Dorita Chang: “Reorganization of depth-related cortical processing with learning as shown by human rTMS”
CiNet B1F Seminar Room
Visual judgments require both the detection and segregation of meaningful signals from noise (coarse judgments), and fine discrimination between highly similar items (fine judgments). These two processes are thought to rely on dorsal and ventral processing, respectively. In this talk, I first review psychophysical work showing the role of perceptual learning in supporting signal-in-noise and fine discrimination processes. I will show initially, using binocular disparity perception as an example, that the two processes are facilitated by learning, but transfer in highly specific manner (i.e., learning transfer is best supported with fine discrimination training). I will show surprising consequences with signal-in-noise training however, with large learning transfer across different visual features (e.g., disparity, motion, or orientation). Next, I will review subsequent work where we used online repetitive transcranial magnetic stimulation (rTMS) to test whether training reorganizes signal-in-noise and fine discrimination processing in the human brain. Again using binocular disparity perception as an example, I will show that performance on the signal-in-noise task is significantly affected by parietal but not ventral stimulation. By contrast, judging fine visual differences is impaired by ventral stimulation, but not parietal TMS.
Intriguingly, this anatomical specificity is reconfigured following training.
These findings suggest that perceptual learning does not necessarily implicate changes in early sensory cortex. Instead, learning generalisation involves changes in the functional contributions of high-level sensory processing.