Paper Info Reviews Author Feedback Meta-Review Back to top
List of Papers Browse by Subject Areas Author List

Abstract

Electroencephalography (EEG)-based attention disorder research seeks to understand brain activity patterns associated with attention. Previous studies have mainly focused on identifying brain regions involved in cognitive processes or classifying Attention-Deficit Hyperactivity Disorder (ADHD) and control subjects. However, analyzing effective brain connectivity networks for specific attentional processes and comparing them has not been explored. Therefore, in this study, we propose multivariate transfer entropy-based connectivity networks for cognitive events and introduce a new similarity measure, “SimBrainNet”, to assess these networks. A high similarity score suggests similar brain dynamics during cognitive events, indicating less attention variability. Our experiment involves 12 individuals with attention disorders (7 children and 5 adolescents). Noteworthy that child participants exhibit lower similarity scores compared to adolescents, indicating greater changes in attention. We found strong connectivity patterns in the left pre-frontal cortex for adolescent individuals compared to the child. Our study highlights the changes in attention levels across various cognitive events, offering insights into the underlying cognitive mechanisms, brain dynamics, and potential deficits in individuals with this disorder.

Links to Paper and Supplementary Materials

Main Paper (Open Access Version): https://papers.miccai.org/miccai-2024/paper/2277_paper.pdf

SharedIt Link: pending

SpringerLink (DOI): pending

Supplementary Material: N/A

Link to the Code Repository

https://github.com/DDasChakladar/SimBrainNet

Link to the Dataset(s)

N/A

BibTex

@InProceedings{Das_SimBrainNet_MICCAI2024,
        author = { Das Chakladar, Debashis and Simistira Liwicki, Foteini and Saini, Rajkumar},
        title = { { SimBrainNet: Evaluating Brain Network Similarity for Attention Disorders } },
        booktitle = {proceedings of Medical Image Computing and Computer Assisted Intervention -- MICCAI 2024},
        year = {2024},
        publisher = {Springer Nature Switzerland},
        volume = {LNCS 15002},
        month = {October},
        page = {pending}
}

Reviews

Review #1

  • Please describe the contribution of the paper

    In this study, the author proposed multivariate transfer entropy-based connectivity networks for cognitive events and introduce a new similarity measure, “SimBrainNet”. They found strong connectivity patterns in the left pre-frontal cortex for adolescent individuals compared to the child.

  • Please list the main strengths of the paper; you should write about a novel formulation, an original way to use data, demonstration of clinical feasibility, a novel application, a particularly strong evaluation, or anything else that is a strong aspect of this work. Please provide details, for instance, if a method is novel, explain what aspect is novel and why this is interesting.

    The paper highlights two novel approaches: (a) constructing MTE-based effective connectivity brain networks for two cognitive events, and (b) building a robust similarity measure (SimBrainNet) that computes the similarity between those brain networks using spatial neighbors of EEG channels.

  • Please list the main weaknesses of the paper. Please provide details, for instance, if you think a method is not novel, explain why and provide a reference to prior work.

    aThe introduction of SimBrainNet as a measure for evaluating similarities in brain networks is a novel contribution. However, the validation of this measure appears to be limited within the study. It would be advantageous to see a more robust validation against established neurophysiological or behavioral benchmarks to substantiate the claim that lower similarity scores correspond to greater attention variability. In additon, they handled small sample size of 12 individuals, segmented into two age group. Further experiments are needed by further utilizing other public DB datasets.

  • Please rate the clarity and organization of this paper

    Satisfactory

  • Please comment on the reproducibility of the paper. Please be aware that providing code and data is a plus, but not a requirement for acceptance.

    The authors claimed to release the source code and/or dataset upon acceptance of the submission.

  • Do you have any additional comments regarding the paper’s reproducibility?

    N/A

  • Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review. Pay specific attention to the different assessment criteria for the different paper categories (MIC, CAI, Clinical Translation of Methodology, Health Equity): https://conferences.miccai.org/2024/en/REVIEWER-GUIDELINES.html

    In this study, the author proposed multivariate transfer entropy-based connectivity networks for cognitive events and introduce a new similarity measure, “SimBrainNet”. They found strong connectivity patterns in the left pre-frontal cortex for adolescent individuals compared to the child. The results are interesting. However, some critical comments need to be addressed to further improve the manuscript.

