Brain Talks -seminar: Iiro Jääskeläinen

Professor Iiro Jääskeläinen will give a lecture on classification of EEG signals to disclose specific brain states under naturalistic stimulus/task conditions. Jääskeläinen works in the Department of Neuroscience and Biomedical Engineering in Aalto University.

"Previous studies have shown that multivariate pattern analysis (MVPA) of fMRI data is highly useful in disclosing distributed activity patterns of brain activity underlying perception, cognition and emotion including under naturalistic stimulus conditions, however, fMRI is cumbersome for many experimental designs, it is costly to operate, and is limited in temporal resolution. EEG, while lacking the fine-grained spatial resolution, is less costly, offers superior temporal resolution, and allows for task such as persons interacting in naturalistic settings. 

In this talk, I present results from two studies where MVPA analysis of EEG has been used to disclose complex cognitive states during naturalistic stimulus and task conditions. In the first study, EEG data were used to successfully identify distributed patterns of EEG signals that predicted success of a video recording of health expert persuasion to avoid unhealthy sugary foods. In the second study, EEG data was used to identify distributed patterns of EEG signals that classified between different types of learning vs. control tasks. Taken together, the results from these two studies suggest that machine learning approaches can be used to classify higher cognitive states and -events based on dynamic EEG signals. Tentatively, this can be a useful for many types of experimental paradigms where EEG is the most suitable methodology for measurement of brain activity such as participants interacting with others in natural environments and tasks." 

Read more on Jääskeläinen's research interests here.