For best experience please turn on javascript and use a modern browser!
You are using a browser that is no longer supported by Microsoft. Please upgrade your browser. The site may not present itself correctly if you continue browsing.

Studying vision in humans and computers

‘You might think that seeing the world is a really easy matter,’ says psychobiologist Iris Groen. ‘But it’s actually a complex process that involves a large part of the brain.’ In September 2020 she was appointed as a MacGillavry fellow at the Informatics Institute of the University of Amsterdam. The big question she aims to answer with her research is: How does the human brain convert the rays of light that fall on your retina into the image in your mind?

Fascination for human biology

Groen works on visual perception. This means she studies how people actually see. ‘Each side of the brain has four lobes, one of which is devoted completely to visual perception,’ she explains.

‘The brain comprises a mass of cells that communicate with each other via electric impulses. I study how people’s brains respond to simple tasks in which they see images. Some areas of the brain become active when you see lines or black and white images, other areas react to parts of objects, and yet others to entire objects, such as faces or houses. But how do these areas know what they are seeing? That’s what I find so fascinating.’

Iris Groen
Iris Groen

The first time she started wondering about the functionality of brains was at secondary school, when she was writing an essay about the chemicals that are released when one falls in love. Groen recounts: ‘People often find it unromantic to think that being in love is something biological, because they feel this makes it less magical. But to me, the fact that being in love has a biological cause actually makes it more real.’

After this essay she became interested in philosophy-of-mind questions, such as ‘what does it mean to think about something?’ and ‘what is consciousness?’ So Groen decided to enrol in the interdisciplinary Natural & Social Sciences bachelor at the UvA, with a major in Psychobiology, followed by a Master’s degree in Brain and Cognition Sciences.

Real world images

For her PhD project – conducted at UvA’s Psychology department – Groen studied what happens in the brains of test subjects when they see images. She did this while using photos with scenes from the real world. This was a new approach in her field; until then, it was customary to present the test subjects with simplified images.

Groen: ‘Generally speaking, images in this kind of research are reduced to lines or are cut out or set against a neutral background. The researchers can then adjust certain elements of these drawings to see whether this makes a difference for the brain. But I always felt that this simplification wasn’t correct when compared to the real world, because you never see a face floating in front of a grey background. You always see things in a context.’

It turned out that using these more complex images was indeed useful. ‘My main research finding was that crowded images are predictive for much higher activity in the brain. If an image is chaotic and unstructured, it takes the brain longer to react. I developed a computational model that, based on the structure of an image, can predict what the brain responses will be,’ says Groen.

Two body problem

After her PhD Groen was confronted for the first time with the “two body problem”, something many couples face when both work in academia. It’s often difficult for the two partners to find research positions that are close together, meaning they are forced to maintain a long-distance relationship.

Groen and her partner – an astronomer – both wanted to do a postdoc in the United States, but also be able to live together. ‘Towards the end of our respective PhD projects we did a tour of labs on the East and West coast in search of post-doc positions. We visited people we knew from conferences and we wrote to labs, asking if we could give a presentation of our work.’

Their efforts were not in vain: they found jobs near one another. Groen took on a postdoc position at the National Institute of Mental Health, a research institute run by the US government in Bethesda, Maryland. Here, too, her work involved presenting test subjects with images and studying the response in the brain using EEG and fMRI; just like she did during her PhD research. But now, her research involved brain stimulation with magnetic pulses, also known as Transcranial Magnetic Stimulation (TMS).

TMS and ECoG

With TMS magnetic pulses are transmitted to a certain area of the brain which is then temporarily disrupted. It is regarded as a promising therapy against depression and other neuropsychological disorders – but exactly how it works is still a question mark for scientists. Groen set out to try and solve part of this puzzle by studying what effect TMS has on the parts of the brain that we use for visual processing.

