Multi-Modal EEG–Fusion Neurointerface Wheelchair Control System

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Introduction needs a lot of improvement to include much more references and comparisons.

Line 73: EEG is already defined in line 32. Please used it as EEG not as electroencephalogram

Line 75: Please you either explain what is Jetson Nano Platform, or put a reference for it.

Line 96: Space missing                             

Line 119: What is MCU and PWM? Write the full long form for the first time

Line 119-120: What Infineon full-bridge motor driver chips?

Line 129: mention a reference and refer to the picture of the 64-channel system.

Line 5-7: you mentioned you used dry electrodes which is against what’s written in Line 130.

Line: 139 – 146: You need to justify with evidence and references for these assignments for the electrodes.

Figure5: Is it FP1 or FP2? Why no labels on the axes?

Line190: Decision Fp1 and Fp2 are blink-related based on what? Please explain or at least put references

Line193: What are A1 and A2 electrodes?

Line 199: Figure reference error.

Line206: Are ending the blink when T_lockout passes or when the amplitude goes below T_h?

Algorithm1: Please explain the variables mentioned in the algorithm.

Line 211: what is ERD and ERS?

Line 211: I believe there is a typo

Line 212: I believe there is a typo

Line211-214: Why you call it rhythm?

Line 215: you need to define MI first.

Line: 216: ERD and ERS are defined here but used earlier

Line 219: Here you call it motor imagery not MI.

You need to recheck the whole paper please. In general it needs improvements

Also please don’t use  words like “superior performance”. Please use better words that compare your work to some baseline.

Author Response

Thank you very much for your valuable suggestions and detailed feedback on this research paper. The issues you pointed out are very instructive, and we have sorted out and completed all the corrections one by one, and organized the complete corrections in a Word document for easy review

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

In this paper, the authors describe a "fusion" control system for electric wheelchairs, which assumes combining electroencephalogram (EEG) signal analysis with other sensory information. The manuscript is well-written and comprehensive, and its theoretical assumptions correspond well with experimental work. However, in my opinion, the reported results are far from real practical applicability. The Authors need to invite some medical researchers, physicians, and rehabilitation experts into their team to better understand the limitations of disabled people. My more detailed comments are as follows:

 

1. The term "abstract" is repeated twice in the corresponding section.

2. Physical functional impairments are often supplemented with mental limitations, which increase the complexity of EEG-based control system design. The most important part here is tuning the EEG (and other sensory subsystems) to a certain patient. The repeatability of such systems is usually really low.

3. How well are the complex operations (e.g. ramp movement, elevator/escalator interactions) performed by your system? Avoiding obstacles?

4. • Please, add some references into line 31 to prove that "existing brain-controlled wheelchair technologies provide some support…" Please note the existing EEG-controlled electric wheelchair systems here, which are currently in relatively broad usage and serially produced.

5. An electric wheelchair is a unique and expensive life-supporting device, and it must be extremely reliable. The aim to "reduce costs" may be a bit contrary to retaining the overall robustness of the system. In other words, other properties of the device should not suffer from cost reduction.

6. It is known that more complex sensory systems are more prone to failures and are harder to calibrate. How does this affect the overall usability of your system?

7. How can the EEG signals be "continuously acquired" while the ADC in your system produces digital output? Please, correct the terms here.

8. Even non-invasive EEG interfaces are uncomfortable to wear in long-term operation and are prone to various environmental changes, e.g. sweating. How does your system take it into account?

9. Is the proposed system compatible with medical sensors, which are built-in in some wheelchair models?

10. Fig. 6 is too simple and can be reformulated as an algorithm in text form.

11. The title can be corrected as follows: "Multi-modal EEG-Fusion Neurointerface Wheelchair Control System"

12. The authors declare that only 3 test subjects were involved in the real-world experiments. How long did it take to readjust the system after the change of operator? Were any disabled \old people among the test subjects? Is the system easy to learn and operate?

Nevertheless, I highly value the signal processing and classification part of this paper. I believe it can be reconsidered for publication after major revisions.

Author Response

Thank you very much for your valuable suggestions and detailed feedback on this research paper. The issues you pointed out are very instructive, and we have sorted out and completed all the corrections one by one, and organized the complete corrections in a Word document for easy review

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The authors propose a multimodal EEG-fusion neurointerface wheelchair system integrating motor imagery, intentional blink detection, and attention-level analysis to enable precise and adaptive wheelchair control.

Questions:

1. The literature review needs improvement. Few recent articles are cited, and there is little discussion of existing work. What are the weaknesses of existing methods, and what do the authors aim to improve or address?

2. The abbreviation EEG is not defined in the abstract or keywords, and there is no list of abbreviations in this article. Please define this abbreviation.

3. Some tables are going beyond the right side margin like Tables 1 and 4.

4. The path of the designed device seems simple and requires a large database for machine learning. So, the more possible paths, the larger the database? Is this feasible? It would be interesting to discuss these limitations.

5. Would it be possible for the authors to present a second case study with a different trajectory for the designed equipment? A large number of case studies reinforces the conclusions.

6. Why did the authors use two different machine learning techniques: Random Forest and Support Vector Machine? Why were these two techniques chosen among so many machine learning techniques?

7. The conclusions need to be improved by pointing out what the future steps would be to make the project viable and applicable to any route taken by the device.

