or synthetic data generation.

2. Methodological Clarity:

o Fairness Mitigation Pipeline: Provide explicit details on the "optimized preprocessing fairness mitigation" algorithm (e.g., reweighting, adversarial debiasing). Include hyperparameters and code references if available.

o Feature Selection: Clarify the implementation of SULOV and recursive XGBoost (e.g., library versions, selection criteria).

o Adversarial Attacks: Specify perturbation magnitudes (e.g., noise levels in poison attacks) and success rates (e.g., % adversarial examples generated).

3. Data Availability:

o Remove Ambiguity: Delete the redundant "available on request" statement unless legally/ethically justified.

o Share Processed Data: Deposit preprocessed/feature-engineered datasets in a public repository (e.g., Zenodo) to ensure reproducibility.

4. Ethical and Bias Considerations:

o Sensitive Attribute Binarization: Justify thresholds for discretizing age, race, and gender. For example, clarify how age groups were defined (e.g., <60 vs. ≥60).

o Causal Mechanisms: Discuss how adversarial attacks amplify bias (e.g., label leaks correlating with SPD decline) rather than solely reporting correlations.

5. Language and Grammar:

o Correct grammatical errors (e.g., "data was" → "data were," "prepossessing" → "preprocessing").

o Simplify jargon-heavy sections (e.g., "quantum adversarial machine learning") with brief definitions or examples.

o Reduce redundancy (e.g., repetitive mentions of "fairness-accuracy trade-offs" in the Discussion).

6. Formatting Consistency:

o Ensure uniform citation style (e.g., "[13]" vs "as shown in [13]").

o Standardize mathematical notation (e.g., italics for variables: FPR vs. FPR).

7. Result Presentation:

o Tables 3–6: Reformat to avoid split tables across pages and ensure alignment.

o Figures: Improve resolution and label clarity (e.g., axis labels in Figures 5–8).

8. Discussion Limitations:

o Explicitly state the exclusion of deep learning models as a limitation and its impact on generalizability.

o Discuss potential solutions to accuracy-fairness trade-offs (e.g., hybrid models, post-hoc calibration).

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Dheeb Albashish

Reviewer #3: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PONE-D-25-14929R1

Trustworthy AI for medical decisions: Adversarial robust and fair machine learning prediction for Parkinson’s disease

PLOS ONE

Dear Dr. Rahim,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Sep 19 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Rizik M. H. Al-Sayyed, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

1. If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise.

2. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Additional Editor Comments:

The manuscript still needs some minor comment. Please address the comments made by reviewer 2 carefully and res-submit.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2:

The authors had addressed the majority of the comments; however, some points still require attention in the article."

Abstract: The methodology remains unclear. The authors are encouraged to clarify the approach used in the study.

Introduction:

The introduction includes a list of six primary contributions, which is quite extensive. It is recommended that the authors narrow it down to the single most significant and novel contribution of the study.

Table 7: This table should be relocated to the literature review section. Additionally, the authors should analyze and discuss the reviewed papers listed in Table 7 within the context of the literature review.

Results Section:

It would be beneficial to add a discussion section to interpret and contextualize the findings.

The authors are also advised to perform statistical tests to assess the significance of the results, especially in comparison with the most recent studies listed in Table 7.

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Partly

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: N/A

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: (No Response)

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: The authors had addressed the majority of the comments; however, some points still require attention in the article."

Abstract: The methodology remains unclear. The authors are encouraged to clarify the approach used in the study.

Introduction:

The introduction includes a list of six primary contributions, which is quite extensive. It is recommended that the authors narrow it down to the single most significant and novel contribution of the study.

Table 7: This table should be relocated to the literature review section. Additionally, the authors should analyze and discuss the reviewed papers listed in Table 7 within the context of the literature review.

Results Section:

It would be beneficial to add a discussion section to interpret and contextualize the findings.

The authors are also advised to perform statistical tests to assess the significance of the results, especially in comparison with the most recent studies listed in Table 7.

