ADSMI: MICCAI Workshop on Advancing Data Solutions in Medical Imaging AI

In the rapidly evolving field of Medical Imaging AI, the quest for robust, fair, and efficient computational models is more critical than ever. With the advent of advanced imaging techniques and the exponential growth in the volume and complexity of medical data, traditional machine learning approaches face significant hurdles in data solutions. The MICCAI Workshop on Advancing Data Solutions in Medical Imaging AI (ADSMI) aims to provide a venue for discussions and innovations at the confluence of medical imaging, artificial intelligence, and data science, addressing the nuanced challenges of data scarcity, quality, and interoperability in medical imaging AI.

ADSMI is the joint workshop of Data Augmentation, Labeling, and Imperfections (DALI), Big Task Small Data, 1001-AI (BTSD-1001AI), and Medical Image Learning with Limited and Noisy Data (MILLanD).

Important Dates

Paper Submission Opens: April 18, 2024
Paper Submission Deadline: ~~June 24~~ June 29, 2024
Notification to Authors: July 15, 2024
Camera Ready Deadline: August 2, 2024
Workshop Day: October 10, 2024, Marrakesh, Morocco

Call For Papers

At a critical juncture in medical imaging AI, the integration of expansive datasets, foundational models, and advanced methodologies holds the promise of transforming diagnosis, treatment planning, and patient care. Yet, this potential is frequently curtailed by inherent data-related challenges, including limited annotations, privacy issues, and the long-tail distribution of medical conditions. Furthermore, the ethical dimensions of AI—encompassing fairness, equity, and bias—demand a shift towards more inclusive, transparent, and accountable AI systems.

ADSMI is committed to nurturing a multidisciplinary dialogue that narrows the gap between technological progress and clinical application. Our agenda covers a wide array of topics, ranging from novel data synthesis methods to the compilation of comprehensive, multimodal datasets mirroring patient population diversity. We delve into the exploration of foundational model training, utilizing the capabilities of large neural networks to explore new potentials in medical image analysis, diagnosis, and tailored medicine.

The workshop also seeks contributions that advance data curation practices, algorithmic fairness, privacy protection techniques, and the establishment of benchmarks to enable fair comparisons among emerging technologies. Special interest is directed towards innovative solutions for handling noisy or incomplete data and annotations, domain adaptation, incorporation of domain-specific knowledge within AI models, and foundation model development and applications. Our objective is to foster the development of advanced AI solutions that are also practically useful and in alignment with the goals of public health and patient welfare.

We extend an invitation to researchers, clinicians, technologists, and industry stakeholders to share their insights, research findings, and future visions for data solutions in medical imaging AI. Through keynote talks, paper presentations, and interactive sessions, ADSMI endeavors to stimulate the development of AI solutions that are technologically sound, ethically grounded, and universally accessible. Our topics of interest include but not limited to:

Innovations in semi-, weakly-, and self-supervised learning.
Multi-modal and synthetic data utilization.
Domain adaptation, generalization, and learning from limited datasets.
Enhancements in data annotation efficiency and strategies for managing label imperfections.
Development and evaluation of new datasets aligned with the workshop’s themes.
Foundation model development and applications for medical image analysis.
Exploration of zero-shot learning in medical image analysis for recognition, detection, and segmentation.
Efficient annotation methodologies employing one-shot or few-shot learning.
Techniques for synthesizing medical images across diverse modalities, imaging sites, and from textual descriptions.
Ethical considerations in AI-driven medical imaging, focusing on fairness, equity, and privacy.

Submissions to our workshop will be managed using the same platform as the main MICCAI conference, using Microsoft CMT. Workshop paper submission website is at: https://cmt3.research.microsoft.com/ADSMI2024

The ADSMI workshop will employ the same reviewing standards as the main conference. ADSMI workshop paper submissions should be anonymized to accommodate a double-blind review. Papers should be formatted using LaTeX or MS Word templates available at Lecture Notes in Computer Science. Manuscripts should be up to 8 pages (text, figures, and tables) plus up to 2 pages of references. In submitting a paper, authors implicitly acknowledge that no paper of substantially similar content has been or will be submitted to another conference or workshop until the decisions have been made by our workshop. Supplemental material submission is optional, which may include:

Videos of results that cannot be included in the main paper
Anonymized related submissions to other conferences and journals
Appendices or technical reports containing extended proofs and mathematical derivations that are not essential for the understanding of the paper

Contents of the supplemental material should be referred to appropriately in the paper, and reviewers are not obliged to look at it.

Accepted ADSMI workshop papers will be published with MICCAI 2024 Proceedings in the Springer Lecture Notes in Computer Science (LNCS) series.

Camera Ready Submission Guidelines

Please carefully address the feedback provided by the reviewers. Submit the revised materials to the ADSMI CMT site as a single zip archive, named in the format adsmi24_id-X.zip, with “X” being replaced by your unique paper ID.

Your submission should include:

Manuscript: A maximum of 8.5 pages, inclusive of text, figures, and tables, with an additional allowance of up to 2 pages for references. The file should be named manuscript.pdf.
Supplementary Material (Optional): Name the file supplementary_material.pdf. Note that source files for supplementary materials aren’t mandatory.
Changes Document: A detailed list of modifications made post-review. Name the file changes_after_review.pdf.
Copyright Form: Download and fill out the copyright form. The form should be signed by the corresponding author. Digital signatures will not be accepted. Save the file as copyright.pdf.
Source Files: Include a folder named src/, which houses the source files for your manuscript (e.g., .tex, .bib, .docx).

To ensure your paper is presented at MICCAI ADSMI 2024, a minimum of one paper author must register to attend on the second workshop day, October 10. Per MICCAI guideline, this registration must be an “in-person” registration. The camera-ready submission portal will prompt you to provide the registration number of the author who will be presenting your work.

Program

Location

Workshop: Oliveraie, Conference Center
Poster Session: The main poster room, the same boards as the main conference

Keynote Speakers

Christian Ledig, University of Bamberg, Germany [Keynote Info]
Ulas Bagci, Northwestern University, USA [Keynote Info]

Schedule

1:30 - 1:40 PM WEST Opening remarks and introduction

Keynote Session 1

1:40 - 2:15 PM WEST Keynote: From Images to Embeddings: Is End-to-End Learning Coming to an End?, Christian Ledig (University of Bamberg); [Details]

Oral Session 1

2:15 - 2:55 PM WEST

Physics-informed Unsupervised Test-time Adaptation for MRI Super-Resolution, Weitong Zhang (Imperial College London)
LoGex: Improved tail detection of extremely rare histopathology classes via guided diffusion, Maximilian Mueller (University of Tübingen)
Label Sharing Incremental Learning Framework for Independent Multi-Label Segmentation Tasks, Deepa Anand (GE HealthCare)

Coffee Break/Poster Session

2:55 - 3:30 PM WEST Poster Session
3:30 - 4:00 PM WEST Coffee Break/Poster Session

Keynote Session 2

4:00 - 4:30 PM WEST Keynote: A Wolf in the Sheep’s Clothing: Diagnosis and Risk Stratification of Pancreatic Cysts with Explainable AI, Ulas Bagci (Northwestern University); [Details]

Oral Session 2

4:30 - 5:20 PM WEST

MedMNIST-C: Comprehensive benchmark and improved classifier robustness by simulating realistic image corruptions, Francesco Di Salvo (University of Bamberg)
Learning from Similarity Proportion Loss for Classifying Skeletal Muscle Recovery Stages, Yu Yamaoka (Osaka University)
Selective Test-Time Adaptation for Unsupervised Anomaly Detection using Neural Implicit Representations, Sameer Ambekar (Technical University Munich)
DermDiff: Generative Diffusion Model for Mitigating Racial Biases in Dermatology Diagnosis, Nusrat Munia (University of Kentucky)
5:20 - 5:30 PM WEST Closing and award announcement

Poster List

01 Physics-informed Unsupervised Test-time Adaptation for MRI Super-Resolution
02 Variational multimodal distillation for diagnosing plaque vulnerability in carotid 3D MRI
03 MedMNIST-C: Comprehensive benchmark and improved classifier robustness by simulating realistic image corruptions
04 CoMoTo: Unpaired Cross-Modal Lesion Distillation Improves Breast Lesion Detection in Tomosynthesis
05 Learning from Similarity Proportion Loss for Classifying Skeletal Muscle Recovery Stages
06 LoGex: Improved tail detection of extremely rare histopathology classes via guided diffusion
07 SAM Carries the Burden: A Semi-Supervised Approach Refining Pseudo Labels for Medical Segmentation
08 Unsupervised Domain Adaptation for Pediatric Brain Tumor Segmentation
09 Rethinking Annotator Simulation: Realistic Evaluation of Whole-Body PET Lesion Interactive Segmentation Methods
10 Federated Self-supervised Domain Generalization for Label-efficient Polyp Segmentation
11 Selective Test-Time Adaptation for Unsupervised Anomaly Detection using Neural Implicit Representations
12 AMAES: Augmented Masked Autoencoder Pretraining on Public Brain MRI Data for 3D-Native Segmentation
13 Enhancing Single-Slice Segmentation with 3D-to-2D Unpaired Scan Distillation
14 Enhancing the automatic segmentation and analysis of 3D liver vasculature models
15 Adaptive Pseudo Label Selection for Individual Unlabeled Data by Positive and Unlabeled Learning
16 Label Sharing Incremental Learning Framework for Independent Multi-Label Segmentation Tasks
17 The Effect of Lossy Compression on 3D Medical Images Segmentation with Deep Learning
18 DermDiff: Generative Diffusion Model for Mitigating Racial Biases in Dermatology Diagnosis

