The brain represents categories of visual stimuli through distinct patterns of neural activity. Physiological methods such as scalp electroencephalography (sEEG) provide direct measures of the brain's electrical activity with high temporal resolution, making sEEG a potential tool to elucidate how memory representations change over time in the human brain. However, it is not known how reliably sEEG data can be used to track the reactivation of memory representations of visual stimulus categories over time. To address this, we analyzed data collected from five participants while they encoded and retrieved word-object and word-place image pairs presented on a computer. We applied two multivariate methods to determine whether distinct patterns of neural activity during encoding were associated with object vs. place trials. We found that participant-specific logistic regression and support vector machine classifiers could differentiate between object and place trials in three of four participants using evoked sEEG activity from encoding with significant accuracy. We then applied these trained classifiers to retrieval data to examine whether category-specific patterns of activity were reactivated. Classifier accuracy was only significant for one of four participants which may have been due to poor overall performance on the behavioral task. Therefore, we altered the structure of the paired associate task and carried out a behavioral pilot experiment. We found that participants' memory for word image pairs was significantly greater than chance. Our analysis shows that sEEG data can be used to classify an individual brain's response to visual stimuli and reactivation of memory representations associated with those stimuli.
"},{"label":"Degree","value":["Bachelor of Arts with Departmental Honors"]},{"label":"Administrative Note","value":"Drupal Node ID: 23790