Analysis and Classification of Motor Dysfunctions in Arm Swing in Parkinson’s Disease

Round 1

Reviewer 1 Report

This paper describes a wearable system for discriminating normal subject from subjects with motion dysfunctions.

This paper lacks of details and some of the authors’ choices are not clear. For example, why did the author choice the Morlet Wavelet? How did they designed the CNN layers dimensions?

The text is not always clear and there are many sentences that should be rephrased.

Concerning the paper organization, section 3.9 is a repetition of the information contained in the previous section.

The results are reported as tables, but no comments are provided. It is mandatory to extensively comment the results.

Moreover, it is of critical importance to compare the proposed solution with the state of the art in the discussion section, for example with a table that summarizes all the results.

Finally, I suggest to check the references, since the papers from [23] to [39] are not referenced in the main text. Also, some references are not ordered in the text. References [30], [32] and [35] are in German.

Author Response

Reviewer 1
Thanks for your review. Your comments would be very helpful to improve the quality of the paper. In the following, you can see your comments and how we realized them in the new version:

This paper lacks of details and some of the authors’ choices are not clear. For example, why did the author choice the Morlet Wavelet? [A]: We have corrected and edited the text.

 How did they designed the CNN layers dimensions? [A]: We have corrected and edited the text.

 The text is not always clear and there are many sentences that should be rephrased. [A]: Yes, you are right. The following issues were checked and reworked:
o Sorting references in ascending order.
o Checking the correct wording of shortcuts.
o Checking space between value and unit.
o Grammar and vocabulary control of the whole text.
o New title, which is more related to motor dysfunctions.
o Reorganization of the introduction. First all stationary gait analysis systems. Then the wearable ones.

 Concerning the paper organization, section 3.9 is a repetition of the information contained in the previous section. [A]: I disagree with you. We separated methods and methodology. In Methods we only described our used methods. In methodology we describe our whole process.

 The results are reported as tables, but no comments are provided. It is mandatory to extensively comment the results. [A]: The results were better commented.

 Moreover, it is of critical importance to compare the proposed solution with the state of the art in the discussion section, for example with a table that summarizes all the results. [A]: A new paragraph has been inserted for this purpose.

 Finally, I suggest to check the references, since the papers from [23] to [39] are not referenced in the main text. Also, some references are not ordered in the text. References [30], [32] and [35] are in German. [A]: We deleted all references that weren’t used.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper proposes a classification method of Parkinson disease using CNN. Data is taken from commercial wristband sensor. The paper is well written.

Below are some comments:

How the data set was divided for training, validation, and test? What was the training time, configuration of the computer for training,  and the size of the model? In the Result section, a comparison table with other works need to be presented. Some typo: Line 83: "is low cost"; Line 143: "Python library SciPy" 

 

Author Response

Thanks for your review. Your comments would be very helpful to improve the quality of the paper. In the following, you can see your comments and how we realized them in the new version:

 How the data set was divided for training, validation, and test?

[A]: The text has been modified.

 What was the training time, configuration of the computer for training, and the size of the model?

[A]: The values for training time, configuration of the computer were added.

 In the Result section, a comparison table with other works need to be presented.

[A]: A new paragraph has been inserted for this purpose.

 Some typo: Line 83: "is low cost"; Line 143: "Python library SciPy"

[A]: Both have been corrected.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Authors significantly improved the paper. The structure of the paper and the contribution are now well explained.

I suggest only a grammar check since some typos are still present, for example:

- line 49 “which model method provides…”

- line 65 “DecisionTtrees”

- line 75-76 rephrase

- the bullet list points a and f should be rephrased because line 83 has a “that” as last word.

- line 95 “mythology”

- line 137 “The result of Cleanup is in Figure 4 can be seen”