By Amy Leson
This March I had the opportunity to attend the ComPh Modelling Week hosted by Simula and the Helmholtz Zentrum München in Fornebu, Norway at the beautiful Technopolis building. I would like to thank the MI-NET, Mathematics for Industry Network, for sponsoring me to attend this event.
Goal of the Week: “…designing and implementing computational tools for modelling and data analysis from existing and developing experimental modalities in various biomedical fields…”
A Crash Course: From RNA Sequencing to Disease Prevention
A combination of stimulating lectures and hands-on projects were presented throughout the week. As an engineer, the lectures that included “biology” in the title instilled fear in me. However, I was luckily situated between two Biologists doing their PhD’s. Soon, I found myself being enlightened and thoroughly interested in the current development towards improved personalized care through Polygenic Risk scores and RNA sequencing. Additional lectures continued to focus on the transferability of methods and techniques to clinical use and the exciting prospects.
Maths + Modelling + Industry = Excitement!
As a key feature of this conference was to further apply mathematics to industry, T. Kessler from Alacris and E. Samset from GE Healthcare Ultrasound Division gave talks about the applications of modelling, imaging and machine learning to further improve healthcare. With many advanced imaging courses coming up in my semester, I was glued to these talks and am keen to further apply my learning in order to contribute to these fields.
Predicting Aneurysm Rupture is Difficult
Time not spent in lectures were spent working on a variety of projects. With my background in the cardiovascular field, I couldn’t wait to get my hands on my project titled: “Accurate biophysical simulations with machine learning techniques” led by K. Valen-Sendstad. Our team, consisting of a multitude of academic backgrounds, chose to focus on the process between the “Image” and the “3D Model” which is highly subjective when done manually. Utilizing our knowledge from the “Math Guru” V. Naumova on dictionary learning, we denoised and inpainted images with the aim of being able to completely automate the process from Image to Risk Score. It was great fun, and I am excited to utilize some of these tools in my projects in imaging and 3D printing!
What do you get when you put biologists, mathematicians and engineers in one room?
Some of my takeaways:
- The future of modelling in the healthcare field is exciting and relevant for more personalized treatments.
- “All models are wrong, some are useful” – J. Hasenauer emphasizing that models are context specific and the importance of validation.
- If you get a group of biologists, mathematicians and engineers together they will have tons of fun and nerdy jokes. We enjoyed some delicious pizza with radius “z” and thickness “a” whose volume was pi(z*z)a.