    • Please provide more specific technical reason into the decision to use MTE over other potential entropy-based measures? What specific advantages does MTE offer in the context of analyzing brain networks for attention disorders compared to other entropy measures such as simple Transfer Entropy or Conditional Entropy?

    • Please discuss the rationale on the criteria used for selecting the 13 EEG channels based on the Brodmann areas? How do these specific areas relate to the cognitive tasks performed in the study, and what is the justification for excluding other regions?

    • In order to prove the supority of the proposed network, an experiment was conducted using a public dataset. However, it is a small number of data sample to prove novelty due to using only 12 subjects. Further experiments are needed by further utilizing other public DB datasets.

    • The authors mentioned that the EEG artifacts were removed using ICA. Which artifacts were removed? Eye movement? Body movement? Then, which channel was used as a reference to extract the Independent Component?

    • And, is there a lot of noise when performing the “surround suppression” task? It is necessary to write more clearly for EEG pre-processing.

    • What is the computational cost as this brain network similarity is trained and predicted?

    • What are the next steps for further development of SimBrainNet? Are there specific aspects of the model that you plan to improve or expand? How might future research better integrate this similarity measure into broader neuroscientific or clinical research frameworks?

    • The authors mentioned the significant differences in MTE connectivities between two age groups using t-tests. What about without MTE connectivity? and Please describe about statistical analysis such as kinds of t-test, normailty test, homoskedasticity, t-value…it is few samples..

    • It is recommended to make a Discussion section to discuss the proposed technique and experimental results, study limitations, and future work.

  • Rate the paper on a scale of 1-6, 6 being the strongest (6-4: accept; 3-1: reject). Please use the entire range of the distribution. Spreading the score helps create a distribution for decision-making

    Reject — should be rejected, independent of rebuttal (2)

  • Please justify your recommendation. What were the major factors that led you to your overall score for this paper?

    Novelty, experiements, evaluation

  • Reviewer confidence

    Confident but not absolutely certain (3)

  • [Post rebuttal] After reading the author’s rebuttal, state your overall opinion of the paper if it has been changed

    Weak Accept — could be accepted, dependent on rebuttal (4)

  • [Post rebuttal] Please justify your decision

    In the revised manuscript, the authors answered the questions which I concerned. I really appreciate the authors reflecting on my comments carefully.



Review #2

  • Please describe the contribution of the paper

    They propose a multivariate transfer entropy-based connectivity network based on 13 EEG channels. Using a public EEG dataset, they analyzed 12 individuals with attention disorders (7 children and 5 adolescents). They found a higher similarity score in the adolescents group than in the children’s group, which they though was due to greater changes in attention in the younger participants.

  • Please list the main strengths of the paper; you should write about a novel formulation, an original way to use data, demonstration of clinical feasibility, a novel application, a particularly strong evaluation, or anything else that is a strong aspect of this work. Please provide details, for instance, if a method is novel, explain what aspect is novel and why this is interesting.

    Interesting idea. Got some positive results.

  • Please list the main weaknesses of the paper. Please provide details, for instance, if you think a method is not novel, explain why and provide a reference to prior work.

    The methodology could have been better described. They also discussed that 15 subjects with attention disorder were selected randomly from each version based on age groups of G1 and G2, but that doesn’t say whether these subjects are the same as the 12 used in the model, or different.

    Small sample size.

    The authors also need to review their papers in order to check from grammatical mistakes

  • Please rate the clarity and organization of this paper

    Satisfactory

  • Please comment on the reproducibility of the paper. Please be aware that providing code and data is a plus, but not a requirement for acceptance.

    The authors claimed to release the source code and/or dataset upon acceptance of the submission.

  • Do you have any additional comments regarding the paper’s reproducibility?

    N/A

  • Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review. Pay specific attention to the different assessment criteria for the different paper categories (MIC, CAI, Clinical Translation of Methodology, Health Equity): https://conferences.miccai.org/2024/en/REVIEWER-GUIDELINES.html

    Interesting idea to look at a multivariate transfer entropy-based connectivity network based on 13 EEG channels in Attention-Deficit Hyperactivity Disorder. Using a public dataset, they were able to show similarity scores with some differences between the children and the adolescents groups.

  • Rate the paper on a scale of 1-6, 6 being the strongest (6-4: accept; 3-1: reject). Please use the entire range of the distribution. Spreading the score helps create a distribution for decision-making

    Weak Accept — could be accepted, dependent on rebuttal (4)

  • Please justify your recommendation. What were the major factors that led you to your overall score for this paper?

