ArsenicNet: An efficient way of arsenic skin disease detection using enriched fusion Xception model

Md Humaion Kabir Mehedi; Kh. Fardin Zubair Nafis; Krity Haque Charu; Jia Uddin; Md Golam Rabiul Alam; M.F. Mridha

doi:10.1371/journal.pone.0322405

Peer Review History

Original SubmissionOctober 7, 2024
7 Oct 2024 Author Response https://doi.org/10.1371/journal.pone.0322405.r001
19 Nov 2024 Decision Letter - Asadullah Shaikh, Editor PONE-D-24-44589ArsenicNet: An Efficient Way of Arsenic Skin Disease Detection Using Enriched Fusion Xception ModelPLOS ONE Dear Dr. Mridha, 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 03 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, Asadullah Shaikh, 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. [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: Partly Reviewer #2: Partly Reviewer #3: Partly ******** 2. Has the statistical analysis been performed appropriately and rigorously? Reviewer #1: Yes Reviewer #2: Yes 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: Yes Reviewer #3: Yes ****** 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: No Reviewer #3: Yes ****** 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: The manuscript introduces "ArsenicNet," a novel fusion model for arsenic-induced skin disease detection that combines the Xception model with an Inception module. The paper aims to address a significant public health issue by providing an accurate and efficient approach to detect arsenic contamination-related skin conditions, particularly for populations in Bangladesh who rely heavily on groundwater. With an impressive accuracy of 97.69%, the model outperformed five state-of-the-art deep learning models. Using the "ArsenicSkinImageBD" dataset, the authors argue that ArsenicNet is a promising tool for public health diagnostics. Major Comments: 1. Model Justification and Architecture: Comment: The integration of Xception with an Inception module is creative, yet the manuscript lacks a detailed rationale for choosing this specific combination. While the authors provide a block diagram and description, additional explanation of how each module contributes to performance and why they were chosen over other architectures is necessary. Recommendation: Include a more thorough discussion on why the Xception-Inception fusion is preferred, specifically highlighting the roles of multi-scale feature extraction and depthwise separable convolutions in improving accuracy for arsenic skin disease detection. 2. Dataset Limitations and Generalizability: Comment: The dataset used in this study, "ArsenicSkinImageBD," is highly specific to Bangladesh. While appropriate for this study, this limited scope risks reducing the model's generalizability across broader populations and environments. Arsenic contamination affects regions beyond Bangladesh, and the model might face performance drops in other contexts due to demographic and environmental variations in skin disease manifestation. Recommendation: Address this limitation by discussing potential generalizability issues and the need for future external validation. If feasible, suggest synthetic augmentation techniques or access to additional datasets for a more comprehensive evaluation. 3. Lack of Explainability in Model Predictions: Comment: In a clinical setting, explainability is crucial. The paper currently lacks any application of Explainable AI (XAI) methods to enhance the interpretability of the model’s predictions, especially given the model’s sensitivity to complex patterns in skin images. Recommendation: Incorporate XAI tools, such as SHAP or Grad-CAM, to illustrate how the model makes decisions. Visual representations of feature importance or heatmaps over skin images could be particularly beneficial in understanding the model’s decision-making process and would enhance trust in clinical applications. 4. Comparison with Modern Techniques: Comment: While the model achieves significant accuracy, the comparative analysis is limited to relatively established deep learning architectures like InceptionV3, VGG19, and ResNet152V2. Given the rapid advancements in computer vision, including the adoption of Vision Transformers (ViT) and ensemble learning methods, the benchmarks used in this study might not fully demonstrate the model's relative performance. Recommendation: Test the model against Vision Transformers, ensemble learning techniques, or hybrid models that have recently shown promising results in image classification tasks. This would provide a more comprehensive assessment and enhance the study's relevance. 5. Performance Metrics and Thresholds: Comment: The paper reports high accuracy, but it does not provide adequate detail on other critical metrics, such as sensitivity, specificity, and F1-scores, across different thresholds. These metrics are especially relevant in medical applications, where minimizing false negatives (missed disease cases) is often more critical than maximizing accuracy alone. Recommendation: Expand the performance analysis to include a discussion of sensitivity, specificity, and the implications of various classification thresholds. An ROC curve and AUC values provide an initial sense of balance, but concrete metrics on false negatives and false positives at different thresholds are necessary to evaluate the model's robustness and clinical applicability. 