Some mistakes:

8. "Abstract: Abstract"

9. Line 199: "As illustrated in Figure ??,"

Author Response

Thank you very much for your valuable suggestions and detailed feedback on this research paper. The issues you pointed out are very instructive, and we have sorted out and completed all the corrections one by one, and organized the complete corrections in a Word document for easy review

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The Authors propose an EEG-based wheelchair control system for physically impaired individuals. The application seems interesting and novel, however the article seems to be in need of better contextualization and system description. Here, my comments:

- The Authors should make examples of existing Brain-Controlled wheelchair systems and produce references to back up the claims of instability for these kinds of systems.

 

- References for the cons of existing Brain-Controlled wheelchair systems should also be provided by the Authors (lines 50 through 59).

 

- Devices elements and pre-processing techniques should be briefly described in the “Control principle” paragraph to aid readers’ comprehension.

 

- Please define the speed corresponding to the predefined levels mentioned in the paragraph “Control principle”.

 

- The Authors should mention why the Smrting 64 Electrodes Cap was used in this system, and why such a spatial resolution was chosen instead of favoring smaller and less bulky EEG headsets, especially when using only 8 electrodes for the data acquisition.

 

- Pictures of the fabricated modules should be added by the Authors, and not only their schematics.

 

- Authors should describe the interpolation techniques they used in the pre-processing pipeline.

 

- At line 199 the figure reference should be corrected.

 

- The Authors should better define the blinking detection thresholds and how they were calculated.

 

- I believe that the Authors should better differentiate between the pre-processing steps described in the “Lightweight Preprocessing for Blink Signal Detection” and those in the “Signa Processing Pipeline for Motor Imagery” paragraphs.

 

- At line 239 the paragraph’s title seems to be the same as the previous one, please revise.

 

- The Authors should mention the delay introduced by the pre-processing and processing pipelines, including the classification steps.

 

- Please provide references for the claim made at lines 394 through 397 and provide examples for the one made at lines 400 through 402.

 

- Please provide more details of the used SVM model.

 

- The Authors should mention the stability of the used electrode configuration, especially giving more information on the electrodes used and their contact impedance change through time.

 

- Authors should better contextualize the achievements of their proposed system with a more in-depth state-of-the-art analysis.

 

- Please unify reference format.

Author Response

Thank you very much for your valuable suggestions and detailed feedback on this research paper. The issues you pointed out are very instructive, and we have sorted out and completed all the corrections one by one, and organized the complete corrections in a Word document for easy review

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Sorry the paper still needs improvement. You need much more references.

Author Response

Thank you very much for your valuable suggestions and detailed feedback on this research paper. The questions you pointed out were very enlightening, and we sorted out and completed the reference corrections one by one, and organized the full corrections in a Word document for easy viewing

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you very much for revising your manuscript following my suggestions. I am impressed by overall depth of your research and the comprehensiveness of the point-by-point reply letter. I am satisfied with the revised version and believe it can be accepted for publication in its present form.

Author Response

Thank you for taking the time out of your busy schedule to review this article, your valuable comments and pertinent suggestions are very enlightening and provide important guidance for the improvement of the paper

Reviewer 3 Report

Comments and Suggestions for Authors

The authors propose a multimodal EEG-fusion neurointerface wheelchair system integrating motor imagery, intentional blink detection, and attention-level analysis to enable precise and adaptive wheelchair control.

The article has been improved, the contribution is good and all questions have been effectively answered.

Author Response

Thank you for taking the time out of your busy schedule to review this article, your valuable comments and pertinent suggestions are very enlightening and provide important guidance for the improvement of the paper

Reviewer 4 Report

Comments and Suggestions for Authors

The Authors have addressed all the Reviewers' comments.

Author Response

Thank you for taking the time out of your busy schedule to review this article, your valuable comments and pertinent suggestions are very enlightening and provide important guidance for the improvement of the paper

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

Line 55: Recent Reviews?? What reviews are you talking about here?

Line 57: “undermines real-world reliability” What does this mean?

Table 1: Why do you call it Glossary Terms? Why do you need it here and it is not even complete.

Line 71: What is ADC?

Line 73: What is the sampling rate? How did you call it high? What is the baseline here?

Line 78: What is Cuda parallel computing?

Line 90: “are manually inspected and removed” How will it be manually done for a live system?

Line 91: What is InfoMax algorithm?

Line 105: The microcontroller is not defined, so you can’t call it the microcontroller.

Line 107: Moving State ( Why capital letters? )

Line 124: What is TCN?

Line 131: “integrated system design” Should be capital letter

Line 141: “power” repeated

Line 150: which headset? 64 channels but which one?

Line 190: What is I2C bus?

Line 205: EOG has already been defined before

Line 218: eg FP1, FP2? You should know it by know it is not unknown

Line 244: What is the high and low threshold? What is +3 and +1 means?

Line 273: EMG already defined before

Line 275: Why do you say “such as”? Isn’t it selected and tested already? And how did you select C3 and C4? Also, why you need 64 channels headset while you are not using most of them?

Line 278: ICA already defined earlier.

Line 302: ERS and ERD already defined.

Line 289: I’m not quite sure what’s the use of “Experimental Protocol for MI Tasks”. What exactly are you doing here?

Line 369: how did you choose these electrodes?

Line: 328: What dataset? What does it contain? Was it collected by you or available online?

 

You need to check your paper again please

Author Response

Thank you for taking the time out of your busy schedule to review this article, and your professional opinions and detailed suggestions not only accurately point out the direction to be improved in the paper, but also have important guiding significance for improving the rigor and academic value of the research. We have carefully sorted out each of your feedback and will implement it one by one in subsequent revisions. Regarding your specific review opinions, detailed responses (including revision ideas, supplementary arguments, data explanations, etc.) have been uniformly organized into a Word document Once again, I would like to express my sincere gratitude to you for your hard work and professional control! 

Author Response File: Author Response.pdf