Combining neural oscillations, physiology and privacy-preserving LiDAR/millimeter wave sensing technology to track attention states in natural contexts

University Of California Los Angeles, Los Angeles CA

Investigators

Abstract

Project Summary Impairments of attention are common across neurodevelopmental disorders (e.g., ADHD), and contribute to negative life outcomes such as in academic achievement. Critically, leading treatment strategies are ineffective at improving educational outcomes. Lacking is assessment of attention deficits in real-life contexts, where visuospatial attention circuitry is influenced by interactions with regulatory systems including arousal and motivation. As such, tracking of neuro-behavioral attention states in natural environments has been increasingly recognized as crucial to: (i) understanding the interaction between neural circuitry of attention and regulatory influences, and (ii) identifying pathways to improving behavioral outcomes. Several technical challenges in this effort exist, however, and include quantification and synchronization of objective measures of such interacting systems with neural indicators of attention, portability of such multi-modal assessment systems, and increasingly, protection of privacy, with video recordings a gold standard in the field. The objective of this R61/33 proposal is to address these challenges. We will (Aim 1) develop an integrated, portable sensor suite for concurrent recording of neural activity, physiological arousal, motor signals, and physical interactions in social environments. We will integrate our previously developed tracking of neural oscillatory features of visual attention with additional sensors (heart rate, motor, and novel LiDAR/millimeter wave sensing) to extract physiological and movement-derived features of arousal and motivation. Project milestones quantify our aims to (i) optimize synchronization & portability, while (ii) introducing privacy-preservation technology to eliminate reliance on video recordings, and (iii) achieve above-chance classification of system states, and overall attention state as expressed in behavior. Next, we will deploy this technology to explain and predict neuro-behavioral attention states during mock- classroom and real-classroom learning activities, while manipulating contextual variables such as degree of attention support (group vs individual work), motivation (rewarded activities) and arousal (testing context). We will test the hypotheses (Aim 2) that (i) a multi-dimensional profile of attention states will improve prediction of neural activity and behavioral attention states and delineate how visual attention circuitry and regulatory influences interact in natural contexts, while also (ii) accounting for individual variability in inattention by differentiating between causal pathways to inattention, and (iii) identifying individual learning contexts that optimize an individual’s attention. The result of the project will be a scalable, integrated, portable technology designed to improve the accuracy of determining sources of inattention on individual basis, thus allowing for more targeted treatments.