Awards and Sponsors

We are pleased to announce that a prestigious award will be presented at the upcoming DALI workshop, and we are grateful to our generous sponsor. The Best Paper Award with an amount of $500 will be sponsored by a leading companies in the industry: NetMind.

We are thrilled to have NetMind as the sponsor for the ADSMI workshop and are grateful for their support in recognizing outstanding contributions in medical AI research. We encourage all attendees to take the opportunity to learn more about the decentralized AI platform.

People

General Chairs:

Yuan Xue, Ohio State University, USA
Chen Chen, University of Sheffield, UK
S. Kevin Zhou, University of Science & Technology of China, China
Rama Chellappa, Johns Hopkins University, USA
Sameer Antani, National Institutes of Health, USA
Zhiyun Xue, National Institutes of Health, USA

Executive Chairs

Shuo Zhou, University of Sheffield, UK
Mohammod Naimul Islam Suvon, University of Sheffield, UK
Chao Chen, Stony Brook University, USA
Esther Puyol-Antón, King’s College London, UK
Le Zhang, University of Birmingham, UK
Bo Zhou, Northwestern University, USA
Xueqi Guo, Yale University and Siemens Healthineers, USA
Marleen de Bruijne, University Medical Center Rotterdam, Netherlands and the University of Copenhagen, Denmark
Qingsong Yao, Chinese Academy of Science, China
Sivaramakrishnan Rajaraman, National Institutes of Health, USA
Zhaohui Liang, National Institutes of Health, USA
Ghada Zamzmi, Food and Drug Administration, USA
Marius George Linguraru, Children’s National Hospital, USA

Program Committee

Amogh Subbakrishna Adishesha, Penn State University
Anabik Pal, IISER Berhampur
Anqi Feng, Johns Hopkins University
Bin Wang, Northwestern University
Cheng Ouyang, Imperial College London
Christoph M. Friedrich, University of Applied Sciences and Arts Dortmund
Chunjiang Wang, University of Science and Technology of China
Eung-Joo Lee, University of Arizona
Fan Wang, Stony Brook University
Farheen Ramzan, University of Sheffield
Fenghe Tang, University of Science and Technology of China
Fuping Wu, University of Oxford
Gilbert Lim, SingHealth
Han Li, University of Science and Technology of China
Haobo Zhu, University of Oxford
Haoran Lai, University of Science and Technology of China
Jiazhen Pan, Technical University of Munich
Jinkui Hao, Cixi Institute of Biomedical Engineering, Ningbo Institute of Industrial Technology, Chinese Academy of Sciences
Jun Li, Institute of Computing Technology, Chinese Academy of Sciences
Kexin Ding, University of North Carolina at Charlotte
Le Zhang, University of Birmingham
Liu Li, Imperial College London
Mingyue Zhao, University of Science and Technology of China
Mohammod Naimul Islam Suvon, University of Sheffield
Muchao Ye, The Pennsylvania State University
Nicha Dvornek, Yale University
Nicholas Heller, University of Minnesota
Ning Bi, University of Oxford
Peiyu Duan, Yale University
Peng Guo, Massachusetts General Hospital
Prasanth Ganesan, Stanford Medicine
Qingsong Yao, Institute of Computing Technology, CAS
Qiong Liu, Yale University
Ruochen Wang, AbleTo, Inc.
Sameer Ambekar, Technical University Munich
Samira Zare, University of Houston
Shuo Zhou, University of Sheffield
Sivaramakrishnan Rajaraman, National Library of Medicine
Weitong Zhang, Imperial College London
Wenrui Fan, The University of Sheffield
Xueming Fu, University of Science and Technology of China
Xueqi Guo, Siemens Healthineers
Yinchi Zhou, Yale University
Yiying Wang, University of Oxford
Yubo Fan, Vanderbilt University
Yujia Li, Institute of Computing Technology, Chinese Academy of Sciences
Yuli Wang, Johns Hopkins University
Zeju Li, Imperial College London
Zhangxing Bian, Johns Hopkins University
Zhaohui Liang, National Library of Medicine
Zuhui Wang, Stony Brook University

ADSMI @ MICCAI 2024