Who’s who in the Zoo – Bart Elen

In the first Who’s who in the Zoo of 2020, meet Bart from the Eye for Diabetes project

– Helen

 


Bart - Bart Elen

Project: Eye for Diabetes

Researcher: Bart Elen

Location: Health department, VITO, Belgium

 

What are your main research interests?

Data science, deep learning, analysis of retinal images, Internet of things

 

Who else is in your project team? What are their roles?

Patrick De Boever (VITO), Project Leader

Carina Veeckman (VUB), User Engagement

Sven De Boeck (VUB), Science Communication

Luk Buyse, President of Diabetes Liga

 

Tell us more about the data used in your project

We use retinal fundus images collected by the EyePACS screening network for Diabetic Retinopathy.

 

How do Zooniverse volunteers contribute to your research? 

The Zooniverse volunteers look for lesions on the fundus images, caused by Diabetic Retinopathy.

 

What have been the biggest challenges in setting up your project?

It was challenging to design our project to enable everyone to perform the challenging labelling task of finding the lesions caused by Diabetic Retinopathy. We took a number of approaches to make this possible, including writing a good tutorial, a training module and augmenting our images.

 

What discoveries, and other outputs, has your project led to so far?

The volunteer labels have led to an improvement in our algorithm for automatically detecting lesions. Our project has also had a lot of media coverage in Flanders.

 

Once you’ve finished collecting data, what research questions do you hope to be able to answer?

The new data labels will enable the creation of better computer models for the automatic detection of lesions caused by Diabetic Retinopathy. We hope that these models will enable a more consistent diagnoses of Diabetic Retinopathy by ophthalmologists, and a better follow-up of the disease progress.

 

What’s in store for your project in the future?

In the future, there will be more labeled data, the creation of a better computer model for the automated detection of Diabetic Retinopathy lesions, and the release of the newly created label dataset to the research community.

 

What are your favourite other citizen research projects and why?

The “Radio Meteor Zoo” project on Zooniverse. Speaking to the creators of this project convinced us to choose the Zooniverse platform for our own citizen science project.

 

What guidance would you give to other researchers considering creating a citizen research project?

Contact the teams running existing citizen science projects. You will learn a lot from their experiences.

 

And finally, when not at work, where are we most likely to find you?

I like go running, follow MOOCS, play board games, and watch tv.

 


 

You can learn more about Eye for Diabetes here (note, this website is in Dutch)