6. Reproducibility and Model Training Details: Comment: Although the manuscript includes some hyperparameter settings, it lacks comprehensive details on model training, such as batch size, data preprocessing specifics, data augmentation techniques, and hardware configurations. These are essential for replicating the study and ensuring its robustness. Recommendation: Add a dedicated section that details all preprocessing steps, including data normalization, resizing, augmentation parameters, and specific hardware configurations used during training. Provide detailed values for all hyperparameters, learning rates, and decay schedules to facilitate reproducibility. 7. Broader Implications and Ethical Considerations: Comment: The health implications of arsenic detection are significant, and while the study highlights the potential of ArsenicNet for public health, it lacks a broader discussion on ethical considerations, such as privacy concerns, data security, and the implications of false positives and negatives in a medical setting. Recommendation: Address the ethical considerations of deploying such models in healthcare, including privacy safeguards, the potential impacts of misdiagnosis, and plans to manage data confidentiality in clinical applications. Minor Comments: Abstract: The abstract should briefly highlight the limitations of the dataset and the need for broader validation. Figures: Figures related to the model architecture could be made clearer with improved annotations on each component. Terminology Consistency: Ensure consistency in terminology, especially regarding technical terms like "false positives," "false negatives," "precision," and "sensitivity." Reviewer #2: The paper presents an interesting topic. However, it needs general language and structural editing to enhance its readability. For instance, in the related work, a statement such as "In this paper, authors tried to get information of measuring arsenic in groundwater [5]. Authors wanted to draw a concern regarding..." is a bit low for this kind of research paper. I recommend proper language editing to enhance the paper's readability and resubmission for proper review. Reviewer #3: 1.Ensure a logical flow from data preparation, model architecture, hyperparameter tuning, experimental setup, and result analysis. This will help readers follow your process easily. 2.Provide clear justifications for why Xception and an inception module were combined and how this particular choice impacts classification. 3.Clearly justify the choice of data split (80-10-10 vs. 70-15-15). Discuss how this choice impacts model generalization and robustness, ideally with cross-validation results or a graph showing performance for each split. 4.Provide more in-depth information on the inception module, explaining how each filter size contributes to multi-scale feature extraction and its impact on model performance. 5.Provide more context around why certain models perform better on this dataset. If available, show how similar architectures performed on similar datasets. 6.Include a more in-depth analysis of confusion matrix results and the ROC curves for each model. Highlight why certain models had higher false positives or false negatives. 7.Use bar graphs or spider charts to compare models across multiple metrics (accuracy, F1, precision, recall). This will offer a clearer view of each model’s strengths and weaknesses. 8. Acknowledge any limitations, such as dataset size, imbalance, or possible overfitting issues, and discuss how these could affect the model's performance in real-world scenarios. 9.Propose specific future work directions, like exploring other architectures, optimizing additional hyperparameters, or testing on larger datasets. ****** 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: No 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. https://doi.org/10.1371/journal.pone.0322405.r002
Revision 1
26 Jan 2025 Author Response Reviewer#1: The manuscript introduces "ArsenicNet," a novel fusion model for arsenic-induced skin disease detection that combines the Xception model with an Inception module. The paper aims to address a significant public health issue by providing an accurate and efficient approach to detect arsenic contamination-related skin conditions, particularly for populations in Bangladesh who rely heavily on groundwater. With an impressive accuracy of 97.69%, the model outperformed five state-of-the-art deep learning models. Using the "ArsenicSkinImageBD" dataset, the authors argue that ArsenicNet is a promising tool for public health diagnostics. Reviewer#1, Concern # 1 (please list here): Model Justification and Architecture: Comment: The integration of Xception with an Inception module is creative, yet the manuscript lacks a detailed rationale for choosing this specific combination. While the authors provide a block diagram and description, additional explanation of how each module contributes to performance and why they were chosen over other architectures is necessary. Recommendation: Include a more thorough discussion on why the Xception-Inception fusion is preferred, specifically highlighting the roles of multi-scale feature extraction and depth wise separable convolutions in improving accuracy for arsenic skin disease detection. Author response: Thanks for the comment. Author action: Xception uses depth wise separable convolutions to extract strong spatial features uniformly across the feature map where a multi-scale processing module such as Inception helps the model learn features of different