Automatic Facial Palsy, Age and Gender Detection Using a Raspberry Pi

Round 1

Reviewer 1 Report

Dear Authors,

Major comments:

It is mentioned in lines 94 onwards "However, previous methods for diagnosing FP or detecting facial features had problems, including considering older methods and often using training data from spoofed FP, in addition to insufficient or non-existent quantitative results and low accuracy in many cases."
Such statements must be supported with references.

This manuscript proposed " a high-accuracy diagnostic system to detect FP and the patient's gender and age in real-time by using deep learning algorithms based on CNN."  However, the authors have failed to compare the performance of the proposed algorithm with the previous model.

It is recommended that authors provide better justification on why the model is better in performance compared to similar studies.

Minor comment:

1. Line 116 states, The other part of the data was obtained from websites on the "Internet", please specify the data source.

2. Please provide codes and specific parameters used in the model. It is recommended to provide codes and parameters in GitHub.

 

Extensive editing of the English language is required.

Author Response

Dear Reviewer 1,

Thank you very much for your valuable comments. We are so glad to follow your suggestions supporting our manuscript. We have addressed your comments as explained below and corrected the manuscript as that possible and believe that the revised manuscript can meet the publication's requirements.

The amendments in the manuscript were highlighted using track change.

 

Many thanks for your time and effort,

 

 

Authors

 

Comments and Suggestions for Authors

Major comments:

It is mentioned in lines 94 onwards "However, previous methods for diagnosing FP or detecting facial features had problems, including considering older methods and often using training data from spoofed FP, in addition to insufficient or non-existent quantitative results and low accuracy in many cases."
Such statements must be supported with references.

# We highly appreciate the references to this recommendation and have modified the manuscript and added references (please see page 3).

This manuscript proposed " a high-accuracy diagnostic system to detect FP and the patient's gender and age in real-time by using deep learning algorithms based on CNN."  However, the authors have failed to compare the performance of the proposed algorithm with the previous model.

It is recommended that authors provide better justification on why the model is better in performance compared to similar studies.

# We have followed this recommendation and added a comparison table with previous works and clarified the difference between them and the proposed system (please see page 11).

Minor comment:

1. Line 116 states, the other part of the data was obtained from websites on the "Internet", please specify the data source.

# Thanks for your valuable comments and we have modified the manuscript accordingly and added the site as a reference (please see page 3 and page 15).

 

2. Please provide codes and specific parameters used in the model. It is recommended to provide codes and parameters in GitHub.

# Thank you for these comments We are determined to upload the code in GitHub as soon as the research is published.

Comments on the Quality of English Language

Extensive editing of the English language is required.

# Thank you for taking the time to look at our manuscript and we have modified the manuscript according to this valuable recommendation. The manuscript has been sent to an English-speaking person to fix grammatical errors.

Reviewer 2 Report

Manuscript Number: 2371317

Comments:

The research paper by Ali Saber Amsalam and colleagues entitled "Automatic Facial Palsy, Age and Gender Detection Using a 2 Raspberry Pi" provides good information. The description is in good shape to prove the current data and is helpful for the field. There are a couple of corrections that should be made and some suggestions to improve clarity:

 

1.    The study used a dataset of 20,600 images containing 19000 normal images and 1600 FP images to achieve an accuracy of 98 %. So my question is whether the number of participants was examined from the same area, or does it also cover other countries’ data?

 2.    Patient no. should increase by at least 30 with a wide range of areas covered to compare the data.

3.    I would like to suggest that the authors have bioinformatics analysis, as they mentioned artificial intelligence. Having some analysis would strengthen the paper significantly.

4.    A significant challenge currently not described/considered is the importance of the potential applications. The authors could consider including this and perhaps briefly mentioning the challenges to investigating this.

 

5.    The overall review description is concise. I recommend that the author elaborate briefly for the reader to understand the scope better and try to include recent publications as a reference. 

English Proof reading is mandatory before publication.

Author Response

Dear Reviewer 2,

Thank you very much for your valuable comments. We are so glad to follow your suggestions supporting our manuscript. We have addressed your comments as explained below and corrected the manuscript as that possible and believe that the revised manuscript can meet the publication's requirements.

The amendments in the manuscript were highlighted using track change.

 

Many thanks for your time and effort,

 

 

Authors

 

Comments and Suggestions for Authors

 

Comments:

The research paper by Ali Saber Amsalam and colleagues entitled "Automatic Facial Palsy, Age and Gender Detection Using a 2 Raspberry Pi" provides good information. The description is in good shape to prove the current data and is helpful for the field. There are a couple of corrections that should be made and some suggestions to improve clarity:

# Thank you very much for your encouraging and valuable comments.

 

1. The study used a dataset of 20,600 images containing 19000 normal images and 1600 FP images to achieve an accuracy of 98 %. So, my question is whether the number of participants was examined from the same area, or does it also cover other countries’ data?

# Thank you very much for your valuable comments. After being trained on the training data, the system was tested on a group of real patients at Al-Rifai General Hospital from the Department of Physiotherapy in Dhi Qar Health of the Iraqi Ministry of Health.

2. Patient no. should increase by at least 30 with a wide range of areas covered to compare the data.

 # Thank you for these comments. Due to the difficulty of collecting data and convincing patients to undergo imaging because facial paralysis is a disease that causes embarrassment to patients; however, four more patients were added to bring the total number to 20 participants (please see page 3).

3. I would like to suggest that the authors have bioinformatics analysis, as they mentioned artificial intelligence. Having some analysis would strengthen the paper significantly.

# Thank you for your feedback on our paper. While we appreciate your suggestion, we respectfully disagree with your assertion that conducting bioinformatics analysis would significantly strengthen our paper. We believe that our paper provides a comprehensive and thorough analysis of the topic at hand, which is focused on the application of artificial intelligence in a specific domain. While bioinformatics analysis may be a useful approach in other contexts, we do not see how it would contribute to the goals and scope of our paper. Thank you again for your input.

4. A significant challenge currently not described/considered is the importance of the potential applications. The authors could consider including this and perhaps briefly mentioning the challenges to investigating this.

 # Thanks for your valuable comments. Due to the difficulty of collecting data and convincing patients to undergo imaging because facial paralysis is a disease that causes embarrassment to Patients; However, four more patients were added to bring the total number to 20 participants.

5. The overall review description is concise. I recommend that the author elaborate briefly for the reader to understand the scope better and try to include recent publications as a reference.

# We highly appreciate the references to this recommendation and have modified the manuscript and added references and the proposed system was compared with previous studies (please see pages 2, 3 and 11).

 

Comments on the Quality of English Language

English Proof reading is mandatory before publication.

# Thank you for taking the time to look at our manuscript and we have modified the manuscript according to this valuable recommendation. The manuscript has been sent to an English-speaking person to fix grammatical errors.

Round 2

Reviewer 1 Report

For reproducibility, authors must provide the codes and specific hyper-parameters used during testing and validation, along with sample dummy data.

moderate editing required.

Author Response

Dear reviewer, 

I would like to express my sincerest gratitude for your valuable feedback and suggestions on my manuscript. Your thorough evaluation and constructive comments have undoubtedly contributed to enhancing the quality and rigor of my work.

Please follow the link below to access the code

https://drive.google.com/file/d/1u_IiwL52pZCufR5u0gIelGjuCSixe5ei/view?usp=share_link

 

Best regards,