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

Abstract

Histology slide digitization is becoming essential for telepathology (remote consultation), knowledge sharing (education), and using the state-of-the-art artificial intelligence algorithms (augmented/automated end-to-end clinical workflows). However, the cumulative costs of digital multi-slide high-speed brightfield scanners, cloud/on-premises storage, and personnel (IT and technicians) make the current slide digitization workflows out-of-reach for limited-resource settings, further widening the health equity gap; even single-slide manual scanning commercial solutions are costly due to hardware requirements (high-resolution cameras, high-spec PC/workstation, and support for only high-end microscopes). In this work, we present a new cloud slide digitization workflow for creating scanner-quality whole-slide images (WSIs) from uploaded low-quality videos, acquired from cheap and inexpensive microscopes with built-in cameras. Specifically, we present a pipeline to create stitched WSIs while automatically deblurring out-of-focus regions, upsampling input 10X images to 40X resolution, and reducing brightness/contrast and light-source illumination variations. We demonstrate the WSI creation efficacy from our workflow on World Health Organization-declared neglected tropical disease, Cutaneous Leishmaniasis (prevalent only in the poorest regions of the world and only diagnosed by sub-specialist dermatopathologists, rare in poor countries), as well as other common pathologies on core biopsies of breast, liver, duodenum, stomach and lymph node. Upon acceptance, we will release our code, datasets, pretrained models, and cloud platform for uploading microscope videos and downloading/viewing WSIs with shareable links (no sign-in required) for telepathology and knowledge sharing.

Links to Paper and Supplementary Materials

Main Paper (Open Access Version): https://papers.miccai.org/miccai-2024/paper/2279_paper.pdf

SharedIt Link: pending

SpringerLink (DOI): pending

Supplementary Material: N/A

Link to the Code Repository

https://github.com/nadeemlab/DeepLIIF

Link to the Dataset(s)

https://github.com/nadeemlab/DeepLIIF

BibTex

@InProceedings{Zeh_Rethinking_MICCAI2024,
        author = { Zehra, Talat and Marino, Joseph and Wang, Wendy and Frantsuzov, Grigoriy and Nadeem, Saad},
        title = { { Rethinking Histology Slide Digitization Workflows for Low-Resource Settings } },
        booktitle = {proceedings of Medical Image Computing and Computer Assisted Intervention -- MICCAI 2024},
        year = {2024},
        publisher = {Springer Nature Switzerland},
        volume = {LNCS 15004},
        month = {October},
        page = {pending}
}

Reviews

Review #1

  • Please describe the contribution of the paper

    This paper demonstrates a pathology workflow using a low cost components, and focusing on disease prevalent in developing world. The key idea is to perform stitching of frames from videos, and them image restoration to restore images from low-cost component to be on par with high quality component. Some paired, low-cost - high-cost images were used to train the image restoration model. Some qualitative analysis of the resulting image on diagnostic tasks shows mixed results.

  • 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 problems faced by a low resource environment were comprehensively considered by the authors. Digital pathology requires many expensive components to implement effectively, and the paper has done a reasonable job at focusing on utilizing cheaper components while generating high quality images.
    • The implementation of the solution is convincing, and the result looks reasonable to my non-pathologist eyes.
    • The paper is well-written and relatively easy to follow.
  • 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.
    • My main concern is that the evaluation of the pipeline is still relatively lacking. Most of the evaluation is how well the image restoration works. However, to really evaluate the effectiveness of the low cost pipeline, the authors should have focused on a few specific diagnostic tasks, and compare pathologist performance on each platform, similar to what figure 4 is showing (though fig 4 is showing that low cost component works just as well as high end, and therefore image restoration is not necessary).
    • Since cost is an objective measure here, it could be helpful to estimate the cost for both high end and low-end pipeline. There are only a few mentioning of component costs, but the technicians’ and pathologists’ times have not been considered.
  • 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 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

    Please see the weakness section for questions.

  • 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?

    Despite looking at an important problem and relatively well-execution of the idea, I think the paper evaluation still misses the points, so I am recommending a weak accept.

  • 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



Review #2

  • Please describe the contribution of the paper

    Describes a free open-access cloud-based slide digitization workflow to create, view, and share scanner-quality stitched whole-slide images from uploaded 10X manual scan videos, acquired from low cot microscopes with built-in digital cameras. Could be used in low-resource areas.

  • 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.

    Innovative idea on how to improve access to WSI technology and related benefits and seems like could truly be done at low cost. Results are early but positive.

  • 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.

    Number of cases very small limiting statistical power. Although use traditional metrics like calculated similarity indices etc. but no observer studies using clinicians who would have to use these images and/or whether AI scheme performance would be impacted in any way.

  • 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 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?

    no

  • 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

    Number of cases very small limiting statistical power. Although use traditional metrics like calculated similarity indices etc. but no observer studies using clinicians who would have to use these images and/or whether AI scheme performance would be impacted in any way.

  • 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?

    Solid innovative idea with practical possible solution for low resource users. Good data but need to acknowledge low sample size.

