Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Paper: Parkinson's Classification and Feature Extraction from Diffusion Tensor Images 2019 #479

Open
wants to merge 9 commits into
base: 2019
from

Conversation

Projects
None yet
5 participants
@raj-sivakumar
Copy link

commented May 23, 2019

If you are creating this PR in order to submit a draft of your paper,
see http://procbuild.scipy.org/ for logs generated by the build
process.

See the project readme
for more information.

@aweinstein

This comment has been minimized.

Copy link

commented May 27, 2019

Hi: I was asked to review this paper. The following are my comments.

In general the paper is very readable and understandable, and the results are interesting.

Typos:

  • I think the correct spelling is "Levodopa", not "Levdopa."
  • The heading for the reference section is missing.
  • Repeated "or" in sentence "We are thus able to use ..."
  • Caption of the tables says "labeltable1" and "labeltable2".

Comments:

  • It would be great if you can share the code necessary to reproduce the results of the paper.
  • In the introduction, when you say "A DTI record is a four-dimensional ...", explain briefly why DTI includes a time series, considering that white matter can be described by a 3D image.
  • I find this sentence awkward: "We can generalize this premise to machine learning, when model complexity of models, often result in exponential increases in number of computations." Please try to reword it.
  • The paragraph "Similarly to SVD, it is ...", seems disconnected with the text. It is not clear what "it" is. I guess is the Tucker method, but is ambiguous. Perhaps it should not be in a different paragraph.
  • It is not clear what object is represented by a tensor in your experiments. I know it the DTI image, but it would be better if you make this explicit.
  • Experiment II is not clear to me. In which sense you are dealing with a linear dynamical systems? Please explain.
  • Figures and tables are not referenced in the text.
  • I suggest to add a Discussion and Conclusion section.
@alberto-antonietti

This comment has been minimized.

Copy link

commented Jun 8, 2019

Hi! I have been assigned to review this paper too.

I read it before looking at @aweinstein, but I run across the same typos, and I agree with all his comments.

First of all, congratulations on the fascinating work that have important implications for early recognition of PD cases.

In addition to the issues raised by @aweinstein, I would add the following comments:

  • References should be added to the main text when it is appropriate to reference previous works, while they are currently present at the end of the manuscript.
  • The acronyms used should be defined when they are first used, even if they are common knowledge for the topic (e.g., MRI, fMRI, CT, SPECT, etc.).
  • With the sentence: "DTI has been shown to be a promising biomarker in Parkinsonian symptoms and can provide unique insights into brain network connectivity" did you mean that some indexes/features extracted from DTI can be used as biomarkers? Because I am not sure if DTI itself can be a biomarker.
  • Page 2. " a balance between the two analysis", it is not clear to me which are the two analyses that have to be considered among the ones cited above.
  • Page 2. Ax=b can be used to represent uniquely a system of linear equation; this should be specified.
  • Page 2. SVD should be defined before the equation A=USV^T
  • Page 2. VT is miswritten after the equation (the T is not put in superscript).
  • Page 2. The Tucker decomposition is one of the main methods of the paper, and it should be explained, at least briefly (e.g., making use of Figure 1).
  • In the description of both experiments, it is not clear at which point the dimensionality reduction (PCA, LDA, Kernel PCA) is applied.
  • Methods, Experiment II. Some details should be added, since the "scale down" operation is not sufficiently described. What do the Authors mean with scaling down the coronal slice?
  • It has not specified the division of the dataset between training and testing set, and if the reported accuracies and F-measures are computed for the training or the testing phase. If no separation of the dataset has been performed, the Authors should discuss this limitation.
  • Tables 1 and 2: "Kernel PcA" should be "kernel PCA".
  • Page 3. "Compare to present results Cole et al." -> "Compared to present results, Cole et al."

I hope that my comments can help you in improving your manuscript!

@raj-sivakumar

This comment has been minimized.

Copy link
Author

commented Jun 10, 2019

Thanks both for your comments! I will make changes per your recommendation and submit an update this week!

@himito

This comment has been minimized.

Copy link

commented Jun 11, 2019

Hi! I have been assigned to review this paper too.

I read today the paper and I run across the same typos.

I agree with all the comments of @aweinstein and @alberto-antonietti. In addition to the issues raised by them, I would add the following comments:

  • Add a "references" section
  • V^T instead of VT
  • Add the websites of the used python packages.
  • Add fMRI works on Parkinson detention in the Existing Work.
  • I would appreciate if you could use reproducible research methods (e.g., share the code, the datasets, notebooks, etc.)
  • duplication of "the" word
@deniederhut

This comment has been minimized.

Copy link
Member

commented Jun 25, 2019

Hi! Just checking in. @raj-sivakumar, the approval deadline for this paper is coming up fast. Have you had time to address the issue raised by your reviewers?

@raj-sivakumar

This comment has been minimized.

Copy link
Author

commented Jun 25, 2019

Hi! Just checking in. @raj-sivakumar, the approval deadline for this paper is coming up fast. Have you had time to address the issue raised by your reviewers?

Hi @deniederhut Thank you for following up. I am addressing the comments and planning to push tomorrow morning.

@raj-sivakumar

This comment has been minimized.

Copy link
Author

commented Jun 25, 2019

If you had any help to offer on adding bibtex citations though I would appreciate it, I keep running into the same errors that stops the pdf build.

@raj-sivakumar

This comment has been minimized.

Copy link
Author

commented Jun 25, 2019

So as I was attempting to submit I noticed that my table captions that match the format of the examples are stopping the pdf build. This wasn't previously an issue, should I be doing something different from the 00_bibderwalt example?

@deniederhut

This comment has been minimized.

Copy link
Member

commented Jun 26, 2019

So the :bibliography: mybib tag needs to go in your paper, not in the 00_bibderwalt example.

It also looks like your paper is missing the :corresponding: label.

Could I ask you to remove the 00 from your paper's name? So that the file is called sivakumar.rst?

As for the table, the rst format requires the table part to be indented. That said, you only need the rst table or the latex table -- not both.

You're missing a space between the citation display name and the details here in line 421.

Hope this helps!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
You can’t perform that action at this time.