EEG Biomarkers Attention and Memory

Attention varies over time, and this affects episodic encoding.  Prior studies have identified neural subsequent memory effects, patterns of brain activity that can help predict whether subjects will remember information about a particular event, representing neural states of “readiness to encode”. To investigate whether this signal is sensitive to temporal selection, I conducted another free recall study where participants encoded words while performing target detection, recording scalp EEG activity throughout. The study found that the relationship between spectral EEG activity and memory encoding varied depending on whether participants responded to a target in the detection task. Importantly, I observed these effects both before and after stimulus presentation, indicating that a person’s brain state prior to encoding can interact with the attentional demands of a secondary task to influence memory formation. These findings, currently under peer review, suggest that “readiness to encode” information depends on the brain’s readiness to respond to important stimuli in the environment.

In my postdoctoral research at the University of Pennsylvania, I have expanded on my graduate work by studying the neural biomarkers of attention and memory in older adults. Do young and older adults display similar biomarkers of memory encoding, and are age-related differences in these biomarkers sensitive to task demands? To address this, I analyzed data from a multi-session EEG study of free recall in young (age 18-30) and older (age 61-85) adults. Older adults exhibited distinct subsequent memory effects, which varied depending on whether they performed a secondary task at encoding. This study presented the opportunity to use machine learning to identify neural features associated with age-related memory change. Using multivariate regression, I trained classifiers to predict encoding success at the single-trial level by aggregating thousands of features of neural activity across multiple sessions. Interestingly, although classifiers predicted encoding success with equal accuracy across young and older adults, they identified unique neural features related to memory in each age group (Fig. 2). The findings, which are under review [5], indicate that neural biomarkers of encoding change with age, highlighting important differences in how young and older brains process and retain information.