  • 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



Review #3

  • Please describe the contribution of the paper

    The paper presents a significant contribution to the field of histology digitization by introducing a cost-effective workflow utilizing a low-cost yet clinically viable microscope. By addressing the expense associated with digitizing histology slides, the authors tackle a crucial barrier in global healthcare, particularly in regions with limited access to expert dermatopathologists. Focusing on the diagnosis of Cutaneous Leishmaniasis, a neglected tropical disease, they highlight the urgent need for accessible diagnostic tools. The proposed workflow facilitates telepathology and enhances disease detection and patient care. The integration of a deep learning approach for deblurring, upsampling, and reducing stain variations further strengthens the methodology’s efficacy. Through meticulous experimentation and comparison with recent methods, the authors demonstrate the feasibility and effectiveness of their approach, offering a promising solution to enhance digitization and diagnoses.

  • 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 paper is commendably well-motivated, structured, and consistently clear in its presentation, facilitating a comprehensive understanding of the research objectives and findings.

    2. Addressing the cost-effectiveness challenge in histology digitization, particularly significant for underprivileged regions, underscores the authors’ commitment to advancing global healthcare. Their development of a low-cost solution holds promise for widespread adoption and the advancement of telepathology.

    3. Drawing attention to Cutaneous Leishmaniasis as a poignant example of a disease prevalent in impoverished areas, the authors effectively highlight the urgent need for accessible diagnostic tools. This aligns with their endeavor to bridge healthcare disparities exacerbated by the scarcity of specialized dermatopathologists.

    4. The integration of a deep learning approach for deblurring, upsampling, and stain variation reduction is rigorously evaluated against three established methods for image restoration. Notably, the inclusion of error bars enhances the robustness of performance assessment, affirming the reliability of the proposed methodology.

    5. The authors adeptly visualize the performance metrics of each model, enhancing the clarity and accessibility of the results, thereby facilitating comprehension and interpretation.

    6. A commendable aspect of the paper is the thorough acknowledgment of its limitations, coupled with proposed avenues for future research. This demonstrates a conscientious approach to scientific inquiry and sets a precedent for ongoing refinement and advancement in the field.

  • 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 introduces two key contributions, the first being the generation of scanner-quality Whole Slide Images (WSIs) from low-quality videos. However, there’s a lack of detailed information on the practical deployment of this approach in resource-constrained hospital settings, which is crucial given the paper’s emphasis on adapting slide digitization workflows for such environments. While the utilization of a low-cost microscope is highlighted, additional insights into the implementation of this workflow within histology labs to facilitate telepathology would enhance the practicality and applicability of the proposed solution.

    2. The second aspect of the paper focuses on improving image quality through deblurring, upsampling, and reducing contrast variations. While this methodology isn’t novel, as evidenced by recent work (Rong, Ruichen, et al., 2023), the application of these techniques to a new dataset represents a novel contribution. However, the paper could benefit from a discussion on potential failure modes of this method. It would be valuable to explore scenarios in which this method may not yield optimal results and outline strategies for addressing these errors to ensure the reliability and robustness of the proposed approach.

  • 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 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

    1. The paper’s emphasis on adapting slide digitization workflows for low-resource settings is commendable. However, to enhance the practical utility of the proposed solution, it would be beneficial for the authors to provide more detailed guidance on the implementation of their workflow within histology labs in countries with limited resources. This could include insights into infrastructure requirements, potential challenges, and strategies for overcoming barriers to adoption.

    2. While the paper highlights the potential benefits of the proposed methodology for telepathology and disease diagnosis, a more thorough evaluation of its clinical feasibility would strengthen the overall contribution. This could involve conducting pilot studies or collaborating with healthcare institutions in underserved regions to assess the practicality and effectiveness of the workflow in real-world settings. Additionally, incorporating feedback from pathologists and other healthcare professionals could provide valuable insights into the usability and impact of the proposed solution.

    3. The authors have effectively acknowledged the limitations of their methodology, particularly regarding image quality improvement techniques. To further enhance the reliability of their approach, it would be beneficial for the authors to explore potential failure modes of the image restoration method and develop strategies for error handling. This could involve conducting sensitivity analyses or identifying common scenarios in which the method may not yield optimal results, thereby informing the development of robust error mitigation strategies.

    4. While the authors have compared their methodology with recent methods for image restoration, providing a more comprehensive comparison with existing solutions would strengthen the paper’s contribution. This could involve benchmarking the proposed workflow against a broader range of state-of-the-art techniques and discussing the relative advantages and limitations of each approach. Additionally, highlighting the unique strengths of the proposed methodology compared to existing solutions would provide valuable context for readers and enhance the paper’s impact.

    5. The authors have briefly mentioned potential avenues for future research, such as addressing limitations and expanding the application of their methodology. To enrich the discussion and provide actionable insights for the research community, it would be beneficial for the authors to elaborate on these future directions and their implications. This could include outlining specific research questions, proposing experimental designs, and discussing the potential impact of addressing these challenges on the field of digital pathology and healthcare delivery.

  • 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?

    The paper receives a positive overall score recommendation due to its clear organization, significant contribution to addressing histology digitization challenges in low-resource settings, and thoughtful acknowledgment of limitations and future directions. However, a more thorough evaluation of the methodology’s clinical feasibility and robustness, along with a broader comparison with existing solutions, would further enhance the paper’s impact and rigor. Overall, the paper presents valuable insights and potential avenues for advancing digital pathology and improving healthcare access worldwide.

  • 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

N/A



Meta-Review

Meta-review not available, early accepted paper.



back to top