sizes and complexity. The fine-grained features of Xception are enhanced by the Inception module, making the model more adaptable to different datasets. The modular design of Inception and Xception's computational efficiency makes the architecture lightweight. The trainable parameters for Xception were 51,512,578 whereas the proposed combination had 814,962. The fixed Xception architecture can be made flexible by changing the filter sizes or branch configurations in the Inception module to suit the dataset. Medical images, such as skin disease images, benefit from this combination because the Xception backbone extracts strong features, and the Inception module captures context and scale-specific information. Without the Inception module the Xception test accuracy is 94.62% and with the Inception module 97.69%. The relevant text has been added in Section 3.4, page no 8 and all the revised text is highlighted in yellow colour.________________________________________ Reviewer#1, Concern # 2 (please list here): Dataset Limitations and Generalizability: Comment: The dataset used in this study, "ArsenicSkinImageBD," is highly specific to Bangladesh. While appropriate for this study, this limited scope risks reducing the model's generalizability across broader populations and environments. Arsenic contamination affects regions beyond Bangladesh, and the model might face performance drops in other contexts due to demographic and environmental variations in skin disease manifestation. Recommendation: Address this limitation by discussing potential generalizability issues and the need for future external validation. If feasible, suggest synthetic augmentation techniques or access to additional datasets for a more comprehensive evaluation. Author response: Thank you for your concern. Author action: We used “ArsenicSkinImageBD” dataset for our study. The dataset sample has been collected from 4 different villages in Bangladesh. We did not find any other public arsenic skin image based dataset. So we could not do any external validation, but in the future we will try to incorporate other skin disease datasets with our used dataset in this study. We believe our proposed model may not drop performance in other contexts due to demographic and environmental variations in skin disease manifestation. ________________________________________ Reviewer#1, Concern # 3(please list here): Lack of Explainability in Model Predictions: Comment: In a clinical setting, explainability is crucial. The paper currently lacks any application of Explainable AI (XAI) methods to enhance the interpretability of the model’s predictions, especially given the model’s sensitivity to complex patterns in skin images. Recommendation: Incorporate XAI tools, such as SHAP or Grad-CAM, to illustrate how the model makes decisions. Visual representations of feature importance or heatmaps over skin images could be particularly beneficial in understanding the model’s decision-making process and would enhance trust in clinical applications. Author response: Thanks for the observation. Author action: We applied Grad-Cam to understand the decision-making process of the model and would enhance trust in clinical applications. Section “4.4 Visualization with Explainable AI Grad-CAM” added with detailed explanations and examples. The revised text has been highlighted with in yellow color on page no 17. Reviewer#1, Concern # 4 (please list here): Comparison with Modern Techniques: Comment: While the model achieves significant accuracy, the comparative analysis is limited to relatively established deep learning architectures like InceptionV3, VGG19, and ResNet152V2. Given the rapid advancements in computer vision, including the adoption of Vision Transformers (ViT) and ensemble learning methods, the benchmarks used in this study might not fully demonstrate the model's relative performance. Recommendation: Test the model against Vision Transformers, ensemble learning techniques, or hybrid models that have recently shown promising results in image classification tasks. This would provide a more comprehensive assessment and enhance the study's relevance. Author response: Thank you for the comment. Author action: As we already know, the transformer model requires a large amount of data for training, whereas our dataset contains only 1287 samples. We have tried different transformer models such as ViT, Swin and Deit. However, without ViT, none could perform significantly well. The ViT results are reported in Table 5 Reviewer#1, Concern # 5 (please list here): Performance Metrics and Thresholds: Comment: The paper reports high accuracy, but it does not provide adequate detail on other critical metrics, such as sensitivity, specificity, and F1-scores, across different thresholds. These metrics are especially relevant in medical applications, where minimizing false negatives (missed disease cases) is often more critical than maximizing accuracy alone. Recommendation: Expand the performance analysis to include a discussion of sensitivity, specificity, and the implications of various classification thresholds. An ROC curve and AUC values provide an initial sense of balance, but concrete metrics on false negatives and false positives at different thresholds are necessary to evaluate the model's robustness and clinical applicability. Author response: Thank you for the comment. Author action: Our model not only achieved higher accuracy but also a higher recall/sensitivity