info@bigdatarepublic.nl
menu

Online meetup: Treating Covid-19 with data

Improving Covid-19 treatment with interpretable Machine Learning
How can we use machine learning to predict whether a Covid patient is in a good enough state for mechanical ventilation to be stopped? Or what is the likelihood of a Covid patient surviving his or her stay in the ICU? These questions and their implications are addressed in this online meetup.

Machine learning applied to Dutch ICUs
Michele Tonutti from Pacmed and Tom de Ruijter from BigData Republic will be speaking to you about the Dutch ICU Data Warehouse: a large-scale collaboration of Dutch ICUs, initiated by Amsterdam UMC, that started with sharing data to improve Covid-19 treatment. They will share their experience in developing the ML models for this project, as well as discuss the importance of model interpretability in medicine.

Join the meetup!

Sept 23 2021, 16.30 Amsterdam time (CEST)

Agenda:

  • 16.30 Welcome & Introduction
  • 16.45 - 17.00 The Dutch ICU Data Warehouse
  • 17.00 - 17.20 Interpretable ML for better treatment of COVID in the ICU by Michele Tonutti
  • 17.20 - 17.40 Safe extubation of patients through interpretable Machine Learning by Tom de Ruijter
  • 17.40 - 18.00 Panel with Luca Roggeveen, Medical Doctor at Amsterdam UMC and Data Scientist at Pacmed, and Rik Kraan, Medical Doctor and Data Scientist at Vantage AI.

Meet the speakers

Michele Tonutti - Michele is a senior data scientist at Pacmed. His interests include academical, medical and social issues. He has extensive experience in both data science and data engineering. At Pacmed Michele applies both disciplines to build scalable, production-ready data pipelines. He also writes reusable code to train interpretable and reproducable state-of-the-art machine learning models.

Tom de Ruijter - Tom is a senior data scientist and consultant at BigData Republic. He loves coding and a well constructed program. His background in computer science with a specialization in AI and ML came in very handy during a 6-month collaboration with Pacmed last year. Moreover, he enjoys explaining his work to crowds, small and large, which is exactly what he'll be doing during this meetup.