Luminance-polarity distribution across the symmetry axis affects the electrophysiological response to symmetry

Abstract

Electrophysiological studies of symmetry have found a difference wave termed the Sustained Posterior Negativity (SPN) related to the presence of symmetry. Yet the extent to which the SPN is modulated by luminance-polarity and colour content is unknown. Here we examine how luminance-polarity distribution across the symmetry axis, grouping by luminance polarity, and the number of colours in the stimuli, modulate the SPN. Stimuli were dot patterns arranged either symmetrically or quasi-randomly. There were several arrangements: 'segregated'-symmetric dots were of one polarity and randomly-positioned dots were of the other; 'unsegregated'-symmetric dots were of both polarities in equal proportions; 'anti-symmetric'-dots were of opposite polarity across the symmetry axis; 'polarity-grouped anti-symmetric'-this is the same as anti-symmetric but with half the pattern of one polarity and the other half of opposite polarity; multi-colour symmetric patterns made of two, three to four colours. We found that the SPN is: (i) reduced by the amount of position-symmetry, (ii) sensitive to luminance-polarity mismatch across the symmetry axis, and (iii) not modulated by the number of colours in the stimuli. Our results show that the sustained nature of the SPN coincides with the late onset of a topographic microstate sensitive to symmetry. These findings emphasise the importance of not only position symmetry, but also luminance polarity matching across the symmetry axis.

Keywords: Colour; EEG; ERP; Luminance-polarity; Microstates; Sustained Posterior Negativity; Symmetry.

Fig. 1

Examples of 100% position-symmetric patterns with different luminance-polarity arrangements. (A) Symmetric pattern in which elements across the vertical axis are matched in luminance polarity. (B) Anti-symmetric pattern in which the matched pairs across the symmetry axis have opposite luminance-polarity. (C) Polarity-grouped anti-symmetric pattern in which each half of the pattern is of different luminance-polarity. Note, each of the patterns contains the same amount of positional information but they differ in regard to their luminance-polarity arrangement.

Fig. 2

Examples of 100% position-symmetric stimuli used in Experiment 1 (A–C), 50% position-symmetric stimuli used in Experiment 2 (D–H), and chromatic stimuli used in Experiment 3 (I–K). A) Anti-symmetric patterns in which position-symmetric dots were of opposite luminance-polarity across the symmetry axis; B) Polarity-grouped – the same as the anti-symmetric condition but with one half of the pattern of one luminance polarity and the other of opposite polarity; C) Unsegregated – symmetric-pairs were the same luminance polarity, with an equal number of black and white dot pairs. (D–H) Examples of 50% position-symmetric patterns used in Experiment 2: (D) Anti-symmetric; (E) Polarity-grouped anti-symmetric; (F) Unsegregated; (G) Segregated in which symmetrical pairs are black and noise dots are white; (H) Single polarity patterns. In this experiment, we also used single polarity patterns with 100% position symmetry (not shown). (I–K) Examples of 100% position-symmetric chromatic stimuli made of two (I), three (J) or four (K) colours.

Fig. 3

Results for Experiment 1. (A) Performance (% correct answers) in the symmetry detection task with symmetric, anti-symmetric and luminance-polarity grouped patterns containing 100% position symmetry. (B) Grand-average ERPs for anti-symmetric (red), polarity-grouped anti-symmetric (pink), unsegregated (burgandy), noise (blue) and polarity-grouped noise (light blue) conditions. (C) Difference waves (Symmetry – Noise) for each condition. Waveforms depict the average of electrodes PO7 and PO8. (D–F) Topographic difference map for anti-symmetric condition (D), polarity-grouped anti-symmetric (E) and unsegregated (F). Each topographic difference map shows the difference between symmetry and noise in the 200–600 ms time window. Black dots indicate the position of electrodes PO7 and PO8.

Fig. 4

TANOVA and microstate segmentation analysis for Experiment 1. (A) Grand average ERPs at electrode sites PO7 & PO8 with light grey areas showing periods of stable topographic differences determined by the TANOVA. (B) Onset and offsets of the stable topographic microstates between 200 and 600ms and Global Field Power waveforms for each condition. Global Field Power waveforms, a measure of differences in the scalp electric field strength, are displayed for a visual comparison to topographic microstates which are independent of field strength. The higher GFP the more stable EEG topography and the higher global excitation. Each map is represented by a different colour. (C) Topographic microstates derived from the segmentation procedure for four maps which best fit the individual subject data between 200 and 600 ms. Topographic maps show the head from above with nasion plotted upward.

Fig. 5

Results for Experiment 2. (A) Performance (% correct answers) in the symmetry detection task with segregated, unsegregated, anti-symmetric, polarity-grouped anti-symmetric and single polarity patterns containing 50% position symmetry. (B) Grand-average ERPs (left) and SPN (right) difference wave for anti-symmetric, polarity-grouped anti-symmetric, segregated and unsegregated patterns; (C) Grand-average ERPs (left) and the SPN difference wave for single-polarity patterns containing 50% and 100% position symmetry. Waveforms depict the average of electrodes PO7 and PO8. (D–H) Topographic difference map for anti-symmetric (D), polarity-grouped anti-symmetric (E), unsegregated (F), segregated (G) and single polarity patterns containing 50% and 100% position symmetry (H). Each topographic difference map shows the difference between symmetry and noise in the 200–600 ms time window. Black dots indicate the position of electrodes PO7 and PO8.

Fig. 6

Topographic and microstate segmentation analysis for Experiment 2. (A) Grand average ERPs represent the average of posterior electrodes PO7 & PO8, with light grey areas showing periods of stable topographic differences determined by TANOVA. (B) Onset and offsets of topographic microstates between 200 and 600 ms and Global Field Power waveforms for each condition. Global Field Power waveforms, a measure of differences in the scalp electric field strength, are displayed for a visual comparison to topographic microstates which are independent of field strength. The higher GFP the more stable EEG topography and the higher global excitation. Each map is represented by a different colour. (C) Topographic microstates derived from the segmentation procedure for four maps which best fit the individual subject data between 200 and 600 ms. Topographic maps show the head from above with nasion plotted upward.

Fig. 7

Results for Experiment 3. (A) Performance (% correct answers) in the symmetry detection task with two, three and four colours. (B) Grand-average ERPs for two colour symmetry (red) and noise (blue), three colour symmetry (pink) and noise (light blue) and, four colour symmetry (dark red) and noise (cyan) conditions. (C) SPN difference waves (Symmetry–Noise) for each condition. Waveforms depict the average of electrodes PO7 and PO8. (D–F) Topographic difference map for two (D), three (E) and four (F) colours. Each topographic difference map shows the difference between symmetry and noise in the 200–600 ms time window. Black dots indicate the position of electrodes PO7 and PO8.

Fig. 8

TANOVA and microstate segmentation analysis for Experiment 3. (A) Grand average ERPs represent the average of posterior electrodes PO7 & PO8, with light grey areas showing periods of stable topographic differences determined by TANOVA. (B) Onset and offsets of the stable topographic microstates between 200 and 600 ms and Global Field Power waveforms for each condition. Global Field Power waveforms, a measure of differences in the scalp electric field strength, are displayed for a visual comparison to topographic microstates which are independent of field strength. The higher GFP the more stable EEG topography and the higher global excitation. Each map is represented by a different colour. (C) Topographic microstates derived from the segmentation procedure for four maps which best fit the individual subject data between 200 and 600 ms. Topographic maps show the head from above with nasion plotted upward.