    It seems that 12 (or 15?) individuals is a small dataset, especially if it is using a public dataset where you would expect more subjects to be available. The methodology could be better described.

  • Reviewer confidence

    Somewhat confident (2)

  • [Post rebuttal] After reading the author’s rebuttal, state your overall opinion of the paper if it has been changed

    Weak Accept — could be accepted, dependent on rebuttal (4)

  • [Post rebuttal] Please justify your decision

    Authors have addressed some points from reviewers



Review #3

  • Please describe the contribution of the paper

    This paper proposes a novel method called “SimBrainNet” to evaluate the similarity between brain connectivity networks derived from electroencephalography (EEG) data. The method is applied to assess attention disorders in children and adolescents by comparing their brain networks during cognitive events. The similarity score provides insights into the underlying cognitive mechanisms and brain dynamics related to attention deficits.

  • Please list the main strengths of the paper; you should write about a novel formulation, an original way to use data, demonstration of clinical feasibility, a novel application, a particularly strong evaluation, or anything else that is a strong aspect of this work. Please provide details, for instance, if a method is novel, explain what aspect is novel and why this is interesting.
    1. The proposed “SimBrainNet” method is a novel formulation for quantifying the similarity between brain connectivity networks. It uses a combination of substitution, insertion, and deletion algorithms to transform one network into another, accounting for differences in node and edge structures.
    2. It is interesting to note that this method has been shown to make a difference in assessing attention disorders in children and adolescents. By comparing brain networks during cognitive events, this method provides insight into the underlying cognitive mechanisms involved in attention variability and attention deficits.
  • Please list the main weaknesses of the paper. Please provide details, for instance, if you think a method is not novel, explain why and provide a reference to prior work.
    1. The paper does not provide a comprehensive comparison with existing brain network similarity measures, such as those mentioned in the related work section (e.g., SimiNet [Mheich et al., 2017], graph matching-based methods [Osmanlıoğlu et al., 2018]). A more detailed comparison and discussion of the advantages and limitations of the proposed method would strengthen the novelty claim.
    2. While the clinical feasibility of the proposed method is demonstrated on attention disorder data, the paper does not provide a clear pathway or discussion on how the findings could be translated into clinical practice or interventions. Additional discussion on potential clinical implications and future directions would be beneficial.
    3. The paper does not provide a detailed analysis or discussion of the computational complexity and scalability of the proposed method, particularly for larger brain networks or higher-density EEG data. This aspect could be a potential limitation and should be addressed.
  • Please rate the clarity and organization of this paper

    Satisfactory

  • Please comment on the reproducibility of the paper. Please be aware that providing code and data is a plus, but not a requirement for acceptance.

    The submission does not mention open access to source code or data but provides a clear and detailed description of the algorithm to ensure reproducibility.

  • Do you have any additional comments regarding the paper’s reproducibility?

    While providing the code is not a strict requirement, making the code openly available would greatly facilitate understanding and replicating your proposed method. I strongly encourage you to consider open-sourcing your implementation, as it would substantially increase the impact and adoption of your contributions by the research community.

  • Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review. Pay specific attention to the different assessment criteria for the different paper categories (MIC, CAI, Clinical Translation of Methodology, Health Equity): https://conferences.miccai.org/2024/en/REVIEWER-GUIDELINES.html
    1. Provide a more comprehensive comparison with existing brain network similarity measures, such as SimiNet [Mheich et al., 2017] and graph matching-based methods [Osmanlıoğlu et al., 2018]. A detailed discussion of the advantages and limitations of your method compared to these existing approaches would help to better establish its novelty and potential impact.
    2. Consider expanding the discussion on the clinical translation and potential implications of your findings. While you have demonstrated the feasibility of your method on attention disorder data, it would be beneficial to provide more insights into how these findings could be translated into clinical practice or inform interventions for patients with attention deficits.
    3. Address the computational complexity and scalability of your method, particularly for larger brain networks or higher-density EEG data. Provide an analysis or discussion on the potential limitations or challenges in terms of computational resources and time complexity.
    4. Consider the possibility of releasing the code for your method as open-source. This could not only enhance the reproducibility of your work but also facilitate wider adoption and potential extensions by the research community.
  • Rate the paper on a scale of 1-6, 6 being the strongest (6-4: accept; 3-1: reject). Please use the entire range of the distribution. Spreading the score helps create a distribution for decision-making

