Registration
00

day

00

hr

00

min

00

sec

Time to team-up deadline

310 / 249

participants / team

Register

Topic provider

NTU Med God-A data platform for precision medicine

2020/07/24 Team-up procedure

Step 1: Input team name in the box. Before this, the default team name is your username. Step 2: Click the "request merge" button, then your request will be sent to the team leader immediately. Notes: 1. Make sure the member you want to invite has an AIdea account. 2. Make sure the member you want to invite has registered this challenge.

Introduction

CyberKnife uses a high-energy X-ray machine on a robotic arm to precisely deliver radiation beams that destroy tumor cells and stop tumor growth while avoiding damage to healthy tissue.

Before treatment, need to create a personalized treatment plan to evaluate the unique shape, size and location of the tumor, as well as the tissue and structures that need to be protected.

Clinical challenge:
1-CyberKnife mostly is used for tiny tumor
2-It is a very time-consuming process, It will take hours for two professionals (one doctor and one medical physicist) to perform the computer planning, and for complicated case, it will take a day to create the plan for a patient
In order to enhance the workflow, we want to train the AI to identify the location of brain tumors and contour it.

Through this challenge, we hope to encourage participants to learn from each other and optimize the training process to get the best model performance as possible.
Welcome to join us! Action now!

Organizer:

台大醫神-精準醫療人工智慧輔助決策系統(NTU Med God-A data platform for precision medicine )

Co-organizers
科技部補助全幅健康照護中心(MOST All Vista Healthcare Center)
商之器科技(EBM Technologies )
台大醫院神經外科(The Neurosurgery in NTUH)
國立台灣大學生醫電子與資訊學研究所(Graduate Institute of Biomedical Electronics and Bioinformatics, National Taiwan University)
威強電工業電腦(IEI Integration Corp.)
威聯通科技(QNAP Systems, Inc.)

The Ministry of Science and Technology All Vista Healthcare Center (MAHC), a research center launched in March 2018 and fully supported by Taiwan government, is dedicated to AI applications in medicine and healthcare. The MAHC aims to establish a world level platform to facilitate explorations of AI model-driven or data-driven technology for medical and healthcare, through which any technology targeted for medical AI applications can be quickly developed and validated. The MAHC is expected to lead and promote domestic research teams specialized in biomedical and healthcare fields and in turn to improve the health and well-being of people in Taiwan.

There are currently 11 project teams, affiliated with top academics and medical centers in Taiwan, under the MAHC. Based on the research topics, the projects can be categorized into five areas: mental healthcare, medical imaging, decision support, precision drug use, and AI-ethical, legal, and social impact (AI-ELSI).

In addition to working with research groups from domestic institutions in Taiwan, the MAHC is forming collaborations/partnerships with internationally renowned academics such as Stanford University, Harvard Medical School and its affiliates, and University of Toronto (Canada) for the advancement of medical AI research. Moreover, the MAHC is also working with industrial partners to promote commercialization and entrepreneurship.

We anticipate that the MAHC become a top-notch cluster for digital health and medicine so that medical achievements enhanced by the AI techniques in Taiwan can be more easily recognized worldwide.

Prize

1st  place: One team $USD 2,000 (tax included)+ QNAP NAS*1 (value USD300)

2nd place: One team $USD 1,700 (tax included)+ QNAP NAS*1 (value USD300)

3rd  place: One team $USD 1,300 (tax included)+ QNAP NAS*1 (value USD300)

4th place: One team $USD 1,000 (tax included)+ QNAP NAS*1 (value USD300)

Remark1: After submission of AI model and inference source code,the top 4 winners will receive the prize money after the deduction of tax. (vary by country)

Activity time

Start from July 17, 2020

DATES

Events

July.17

Release of training data *1500 (Brain MRI image 3D T1+C )

July.24

Release of validation data*50 (Brain MRI image 3D T1+C )

July.24-Sept.26

Submit answers to the online score board to get score and ranking (Dice score)

Sept.18

Release of testing data*100 (Brain MRI image 3D T1+C )

Sept.25

Freeze the leaderboard

Sept.26

Award Announcement Day via google meet

Sept 26 Award Announcement Day
pm1:00~3:30       Taiwan time
am10:30~13:00    India time
pm2:00~4:30       Japan time

 

Evaluation Criteria

Evaluation method: The final score is mean DICE

 

If any question, you may leave a message in this forum,

or contact iEi PM Jessica <jessicalin@ieiworld.com>

Rules

DATES

Events

July.17

Release of training data *1500 (Brain MRI image 3D T1+C )

July.24

Release of validation data*50 (Brain MRI image 3D T1+C )

July.24-Sept.26

Submit answers to the online score board to get score and ranking (Dice score)

Sept.18

Release of testing data*100 (Brain MRI image 3D T1+C )

The final score is only based on testing dataset.

Sept.25

Freeze the leaderboard (*remark 1)

Sept.26

Award Announcement Day via google meet (*remark 2)

 

Remark1: Freeze leaderboard while challengers can still summit results to update scores. The winner will be revealed on Award Announcement Day!

Remark2: Top 10 teams: 8-minute oral presentation (pls send slides before Sept.25)

Rules:

  • The maximum number of uploading is 5 times per day

  • Max 5 challengers in one team, at least one student

  • The evaluation result would be based on the final version that you upload. If two teams (or more) got the same evaluation scores, the time they upload would determine the ranking.
  • When using external data sets, participants are not allowed to use future data as the basis of the test results. Data sets instructions and references should be stated in the forum for references.
  • If there is cheating or fraud during the activity, the team that cheats will be disqualified from the activity and the vacancy would be filled up by other teams in the ranking order.
  • Each person can only be in one team, and every single person or team can only hold one account. In other words, duplication participation is not allowed, if found, the user will be disqualified.
  • According to the request which is given by the issue provider, only teams of one are allowed to participate in the activity. If the same answer sheet are found during the examination, the participants who owned them will be disqualified from the issue, and all the uploaded data will be deleted.
  • Please make sure to keep the result safe so that it will not be copied or stolen by anyone else and thus will not cause the invasion of individual rights.
  • If there is a dispute in the future, the organizer reserves the right to make the final decision.