A machine learning approach to predict perceptual decisions: an insight into face pareidolia.

The perception of an external stimulus not only depends upon the characteristics of the stimulus but is also influenced by the ongoing brain activity prior to its presentation. In this work, we directly tested whether spontaneous electrical brain activities in prestimulus period could predict perceptual outcome in face pareidolia (visualizing face in noise images) on a trial-by-trial basis. Participants were presented with only noise images but with the prior information that some faces would be hidden in these images, while their electrical brain activities were recorded; participants reported their perceptual decision, face or no-face, on each trial. Using differential hemispheric asymmetry features based on large-scale neural oscillations in a machine learning classifier, we demonstrated that prestimulus brain activities could achieve a classification accuracy, discriminating face from no-face perception, of 75% across trials. The time-frequency features representing hemispheric asymmetry yielded the best classification performance, and prestimulus alpha oscillations were found to be mostly involved in predicting perceptual decision. These findings suggest a mechanism of how prior expectations in the prestimulus period may affect post-stimulus decision making.