Reviewer #3: The revised version is technically sound and the data support the conclusions. All concerns in the previous review have been addressed in the revised version. Explanation of the methods is greatly improved, justifications of claims are better supported and statistical rigor and clarity of analyses are improved. The models used are standard and appropriate (Decision Tree, Random Forest) and the authors use standard fairness mitigation techniques. The study is validated with various performance metrics and the comparisons with state-of-the-art models bring robustness to the work. The limitations mentioned by the authors, such as not including deep learning models and a simplified view on sensitive attributes are not crucial for the main findings. All the data underlying the results are included in the public GitHub repository. The revised version is submitted, and grammatical accuracy and structural clarity have been improved. The paper is well written and accessible for a broad audience.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: Yes: Dheeb Albashish

Reviewer #3: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PONE-D-25-14929R2

Trustworthy AI for medical decisions: Adversarial robust and fair machine learning prediction for Parkinson’s disease

PLOS ONE

Dear Dr. Rahim,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Nov 13 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Rizik M. H. Al-Sayyed, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise.

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Additional Editor Comments:

Manuscript strong points:

The manuscript includes an SoTA comparison in Table1 (2021–2025) with methods, datasets, and headline metrics, and situates the contribution in that context.

The Methods are described with dataset splits, cross-validation, metrics, and statistical testing; hyperparameter grids and attack settings are summarized (Table 8).

The paper is transparent about limitations (no DL baselines in experiments, preprocessing-only mitigation, limited attack types, discretized sensitive attributes, single-dataset evaluation), which is exactly the right section to acknowledge before acceptance.

I would be satisfied to proceed with “Accept” (after minor revision) contains:

Direct, like-for-like baselines on your splits. The SoTA table is useful, but adding head-to-head results (accuracy and fairness metrics) for at least a strong classical model (e.g., XGBoost/SVM) and one representative DL baseline to match the works you cite.

Fairness method baselines. Compare “optimized preprocessing” against at least one in-processing (e.g., adversarial debiasing) or post-processing (e.g., equalized odds) approach on the same pipeline.

Reproducibility details. Report the final chosen hyperparameters (not just ranges), random seeds, and fold assignments; and provide code or a repository link/script to reproduce preprocessing, fairness metrics, and attack generation (poison rate, leakage construction).

Generalization check (lightweight). If external validation isn’t feasible now, add a sensitivity analysis (e.g., varying poison rate beyond 5% and reporting CIs across folds) to strengthen the robustness story you’re telling.

[Note: HTML markup is below. Please do not edit.]

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PONE-D-25-14929R3

Trustworthy AI for medical decisions: Adversarial robust and fair machine learning prediction for Parkinson’s disease

PLOS ONE

Dear Dr. Rahim,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Jan 25 2026 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Rizik M. H. Al-Sayyed, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise.

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

Reviewer #4: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

Reviewer #4: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

Reviewer #4: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

Reviewer #4: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

Reviewer #4: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: The authors have successfully addressed all the listed comments. Congratulations!

The modifications made to the abstract, introduction, and results sections demonstrate a clear improvement in the manuscript.

Reviewer #4: 1- Some sentences in the abstract are long and dense. Consider breaking them into shorter statements to enhance readability.

2- The introduction contains abrupt transitions. Adding linking sentences would improve narrative flow.

3- The early introduction lists PD symptoms with several citations grouped together. Consider integrating citations more smoothly after each relevant clinical statement.

4- The description of the PPMI dataset's demographics could be expanded slightly for clarity.

5- SPD and EOD are introduced later in the abstract without explanation. Either define them briefly or ensure they appear earlier in the text before being referenced.

6- The methodology appears scattered across subsections. Adding a short roadmap paragraph at the beginning of the Methods section would help guide the reader.

7- Ensure every figure and table is explicitly cited in the text in sequential order.

8- There are occasional grammatical issues.

9- The conclusion summarizes findings but could benefit from a clearer statement of the study’s practical impact for clinicians and system designers.

10- Although the results are strong, the manuscript would benefit from a brief section acknowledging limitations and future research directions.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: Yes: Dheeb Albashish

Reviewer #4: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

To ensure your figures meet our technical requirements, please review our figure guidelines: https://journals.plos.org/plosone/s/figures

You may also use PLOS’s free figure tool, NAAS, to help you prepare publication quality figures: https://journals.plos.org/plosone/s/figures#loc-tools-for-figure-preparation.

NAAS will assess whether your figures meet our technical requirements by comparing each figure against our figure specifications.

Trustworthy AI for medical decisions: Adversarial robust and fair machine learning prediction for Parkinson’s disease

PONE-D-25-14929R4

Dear Dr. Rahim,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. For questions related to billing, please contact billing support.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Rizik M. H. Al-Sayyed, Ph.D.

Academic Editor

PLOS One

Additional Editor Comments (optional):

Reviewers' comments: