Engaging students through collaborations and competition
The conference featured an exciting program aimed at engaging students – and engaged they were. They took an active part in all activities, including a data competition, scientific talks, poster sessions, and an Oxford-style debate. Several excellent talks and poster presentations were delivered. The awards for the best oral presentation and best poster, along with a travel grant of NOK 5000, went to Mathilde Gajda Faanes for her presentation Automatic Brain Tumor Segmentation in Intra-operative Ultrasound Images Using Deep Learning, and to Artem Chernyshov for his poster Automated Contouring and Quantification of the Right Ventricle in 2D Echocardiography.
|
|
Photos: Kushtrim Visoka (NORA.ai) |
|
The dataset competition, organized by Steven Hicks from Simula, was a scaled-down version lasting one hour and was won by the team HelloAI (student group 1), with an F1 score of 0.7091.
In the Oxford-style debate, four teams with two brave debaters per team competed, arguing for and against the motions 'AI developers should legally be responsible for diagnostic errors' and 'AI-enhanced radiology will increase healthcare inequalities.' The teams arguing for the motions shifted the most votes in their favor, leading them to victory in their debates.
Social events and networking
No conference is complete without social events and networking, which were integral to the program. The highlight was the scenic walk along the beautiful shores of Brekastad, combined with a fun historic quiz about Ørland. The winning group, the Defenders of the Coast, scored a whopping 9/9 on the quiz, for which they received some awesome NORA T-shirts. The TEAMS were also challenged to provide a short testimonial about the PRESIMAL Autumn School, which they did in style.
| Team | Testimonial |
|---|---|
| Nontiktok | “It's nice to meet nice people in a great environment” |
| Two Syllable Squad | “I came to PRESIMAL to build bridges, Across interdisciplinary ridges, We've come a lot farther, And become a lot smarter, And now we look forward to dinner” |
| Shallow Learning | “Radiology, Its the way it got to be, We love deep learning” |
| HelloAI | “PRESIMAL: Updating our weights and biases!” |
| The snowflakes | “Uniting bright minds for the improving future patient care🚀” |
| Defenders of the coast (winners) | "The research school has been a fully connected experience, where we have shared our deepest layers to create a centralized community for active federated learning in medical imaging and AI." |
The Science
Wednesday, Sept. 18th – Kicking Off Medical AI
The day began with Leif Oltedal (MMIV) highlighting the need for a medical AI research school, followed by Keyur Radiya (UNN) discussing "Software as a Medical Device." Samuel Kuttner (UNN) dove into imaging physics for AI scientists, and Frank Riemer (HUH) explored how incomplete data reveals the human side of imaging. Cecilia Marie Futsæther (NMBU) wrapped up with an intro to deep learning in medical imaging, and Steven Hicks (Simula) led a workshop on datasets for AI research.
Thursday, Sept. 19th – Clinical AI in Action
Atle Bjørnerud (OUS) shared lessons from Norway's Radiology AI-LAB, while Mattijs Elschot (NTNU) discussed AI for MRI and PET imaging. Charlotta Lindvall (Harvard Medical School) explained how natural language processing can transform cancer care, and Frederik Baastrup (HUMAN BYTES) showcased real-world AI case studies. The day ended with a medical data contest, led by Steven Hicks (Simula).
Friday, Sept. 20th – AI Solutions in Practice
Prabu Ramprabananth (Vestre Viken) kicked off the final day with his experience implementing AI in hospital workflows. Tone Bathen (NTNU) presented on AI for prostate cancer detection, followed by a panel debate and awards ceremony, closing the event.
A big thanks
The organising community would like to thank all the students, the presenters and Ørland Kysthotell for making The PRESIMAL Autumn Research School a fun, educational, and exciting experience!
About PRESIMAL
PRESIMAL unites Norway’s leading research environments in precision imaging and machine learning. The network advances research on machine learning in precision medicine, with a focus on medical imaging across various methods and diseases. It aims to develop more accurate diagnostic tools for improved prognostication and treatment. By sharing imaging data, methods, infrastructure, and interdisciplinary expertise, PRESIMAL accelerates progress in precision imaging. Additionally, the network promotes the creation of national digital platforms for testing machine learning algorithms, driving their translation into clinical practice.