Multi-site, Multi-Domain Airway Tree Modeling (ATM’22) Banner
Join

Warm Reminder:

Dear participants, the online long-term validation phase is still running. We do not limit the submission, however,  the participants are expected to design innovative approaches that deal with the airway tree-like structures or other tubular strutctures, and then validate on this online long-term validation phase. 


Take the consideration of the online maintenance cost, parameters tuning is not the primary goal of this online long-term validation phase. Further, to validate the proposed method, the Docker or API submission via e-mail for the test phase is also welcome.



Important News:

1. We are glad to announce our challenge paper "Multi-site, Multi-domain Airway Tree Modeling" has been accepted for publication in Medical Image Analysis! If you are interested in our challenge or using our data, please cite this paper!


2. ATM-22-Related-Work is an official repository of ATM'22 which collects our challenge information, results, and related works that have been established and updated occasionally!


How to Participate?:

As our challenge is an open call and new participants can still register via this challenge website. If you have participated in our challenge, it is deemed that your organization automatically agrees to the data agreement, regardless of whether you are participating anonymously or whether the competition ID is real. Please complete the data agreement file to send the signed document back to us to finish the registration! 



NEWS


2022/08/31: The final test phase is finished! Thank you to all participants.

2022/08/18: The Test Phase and the Validation Phase 1 (Live Leadboard) of ATM'22 Openning! The participants, who successfully took part in the previous stage, please submit your result on this new Validation Phase 1 (Live Leaderboard)! Besides, we will also download the previous validation phase best result for our local test. And the final rankings depend on the Test phase and the averaged metrics of TD, BD, DSC, and Precision.

2022/08/17: The  validation submission phase 1 has completed! Thanks for your participation.

2022/08/02: We have created the Test Docker Submission Guideline and all participants that successfully participated in validaition submission phase 1 can prepare your test docker and paper, we will take care of your submission! All participants should  E-mail to IMR-ATM22@outlook.com with a downloadable link to the zipped docker image, along with a  qualified short paper that describes your methods and experiment settings. If we don't receive the e-mails before  31/08/2022, 23:59, Time zone in Beijing, China (GMT+8), your participation is unfortunately unsuccessful!

2022/07/19: We have allowed ten submissions in the  validation submission phase 1! Please take part in this stage!

2022/07/11: The evaluation for validation is open! Please refer to validation submission phase 1! Please pay attention to two things: 1) The saved prediction results should have the same origin and spacing as those of corresponding input NIFTI CT images. 2) Only the largest component of the binary results are considered,  please guarantee your results are processed with the largest component extraction. 

2022/06/01: Early submission for validation is open! Please refer to validation submission phase 1!

2022/05/30: The training data have been released! Please refer to the dataset page! All participants must  send the signed document to IMR-ATM22@outlook.com to get the data access and obey the challenge rule!

2022/05/20: The registration has been open! Some participants have not  send the signed document to IMR-ATM22@outlook.com, please ensure that you have completed the full registration procedure!

2022/05/01: Please refer to the Registration page for detailed registration instructions.

2022/04/19: Challenge Website Open.  Alternative personal website: https://puzzled-hui.github.io/ATM/


  Background and Clinical Significance

Airway segmentation is a crucial step for the analysis of pulmonary diseases including asthma, bronchiectasis, and emphysema. The accurate segmentation based on X-Ray computed tomography (CT) enables the quantitative measurements of airway dimensions and wall thickness, which can reveal the abnormality of patients with chronic obstructive pulmonary disease (COPD). Besides, the extraction of patient-specific airway models from CT images is required for navigation in bronchoscopic-assisted surgery. Due to the fine-grained pulmonary airway structure, manual annotation is however time-consuming, error-prone, and highly relies on the expertise of clinicians.

