Abstract

Medical images captured in less-than-optimal conditions may suffer from quality degradation, such as blur, artifacts, and low lighting, which potentially leads to misdiagnosis. Unfortunately, state-of-the-art medical image enhancement methods face challenges in both high-resolution image quality enhancement and local distinct anatomical structure preservation. To address these issues, we propose a Clinical-oriented High-resolution Lightweight Medical Image Enhancement Network, termed CHLNet, which proficiently addresses high-resolution medical image enhancement, detailed pathological characteristics, and lightweight network design simultaneously. More specifically, CHLNet comprises two main components: 1) High-resolution Assisted Quality Enhancement Network for removing global low-quality factors in high-resolution images thus enhancing overall image quality; 2) High-quality-semantic Guided Quality Enhancement Network for capturing semantic knowledge from high-quality images such that detailed structure preservation is enforced. Moreover, thanks to its lightweight design, CHLNet can be easily deployed on medical edge devices. Extensive experiments on three public medical image datasets demonstrate the effectiveness and superiority of CHLNet over the state-of-the-art.

Links to Paper and Supplementary Materials

Main Paper (Open Access Version): https://papers.miccai.org/miccai-2024/paper/2774_paper.pdf

SharedIt Link: pending

SpringerLink (DOI): pending

Supplementary Material: https://papers.miccai.org/miccai-2024/supp/2774_supp.pdf

Link to the Code Repository

N/A

Link to the Dataset(s)

N/A

BibTex

@InProceedings{Wan_AClinicaloriented_MICCAI2024,
        author = { Wang, Yaqi and Chen, Leqi and Hou, Qingshan and Cao, Peng and Yang, Jinzhu and Liu, Xiaoli and Zaiane, Osmar R.},
        title = { { A Clinical-oriented Lightweight Network for High-resolution Medical Image Enhancement } },
        booktitle = {proceedings of Medical Image Computing and Computer Assisted Intervention -- MICCAI 2024},
        year = {2024},
        publisher = {Springer Nature Switzerland},
        volume = {LNCS 15003},
        month = {October},
        page = {pending}
}


Reviews

Review #1

  • Please describe the contribution of the paper

    (1) The paper introduces CHLNet, a network designed to enhance high-resolution medical images while preserving detailed anatomical structures in a lightweight framework suitable for edge devices. (2) CHLNet addresses quality issues like blur and low lighting, and outperforms existing methods in tests on three public datasets.

  • 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) CHLNet introduces a novel dual-component approach to medical image enhancement. (2) The design of CHLNet is lightweight, making it not only efficient in processing but also suitable for deployment on medical edge devices. (3) The paper demonstrates the clinical feasibility and superiority of CHLNet through extensive experiments on three public medical image datasets.

  • 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 paper demonstrates the technical efficacy of CHLNet through evaluations on public datasets, however, it lacks evaluations or case studies on the actual deployment in clinical settings, or clinical datasets gathered from real clinical environment.

  • 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

    The reviewer recommends the authors to increase the font size of Fig. 1, 2.

  • 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?

    The evaluation results show that the proposed CHLNet outperformed other state-of-the-art works, in three public datasets. However, the improvement is not significant compared to the second-best scores. It might be better to test the proposed CHLNet with one more private, real clinical dataset, to better demonstrate its novelty.

  • 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



Review #2

  • Please describe the contribution of the paper

    AI-based image high-resolution enhancement is challenging. Quality degradation, such as blur, artifacts, and low contrast can obscure anatomical details that are required for sufficient diagnosis. The present paper introduces a two step approach that tackles this problem that includes a High-resolution Assisted Quality Enhancement Network and a High-quality-semantic Guided Quality Enhancement Network. This is applied to three distinct publicly available datasets and superiority over state-of-the-art approaches is shown.

  • 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. Innovative and novel approach to improve quality of high-resolution images
    2. 3 distinct datasets obtained with different techniques
    3. Comparison with existing approaches showing superiority over state-of-the art
    4. Lightweight design and high performance
    5. Well written and presented
  • 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. Details on the images are lacking. Although datasets are publicly available details of imaging procedures and patients characteristics are lacking
  • Please rate the clarity and organization of this paper

    Very Good

  • 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?

    None

  • 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

    Provide details of imaging procedures and patients characteristics in the supplement

  • 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

    Accept — should be accepted, independent of rebuttal (5)

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

    Innovative and well written. Superiority over other approaches in multimodal images. Straightforward application into clinical praxis.

  • 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



Review #3

  • Please describe the contribution of the paper

    The authors introduce CHLNet, a promising clinical-oriented high-resolution medical image enhancement network. Guided by high-quality and high-resolution images, CHLNet ensures global image quality while preserving local anatomical landmarks and disease-related lesion information. Extensive quantitative and qualitative comparative experiments on three medical imaging datasets demonstrate CHLNet’s superiority over previous best methods.

  • 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. Innovation: The High-resolution Medical Image Enhancement, compared to existing methods, enhances images and preserves structures globally and locally through two components. It removes various low-quality factors while capturing global pathological structures and effectively models and transfers rich local high-quality semantic knowledge through knowledge distillation, reconstructing local pathological details in images.
    2. Clear Background Introduction: The background is clearly introduced, with a comprehensive grasp of existing methods.
    3. Detailed Algorithm Description: The algorithm is described clearly and in detail.
    4. Essential Ablation Experiments: The authors provide detailed experimental results for all key components involved in the experiments, including loss functions, key modules, and resolution sizes.
    5. Comprehensive Experimental Evaluation and Comparison: A comprehensive quantitative and qualitative comparison with state-of-the-art medical image enhancement methods is conducted, including traditional image enhancement methods and several deep learning approaches: supervised, unsupervised, semi-supervised, and diffusion model-based methods. Visualizations, heatmaps, and enhanced images of segmentation results for different methods are provided.
    6. Lightweight: CHLNet retains relatively fewer parameters and flops, making it suitable for clinical applications.
    7. Strong Reproducibility: The authors provide detailed code, facilitating the replication of their work by readers.
  • 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. Lack of Computational Complexity Analysis: An analysis of the computational complexity of the proposed method is missing, which is an important consideration for computational resource requirements and real-time demands in practical applications.
    2. Insufficient Discussion: The paper does not sufficiently discuss the limitations of the proposed method, such as potential issues when dealing with specific types of pathological images or particular diseases.
  • 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?

    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

    Please improve your paper based on the comments.

  • 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?

    based on a thorough evaluation of the paper

  • Reviewer confidence

    Very confident (4)

  • [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

Thank you for approving our work. We are deeply grateful for your constructive feedback and will refine the article according to your insightful suggestions.




Meta-Review

Meta-review not available, early accepted paper.



back to top