Next Article in Journal
CNN-LSTM Prediction Method for Blood Pressure Based on Pulse Wave
Previous Article in Journal
Deep Learning Techniques for the Classification of Colorectal Cancer Tissue
 
 
Article
Peer-Review Record

Mechanism of Motion Direction Detection Based on Barlow’s Retina Inhibitory Scheme in Direction-Selective Ganglion Cells

Electronics 2021, 10(14), 1663; https://doi.org/10.3390/electronics10141663
by Mianzhe Han 1, Yuki Todo 1,* and Zheng Tang 2,*
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3: Anonymous
Electronics 2021, 10(14), 1663; https://doi.org/10.3390/electronics10141663
Submission received: 12 May 2021 / Revised: 9 July 2021 / Accepted: 10 July 2021 / Published: 13 July 2021
(This article belongs to the Section Bioelectronics)

Round 1

Reviewer 1 Report

Dear authors,

The manuscript is very interesting but there is no discussion. 

There are also some minor errors like:

line 59: Barlow directionally selective gunglion cells (ganglion)

line 92: In this section, we introduce the more complex situation (a more complex)

Author Response

Thank you for your review.

Our discussion is in the final Conclusion part, I amended the title of this part, and then gave more detailed discussion .

We corrected minor errors in the article.

Reviewer 2 Report

The manuscript by Mianzhe Han et al. sets out to simulate and investigate the mechanism of motion direction detection in direction-selective ganglion cells. By simulating the cell sensory in the Barlow’s retina inhibitory scheme, authors have identified a sensory model for the local motion sensing as well as global motion sensing. The model was further confirmed with the large volume image data, and has suggested comparable (even better) performance as the CNN based approaches. In general, this is an interesting topic and the paper is very well written, the presented data is convincing, the methodology is sound, and the experiments are well suited. I very much enjoyed following the experiments and analyses conducted by the authors.

Author Response

Thank you for your review. Thanks for your comments and suggestions.

We check our mistake in line 95 & 62. And gave more detailed explanation in discussion and conclusion.

Reviewer 3 Report

The article is both well written and presents interesting results.  I think the manuscript can be published with very few modifications.  The following suggestions represent minor revisions at most.

-The introduction is very tightly packed, and the style of writing occasionally hides the connections between statements the authors are trying to make. I think a few tricks may improve flow and readability, for example, page 1 line 30 might be a good place for a paragraph break to give the reader a pause between ganglion cells broadly interpreted, and the role of starburst cells specifically

-There are new ideas being presented in the materials and methods section that should have been clearly laid out in the introduction and hypothesis.  For example “null direction from up to

down, which will reject the null stimulus by veto”.  The reader has no concept of these terms yet, and it’s only later in section 1.2 that we learn this figure (1) represents light over the receptive field.  I think all the information is here, but the authors need to rearrange so that an easily readable description of the process precedes the details (null direction, etc).  Likewise, line 81 would be very welcome as extra information in the figure caption.

-Section 1.3 requires better formatting.  Perhaps readability would be improved if the authors treat some of the more mathematical functions as formulas, centered on the page.  Alternately, the list that is currently used could be indented.

-Table 1.  Noise should not be plural

-discussion: I think the results are already interesting, but could be improved further by discussing whether biological vision in relation to the model.  For example, does biological vision also suffer similar accuracy drops with the noise levels used?  Also, motion is a temporal phenomenon, do the authors think a recurrent network would better match the results of their model?

Author Response

Thank you for your review. Thanks for your comments and suggestions, we made the following changes

-We have give a break between page 1 line 30 

-We write more clearly in the introduction part in line 48~50 page 2.

-We have added a more detailed explanation in the caption of Figure 1&2.

-In Section 1.3, we display the list separately to make it more readable

-We amended noises to noise

-In conclusion and discussion, we gave a more detailed explanation of the biological model, RNN, CNN, and our proposed mechanism.

Round 2

Reviewer 1 Report

Dear authors,

There is not a real discussion but a conclusion, in fact there is not any reference.

Author Response

Thank you for your reply.

We added the discussion in the results section.
About the references, we use Barlow’s scheme as the basis to conduct research on the mechanism of direction selection.

Round 3

Reviewer 1 Report

Dear authors,

even if still bare, there is now an outline of a discussion combined to results.

Part of the conclusion section may be moved to "results and discussion" to better nourish the discussion.

References are still lacking. As stated in the "Instructions for Authors", the results should be discussed and interpreted in perspective of previous studies and of the working hypotheses. 

 

Author Response

Dear reviewer:

Thank you for your review and comments.

We expand the discussion part and add previous study in this part. We compare our mechanism and previous study,  point out the deficiencies of previous research and the improvements we have made.

Back to TopTop