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

Abstract

Positron emission tomography (PET) is widely recognized as the most sensitive molecular imaging modality, enabling the in vivo visualization of molecular pathways. Despite its exceptional utility, concerns about ionizing radiation expo-sure have limited its broader application. A recent breakthrough in total-body PET imaging addresses this limitation by significantly increasing geometric cov-erage and sensitivity. This innovation reduces radiation exposure to levels compa-rable to the dose received during a transatlantic flight, achieved through advanced computational techniques. To accelerate progress in this field, we have curated a benchmark dataset specifically designed for developing ultra-low dose PET imag-ing methodologies. This dataset was pivotal in the Ultra-Low Dose PET Imaging Challenge held in 2022, 2023, and 2024. The challenge aimed to foster innovative computational algorithms capable of recovering high-quality imaging from low-dose scans acquired on total-body PET systems. The dataset includes both stand-ard-dose and simulated low-dose total-body PET images from 1,447 patients. These were acquired using Siemens Biograph Vision Quadra PET/CT and Unit-ed Imaging uExplorer PET/CT scanners. In addition, we have developed a cus-tomized evaluation system to assess the performance of algorithms in recovering image quality from low-dose scans. This paper provides a comprehensive de-scription of the benchmark dataset and evaluation framework, aimed at driving future advancements in ultra-low dose PET imaging. The dataset is available at https://udpet-challenge.github.io, subject to the completion of a signed Data Transfer Agreement.

Links to Paper and Supplementary Materials

Main Paper (Open Access Version): https://papers.miccai.org/miccai-2025/paper/0232_paper.pdf

SharedIt Link: Not yet available

SpringerLink (DOI): Not yet available

Supplementary Material: Not Submitted

Link to the Code Repository

https://udpet-challenge.github.io/

Link to the Dataset(s)

https://udpet-challenge.github.io/

BibTex

@InProceedings{XueSon_UDPET_MICCAI2025,
        author = { Xue, Song and Wang, Hanzhong and Chen, Yizhou and Liu, Fanxuan and Zhu, Hong and Viscione, Marco and Guo, Rui and Rominger, Axel and Li, Biao and Shi, Kuangyu},
        title = { { UDPET: Ultra-low Dose PET Imaging Challenge Dataset } },
        booktitle = {proceedings of Medical Image Computing and Computer Assisted Intervention -- MICCAI 2025},
        year = {2025},
        publisher = {Springer Nature Switzerland},
        volume = {LNCS 15972},
        month = {September},
        page = {616 -- 623}
}

Reviews

Review #1

  • Please describe the contribution of the paper

    This paper introduces UDPET, the largest benchmark to date for evaluating low-dose PET imaging methods. The dataset comprises 1,447 paired total-body FDG PET scans (low-dose and high-dose) acquired from two different scanner vendors. Additionally, the authors developed a dedicated metric for quantitative assessment of imaging quality.

  • Please list the major strengths of the paper: you should highlight 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 largest dataset for low-dose PET image enhancement/reconstruction.
    2. Total-body FDG PET images acuiqred by different vendors with varying total doses.
    3. A common evaluation metric for the comparisons across different methods.
  • Please list the major weaknesses of the paper. Please provide details: for instance, if you state that a formulation, way of using data, demonstration of clinical feasibility, or application is not novel, then you must provide specific references to prior work.

    1. Although the largest benchmark, the images were provided without sinogram data, which may hamper its usage in developing reconstruction methods.

    2. The dataset only contains PET images with the 18F-FDG tracers, which may hamper its usage in developing reconstruction/enhancement methods with promising generalizability.

    3. The designed evaluation metric contains varying weights across different components, while without convincing explanation and justification.

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

  • Optional: If you have any additional comments to share with the authors, please provide them here. Please also refer to our Reviewer’s guide on what makes a good review and pay specific attention to the different assessment criteria for the different paper categories: https://conferences.miccai.org/2025/en/REVIEWER-GUIDELINES.html

    N/A

  • 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.

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

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

    UDPET undoubtedly represents a valuable contribution to the development of low-dose PET imaging methods. However, the current dataset has several limitations, including the absence of real sinograms, the lack of diverse radioactive tracers, and insufficient detail regarding the proposed evaluation metric.

    Additionally, Figure 1 could be enhanced by including sample images from the uExplorer scanner to better illustrate inter-scanner variability.

  • Reviewer confidence

    Very confident (4)

  • [Post rebuttal] After reading the authors’ rebuttal, please state your final opinion of the paper.

    N/A

  • [Post rebuttal] Please justify your final decision from above.

