Scientists at the University of Toronto have developed a new way to investigate how we learn by combining two brain imaging techniques and zeroing in on what they call “pathway footprints.”
With it, they have found the first pieces of evidence in the human brain showing that two pathways within the hippocampus—the monosynaptic pathway (MSP) and the trisynaptic pathway (TSP)—play distinct roles in learning. These pathways are like telephone wires within the brain that facilitate communication inside and between its different parts.
“We imagine a messenger walking through the hippocampus carrying packets of information,” says Michael Mack, professor of psychology in the Faculty of Arts & Science. “We don’t know what’s in those packets, or what happens to the message as it goes through different subregions of the brain, but we can potentially track where the messenger is going.”
“Moment to moment, there are multiple messengers. The more messengers that go through a particular pathway, the stronger the footprint is going to look.”
Mack and Melisa Gumus, who completed her PhD at the University of Toronto, detail their work in “Distinct contributions of hippocampal pathways in learning regularities and exceptions revealed by functional footprints,” published by the Proceedings of the National Academy of Sciences.
The pathway footprints method combines advanced imaging techniques—diffusion-weighted imaging and functional Magnetic Resonance Imaging (fMRI)—to estimate the locations of the MSP and TSP endpoints and then measure how active these areas are during different stages of learning.
“This is the first time that we have been able to translate computational models of the hippocampus into testable predictions in humans,” says Gumus, now a postdoctoral researcher at the University of California, Berkeley.
“What we saw is exactly what we would expect from the models and existing theories. One pathway is important in extracting regularities in learning, and the other is important for encoding exceptional items,” says Gumus.
Following the Footprints
For this study, Gumus and Mack designed a rule plus exception learning task to investigate how humans build and adapt knowledge when faced with exceptional information.
They presented participants in an fMRI scanner at the Toronto Neuroimaging Facility (ToNI) with a series of illustrated flowers. Participants were asked to categorize them as flowers that preferred sun or flowers that preferred shade based on one or two notable features (like petal shape or inner circle colour). After each selection, participants were told whether they were right or wrong.
Throughout learning, participants were also presented with flowers that were more difficult to categorize. In fact, for these flowers, those that preferred sun looked like those that preferred shade, and vice versa—they were exceptions to the rules that participants learned quickly. To correctly categorize these exceptional flowers, participants had to consider all their features at once.
The researchers found that the MSP-related footprint was most active during the first stage of the experiment, as participants learned the rules of the task. This stage approximates early learning when concept regularities are established.
Notably, TSP-related footprint activation became relevant later in learning when participants began to correctly categorize exceptions to the rule. In fact, the more a participant activated TSP-related footprints, the better they learned the exceptions.
“It was almost like the MSP was building this knowledge base, the foundation, and then later on in learning the TSP added these exceptional items into a scaffold of knowledge,” Gumus says.
We face exceptions to rules, or concept regularities, all the time. Think about what defines a bird. If it flies and has wings, it is likely a bird, right? But what about bats? The pathway footprints method shows how hippocampal pathways deal with these regularities and exceptions in distinct ways.
Its implications extend beyond the hippocampus as well.
“I'm excited about applying the pathway footprints method to different questions and looking in different regions of the brain,” says Mack. He is currently using the method for a developmental study with children and building a significant dataset to investigate how well it generalizes.
“It is really a tool to be able to look at the brain and other cognitive processes by combining two different neural imaging modalities,” says Gumus. “It opens up a lot of different research pathways.”
Funding
This project was funded by Natural Sciences and Engineering Research council (NSERC) Discovery Grants, a Canadian Institute of Health Research (CIHR) Grant, a Brain Canada Foundation Grant, and a Vanier Canada Graduate Scholarship provided by NSERC.
More Information
To learn more about this study or to speak to its authors, please contact:
- Michael Pereira
Communications Officer, Department of Psychology, University of Toronto
psy.communications.officer@utoronto.ca