Font Size: a A A

Neural transformation from object features to object value

Posted on:2013-08-17Degree:Ph.DType:Dissertation
University:The Johns Hopkins UniversityCandidate:Sasikumar, DennisFull Text:PDF
GTID:1458390008965322Subject:Biology
Abstract/Summary:
Our behavior towards objects in our environment is guided by our perception of their value. However, this value is not represented explicitly in the object. Our brains only have access to the visual image that these objects form on our retinae. Value must be inferred from object shape, typically from sensory features that encode value in a particular category of objects (e.g. color variations tell us whether a banana is edible).;We studied the neural transformation of object information from features to value with microelectrode recording in inferotemporal cortex (IT, the highest level of the ventral visual pathway) and prefrontal cortex. We trained two macaque monkeys on a visual comparison task that required evaluating the reward value of four categories of letter-like shapes based on within-category shape variation. During task performance, information about object category and selective tuning for value-defining feature emerged early in IT, preceding value information in PFC by about 40 ms.;These results reveal a dynamic process through visual and prefrontal cortex that follows the sequential logic of first establishing categorical identity, then applying appropriate evaluative criteria for that category, and finally resolving value. Moreover, by selectively processing features that define value in the current object category, IT provides an initial source of object value information that can be integrated by downstream decision-making processes in higher cognitive areas.
Keywords/Search Tags:Value, Object, Features, Information
Related items