    N/A



Review #2

  • Please describe the contribution of the paper
    • The authors introduce a first-of-its-kind dataset specifically for ultra-low dose PET imaging. This is a significant contribution in the image reconstruction domain where data scarcity has hindered progress.
    • The hybrid evaluation strategy, integrating global metrics (NRMSE, PSNR, SSIM) and localized SUV-based radiomics, is well thought-out and clinically relevant.
  • Please list the major strengths of the paper: you should highlight 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 authors introduce a first-of-its-kind dataset specifically for ultra-low dose PET imaging. This is a significant contribution in the image reconstruction domain where data scarcity has hindered progress.
    • The hybrid evaluation strategy, integrating global metrics (NRMSE, PSNR, SSIM) and localized SUV-based radiomics, is well thought-out and clinically relevant.
  • Please list the major weaknesses of the paper. Please provide details: for instance, if you state that a formulation, way of using data, demonstration of clinical feasibility, or application is not novel, then you must provide specific references to prior work.
    • The lack of included CT scans, although justified, limits the development of multimodal methods.
    • Is there any related work or reference to support this statement ‘Notably, algorithms developed with total-body PET data can potentially be adapted for use with older PET systems, paving the way for advancements over the next decade’?
    • It would be great that the authors provide a few baseline results of some basic reconstruction methods including AI-based and non-AI based models.
  • Please rate the clarity and organization of this paper

    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.

  • Optional: If you have any additional comments to share with the authors, please provide them here. Please also refer to our Reviewer’s guide on what makes a good review and pay specific attention to the different assessment criteria for the different paper categories: https://conferences.miccai.org/2025/en/REVIEWER-GUIDELINES.html

    N/A

  • 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.

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

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

    The manuscript introduces a valuable resource for the PET imaging and AI community.

  • Reviewer confidence

    Very confident (4)

  • [Post rebuttal] After reading the authors’ rebuttal, please state your final opinion of the paper.

    N/A

  • [Post rebuttal] Please justify your final decision from above.

    N/A



Review #3

  • Please describe the contribution of the paper

    The main contribution of this paper is the introduction of a large-scale, multi-center, and multi-scanner benchmark dataset for ultra-low dose PET imaging, specifically designed to support the development and evaluation of computational methods for image quality recovery. The dataset includes 1,447 patients scanned with two state-of-the-art total-body PET/CT systems, with both standard-dose and simulated low-dose images at multiple dose reduction factors. Additionally, the authors present a customized evaluation framework that combines global image similarity metrics and clinically relevant local metrics, addressing the unique requirements of PET imaging. This dataset has already served as the foundation for an international challenge series (UDPET), and its public release (upon DTA) is expected to catalyze further research and clinical translation in ultra-low dose PET imaging.

  • Please list the major strengths of the paper: you should highlight 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. Large and Diverse Dataset: The dataset includes a substantial number of subjects (1,447) from two leading total-body PET/CT systems, enhancing its representativeness and potential for generalization.
    2. Multi-Dose Levels: Simulated low-dose images at five dose reduction factors enable comprehensive benchmarking of algorithms across a range of clinically relevant scenarios.
    3. Perfect Image Alignment: The use of list-mode data for low-dose simulation ensures perfect spatial alignment between low- and standard-dose images, which is critical for supervised learning and evaluation.
    4. Comprehensive Evaluation Framework: The proposed evaluation metrics combine global image quality measures with local, clinically meaningful metrics focused on regions of high uptake, such as tumors and organs.
    5. Clinical Relevance: The dataset addresses a major bottleneck in AI-driven PET imaging, with direct implications for reducing radiation exposure in clinical practice.
    6. Challenge-Driven Validation: The dataset has already been used in multiple years of an international challenge, demonstrating its utility and relevance to the community.
  • Please list the major weaknesses of the paper. Please provide details: for instance, if you state that a formulation, way of using data, demonstration of clinical feasibility, or application is not novel, then you must provide specific references to prior work.

    The manuscript does not adequately discuss potential limitations, such as the representativeness of simulated low-dose data or potential biases introduced by scanner-specific protocols.

  • Please rate the clarity and organization of this paper

    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.

  • Optional: If you have any additional comments to share with the authors, please provide them here. Please also refer to our Reviewer’s guide on what makes a good review and pay specific attention to the different assessment criteria for the different paper categories: https://conferences.miccai.org/2025/en/REVIEWER-GUIDELINES.html

    N/A

  • 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.

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

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

    The paper presents a much-needed benchmark dataset and evaluation framework for ultra-low dose PET imaging, addressing a critical bottleneck in the field. The dataset is large, diverse, and clinically relevant, and the evaluation framework is thoughtfully designed. The paper is well written and organized, and the dataset’s prior use in international challenges demonstrates its value to the community. Addressing the noted weaknesses in a revision would further strengthen the work.

  • Reviewer confidence

    Very confident (4)

  • [Post rebuttal] After reading the authors’ rebuttal, please state your final opinion of the paper.

    N/A

  • [Post rebuttal] Please justify your final decision from above.

    N/A



Author Feedback

N/A



Meta-Review

Meta-review #1

  • Your recommendation

    Provisional Accept

  • If your recommendation is “Provisional Reject”, then summarize the factors that went into this decision. In case you deviate from the reviewers’ recommendations, explain in detail the reasons why. You do not need to provide a justification for a recommendation of “Provisional Accept” or “Invite for Rebuttal”.

    N/A



back to top