  Objective

Bruijne et al. [1] held the ‘Extraction of airways from CT (EXACT'09)’ challenge in 2009 and achieved a great contribution to the field of airway segmentation. They focused on semi-automated and automated algorithms mainly based on multi-threshold, template matching, and region growing, aiming to relieve the burden of manual delineation and help clinicians explore the influence of pneumonia on airways. These traditional algorithms face difficulty in extracting small peripheral bronchi and suffer the risk of airway leakage. With the advance of deep learning methods, fully convolutional networks (FCNs) achieved state-of-the-art performance in the segmentation task of volumetric medical data. Most of the deep learning methods are data-driven while the EXACT’09 contains only 20 CT scans for training and 20 CT scans for testing, which is not sufficient for the artificial intelligence era. 

We collected 500 CT scans from multi-sites. The airway tree structures are carefully labeled by three radiologists with more than five years of professional experience. The intra-class imbalance among the trachea, main bronchi, lobar bronchi, and distal segmental bronchi affects the segmentation performance of peripheral bronchi. In conclusion, we encourage the participating teams to design robust algorithms, which can extract the airway tree structure with high topological completeness and accuracy for clinical use. Our challenge is open call (challenge opens for new submissions after conference deadline).


[1] Lo P, Van Ginneken B, Reinhardt J M, et al. Extraction of airways from CT (EXACT'09)[J]. IEEE Transactions on Medical Imaging, 2012, 31(11): 2093-2107.

How to Submit?


All the registered teams should make a complete submission, containing:

  • The docker container for the Test Phase. The format of the docker file should follow:  Docker Submission Rule ！
  • A short paper that describes your methods and experiment settings, 2 pages at least, preferably no more than 8 pages. Template same as MICCAI: LNCS.
  • ATTENTION: The short paper should contain the detailed methods with the corresponding formula, diagrams and sufficient implementation details.

Only the fully automatic algorithms will be considered. Please send the docker container and the short paper to IMR-ATM22@outlook.com via attachment or give the link to your submission, and then the organizer committee will download it.

[NOTE]: The top-5 teams must make their  code publicly available for verification and reproducible research.


 Citation: 

If using this dataset, you must cite the papers: 

[1]Zhang M, Zhang H, Yang G Z, et al. CFDA: Collaborative Feature Disentanglement and Augmentation for Pulmonary Airway Tree Modeling of COVID-19 CTs[C]//Medical Image Computing and Computer Assisted Intervention–MICCAI 2022: 25th International Conference, Singapore, September 18–22, 2022, Proceedings, Part I. Cham: Springer Nature Switzerland, 2022: 506-516. [paper link]

[2]Zheng H, Qin Y, Gu Y, et al. Alleviating class-wise gradient imbalance for pulmonary airway segmentation[J]. IEEE Transactions on Medical Imaging, 2021, 40(9): 2452-2462. [paper link]

[3]Yu W, Zheng H, Zhang M, et al. BREAK: Bronchi Reconstruction by gEodesic transformation And sKeleton embedding[C]//2022 IEEE 19th International Symposium on Biomedical Imaging (ISBI). IEEE, 2022: 1-5. [paper link]

[4]Qin Y, Chen M, Zheng H, et al. Airwaynet: a voxel-connectivity aware approach for accurate airway segmentation using convolutional neural networks[C]//International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, Cham, 2019: 212-220. [paper link]

You could cite our challenge benchmark manuscipt as well: 

Zhang M, Wu Y, Zhang H, et al. Multi-site, Multi-domain Airway Tree Modeling[J]. Medical Image Analysis, 2023, 90: 102957. [paper link]

BibTex is here:
@article{zhang2023multi,
  title={Multi-site, Multi-domain Airway Tree Modeling},
  author={Zhang, Minghui and Wu, Yangqian and Zhang, Hanxiao and Qin, Yulei and Zheng, Hao and Tang, Wen and Arnold, Corey and Pei, Chenhao and Yu, Pengxin and Nan, Yang and others},
  journal={Medical Image Analysis},
  volume={90},
  pages={102957},
  year={2023},
  publisher={Elsevier}
}

In addition, transfer of the dataset to others is not allowed! We schedule to publish a challenge paper, at most the three authors of the top five performing methods are qualified as authors in the final challenge paper.

The challenge data and results will be free to use after the challenge results publication has been released. But before this, the team should not submit individual papers. The organizers promise the blackout period will not be no longer than six months after the MICCAI-2022 challenge is finished.