Automatic Detection of Dynamic and Static Activities of the Older Adults Using a Wearable Sensor and Support Vector Machines

Round 1

Reviewer 1 Report

General comments:

The paper makes a well-structured and interesting analysis of the use of a Wearable Sensor and Support Vector Machines to autodetect Dynamic and Static Activities of elderly individuals. Although the present study is interesting, some aspects must be improved for its final publication.

My main concern is related to the introduction. These details are developed later. On the other hand, I recommend consistently maintaining the order between Dynamic and Static Activities. For the title, dynamics appear first, but in methods, statics are first analyzed. Also, static activity can lead to confusion, since an activity is understood as something dynamic. I suggest reformulating the idea. In addition, English has to be revised, especially the use of the Oxford comma.

Specific comments:

Introduction.

Two main recommendations I suggest to the authors.

First, the second paragraph where it analyzes the literature, considers that it must be reformulated and changes the detected structure "Najafi et al. Used ... Lee et al. Proposed ..." A possible alternative would be: "Numerous classification algorithms exist to provide human motion classification patterns, such as gyroscope data and the wavelet method, a linear discriminant analysis method, ... "A second sentence would be used for the uses of these methods. "... to analyze the" sit-to-stand "transition in relation to fall risk ...".

Second, the information collected on page 3 and beginning of page 4, I believe should be described in methods. The end of the introduction should focus more on the gap and not on the SVM methodology.

Other minor comments are:

Second paragraph. Cvetković et al. not underlined.

Fourth paragraph. Support Vector Machines change for SVMs.

Materials and Methods

Step 1-5. I suggest putting the title of each step in bold.

Author Response

The paper makes a well-structured and interesting analysis of the use of a Wearable Sensor and Support Vector Machines to autodetect Dynamic and Static Activities of elderly individuals. Although the present study is interesting, some aspects must be improved for its final publication. Thank you for your comment. These insightful comments have helped us to further improve the manuscript. We have tried our best to improve the manuscript as per your suggestions. My main concern is related to the introduction. These details are developed later. On the other hand, I recommend consistently maintaining the order between Dynamic and Static Activities. For the title, dynamics appear first, but in methods, statics are first analyzed. Also, static activity can lead to confusion, since an activity is understood as something dynamic. I suggest reformulating the idea. In addition, English has to be revised, especially the use of the Oxford comma. Thank you for this comment. We agree that order should be followed throughout manuscript and we have revised it as per your suggestion. -We did some research on the word “Static Activity” and found several researchers have used this word and thus we planned to keep the word as it is to bring consistency in the definition. Please find some references below. 1. STATIC AND DYNAMIC PHYSICAL ACTIVITY ARE INDEPENDENTLY CARDIOPROTECTIVE THROUGHOUT ADULTHOOD Maia P. Smith Journal of the American College of Cardiology Volume 73, Issue 9 Supplement 1, March 2019 DOI: 10.1016/S0735-1097(19)32736-6 2. Static and Dynamic Activity Detection with Ambient Sensors in Smart Spaces. Sensors (Basel). 2019 Feb; 19(4): 804. Published online 2019 Feb 16. doi: 10.3390/s19040804 : Introduction. Two main recommendations I suggest to the authors. First, the second paragraph where it analyzes the literature, considers that it must be reformulated and changes the detected structure "Najafi et al. Used ... Lee et al. Proposed ..." A possible alternative would be: "Numerous classification algorithms exist to provide human motion classification patterns, such as gyroscope data and the wavelet method, a linear discriminant analysis method, ... "A second sentence would be used for the uses of these methods. "... to analyze the" sit-to-stand "transition in relation to fall risk ...". Thank you for this comment, we have separated sentences as suggested. Second, the information collected on page 3 and beginning of page 4, I believe should be described in methods. The end of the introduction should focus more on the gap and not on the SVM methodology. Thank you for your comment. We have corrected it now. Other minor comments are: Second paragraph. Cvetković et al. not underlined. Fourth paragraph. Support Vector Machines change for SVMs. Materials and Methods Step 1-5. I suggest putting the title of each step in bold. Thank you for your comment. We have corrected it now.

Reviewer 2 Report

• The use of the word “elderly” has a negative connotation in English. Are the authors targeting “older people” or “elderly”? In section 2.1, the participants' age suggest they are in the older people category. However, the authors do not state if they were healthy or not.  Please make sure to use the correct term in the article.

• The sentence “Finally, Begg et al. used the SVM classifier to analyze the minimum foot clearance owing to aging [17].” The word “finally” suggest that the only research in this area are the few mentioned in this paragraph, which is not the case. Please revise.

• Sec2.1: The data collection was “approved by the Institutional Review Board at Virginia Tech”. Were there other ethical consideration taken into account? If so, please report them.

• Section 3.2 says “Firstly, 60 sets of data were split into training data and test data evenly, and then LibSVM with random SVM parameters was used to obtain a classification accuracy of 60%.” It is not clear what the other 40% was used for. Please clarify.

• English language needs to be revised, there were several minor mistakes.

• The last paragraph of the discussion “ Conclusions based on this study should be considered in the context of its limitations. First…” should be moved to a new section called “limitations". If possible a section "limitation and strengths" would strengthen this article.

• References (in main text and reference section) need to be re-formatted and consistent.  Some article titles are italics, while sometimes the journals are italic and the article title is not.

