Action Prediction Based on Anticipatory Brain

12 Objective. The ability of an automobile to infer the driver’s upcoming actions directly 13 from neural signals could enrich the interaction of the car with its driver. Intelligent 14 vehicles fitted with an on-board brain-computer interface (BCI) able to decode the 15 driver’s intentions can use this information to improve the driving experience. In 16 this study we investigate the neural signatures of anticipation of specific actions, 17 namely braking and accelerating. Approach. We investigated anticipatory slow cortical 18 potentials (SCPs) in electroencephalogram (EEG) recorded from 18 healthy participants 19 in a driving simulator using a variant of the contingent negative variation (CNV) 20 paradigm with Go and No-go conditions: count-down numbers followed by ‘Start’/‘Stop’ 21 cue. We report decoding performance before the action onset using a quadratic 22 discriminant analysis (QDA) classifier based on temporal features. Main Results. (i) 23 Despite the visual and driving related cognitive distractions, we show the presence of 24 anticipatory event related potentials locked to the stimuli onset similar to the widely 25 reported CNV signal (with an average peak value of -8μV at electrode Cz). (ii) We 26 demonstrate the discrimination between cases requiring to perform an action upon 27 imperative subsequent stimulus (Go condition, e.g. a ‘Red’ traffic light) versus events 28 that do not require such action (No-go condition; e.g. a ‘Yellow’ light); with an 29 average single trial classification performance of 0.83±0.13 for braking and 0.79±0.12 for 30 accelerating (area under the curve). (iii) We show that the centro-medial anticipatory 31 potentials are observed as early as 320±200 ms before the action with a detection rate 32 of 0.77±0.12 in offline analysis. Significance. We show for the first time the feasibility of 33 predicting the driver’s intention through decoding anticipatory related potentials during 34 simulated car driving with high recognition rates. 35 Action Prediction Based on Anticipatory Brain Potentials during Simulated Driving 2