Independent Modulators of Ongoing Neural Activity Predict Future Perceptual Performance

Many internal mental processes, including emotional experience, attention, and workload are difficult to measure because they are not well defined by distinct behavioral events. Our research investigates the underlying modulatory processes that influence perception during multitask processing. We Focus specifically on neural processes responsible for 'inattentional deafness,' the tendency for individuals to miss clearly audible alarms under a complex multitask situation such as piloting of an airplane (Callan et al, 2018; Durantin et al, 2017). Such inattentional deafness has been found to be responsible for many aviation accidents. In this study we utilize a technique applying multiple levels of independent component analysis (ICA) to EEG recordings to extract independent modulators of neural activity potentially related to attention and workload. Independent modulators (IMs) are presumed dynamic processes co-modulating spectral activity within various brain networks (Onton & Makeig, 2009; Chuang, et al, 2012). Machine learning (ML) using sparse logistic regression (SLR) was used over pre-stimulus IM activity to predict detection of the audio alarm. Additional correlation analyses between the IMs and various flight parameters were used to determine IM relationships with the piloting task