score, which is very important in medical applications. In Table 5, all performance metrics are listed. Reviewer#1, Concern # 6 (please list here): Reproducibility and Model Training Details: Comment: Although the manuscript includes some hyperparameter settings, it lacks comprehensive details on model training, such as batch size, data preprocessing specifics, data augmentation techniques, and hardware configurations. These are essential for replicating the study and ensuring its robustness. Recommendation: Add a dedicated section that details all preprocessing steps, including data normalization, resizing, augmentation parameters, and specific hardware configurations used during training. Provide detailed values for all hyperparameters, learning rates, and decay schedules to facilitate reproducibility. Author response: Thank you for the suggestion. Author action: We have already provided all the data preprocessing steps, including data normalization, resizing, and augmentation parameters, in Section 3.2. The hardware configurations are listed in Table 4. Moreover, Table 3 provides all hyperparameters. Reviewer#1, Concern # 7 (please list here): Broader Implications and Ethical Considerations: Comment: The health implications of arsenic detection are significant, and while the study highlights the potential of ArsenicNet for public health, it lacks a broader discussion on ethical considerations, such as privacy concerns, data security, and the implications of false positives and negatives in a medical setting. Recommendation: Address the ethical considerations of deploying such models in healthcare, including privacy safeguards, the potential impacts of misdiagnosis, and plans to manage data confidentiality in clinical applications. Author response: Thank you for the comment. Author action: We used a publicly available dataset hosted in Mendeley Data for our study. In addition, the dataset is anonymous, so there is no chance that the patient information will be licked. The authors of the dataset maintained all ethical considerations and privacy of the patient. Reviewer#1, Concern # 8 (please list here): Abstract: The abstract should briefly highlight the limitations of the dataset and the need for broader validation. Author response: Thank you for your comment. Author action: We have added dataset limitations in the abstract. Reviewer#1, Concern # 9 (please list here): Figures: Figures related to the model architecture could be made clearer with improved annotations on each component. Author response: Thank you for your concern. Author action: We have updated Figures 5, 6 and 7. Reviewer#1, Concern # 10 (please list here): Terminology Consistency: Ensure consistency in terminology, especially regarding technical terms like "false positives," "false negatives," "precision," and "sensitivity." Author response: Thank you for the feedback. Author action: We have updated the manuscript. Reviewer#2: The paper presents an interesting topic. However, it needs general language and structural editing to enhance its readability. For instance, in the related work, a statement such as "In this paper, authors tried to get information of measuring arsenic in groundwater [5]. Authors wanted to draw a concern regarding..." is a bit low for this kind of research paper. Reviewer#2, Concern # 1 (please list here): I recommend proper language editing to enhance the paper's readability and resubmission for proper review. Author response: Thank you for the comment. Author action: We have revised the manuscript and proofread by a native speaker. Reviewer#3, Concern #1(please list here): Ensure a logical flow from data preparation, model architecture, hyperparameter tuning, experimental setup, and result analysis. This will help readers follow your process easily. Author response: Thank you for the suggestion. Author action: We have updated according to your suggestions.________________________________________ Reviewer#3, Concern #2(please list here): Provide clear justifications for why Xception and an inception module were combined and how this particular choice impacts classification Author response: Thank you for the feedback. Author action: Xception uses depth wise separable convolutions to extract strong spatial features uniformly across the feature map where a multi-scale processing module such as Inception helps the model learn features of different sizes and complexity. The fine-grained features of Xception are enhanced by the Inception module, making the model more adaptable to different datasets. The modular design of Inception and Xception's computational efficiency make the architecture lightweight. The trainable parameters for Xception were 51,512,578 whereas the proposed combination had 814,962. The fixed Xception architecture can be made flexible by changing the filter sizes or branch configurations in the Inception module to suit the dataset. Medical images, such as skin disease images, benefit from this combination because the Xception backbone extracts strong features, and the Inception module captures context and scale-specific information. The relevant text are added in Section 3.4, page no 8 and all the revised text is highlighted in yellow colour. Reviewer#3, Concern #3(please list here): Clearly justify the choice of data split (80-10-10 vs. 70-15-15). Discuss how this choice impacts model generalization and robustness, ideally with cross-validation results or a graph showing performance for