Shifts in cortical representations predict human discrimination improvement

Abstract

We report experiments combining assessment of spatial tactile discrimination behavior and measurements of somatosensory-evoked potentials in human subjects before and after short-term plastic changes to demonstrate a causal link between the degree of altered performance and reorganization. Plastic changes were induced by a Hebbian coactivation protocol of simultaneous pairing of tactile stimuli. As a result of coactivation, spatial discrimination thresholds were lowered; however, the amount of discrimination improvement was variable across subjects. Analysis of somatosensory-evoked potentials revealed a significant, but also variable shift in the localization of the N20-dipole of the index finger that was coactivated. The Euclidean distance between the dipole pre- and post-coactivation was significantly larger on the coactivated side (mean 9.13 +/- 3.4 mm) than on the control side (mean 4.90 +/- 2.7 mm, P = 0.008). Changes of polar angles indicated a lateral and inferior shift on the postcentral gyrus of the left hemisphere representing the coactivated index finger. To explore how far the variability of improvement was reflected in the degree of reorganization, we correlated the perceptual changes with the N20-dipole shifts. We found that the changes in discrimination abilities could be predicted from the changes in dipole localization. Little gain in spatial discrimination was associated with small changes in dipole shifts. In contrast, subjects who showed a large cortical reorganization also had lowest thresholds. All changes were highly selective as no transfer to the index finger of the opposite, non-coactivated hand was found. Our results indicate that human spatial discrimination performance is subject to improvement on a short time scale by a Hebbian stimulation protocol without invoking training, attention, or reinforcement. Plastic processes related to the improvement were localized in primary somatosensory cortex and were scaled with the degree of the individual perceptual improvement.

Figure 1

Effects of coactivation on discrimination thresholds (n = 16). Dots represent mean thresholds, boxes show the standard errors, and whiskers correspond to the standard deviation. Coactivation period (3 h) is indicated by an arrow. (Left) Shown are results from five consecutive sessions before coactivation. After the fifth session (precondition), coactivation was applied. Testing was continued for two consecutive sessions after coactivation. Note decrease in thresholds after coactivation (session 6, postcondition, P < 0.005 pre-post), but recovery of effects after 24 h of termination of coactivation (session 7) with continuation of stable pretest performance. (Right) Discrimination thresholds obtained for the control finger (index finger of the left hand that was not coactivated) on session 5 (precondition). Note lack of effects after coactivation of the right index finger (session 6, postcondition), indicating finger specificity of the coactivation protocol.

Figure 2

Psychometric functions illustrating the coactivation-induced improvement of discrimination threshold for an individual subject. Correct responses (“two”) in percent (⧫) are plotted as a function of separation distance together with the results of a logistic regression line. (Top) Session 5, precondition before coactivation. (Middle) Session 6, postcondition immediately after coactivation. (Bottom) Session 7, recovery assessed after 24 h of termination of coactivation. 50% level of correct responses is indicated. After coactivation there is a distinct shift in the psychometric functions toward lower separation distances, an improvement that recovers 24 h after coactivation.

Figure 3

Example of SSEP mapping. (Upper) Comparison of the N20-dipole before and after coactivation obtained in the same subject whose psychometric functions are shown in Fig. 2. A spherical head model was used. The points symbolize the electrodes, and the arrow indicates the position and the orientation of the dipole (viewed from the top). (Lower) Distribution of electrodes.

Figure 4

(A) Schematic projection of the average locations (n = 10) of the single equivalent N20-dipoles of the index fingers pre-coactivation (blue symbols) and post-coactivation (red symbols) onto an axial (Left) and a coronar MR slice (Right) of an individual subject. The average difference (pre-post) for the Euclidean distances of the N20 of the index finger of the coactivated and of the control hemisphere are shown (Left). (Right) The average positions of the N20-dipoles are given by the polar angles showing a coactivation-induced shift toward the lateral and inferior aspects of the postcentral gyrus. A comparable effect is lacking on the non-coactivated hemisphere. (B Left) Effects of coactivation on the polar angle of the N20-dipole referred to the z axis recorded in the left hemisphere after electrical stimulation of the right index finger. Dots represent angles, boxes show the standard errors, and whiskers correspond to the standard deviation. Coactivation period (3 h) is indicated by an arrow. Note increase of angle after coactivation (P < 0.005), but recovery after 24 h of termination of coactivation. (Right) Euclidean distance between the N20-dipole location before and after coactivation for the test and the control finger (hemispheric difference, P < 0.05). Note lack of effects indicating finger specificity of the coactivation-evoked SSEP changes.

Figure 5

Correlation between the coactivation-induced changes of (Left) polar angles (pre-post) and the changes of two-point discrimination thresholds for individual subjects (r = 0.765, P = 0.01), and (Right) between Euclidean distances (normalized to left–right) and threshold changes (r = 0.844, P = 0.002). Large gain in spatial discrimination abilities was associated with large dipole shifts and vice versa. Intermediate improvement correlates with intermediate cortical reorganization.