Diagnosis of colour vision deficits using eye movements

Abstract

We set out to develop a simple objective test of functional colour vision based on eye movements made in response to moving patterns. We exploit the finding that while the motion of a colour-defined stimulus can be cancelled by adding a low-contrast luminance-defined stimulus moving in the opposite direction, the "equivalent luminance contrast" required for such cancellation is reduced when colour vision is abnormal. We used a consumer-grade infrared eye-tracker to measure eye movements made in response to coloured patterns drifting at different speeds. An automated analysis of these movements estimated individuals' red-green equiluminant point and their equivalent luminance contrast. We tested 34 participants: 23 colour vision normal controls, 9 deuteranomalous and 2 protanomalous individuals. We obtained reliable estimates of strength of directed eye movements (i.e. combined optokinetic and voluntary tracking) for stimuli moving at 16 deg/s and could use these data to classify participants' colour vision status with a sensitivity rate of 90.9% and a specificity rate of 91.3%. We conclude that an objective test of functional colour vision combining a motion-nulling technique with an automated analysis of eye movements can diagnose and assess the severity of protanopia and deuteranopia. The test places minimal demands on patients (who simply view a series of moving patterns for less than 90 s), requires modest operator expertise, and can be run on affordable hardware.

Figure 1

(Top row) Two superimposed gratings moving in opposite directions, with the contrast of one grating fixed (here at 8%) and the other varying. (middle rows) The observer's optokinetic response to the summed gratings, generally follows the direction of the grating that appears higher-contrast. When the contrasts are perceptually matched the observer perceives only flicker, and the motion of the fixed grating is "nulled". (bottom row) A relatively low contrast (here 8%) luminance-defined grating can null the motion of a high-contrast equi-bright colour grating.

Figure 2

Expected results from the anomaloscope for different CVd groups. The x-axis shows the ratio of red-green in the reference ranging from 0 (pure green) to 73 (pure red). The y-axis indicates the luminance of the yellow test scaled from 0 (black) to 40 (maximum brightness)—that matched the corresponding red-green test. Lines are labelled with the corresponding CV category. Because the red and green primaries of the anomaloscope lie along the red-green axis, the test captures the colours readily confused by dichromats. As a result, extreme dichromats can match the yellow test to red-green references spread across the entire colour mixture range—solid arrows. CVn observers on the other hand see contrast between colours and will match only at near-equal ratios of red and green (40–50). Mild dichromats—dashed lines being less sensitive to either red or green, require a greater amount of that colour to make a luminance match, thus matching values on either side of the CVn range.

Figure 3

Perception of direction in the eye-tracking test. (Top row) Stimuli are comprised of a fixed-contrast (20%) yellow luminance grating added to a R-G coloured grating. (Middle row) When a yellow grating is superimposed on an equiluminant RG grating (R50:G50), a CVn observer will report the direction of the yellow component, because its contrast exceeds C_Eq. Increasing the luminance of either the red or green component first nulls the fixed-contrast component (leading to flicker) and then exceeds it (at which point observers report direction of the R-G grating). (Bottom row) By comparison, an observer with protanomalous colour vision who perceives red as weaker than green—produces responses shifted right along the stimulus axis. Note how the observers require more red (here R62:G38) to achieve equi-brightness.

Figure 4

Schematic DEM gain response for (a) CVn, (b) CVd (deut) and (c) CVd (prot) participants viewing variable-contrast colour gratings in the presence of a fixed-contrast (20%) luminance grating. (a) Expected DEM-gain plot against red/green colour mix of the coloured grating. Points A and E show a strong positive response (in the direction of the coloured grating) when the colour mix is dominated by either red or green. When colour-mix leads to equi-brightness (point C; quantified as B_Red), DEM is poorly driven by the coloured grating and switches to the direction of the luminance-defined grating. At points marked $\otimes$ colour and luminance gratings elicit similar DEM in opposite directions, leading to motion nulling. (b) and (c) With a CVd (deut) and CVd (prot) observer the tip of the “V-function” is shifted horizontally relative to equiluminance (50–50); e.g. (b) leftwards for a deuteranope and (c) rightwards for a protanope. Boxed equations in (b) and (c) explain how equivalent luminance contrast (C_Eq) is estimated from shifting of motion null-points caused (here) by the V-function shifting downwards (Video 2).

Figure 5

Representative ‘V-functions’ for a 20% fixed luminance-contrast component drifting at 16 deg/s. Plots show data from (a) a participant with normal colour vision, (b) a deuteranope and (c) a protanope. In total 8 trials for each red-green condition were carried out: in 4 trials the coloured grating moved left (red symbols), and in 4 it moved right (blue symbols). The V function was fit to the average DEM-gain across trials (small open symbol). Note the left/right shifting of the V functions in (b) and (c), indicating a shift in the red-green equiluminant point. Note also that in (b, c) the end-points of the V-functions are shifted downwards indicating a weak contribution of colour to these participants’ perception of motion. The raw eye-position plots at the bottom of the figure are for conditions corresponding to the circled datapoints in the graphs.

Figure 6

DEM gain (signed for direction of colour component as Fig. 5) measured with a 20% fixed luminance-contrast component drifting at 16 deg/s. Data have been plot on common axes to allow visual comparison of the ‘V’ fits; some data lies outside the axis range. Plots in green, red and white panels show data from deuteranopes, protanopes, and CVn observers respectively. Red and blue symbols show gain-estimates from trials when the coloured grating moved left or right respectively, with the mean-gain across trials indicated by open symbols.

Figure 7

Equibrightness plot against equivalent luminance contrast for (a–c) our 30 participants and (d–f) Cavanagh and Anstis’ participants, run in the three speed conditions. Boundaries of the coloured regions were derived using a K-means algorithm that sought to best separate the three groups. Error rates indicate the percentage of mis-classified participants.

Figure 8

Individual estimates of matching range taken from the anomaloscope test (Appendix A), plot against individual estimates of Euclidean distance from the CVn centroid, for the three speed conditions. Larger matching ranges in the anomaloscope, and larger Euclidean distances are both indicative of more severe CVd. Line of best fit is derived using linear regression.

Figure 9

(a, b) ‘V’ functions of the 3/6 CVn participants that were misclassified (based on results from the 4 deg/s conditions) for both (a) 4 and (b) 16 deg/s conditions. Stimuli moving at 4 deg/s elicit higher but inconsistent DEM gain (as measured by the MSE). More consistent but lower DEM gains are elicited by faster stimuli which leads to more robust fits of the ‘V’ function centred around B_Red = 50%. (c) MSE of the V-fit across speed conditions. The mean MSE at both 8 and 16 deg/s are statistically significantly lower than measures made at 4 deg/s. (d) Box and whisker plot indicate little difference in absolute direction-bias across the different test-speeds. (e) Switching magnitude plot across speed conditions. There is significant difference in mean SD of DEM-gain for both 8 deg/s and 16 deg/s compared to 4 deg/s. Note that outlier analysis has been applied by removing data (indicated by crosses) that is more than three scaled median absolute deviations (MAD) away from the median.