Using movies to probe infant brain development

There’s an old adage, “never work with children or animals,” because it’s known to be notoriously difficult, especially when it comes to certain aspects of science and wanting to generate high quality data from repeated measurements. And that’s where kids can be particularly tricky. But Cameron Ellis, at Stanford University, has partly solved the problem, with the help of movies! He’s interested in how the developing human brain wires itself up and responds to different experiences and stimuli over the initial months of life. When it comes to the visual system, the current gold standard for doing this is to present various defined shapes and colours. But it can be hard to keep subjects still and motivated to remain engaged, which can marr the quality of the data. So can movies, designed as they are to be compelling and immersive, be used instead, he wondered...

Cameron – It's a real curiosity as to why we are born immature. If you look at horses, they can walk within a few minutes, maybe an hour after birth, whereas we take 9 to 12 months to learn how to walk. The same is true for our visual system. Although other animals might have a relatively mature visual system—meaning that the way in which they process visual input is similar in infancy to how it is in adulthood—that’s not the case in humans. Our visual system is really different.

Why it's different is a curious question to which we don’t fully know the answer, but one explanation is that to acquire the kind of expertise we use for processing the visual world in adulthood, it can’t all be genetically pre-programmed. Instead, some of that has to happen through learning during our lifetime. We need to learn how to see the world, and that's part of what happens in that first year of life.

Chris – Are you suggesting there's a rough map there—that things seem to be wired up roughly right—but what we need is the icing on the cake, the finessing of that system, in order to develop the amazing visual system that humans have when they're older?

Cameron – Yes, so when we started this work about five years ago, we didn’t know whether the macrostructure or the general organisation of the visual system in infants was in place. When we found that it was, that was very surprising. In general, the large-scale organisation present in parts of the visual processing system is very similar between infants and adults.
But despite that similarity in the big-picture structure, we still think that over the course of development, there are many changes in the finer details—perfecting the fine-grained aspects.

Chris – So what does using movies bring to the table in this context?

Cameron – Movies are an amazing stimulus for showing to infants and other populations, such as patients who might otherwise be uncomfortable in the neuroimaging environments we use to measure brain activity. The reason is that movies are naturally engaging. When you use a highly engaging stimulus, you're more likely to get better quality data. This is something we’ve seen in our own data—infants show more attention and move less when watching movies.
So in one sense, it’s a data quality issue—we want to see how good the data can be when using movies. But there’s another reason as well: it’s useful to observe how the brain responds to more naturalistic stimuli. While not entirely natural—they are still directed and edited—movies resemble everyday experiences more closely than artificial lab stimuli. Yet, they still evoke brain activity similar to what we see in more controlled experiments where we show specific shapes with precise timing.

Chris – And do your results align when you show movies compared to when you've done more controlled experiments with artificial stimuli like shapes? Do you get broadly the same outcomes?

Cameron – When we use these well-controlled stimuli like colourful shapes, those methods are considered the gold standard for identifying visual organisation in both adult and infant brains. But with movies, we can get rough approximations of those same maps by measuring brain responses. That tells us that the way in which the infant brain processes movies mirrors the structure we would expect to see using those precisely controlled experimental stimuli.

Chris – That must give you confidence to push forward. But also, because you can vary what's in the movie, you can more easily adapt the stimulus, probe different things, and repeat experiments—especially since the children are engaged and actually want to watch the film.

Cameron – Exactly. Beyond proving that movies are a good way of evoking brain responses in less controlled environments, it's also evidence that they are a powerful tool we should be using in a variety of contexts for studying the infant brain. For example, you can use movies to examine how infant brains respond to different languages or to test their understanding of social events and interactions.
You can even study how infants reason about others’ minds. We believe this is a strong case for making movies a central part of the experimentalist’s toolkit for researching the infant mind.

Chris – So what will you go on to explore next using movies, now that you've got the confidence to rely on them?

Cameron – One exciting current direction is using movies to study infants’ brain responses to language. In one experiment, we’re comparing responses to familiar languages—like English for English-speaking infants—against unfamiliar ones like Japanese. We want to see how the brain differs when processing a familiar versus an unfamiliar language.
Another line of research, made possible by this paper, is comparing directly how infant and adult brains process the same movie. That lets us ask: what makes infant brains unique? How do their experiences differ from those of adults?