    Weak Accept — could be accepted, dependent on rebuttal (4)

  • Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
    1. The proposed “SimBrainNet” method is novel and provides a unique approach to quantifying brain network similarity, which is an important problem in neuroimaging and cognitive neuroscience.
    2. The application to attention disorders in children and adolescents is relevant and addresses an important clinical problem. The results demonstrate the potential of the method in capturing age-related differences in attention levels and underlying brain dynamics.
  • Reviewer confidence

    Somewhat confident (2)

  • [Post rebuttal] After reading the author’s rebuttal, state your overall opinion of the paper if it has been changed

    N/A

  • [Post rebuttal] Please justify your decision

    N/A



Review #4

  • Please describe the contribution of the paper

    1.Multivariate Transfer Entropy (MTE)-based Connectivity Networks: It introduces a novel approach for constructing effective brain connectivity networks specific to different cognitive events using MTE. This allows for a more nuanced understanding of how brain regions interact during different attentional tasks. 2.SimBrainNet Similarity Measure: The study proposes a new similarity measure, SimBrainNet, which utilizes the spatial relationships between EEG channels to assess the similarity between these MTE-based connectivity networks. A high SimBrainNet score indicates similar brain dynamics during cognitive events, potentially reflecting less attention variability.

  • Please list the main strengths of the paper; you should write about a novel formulation, an original way to use data, demonstration of clinical feasibility, a novel application, a particularly strong evaluation, or anything else that is a strong aspect of this work. Please provide details, for instance, if a method is novel, explain what aspect is novel and why this is interesting.

    1.Focus on effective connectivity: By analyzing effective connectivity, the study goes beyond simply identifying active brain regions, providing a deeper understanding of the functional interactions between these regions during attention tasks. 2.Novel methodology: The introduction of MTE-based connectivity networks and the SimBrainNet similarity measure offers unique tools for investigating attention-related brain dynamics. 3.Preliminary findings: The study demonstrates the potential of the proposed methods by revealing differences in connectivity patterns between children and adolescents with attention disorders, suggesting a link to changes in attention levels. 4.Good reproducibility: The detailed description of the methodology and data analysis enhances the reproducibility of the research.

  • Please list the main weaknesses of the paper. Please provide details, for instance, if you think a method is not novel, explain why and provide a reference to prior work.

    Sample size: While the current study provides valuable insights, a larger sample size could further strengthen the generalizability of the findings.

  • Please rate the clarity and organization of this paper

    Excellent

  • Please comment on the reproducibility of the paper. Please be aware that providing code and data is a plus, but not a requirement for acceptance.

    The submission has provided an anonymized link to the source code, dataset, or any other dependencies.

  • Do you have any additional comments regarding the paper’s reproducibility?

    Nothing

  • Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review. Pay specific attention to the different assessment criteria for the different paper categories (MIC, CAI, Clinical Translation of Methodology, Health Equity): https://conferences.miccai.org/2024/en/REVIEWER-GUIDELINES.html

    Visualization tools: Developing user-friendly visualization tools for SimBrainNet could facilitate broader adoption of the method by researchers.

  • Rate the paper on a scale of 1-6, 6 being the strongest (6-4: accept; 3-1: reject). Please use the entire range of the distribution. Spreading the score helps create a distribution for decision-making

    Strong Accept — must be accepted due to excellence (6)

  • Please justify your recommendation. What were the major factors that led you to your overall score for this paper?

    This study presents a novel approach for analyzing effective brain connectivity related to attention. The proposed methods hold promise for advancing the field of EEG-based attention research. The additional considerations regarding sample size and visualization tools could be addressed in future work.

