DaT Parkinson's Challenge

Help support early and accurate detection of parkinsonian syndromes by developing models that classify dopamine transporter (DaT) scans as normal or abnormal. #health

€25,000 in prizes
7 weeks left
129 joined

Overview

Parkinsonian syndromes affect millions of people worldwide, yet diagnosing them accurately and early remains challenging. Dopamine transporter (DaT) imaging is an important tool for distinguishing neurodegenerative parkinsonian syndromes from other conditions, but reliable interpretation requires specialized expertise that is not available everywhere.

Each year, more than 20,000 DaT scans are performed in France alone. While many can be classified as normal or abnormal, approximately one in five cases remains difficult to interpret — particularly in early-stage or atypical presentations — delaying diagnosis, complicating treatment decisions, and increasing demands on specialist readers.

Task

In this challenge, we invite data scientists, machine learning engineers, medical imaging researchers, and nuclear medicine specialists to advance AI tools for DaT scan interpretation. Using a unique multicenter dataset of scans collected and annotated by French experts across ten hospital centers in France, your goal is to develop computer vision models that classify DaT scans as normal or abnormal.

Successful models will be released as open source and indexed in the the Bibliothèque Ouverte d'Algorithmes en Santé (BOAS), France's open health algorithm library, while the dataset will be made available as open data on data.gouv. By improving the accuracy and consistency of DaT scan interpretation, winning solutions could help expand access to expert-level diagnostic support and accelerate research into neurodegenerative disorders.

Prizes

Competition End Date:

Sept. 16, 2026, 11:59 p.m. UTC

Place Prize Amount
1st €12,500
2nd €7,500
3rd €5,000
Total €25,000

Note: Prizes delivered by DrivenData in USD, based on the exchange rate on July 22, 2026.

How to compete

  1. Click the "Compete!" button in the sidebar to enroll in the competition.
  2. Get familiar with the problem through the problem description. Additional resources are available on the about page.
  3. Download the data from the data download page.
  4. Create and train your own model.
  5. Package your model files with the code to make predictions based on the runtime repository specification on the code submission format page.
  6. Test your submission locally using the instructions in the runtime repository, and in the smoke test environment.
  7. From the submissions page, submit your code as a ZIP archive for containerized execution. You're in!

Competition rules

Below are a few highlights of the rules. See the full competition rules for complete details. They are designed to promote fair competition and encourage useful, reproducible solutions. If you are ever unsure whether your solution complies with the rules, ask in the competition forum or contact the organizers.

Competition data

Participants must:

  • Use data only for this challenge and only during the challenge period.
  • Delete all local data after the competition ends, unless a separate license allows continued use.
  • Never share, copy, or publish the data. For example, you may not use tools like Codex and ChatGPT that store or retain uploaded data, though you may download model weights and run models locally.

External data and models

External data and pre-trained models are allowed in this competition. Participants may use external data provided they have the legal right to do so. All external data must be shared with the challenge organizers, regardless of prize eligibility, to allow for independent result verification. Additionally, participants may not use tools like Codex or ChatGPT that store or retain uploaded data, but may download model weights and run models locally.

This challenge aims to support open solutions with broad social benefit and real-world applicability. To be eligible for prizes, (1) any external data must be freely and publicly available to all participants; and (2) any external data or pre-trained models used must be licensed so that the resulting model can be released for broad use, in and beyond the competition, including for commercial purposes (no NC, CC NC, or CC BY-NC licenses).

If you have questions about licensing in general or whether specific external data can be used, post in the competition forum or send an email to info@drivendata.org.


Organized by SFMN

This challenge is organized by the French Society of Nuclear Medicine (SFMN), in partnership with the Health Data Hub and GaelO. It is part of the "Health Data Challenges" call for projects, financially supported by the France 2030 plan. The competition dataset was assembled through a nationwide, multicenter collaboration of hospitals across France — learn more on the about page.

SFMN and partner logos


Image courtesy of GaelO