AI against Diabetes

Construct Computer vision models that estimate the risk for diabetic ulcer formation, as well as highlight areas that indicate the risk.


Premium participation: Get 1-on-1 Mentoring & Guidance

Try a free mentoring session

Helping podiatrists with faster diagnosis of diabetic foot ulcers

“The lifetime risk for ulceration in diabetic patients is 15-25%”

Diabetic ulcers are a common complication of diabetes, and can lead to serious health issues if left untreated. Clinicians have a hard time reviewing the tons of images that are sent through the "Voetenchecker" app, so diagnosis of ulcer formation occurs slowly. We can speed up this process by using AI to aid these doctors in recognising and classifying the ulcer formation probabilities faster.

A subset of people diagnosed with diabetes form deep wounds on their limbs, due to the lack of blood circulation. This leads to rapid deterioration, as these wounds tend to become infected with time. This is a very serious problem for diabetic and especially elderly diabetic people, as they often notice these ulcers too late or wait too long to seek medical help. These wounds are difficult to heal and sometimes doctors have to resort to amputation in order to stop an infection from spreading. Our partner for this Challenge, RondOm - a podiatric clinic, has created an app that allows patients to send pictures of their feet to medical professionals for a quick check. With this app RondOm collected a vast dataset of images of diabetic people’s feet and matching labels - indicating the state of the deterioration of the limbs.

Our task for this Challenge will be to construct Computer vision models that estimate the risk for diabetic ulcer formation, as well as to highlight areas that indicate the risk. Let’s tackle this problem together!

GOAL: Construct machine learning models to estimate the risk for diabetic ulcer formation.

The subgoals include:

  1. Perform exploratory and visual analysis of the dataset.
  2. Data cleaning and normalisation.
  3. Build Classification models
  4. Evaluate models with explainable AI

To determine the probability and the risk factors of ulcer formation we will use:

  • Computer Vision
  • Convolutional Neural Networks
  • Object detection
  • Segmentation
  • SHAP

Who are we looking for?

Participants from all levels of experience are welcome!

We are looking for data scientists and AI engineers, preferably experienced in computer vision, segmentation, explainable AI or AI in Health

You will collaborate with a diverse team of up to 25 international collaborators in subteams. You can join as a contributor (8-12 hours per week commitment for 10 weeks) or coach (2-4 hours per week, only for experienced ML professionals)

Did you know

🏥  Preventing diabetic foot ulceration is an under examined field in modern medicine and eHealth development.

⛑️ The lifetime risk for ulceration in diabetic patients is 15-25%

⚕️ The development of an ulcer into a severely life-threatening situation is often rapid, within less than a day.

👣 In the first test of the Voetencheck app, podiatrists identified that 9% of app users had a foot ulcer.

Info session

Our challenge partner

Huisstijl downloads | Fontys

Application deadline

March 26, 2023
To application page


Application Deadline: 26 March 2023

Challenge Kick-off:  27 March 2023 

Midterm Presentation: 2 May 2023

Final Presentations: 1 June 2023

A little more