Thompson Rivers University (TRU) researchers use a high-risk, high-reward innovative approach to assess spontaneous change in infant brain functionality. The traditional approach requires infants to be asleep. With this new method, researchers will attempt to use machine learning to identify brain activity patterns associated with typical and atypical movements in awake and playing infants.
The department of Psychology’s Drs. Jenni Karl, Claudia Gonzalez and Musfiq Rahman, associate professor in Computing Science, welcome the recognition they’ve received for their research from the Social Sciences and Humanities Research Council of Canada (SSHRC) New Frontiers in Research Fund (NFRF) Exploration in the form of a $250, 000 grant.
“In the field of neuroimaging, this research could change the methods people traditionally use to map human brain development,” says Karl.
The grant allows the researchers to unite their disciplines and go beyond the limitations of common interdisciplinary approaches. Although interdisciplinary research can be challenging, the potential for groundbreaking impact makes it worthwhile to take the risk.
“Correlating brain activity with motor behaviour at different stages of development using these data-driven models will be challenging but highly rewarding,” says Karl, who is the head of the Brain and Behaviour Laboratory at TRU.
Karl and Gonzalez will assess the motor behaviour and brain activity of four to 12-month-old infants using functional near-infrared spectroscopy, a form of non-invasive brain imaging. They will use a “naturalistic neuroscience” approach that monitors brain activity in freely-behaving infants. “We aren’t getting the full story of what is happening with our traditional methods,” added Gonzalez, who is the co-principal investigator on the project.
Reaching further using technology
Using machine learning and AI technology, Rahman will attempt to utilize the raw data collected from the assessments and train algorithms to recognize patterns of brain activity related to normal and abnormal motor behaviour.
Future research could incorporate the resulting algorithms in remote health-care settings to help assess the development of infants living in rural or remote locations.
“We will share the annotated dataset with fellow researchers at no cost, allowing them to leverage our data patterns to construct and evaluate their own machine-learning models,” says Rahman.
The research could help to reveal how the human brain’s wiring transforms over the course of early development, making it easier to detect abnormal brain behaviour.
Anyone interested in participating in this research, please email Jenni Karl at email@example.com for more information.