Groen: ‘We wanted to know how well TMS can target specific neural processes. Against expectations, the experiment showed that it didn’t matter which area of the visual cortex was stimulated: activity was reduced in all areas. Groen: ‘This shows that the type of stimulation we used spreads further through the brain than we originally thought.’

Three years later Groen took up a post at New York University. Here she had the opportunity to work with Electrocorticography (ECoG), a technique whereby electrodes are placed in the brain of test subjects. ‘Patients with very serious epilepsy sometimes receive a brain operation if no drugs work for them,’ explains Groen.

‘A surgeon makes a small hole in their skull and places a number of electrodes on the brain. It takes some time before they know which part of the brain is affected by the epilepsy, after which they take out the electrodes again and proceed to surgery. Meanwhile, the patients spend a week or more in the hospital with this bundle of electrodes in their head. This is a unique opportunity for us researchers. If the patients were feeling bored, I asked them if they would like to take part in my research.’

Again, she showed patients images and observed how their brains responded to them, acquiring a new type of information on visual processing in human brains. Groen: ‘The electrodes measure the brain responses with very high resolution, without the distortions that we have with EEG or fMRI.

Because of this, the neural responses to the visual images are extremely clear and reproducible, meaning that they look exactly the same when you show the image for a second time. This makes these data very special and exciting, just beautiful to see.’

Algorithms that can see

At the UvA Faculty of Science, Groen has now joined the Informatics Institute (IvI). This might sound like a surprising career switch, but Groen explains why it isn’t: ‘The artificial intelligence researchers in my new department are creating computer algorithms that should be able to “see”. Think of, for example, algorithms that are used in self-driving cars or that are used to translate a visual image into a sentence, which can be useful for vision-impaired people.

One important question for these researchers is: how do you translate information from the world into code? By showing how the human brain does this, I want to contribute to the further development of these algorithms.’

Part of her research at IvI is funded by a Veni grant from NWO, which she received just after she accepted the position at the UvA. This project focuses on how the brain computes action affordances of the environment, for example where roads and pathways are that you can use to move yourself forward.

Groen is very happy with her MacGillavry Fellowship because she had hoped to return to the Netherlands. It did mean that once again the two body problem raised its head. This time, she negotiated that her partner would also receive an academic post at the University of Amsterdam (although he ended up accepting another position at Leiden University).

Groen: ‘The aim of the MacGillavry Fellowship is to get talented women on track to a professorship. If, as university, you want to recruit and retain women then I think you should try to facilitate that their academic partner also gets a post. That may sound unfair, but it’s reality that this issue plays an important role in women’s career choices.

When I look at other academic couples around me, I often see that the man is already further up the academic ladder than the woman, for various reasons. And this often results in women leaving their academic position, or the academic world, in order to follow their male partner.’

Groen hopes to be a role model for other women in her specialist field. ‘As a student I knew in cognitive terms that a woman could become a professor. But I had very few female professors – I think only one in my first year of university – and I still had this stereotype image of a bearded man in a white lab coat fixed in my brain. I hope I can play an exemplary role for the next generation.’

CV

2003-2006:

BSc in Psychobiology (cum laude), University of Amsterdam, the Netherlands

2006-2009: MSc in Brain and Cognition Sciences (cum laude), University of Amsterdam, the Netherlands & University of Cambridge, United Kingdom
2009-2014: PhD in Visual Perception (cum laude), Department of Psychology, University of Amsterdam, the Netherlands
2015: PhD Thesis prize from the Dutch Society of Psychonomics
2015: Rubicon postdoctoral fellowship grant, Dutch Research Council (NWO)
2014-2017: Postdoctoral fellow, National Institutes of Health, Laboratory of Brain and Cognition, Bethesda, United States
2017-2019: Postdoctoral associate, New York University, Department of Psychology, United States
2019: Veni grant from the Dutch Research Council (NWO)
2020-present: Assistant professor, University of Amsterdam, Informatics Institute, the Netherlands