Purchase this article with an account.
Carly Thurston, Karen Dobkins; Stimulus-specific perceptual learning for chromatic, but not luminance, contrast detection. Journal of Vision 2007;7(9):469. doi: https://doi.org/10.1167/7.9.469.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: Improvement on a perceptual task is dependent, in part, on the plasticity of the visual area underlying the task. Here, we compared perceptual learning on a contrast detection task for chromatic (red/green, isoluminant) and luminance stimuli, subserved by magnocellular (M) and parvocellular (P) subcortical pathways, respectively, to investigate differences in plasticity between the two pathways. To ensure that improvements in sensitivity reflect learning specific to the targeted pathway, subjects were trained on one stimulus type and improvements were compared between the trained vs. untrained stimulus.
Methods: Using standard 2-AFC methods, contrast sensitivity for chromatic (CHR) and luminance (LUM) gratings (0.27cpd, 4 Hz, 7×7 deg, 9 deg eccentric to fixation) was measured using a staircase procedure. The experiment took place over three days: pre-training, training, and post-training. During pre-training and post-training, the two stimulus types were interleaved across trails (150 trials/staircase), whereas training consisted of only one stimulus type for each subject (5 staircases/150 trials each). Learning ratios (LR) were calculated as Sensitivitypost/Sensitivitypre, separately for “trained” and “untrained” stimuli, with LR values [[gt]] 1.0 reflecting learning. Fourteen subjects participated (7 CHR-trained, 7 LUM-trained).
Results: For CHR-trained subjects, LR for the trained stimulus was 1.42 (p=0.032), which was significantly higher (p=0.030) than for the untrained stimulus (1.05, p=0.06), suggesting stimulus specific learning. For LUM-trained subjects, LR for the trained stimulus (1.23, p=0.01) was not significantly higher than for the untrained stimulus (1.04, p=NS). The lack of a specificity effect for LUM-trained subjects was due to a large amount of variability in the LR for the untrained CHR stimulus.
Conclusions: Since CHR-trained, but not LUM-trained, subjects exhibited stimulus specific learning, LUM-training may not selectively tap the M pathway, either because the LUM task invokes the P pathway or because the learning takes place at a higher level in visual processing.
This PDF is available to Subscribers Only