Abstract
Categorization is a process by which behavioral significance is assigned to sensory stimuli. While much known about neural encoding of simple visual features, less is known about how the brain determines stimulus category. We trained monkeys to group random-dot motion stimuli into two categories. 360° of motion directions were divided by a learned “category boundary”. Directions between −45° and 135° were in one category while the remaining directions were in another. Monkeys performed a delayed match-to-category (DMC) task and indicated, by releasing a lever, whether two stimuli (sample and test) were in the same category. After training, the monkeys' categorization performance was excellent: they performed with >80% accuracy, even for stimuli near the category boundary.
We recorded 135 LIP and 40 MT neurons during DMC task performance. A majority of neurons were direction selective (one-way ANOVA, 12 directions, P<0.01) during the sample or delay epoch (LIP: 97/135; MT: 40/40). However, the pattern of direction selectivity differed strikingly between LIP and MT. LIP neurons reflected stimulus category: activity was similar for directions in the same category and differed for directions of different categories. A category-tuning index revealed larger activity differences for directions in different categories and more similar activity for directions in the same category (T-Test, P<0.001). In MT, neurons did not group directions by their category. The category-tuning index revealed that MT neurons showed no influence of the boundary (P>0.5).
This suggests that LIP encodes the behavioral relevance of visual motion, while MT provides a more faithful representation of motion direction.