Solving real-world health challenges through Artificial Intelligence and Machine Learning in 3 days: how did that turn out?
Solving real-world health challenges through Artificial Intelligence (AI) and Machine Learning (ML) in three days: that was the call for applicants for ‘Hacking for Health’, Helmholtz AI’s first health-focused hackathon, which took place at the Helmholtz Diabetes Center in Munich last February 19th, 20th and 21st.
Around 50 hackers coming from all across the Helmholtz Association were divided into eight cross-disciplinary teams who tried to find an innovative solution to a challenge proposed by Helmholtz research teams working in research field health. Challenges included questions related to organ segmentation during surgery, reconstruction of vessel networks, cancer diagnosis through MALDI data, health image data quality, vaccine design, histopathology imaging, and auditory brainstem responses. Of course, there was more to hacking on this three-day event; hackers were also able to enjoy delicious food, have fun, and get the chance to know each other and the Helmholtz AI platform better.
Each challenge was championed by a dedicated mentor, which helped the team members address problems and answer questions. Additionally, a team of Helmholtz AI consultants was present at all times to provide technical support and general guidance. Computational power was kindly provided by Jülich Supercomputing Center (JSC).
At the end of the third day, each hacking team presented their results to an expert panel with researchers Benjamin Schubert, Annalisa Marsico and Matthias Heinig, from the Institute of Computational Biology (ICB), who were extremely impressed with the creativity and solutions displayed by the hackers in such a short timeframe. Further outcomes, coordinated by mentors, are expected to lead to publications in the near future.
All in all, ‘Hacking for Health’ offered young scientists an excellent opportunity to expand their knowledge of applied AI and ML by being part of a cross-disciplinary team mentored by experts. Hackathons such as this one are expected to become a regular tool to nurture and grow the Helmholtz AI network, empowering the next generation of data scientists.