The context

Due to possible complications related to the preterm birth (born before 37 week of pregnancy), preterm infants are admitted to an NICU unit. There is a XGBoost model that uses 3 features for the prediction of the risk that an infant will be diagnosed with sepsis during the admission period in the NICU. Currently, about 20% of the infants will develop a sepsis period during their admission which is related to a higher mortality and adverse long-term outcome What if there was a better, more reliable solution? During this challenge, you will use/ learn about:

The challenge

This was what this challenge is about During this challenge organised by FruitPunch AI For Health x UMC Utrecht, you will be working on improving the existing XGBoost prediction model with more features, exploring advanced models and developing a new model by taking time-series data into account. We are looking for AI enthusiasts based in the Netherlands  with a background in Computer Science/ Data science. Important presentations throughout the challenge will be held in Utrecht .

To achieve this goal, our project will be guided by an expert from the UMC Utrecht and FruitPunch AI. The project will kick off in the first week of February and the team will consist of 15 people that will work on the project for ~12 hours/week for 3 months.

Watch the info session on Friday 21st of January 12:30 CET

Apply for this challenge!

Who we’re looking for

Anyone with an interest in artificial intelligence or machine learning can apply. We expect some experience with programming languages and an interest in machine learning and AI. Some of the techniques you could be using and should spark your interest are, own experience and initiative with prediction is appreciated as well:

  • Medical Analytics
  • Time series
  • Classification
  • Explainable AI

You can join as a contributor (12 hours per week commitment for 2 months), coach (2-4 hours per week, only for experienced ML professionals) and teacher (give one relevant ML / domain masterclass).

During the challenge we will arrange for masterclasses on relevant topics like the use of the cloud computation resources Microsoft, IBM & the HPC Lab have made available for us.


Apply for the challenge HERE!

Challenge format

The challenge will run from the 16th of February until the 12th of May, & you will collaborate with a diverse team of AI enthusiasts and domain experts in subteams, all tackling this problem from different angles.

Some important dates:

  • Info session                             Friday 21/Jan 12:30 CET online
  • Deadline application             23/Jan
  • Challenge Kick-off                 16/Feb 14:30 – 17:30 CET offline
  • Domain Masterclass              16/Feb 14:30 – 17:30 CET (combine) offline
  • Tech Masterclass                    23/Feb 14:30 – 17:30
  • Midterm                                  31/Mar 16:30 – 17:30
  • Final presentation                 12/May 16:30 – 17:30

Related events & content

Watch the info session on Friday 21st of January 12:30 CET

Apply to join the Challenge

Apply for this challenge!

Background info

🥥 Neonatal sepsis is categorized as early
onset in the first 72 hours or late onset after 72 hours.

🍉 1 in 10 infants born in the United
States is born preterm, which comes with many risks like sepsis.

🍊  Around 15% of preterm infants will
develop a late onset sepsis, with an increased chance in the more immature.

🥝 The mortality rate in neonatal sepsis was found to be 20% and the mortality rate was found to be 18.8% in the patients with late-onset sepsis.

🥥 The mortality rate in neonatal sepsis was found to be 20% and the mortality rate was found to be 18.8% in the patients with late-onset sepsis.

🍉 UMC Utrecht’s Neonatology department is currently using an XGBoost model to predict sepsis in newborns.

🍊 Only three variables have been taken into account in the current model: heart frequency, respiration, oxygen saturation.

🥝 The current model is analysed using SHAP to make it more explainable, and doctors receive risk scores of the babies every hour which are effectively the model’s confidence scores for classification.

🥥 The aim of this project is to diagnose sepsis earlier, such that doctors can get notified and take action, reducing mortality in preterm newborns.

Partners in this challenge

The EWUU Alliance



Learn more