Finding the right details in Big Data

2017-03-21

Carolina Wählby and her research team will focus on microscopy data.

Hello there! Carolina Wählby, professor of quantitative microscopy at the Department of Information Technology. You were recently awarded 29 million kronor from the Foundation for Strategic Research in the area of "Big Data and Computational Science," for the project “Hierarchical Analysis of Temporal and Spatial Image Data.” Congratulations to the grant! How does it feel?

– Thanks! We are delighted, obviously, but we also feel great responsibility for carrying out our plans! I think we’ve put together a very good project team with complementary skills, and we have external partners with exciting questions and great interests in research fundamentals and long-term goals, so I'm looking forward to getting started!

The project is called “Hierarchical Analysis of Temporal and Spatial Image Data” and includes your colleague Andreas Hellander and co-applicant Ola Spjuth from the Department of Pharmaceutical Biosciences, as well as partners Astra Zeneca, Vironova and Stockholm University.What is it about?

– Today there are many different imaging systems able to collect large amounts of data, and by using digital image processing and analysis, we can automatically detect and quantify different patterns and processes. But these kinds of analyses are resource-demanding, and the scientifically valuable information is sparse across both time and space. That’s why we want to develop computationally efficient measurements for data description and confidence-driven machine learning that can approximate scientific relevance, and a framework for intelligent information hierarchies that can distribute the data to computing resources and storage options based on its relevance.

– We will focus on microscopy data and work with large-scale time studies of the dynamics of cells and pharmaceutical vesicles in collaboration with AstraZeneca; nanometer-resolved electron microscopy data in collaboration with Vironova; and digital pathology in collaboration with Mats Nilsson at Stockholm University and SciLifeLab.

It sounds like the research project is geared towards the life sciences. Could these methods be useful in other areas as well?

– Naturally, the handling of large amounts of data is not at all unique to the life sciences, many other scientific and industrial areas struggle in dealing with relevant information being sparsely scattered, for example in monitoring and quality control. So yes, I think that there could be many different applications for this type of hierarchical information analysis.

What will you use the 29 million for?

– The funds will primarily be used to recruit key researchers who can help us realize our ideas. Moreover, the money will enable us to meet other researchers, and work within each other's environments, to better understand opportunities, bottlenecks and needs connected to the scientific questions at hand. I am convinced that personal interactions and clear communication are imperative for driving research forward in the best way.

---

Find out more:

Peter Waites