each split. Author response: Thank you for the observation. Author action: For our study, we used different data splits, such as 80-10-10 vs. 70-15-15 but obtained the best performance on 80-10-10. Reviewer#3, Concern #4(please list here): Provide more in-depth information on the inception module, explaining how each filter size contributes to multi-scale feature extraction and its impact on model performance. Author response: Thank you for the comment. Author action: Xception uses depth wise separable convolutions to extract strong spatial features uniformly across the feature map where a multi-scale processing module such as Inception helps the model learn features of different sizes and complexity. The fine-grained features of Xception are enhanced by the Inception module, making the model more adaptable to different datasets. The modular design of Inception and Xception's computational efficiency make the architecture lightweight. The trainable parameters for Xception were 51,512,578 whereas the proposed combination had 814,962. The fixed Xception architecture can be made flexible by changing the filter sizes or branch configurations in the Inception module to suit the dataset. Medical images, such as skin disease images, benefit from this combination because the Xception backbone extracts strong features, and the Inception module captures context and scale-specific information. We have tried with different filter size but got the best outcome with the proposed set given in Fig. 4. Equations 2, 3, 4, 5 and 6 are related to filter and kernels of Inception module. Reviewer#3, Concern #5(please list here): Provide more context around why certain models perform better on this dataset. If available, show how similar architectures are performed on similar datasets. Author response: Thank you for the concern. Author action: Reviewer#3, Concern #6(please list here): Include a more in-depth analysis of confusion matrix results and the ROC curves for each model. Highlight why certain models had higher false positives or false negatives. Author response: Thank you for the remarks. Author action: We have already given it in Section 4.2. Reviewer#3, Concern #7(please list here): Use bar graphs or spider charts to compare models across multiple metrics (accuracy, F1, precision, recall). This will offer a clearer view of each model’s strengths and weaknesses Author response: Thank you for the suggestion. Author action: We have shown multiple metrics (accuracy, precision, recall/ sensitivity and F1 score) in Table 5. Reviewer#3, Concern #8(please list here): Acknowledge any limitations, such as dataset size, imbalance, or possible overfitting issues, and discuss how these could affect the model's performance in real-world scenarios. Author response: Thank you for the concern. Author action: Dataset size with only 1287 samples, is one of the main limitations. As the dataset is small, there is a chance of model overfitting. Therefore, we addressed these issues using early callback function. If we could train on a large dataset, the performance of the model would increase, which would increase its usability in real-world scenarios. Reviewer#3, Concern #9(please list here): Propose specific future work directions, like exploring other architectures, optimizing additional hyperparameters, or testing on larger datasets. Author response: Thank you for the concern. Author action: Future work will focus on expanding the dataset and will combine this work with other skin diseases in different biological and geographical con Attachments Attachment Submitted filename: PLOS ONE-Response-to-Reviewers.pdf https://doi.org/10.1371/journal.pone.0322405.r003
18 Feb 2025 Decision Letter - Asadullah Shaikh, Editor PONE-D-24-44589R1ArsenicNet: An Efficient Way of Arsenic Skin Disease Detection Using Enriched Fusion Xception ModelPLOS ONE Dear Dr. Mridha, 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 Apr 04 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, Asadullah Shaikh, Ph.D. Academic Editor PLOS ONE Journal Requirements: 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 #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: (No Response) 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: Yes 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: I commend the authors for their effort in improving the manuscript based on the previous comments, but I still believe there are still a few things to address. Additional Comments: 1. Language and Structure: Although the authors indicated that the manuscript has been proofread, some sections still require general language and structural editing for improved readability. For example, in the abstract, a statement like "The proposed model achieved 97.69% accuracy, along with 97.63% F1 score and reached the top"; in section 4.3, a statement like "More over, if we notice closely, then we can see that we have got the lowest test accuracy,..."; etc. Recommendation: Professional language editing will perhaps be much better. 2. The figure illustrating the model architecture needs clarity. Adequate annotations explaining each component and its role in the architecture are important. 3. Dataset size: The author acknowledged that the dataset is small. This is a big concern as to the generalisability of the model findings. Recommendation: The author should include a comprehensive discussion (also in the abstract) on the limitations of the research (particularly the chances of model overfitting). Reviewer #3: I recommend to accept the paper for publishing since all my comments were addressed, the modifications has enhanced the paper quality. ****** 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: No 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. https://doi.org/10.1371/journal.pone.0322405.r004
