Figure 6. Training improves discrimination thresholds for trained stimuli, regardless of lesion type. A, In this first experiment, cats were trained to discriminate the left–right direction of motion of sinewave gratings at 0.3 cycles/degree spatial frequency and drifting at 6 Hz. A staircase was used to progressively decrease stimulus contrast in each training session. B, Plot of contrast threshold in cats with 17/18/19 lesions in the intact hemifield (good field), in the impaired hemifield before training (bad field initial), and at the same location post-training (bad field post-training). Training gradually improved contrast thresholds at the trained locations in the contralesional visual hemifield (Fig. 5) until contrast thresholds were no longer significantly different from those in the intact hemifield of vision. C, Cats with LS lesions did not exhibit a significant deficit in contrast sensitivity for direction (compare black and white bars) and additional practice of this task in the impaired hemifield did not further improve the animals' contrast thresholds (gray bar). D, At a different, nonoverlapping location in their impaired hemifield, the same cats were trained to discriminate the left–right direction of motion of random dot stimuli in which the range of dot directions was varied from easier to harder levels in each session. E, Before training, cats with 17/18/19 lesions exhibited significantly raised DR thresholds (black bars). Training improved these abnormal thresholds back to normal levels (gray bars). F, A very similar pattern of postlesion deficit and training-induced recovery of DR thresholds was seen in cats with LS lesions. G, Finally, at yet another location in the impaired hemifield of vision, cats were trained to discriminate the left–right direction of motion of random dot stimuli in which the proportion of coherently moving dots was varied using a staircase procedure in each session. H, Before the onset of training, cats with lesions of areas 17/18/19 exhibited significantly higher coherence thresholds in the contralesional (black bars) versus the ipsilesional hemifield (white bars). Just as in B and E, specific training gradually improved impaired coherence thresholds back to normal levels (gray bars) at the trained locations. I, A similar pattern of postlesion deficit and training-induced recovery of coherence thresholds was seen in cats with LS lesions. The number of cats used in each portion of the experiment is shown as N. All values given are means ± SEM. *p < 0.05, paired Student's t tests relative to intact hemifield values. CS, Contrast sensitivity; Coh, coherence.