  • Reviewer confidence

    Confident but not absolutely certain (3)

  • [Post rebuttal] After reading the author’s rebuttal, state your overall opinion of the paper if it has been changed

    N/A

  • [Post rebuttal] Please justify your decision

    N/A



Author Feedback

Thanks to all reviewers for their valuable comments. 1.[Reviewers #5, #6]: The computational complexity and computational resources/time complexity of the proposed model

  • The overall complexity of SimBrainNet is O(n.m.e), where n, m, and e represent the nodes in G1, G2, and edges in the graphs. Making an MTE graph takes more computational time than SimBrainNet due to the high-dimensional EEG data and intensive probability calculations. After removing non-significant EEG channels using MTE, lower-dimensional MTE graphs are used as input for SimBrainNet. So, the complexity of SimBrainNet will not be increased much with higher-density EEG data. Each MTE graph took 8-10 minutes to create using an 8GB GPU and 32GB of RAM. However, we anticipate decreased graph creation time with more powerful hardware. 2.[Reviewers #5, #6]: How research findings of proposed model could be translated into clinical practice? -The similarity score obtained from the SimBrainNet can enhance the accuracy of ADHD diagnosis and enable the creation of personalized treatment plans tailored to individual needs. These insights can inform the development of neurofeedback protocols and cognitive training programs, particularly age-specific interventions that focus on enhancing prefrontal cortex connectivity. 3.[Reviewers #5, #6]: Discussion section 3.1: Com 1, Reviewer #5 (Comparison of SimBrainNet with existing similarity measures-based brain networks) -Our MTE connectivity graphs are used as inputs for SimBrainNet. MTE connectivity graph takes additional computational time which is not required in other similarity networks [16,20]. However, estimating information flow among brain regions through the MTE graph is crucial for understanding cognitive dynamics. While SimiNet [16] highlights differences between brain networks associated with different picture categories, and the graph matching-based method [20] highlights structural changes in traumatic and control patients, the proposed SimBrainNet emphasizes behavioral changes in attention disorder individuals with different age groups. Thus, our model shows potential for clinical and cognitive applications by measuring human behavioral changes using EEG events. 3.2: Com 9, Reviewer #6 (Limitations & Future works) -Limitations: The size of the experimental dataset is small. Future Work: In the near future, we will evaluate the performance of SimBrainNet using other large-sized open-access ADHD datasets.
    1. [Com 3, Reviewers #6]+ [Weakness of Reviewers #3,#4]: Small sample size of the experimental dataset
  • We used 12 subjects for our experiment. In the near future, we will evaluate our proposed model using ADHD datasets with more number of subjects. Reviewer #4 Com1 (Visualization tools) -In future, we will develop user-friendly visualization tools for SimBrainNet to facilitate broader adoption of the method by researchers. Reviewer #6 Com 1 (Advantages of MTE over other entropy-based measures)
  • Transfer Entropy often leads to spurious connectivity results due to common information from multiple sources. MTE overcomes this issue and provides better information flow in the brain network. Com 2 (Channel selection)
  • Cortical projections of the EEG channels (10-10 electrode placement system) to the Brodmann areas are performed based on the anatomical correlation method. The 13 EEG channels are selected based on the ADHD-specific brain regions mentioned in existing studies [2,13,21]. Com 4,5 (ICA, Preprocessing of EEG)
  • We have removed the eye and muscle movements from EEG using ICA. Pz is used as a reference channel to extract the Independent Component. After preprocessing, we got a smoother signal compared to the raw EEG. Com 8 (Statistical tests without MTE)
  • We didn’t use other connectivity metrics apart from MTE for SimBrainNet, so we didn’t perform statistical tests without MTE connectivity. Reproducibility: We will publish the source code after acceptance of the paper.



Meta-Review

Meta-review #1

  • After you have reviewed the rebuttal and updated reviews, please provide your recommendation based on all reviews and the authors’ rebuttal.

    Accept

  • Please justify your recommendation. You may optionally write justifications for ‘accepts’, but are expected to write a justification for ‘rejects’

    The rebuttal does a good job in addressing the noted concerns. All reviewers are positive about the paper.

  • What is the rank of this paper among all your rebuttal papers? Use a number between 1/n (best paper in your stack) and n/n (worst paper in your stack of n papers). If this paper is among the bottom 30% of your stack, feel free to use NR (not ranked).

    The rebuttal does a good job in addressing the noted concerns. All reviewers are positive about the paper.



Meta-review #2

  • After you have reviewed the rebuttal and updated reviews, please provide your recommendation based on all reviews and the authors’ rebuttal.

    Accept

  • Please justify your recommendation. You may optionally write justifications for ‘accepts’, but are expected to write a justification for ‘rejects’

    N/A

  • What is the rank of this paper among all your rebuttal papers? Use a number between 1/n (best paper in your stack) and n/n (worst paper in your stack of n papers). If this paper is among the bottom 30% of your stack, feel free to use NR (not ranked).

    N/A



back to top