A Comparison of Brain-State Representations of Binary Neuroimaging Connectivity Data. Comment on Samantaray et al. Unique Brain Network Identification Number for Parkinson’s and Healthy Individuals Using Structural MRI. Brain Sci. 2023, 13, 1297

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This comment manuscript provided a nice and comprehensive comment on a previously published paper by Samantary et al. on Brain Science 2023 as well as a  comprehensive comparison of the method proposed by Samantary et al., 2023 with other three conventional methods for the binary brain networks matrix storage and pointed out several potential errors and weakness on the original article. The meticulous examination of these methodologies enhances the understanding of the strengths and weaknesses of each approach. This commentary not only appreciates the original work but also contributes valuable insights to the field, making it a commendable addition to the literature.

I have just several minor comments on this manuscript.

1. The original paper Samantary et al. 2023 proposed the so-called UBNIN using structural MRI. Since the structural brain network from the morphology covariation(regional volume). To get the binary adjacent matrix, we need to use a threshold to binarize the original weighted connectivity matrix. The choice of the cut-off threshold can significantly affect the resulting binary matrix. Binarized the original weighted networks can cause information loss in the brain networks, so the methods would be only applicable to the binary networks (as the title in this comment manuscript). As proposed in this manuscript, the author points out the reproducibility and test-retest reliability of binary brain network construction. I suggest the authors add the cutoff threshold issue as well. 

2. In comparison with other conventional methods in Table 1 and Table 2, UBNIN methods do not show significant advantages compared with those methods. The authors should also comment on this part.

3. The description of the construction of individual brain structural morphology networks in Samantary et al. 2023 was ambiguous. The unit of regional volume was not clearly defined and the correlation definition is ambiguous just by two numbers of regional volume. Even though this method has been used in previous articles, the correlation definition still seems irrational.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Comment to the Manuscript ID brainsci-2776676

Authors have submitted the current manuscript having their comments on the report published by Samantaray et al. 

The information provided about the paper on Unique Brain Network Identification Number (UBNIN) and its comparison with other brain-state representations offers valuable insights. Here are some critical comments:

1. Clarity of Presentation:

   - The text provides a detailed and comprehensive explanation of the UBNIN method, its variations, and the comparison with other representations. However, the complexity of the encoding and decoding processes, especially for UBNIN-C, might pose challenges for readers unfamiliar with the specific methodology.

2. Space Efficiency Analysis:

   - The analysis regarding the space efficiency of UBNIN-C compared to Binary, Base 10, and Hexadecimal is well-detailed. However, a more quantitative comparison, perhaps with numerical examples, could enhance the clarity of the argument.

3. Trade-offs in Representations:

   - The discussion on advantages and disadvantages of UBNIN-C, UBNIN-R, and UBNIN-T is insightful. However, the trade-offs between space efficiency and information loss could be more explicitly discussed. Understanding when to use each variant under different circumstances would add depth to the analysis.

4. Error Detection:

   - The built-in error detection capability of UBNIN-C is highlighted as an advantage. A more detailed exploration of the importance and practical implications of error detection in the context of neuroimaging data would enhance the significance of this feature.

5. Practical Applicability:

   - The paper mentions advantages of UBNIN-C, such as error-checking and encoding/decoding procedures. However, the practical applicability of these features in real-world scenarios or clinical applications is not explicitly discussed. Providing examples or scenarios where UBNIN-C's capabilities could be crucial would strengthen the paper.

6. Rounding and Precision:

   - The discussion on rounding UBNIN values raises important questions about the precision of the method. A more thorough exploration of the impact of rounding on the accuracy of decoded values, especially in the context of neuroimaging, would be beneficial.

7. Uniqueness of UBNIN:

   - The commentary on the term "Unique" in UBNIN raises a critical point about the guarantee of uniqueness. A deeper exploration of the practical implications and potential limitations of UBNIN's uniqueness in real-world neuroimaging scenarios would add depth to the analysis.

8. Visual Representation:

   - The comment on Binary being the only representation capable of visually showing brain connections in print is significant. Discussing the importance of visual representation in the context of neuroimaging studies could be valuable.

In conclusion, while the information provided is comprehensive, a more explicit exploration of the practical implications, potential applications, and the real-world significance of the UBNIN method would strengthen the paper. Additionally, making the text more accessible to a broader audience, including those without an extensive background in neuroimaging or data representation, would enhance its impact.

Thanks,

The Reviewer

Author Response

Please see the attachment.

Author Response File: Author Response.pdf