• The article would strengthen if the authors find more research to cite. SVM and older people activitie recognition is a very common area of research.

Author Response

Comment: The use of the word “elderly” has a negative connotation in English. Are the authors targeting “older people” or “elderly”? In section 2.1, the participants' age suggest they are in the older people category. However, the authors do not state if they were healthy or not. Please make sure to use the correct term in the article. Thank you for your comment. We are sorry, and have corrected it now. The sentence “Finally, Begg et al. used the SVM classifier to analyze the minimum foot clearance owing to aging [17].” The word “finally” suggest that the only research in this area are the few mentioned in this paragraph, which is not the case. Please revise. Thank you for your comment. We have revised it now and the word is removed now. Sec2.1: The data collection was “approved by the Institutional Review Board at Virginia Tech”. Were there other ethical consideration taken into account? If so, please report them. Thank you for your comment. All investigators were CITI certified and followed as per IRB approved protocol to protect the identity of the participants. The details of the IRB are provided in the dissertation document (https://vtechworks.lib.vt.edu/handle/10919/64189 ). Section 3.2 says “Firstly, 60 sets of data were split into training data and test data evenly, and then LibSVM with random SVM parameters was used to obtain a classification accuracy of 60%.” It is not clear what the other 40% was used for. Please clarify. English language needs to be revised, there were several minor mistakes. Thank you for your comment. We agree the sentence was not clear. We have rephrased it now. It reads 60% accuracy instead of 60% dataset. The last paragraph of the discussion “ Conclusions based on this study should be considered in the context of its limitations. First…” should be moved to a new section called “limitations". If possible a section "limitation and strengths" would strengthen this article. Thank you for your comment. We have made limitations section References (in main text and reference section) need to be re-formatted and consistent. Some article titles are italics, while sometimes the journals are italic and the article title is not. Thank you for your comment. We have formatte them consistently now. The article would strengthen if the authors find more research to cite. SVM and older people activities recognition is a very common area of research. Thank you for your comment. we have few articles added now.

Round 2

Reviewer 1 Report

The authors have done a good job of improving the manuscript and have answered all my comments. Congratulations.

Author Response

The authors have done a good job of improving the manuscript and have answered all my comments. Congratulations. We are thankful to the reviewers for their comments. Your comments have helped us significantly improve the quality of manuscript.

Reviewer 2 Report

The authors have done a great job improving the manuscript. I found three minor suggestions for the authors.

• Please check the correcteness of "a SVM" or "an SVM" throughout the whole manuscript.
• Section 2.2: The authors have said that there are more details on the dissertation about the data collection was. After reading the dissertation I could not find them. Nevertheless, it will just be necessary to add a reference to the IRB at Virginia Tech, such as a link to the website, where it's possible to read the ethical protocol.
• Please improve the quality of Figure 3.

Author Response

The authors have done a great job improving the manuscript. I found three minor suggestions for the authors. Please check the correctness of "a SVM" or "an SVM" throughout the whole manuscript. We are thankful to the reviewers for their comments. Your comments have helped us significantly improve the quality of manuscript. We have tried our best to provide satisfying responses to the reviewers. Thank you for this comment we have corrected for this error now. Section 2.2: The authors have said that there are more details on the dissertation about the data collection was. After reading the dissertation I could not find them. Nevertheless, it will just be necessary to add a reference to the IRB at Virginia Tech, such as a link to the website, where it's possible to read the ethical protocol. Thank you, we have updated the link to the IRB. The IRB Number is 11-1088, on page 170 of the dissertation. Thank you https://vtechworks.lib.vt.edu/handle/10919/49248 Please improve the quality of Figure 3. Thank you, we have updated it with new figure now.

Round 3

Reviewer 2 Report

Dear authors, my comment on Secction 2.2 was misunderstood. Nevertheless, you have provided the information that is needed : "IRB Number 11-1088". This should be included in the manuscript since researchers will most likely look only to the manuscript and not the dissertation. You should also add the website of the IRB to the reference list "https://www.research.vt.edu/sirc/hrpp.html".

For example (rewrite as the authors think appropriate):

"All participants signed an informed consent form approved by the Institutional Review Board at Virginia Tech (IRB Number 11-1088 [ref#])".

Where [ref#] points to https://www.research.vt.edu/sirc/hrpp.html in the reference list.

Congratulations on your manuscript.

Author Response

Dear authors, my comment on Secction 2.2 was misunderstood. Nevertheless, you have provided the information that is needed : "IRB Number 11-1088". This should be included in the manuscript since researchers will most likely look only to the manuscript and not the dissertation. You should also add the website of the IRB to the reference list "https://www.research.vt.edu/sirc/hrpp.html". For example (rewrite as the authors think appropriate): "All participants signed an informed consent form approved by the Institutional Review Board at Virginia Tech (IRB Number 11-1088 [ref#])". Where [ref#] points to https://www.research.vt.edu/sirc/hrpp.html in the reference list. Congratulations on your manuscript. Thank you for your reviews. We are sorry, we previously misunderstood the comment. We have corrected the reference list and methods section as per your suggestions. We have updated it as “All participants signed an informed consent form approved by the Institutional Review Board at Virginia Tech (IRB Number 11-1088 [33]).” All consecutive references have been updated as well. Thank you so much.

Round 4

Reviewer 2 Report

Great work, congratulations!