Practising coarse orientation discrimination improves orientation signals in an fMRI-defined region of the macaque posterior inferior temporal cortex

We have recently shown that practising coarse orientation discrimination improves orientation signals in the macaque visual area V4 (Zivari Adab and Vogels, Current Biology, 2011). In this task, monkeys discriminated grating orientations that differed by 90 deg and were masked by noise. V4 neurons showed robust increases in their capacity to discriminate the trained orientations during the course of the training. To identify other areas that are possibly involved in perceptual learning of coarse orientation discrimination, we determined with fMRI which areas were activated by high noise (20% SNR) gratings in the two trained monkeys. Trained and untrained orientations (differing by 45 degree from the trained) were presented at trained and untrained (opposite hemifield) locations using an event-related design during passive fixation. In both animals, activations (high noise grating compared to noise) were present in V2/V3, V4, posterior inferior temporal (PIT) and prefrontal cortex, with the strongest activations in V4. Based on this fMRI study we subsequently recorded single cells in the activated PIT region (close to the anterior tip of the posterior middle temporal sulcus) of both animals. Orientation tuning was assessed during passive fixation with 80% SNR gratings. The majority of neurons preferred one of the two trained orientations and the mean normalized net response was significantly greater for the trained compared to the untrained orientations. These anisotropies were not related to orientation per se but were induced by learning since trained and untrained orientations differed between the two monkeys. Single neurons discriminated more reliably (assessed by ROC analysis) the two trained orientations compared to two untrained orientations differing by 45 degree from the trained ones. This higher discriminability for the trained compared to the untrained orientations was not due to a higher attention to the trained orientations since it was also present when the monkeys were performing a difficult, orthogonal color discrimination task. Previously we showed significant correlations between the monkey’s decision and the neural responses (Choice Probabilities: CPs) in area V4 during coarse orientation discrimination. CPs of the PIT neurons were similar to those observed in V4 and were larger for neurons that discriminated the trained orientations with high accuracy. In conclusion, using a combined monkey fMRI-single cell study we identified an orientation selective region in PIT with response properties that are affected by extensive practising a coarse orientation discrimination task.status: publishe