The time course of pattern discrimination in the human brain.

In electrophysiological experiments on visual pattern discrimination, decision difficulty was manipulated either via the physical characteristics of the test stimuli, or by changing the instruction given to the observer. Visual stimuli were rectangular matrices each composed of 100 Gabor patches having different orientations. Matrices differed in the number of Gabor patches with vertical, or horizontal, orientation. The observers' task was either to discriminate the dominant orientation or to detect collinear elements in the matrix. Relating task difficulty to performance, in the first experimental paradigm (detection of orientation) we obtained the conventional S-like psychometric function but in the second (detection of collinearity) the psychometric function showed a complicated U-curve. Matching between electrophysiological and psychophysical data and image statistical functions allowed us to establish the relative timing of the cortical processes underlying perception and decision making in relation to textural features. In the first 170ms after stimulus onset coding of the low-level properties of the image takes place. In the time interval 170-400ms, ERP amplitude correlated only with complex image properties, but not with task difficulty. The first effects arising from decision difficulty were observable at 400ms after stimulus onset, and therefore this is probably the earliest electrophysiological signature of the decision making processes, in the given experimental paradigm.

[Visual Priming and Perception of the Small Pictures in Scene with Multiscale Objects.]

Biederman and co-authors [1, 2], have shown that the priming effect in the long-tirn priming paradigm does not depend on the difference between the angular sizes of the test stimulus and the primer. However, these two and other similar works (both with long-time and short-fime priming paradigms) studied a small range of the angular sizes of stimuli. In Vakhrameeva et al. [3], it has been shown that there exist two perceptionally different size ranges: perception of the objects with angular size varying from 1-1.5 to 50 deg was found to be invariant, but for the objects which angular size is less than 1-1.5 deg (depending on object class and task) their perception is no longer invariant. In this work we have investigated the presence of priming effect in match-to-sample task with such a difference in the angular sizes of a primer and a test stimuli, when the sizes of the primer (about 4 deg) and the test stimulus (about 0.5 deg) belong to those different physiological size ranges. The sample stimulus was presented with and without the noise superposition. It has been shown that the priming effect is suppressed when the size difference between the primer and the test stimulus is large. A congruent primer can give a positive impact on the recognition of the test objects, but this takes place under viewing conditions complicated by the noise superposition.

[The Spatial-Frequency Characteristics of the Visual System in Schizophrenia].

In order to determine the spatial-frequancy characteristics of the visual system of healthy subjects and patients with schizophrenia, we used the contrast comparison of two Gabor gratings with sinusoidal distribution of brightness. The Gabor gratings have low, medium or high spatial frequencies; the neurons of magnocellular and parvocellular channels are sensitive to these frequencies to different extents. We found an increase in sensitivity to the contrast when comparing the gratings with low frequencies (to which magnocellular channels are most sensitive) in the patients with first-episode schizophrenia who had not receive long-term antipsychotic treatment, as compared with the control group. On the contrary, the sensitivity to the gratings with medium and high spatial frequencies in this group of patients was lower, as well as in patients with first-episode schizophrenia who had received long-term treatment. The patients with chronic schizophrenia showed a decrease in contrast sensitivity in all tested ranges of frequencies. We obtained supplementary evidence of the enhancement of internal noise in the visual system of the patients with schizophrenia. The results help us to explain the clinical data on the development of visual perceptual diorders at different stages of schizophrenia.

[The Effect of the Disease Duration on the Changes of the Visual Evoked Potentials and Contrast Sensitivity in Multiple Sclerosis].

The most sensitive methods to detect pathological changes in the visual system are the method of recording visual evoked potentials and the psychophysical method of measuring contrast sensitivity. Described in the literature features of functional disorders of the visual system in patients with multiple sclerosis are controversial. The results of the study allowed us to make an assumption about the depen-dence of the nature and severity of changes of the evoked potentials and contrast sensitivity and the duration of disease in patients with multiple sclerosis. In some patients with disease duration from 3 to 10 years there are irregularities in the magno-channels (reduced amplitude component P1), in others-- parvo-channels (amplitude reduction N 1) without increasing the latency, in patients with a disease duration of 10 to 14 years--both channels dysfunction (decreased amplitude components P1 and N1) with an increase of latency. The data indicate heterogeneity of pathophysiological changes upon increase of the degree of demyelination and damage of optic nerve fibers in multiple sclerosis.

Perception of fragmented images of three-dimensional objects as the observation angle changes.

Recognition thresholds for incomplete two-dimensional images of three-dimensional objects were measured as the observation angle was changed. A new experimental psychophysical method was developed and programmed for this purpose, this being a modification of the Gollin test, which measures perception thresholds of incomplete outline images. After training to a stimulus alphabet, observers' responses were found to be invariant to changes in the observation angles of three-dimensional objects from 15 degrees to 60 degrees. It is suggested that possible algorithms for the formation of models of three-dimensional images in the human visual system do not operate on the basis of simple section, but involve invariance mechanisms.