Year: 2017
Type: dissertation
Abstract: We spend half of our waking time with thoughts independent of what is going on here and now, which is known as mind wandering.Previous research has characterized mind wandering in relation to brain networks suggesting its roles in cognitive and affective functions, but the neural representation of the thought contents is poorly understood.The ability to read out specific contents from brain activity would allow for much finer characterization as well as for applications to thought-controlled devices.Here, we performed brain decoding of thought contents (object/scene categories and time conditions) during mind wandering by applying machine-learning decoders trained on percept-induced and imagery-induced brain activity patterns (percept-and imagery-trained decoders).Our analyses indicate that objects/scenes viewed during mind wandering can be predicted with moderate accuracies using both of the perception-trained and imagery-trained decoders trained with activity in higher visual cortex and several areas associated with the default mode network (DMN).Furthermore, our analyses also show that time information of mind wandering can be classified using imagery-trained decoders trained with activity in several areas associated with the DMN.These results suggest common neural representations between task-related and task-unrelated thoughts, and open the possibility to predict thought contents of mind wandering from brain activity alone.
Author: 諏訪部, 開
Institution: Nara Institute of Science and Technology
Language: Japanese
Topic: Mind wandering and attention
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7144912313
