Purchase this article with an account.
Li Guo, Alissa Stafford, Susan Courtney, Jason Fischer; The neural encoding of object hardness. Journal of Vision 2018;18(10):1151. doi: 10.1167/18.10.1151.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
In daily life, we continually form intuitions about how the physical world will behave. We "see" that a bulky door will require a hard push to open, or a coffee cup is precariously placed and may fall. To form these predictions about physical behavior, we must at the same time estimate the relevant physical properties of objects and surfaces in the scene (e.g., their hardness, smoothness, and density). Do these two facets of physical scene understanding – representing physical properties and mentally simulating physical dynamics – rely on common neural machinery? Here, we sought to identify brain regions that encode object hardness – a key determinant of behavior in physical interactions – and test whether they coincide with brain regions recruited by physical prediction tasks. In an fMRI experiment, we presented participants with images of everyday objects and tasked them with judging the hardness of each object. In separate scanning runs, we localized functional regions of interest for brain areas engaged when people observe and predict the unfolding of physical events (the "neural physics engine"; Fischer et al., 2016). Within the neural physics engine ROIs, a multivariate pattern analyses revealed parametric encoding of object hardness: the pattern of BOLD response varied smoothly and systematically with incremental changes in hardness. By contrast, we found no such encoding of hardness information in independently localized ventral object-selective regions. A searchlight analysis verified the robust encoding of object hardness within areas corresponding to the neural physics engine, but also uncovered a region in ventral premotor cortex that reliably displayed the most precise encoding of object hardness in the brain. This ventral premotor region may play a complementary role in physical scene understanding to other regions in the neural physics engine, inferring information about physical properties to be used in mental simulations of physical behavior.
Meeting abstract presented at VSS 2018
This PDF is available to Subscribers Only