Revision 2
3 Mar 2025 Author Response Reviewer#2: I commend the authors for their effort in improving the manuscript based on the previous comments, but I still believe there are still a few things to address. Reviewer#2, Concern # 1 (please list here): Language and Structure: Although the authors indicated that the manuscript has been proofread, some sections still require general language and structural editing for improved readability. For example, in the abstract, a statement like "The proposed model achieved 97.69% accuracy, along with 97.63% F1 score and reached the top"; in section 4.3, a statement like "More over, if we notice closely, then we can see that we have got the lowest test accuracy,..."; etc. Recommendation: Professional language editing will perhaps be much better. Author response: Thank you for the comment. Author action: We have revised the manuscript and proofread it by a native speaker. Reviewer#2, Concern #2(please list here): The figure illustrating the model architecture needs clarity. Adequate annotations explaining each component and its role in the architecture are important. Author response: Thank you for the suggestion. Author action: We have clearly drawn and annotated the model architecture figure. We have also provided a detailed description of each component's role. Reviewer#2, Concern #3(please list here): Dataset size: The author acknowledged that the dataset is small. This is a big concern as to the generalisability of the model findings. Recommendation: The author should include a comprehensive discussion (also in the abstract) on the limitations of the research (particularly the chances of model overfitting). Author response: Thank you for the feedback. Author action: As we already said, the dataset is small in size and there are no alternative datasets available. So, there might be a chance of model overfitting. To prevent this issue, we applied different data augmentation techniques, such as horizontal flip, width, and height shift. In addition, we used early stopping function to prevent overfitting issues by monitoring val_loss. Attachments Attachment Submitted filename: Arsenic_PLOS ONE-Response-to-Reviewers 2.pdf https://doi.org/10.1371/journal.pone.0322405.r005
21 Mar 2025 Decision Letter - Asadullah Shaikh, Editor ArsenicNet: An Efficient Way of Arsenic Skin Disease Detection Using Enriched Fusion Xception Model PONE-D-24-44589R2 Dear Dr. Mridha, 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. If you have any questions relating to publication charges, please contact our Author Billing department directly at authorbilling@plos.org. 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, Asadullah Shaikh, Ph.D. Academic Editor PLOS ONE Additional Editor Comments (optional): 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 ******** 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 ****** 3. Has the statistical analysis been performed appropriately and rigorously? Reviewer #2: N/A ****** 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 ****** 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 ****** 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 been able to address all comments raised in the previous review. No further comments. ****** 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: No ******** https://doi.org/10.1371/journal.pone.0322405.r006
Formally Accepted
Acceptance Letter - Asadullah Shaikh, Editor PONE-D-24-44589R2 PLOS ONE Dear Dr. Mridha, I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now being handed over to our production team. At this stage, our production department will prepare your paper for publication. This includes ensuring the following: * All references, tables, and figures are properly cited * All relevant supporting information is included in the manuscript submission, * There are no issues that prevent the paper from being properly typeset You will receive further instructions from the production team, including instructions on how to review your proof when it is ready. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few days to review your paper and let you know the next and final steps. Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. 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. If we can help with anything else, please email us at customercare@plos.org. Thank you for submitting your work to PLOS ONE and supporting open access. Kind regards, PLOS ONE Editorial Office Staff on behalf of Prof. Asadullah Shaikh Academic Editor PLOS ONE https://doi.org/10.1371/journal.pone.0322405.r007

Open letter on the publication of peer review reports

PLOS recognizes the benefits of transparency in the peer review process. Therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. Reviewers remain anonymous, unless they choose to reveal their names.

We encourage other journals to join us in this initiative. We hope that our action inspires the community, including researchers, research funders, and research institutions, to recognize the benefits of published peer review reports for all parts of the research system.

Learn more at ASAPbio .