Top of page

Student/Staff Portal
Global Site Navigation

School of Science

Local Section Navigation
You are here: Main Content

The CAIML 2025 AI Challenge has started!

Tuesday, 07 October 2025

Tags:

Presented by the Centre for AI and Machine Learning (CAIML) for School of Science Coursework Students

Take On Real-World AI in Medical Imaging

The Centre for AI and Machine Learning (CAIML) invites you to join the 2025 AI Challenge, where you will apply machine and deep learning techniques to solve critical problems in healthcare. This year’s competition features two real-world, medical imaging-based problems.

Who Can Participate?

  • All undergraduate and postgraduate coursework students enrolled in the School of Science
  • Must be an onshore student (in Australia)
  • HDR students are not eligible

Prizes

  • First Place in each challenge:
    * $1000 cash prize
    * Certificate of Achievement from CAIML(Only first-place winners will be awarded. One winner per challenge.)

Key Dates

  • Challenge Begins: 6 Oct 2025
  • Final Submissions Due: 30 Nov 2025 11:55 PM (AWST)
  • Winners Announced: 4 Dec 2025

Challenge Tracks

  1. Robust Medical Image Classification (RobustMedCT)
    AI systems can classify organs in abdominal CT scans with impressive accuracy under ideal conditions, but here’s the real challenge:

    In clinical practice, CT images are rarely perfect. They may arrive with noise from scanners, motion blur from patients, or even subtle digital alterations introduced accidentally or maliciously. Under these unpredictable conditions, the reliability of AI becomes a matter of trust, safety, and ultimately, human lives.

    In this challenge, you’ll develop machine and deep learning models that go beyond performance on clean data; models that remain reliable even under challenging, real-world distortions.
    View Challenge →

  2. Bone Mineral Density Prediction
    Every year, millions suffer fractures caused by osteoporosis, a silent disease that weakens bones over time. The clinical gold standard for diagnosis is measuring bone mineral density (BMD) at the hip using a DXA scan. However, DXA machines are expensive, not portable, and often unavailable in remote or resource-limited areas. Imagine a better way: using a simple hand or wrist X-ray, taken by standard machines found in every hospital, and applying AI to predict the patient’s hip BMD. This approach could make osteoporosis screening cheaper, faster, and more accessible, helping doctors identify at-risk patients before fractures occur.

In this challenge, you will develop AI models that predict hip BMD from hand/wrist X-rays, helping transform osteoporosis care.
View Challenge →

What You’ll Gain

  • Experience applying AI to real-world, high-impact problems
  • Boost your portfolio with a standout AI project
  • Win cash, certificates, and CAIML recognition!

How to Participate

  1. Choose your track(s)
  2. Form your team (max 3 students)
  3. Join challenge in Kaggle (links provided above)
  4. Download datasets and guidelines
  5. Build and test your models
  6. Submit by the deadline!

(3 leaderboard submissions allowed per day up until the competition closes)

For questions, contact: Dr Jumana Abu-Khalaf (j.abukhalaf@ecu.edu.au) and Mohammad Al Fawareh (m.alfawareh@ecu.edu.au)

Share

Skip to top of page