The FM-VA test is a new type of visual acuity test that targets the neural substrate of the visual system. The FM prefix stands for "flash" and "mask" and refers to the brief presentation time (50 milliseconds) and the use of partially masked test letters. Conventional acuity tests use unrestricted viewing times and fully formed letters. These latter conditions are useful when targeting optical faults but they are not optimal for neural faults [1]. The present version of the FM-VA test aims for self-monitoring of neural defects and effects of treatment. The test is self-paced and self-scored. It presupposes appropriate correction of any refractive errors.

It is well known that acuity drops monotonically with presentation times shorter than 1 s [2, 3]. An individual who normally maxes out at the 1.0 line (20/20, 6/6, 1') on a letter chart will typically manage no more than 0.5 (20/40, 6/12, 2') with a presentation time of 50 ms. There are likely several factors at play, not the least the constraint that brief exposures impose on the important roles that micro-saccades, drifts, and tremor subserve in vision. The drop effect is likely to be larger than normal in subjects who have lost receptive fields (or their upstream connections) from neural damage because a briefly exposed test target would appear to be more or less fragmented, depending on the degree of damage (see simulation). Notably, fragmentation will not be perceived during long exposures because perpetual eye movements allow a subconscious piecing-together of different fragments to a more complete percept. Interestingly, subjects with neural defects may actually benefit from the addition of movement to test targets [4, 5]. By constraining the roles of movements, flashed presentations can be expected to improve detection and quantification of neural defects. FM-VA uses 50 ms presentations, which can be realized in consumer-grade displays.

Turning to the test targets themselves, it is known that normal subjects tolerate masking by some 50 per cent before acuity begins to drop [5, 6, simulation]. Masks can be viewed as analogues to losses of receptive fields (or their upstream connections) from neural damage. 50 per cent masks spare acuity and so do losses of some 40 per cent of optic nerve axons [7]. These observations point to the existence of an overload of information in conventional acuity targets. Reduction of this overload by masking should enhance sensitivity to neural damage.

The FM-VA test presents 4 conventional test letters in a square layout. The reason for selecting this layout is that the neural substrate should be probed not only in the horizontal direction but also in the vertical one. The test letters are drawn at random from the HOTV subset and are combined with 50 per cent masks. Mask element size is fixed at 1 pixel; the "on" elements are distributed at random. Ideally, the mask elements should subtend a visual angle of 0.7', i e, the size of foveal receptive fields. Some display units produce pixels that are larger than ideal for the intended viewing distance (0.4 m). This may result in a somewhat diminished sensitivity to neural damage. Test results are presented in the format of frequency-of-seeing plots.

The FM-VA test presupposes clear optical media and appropriate correction of any refractive errors. The results are not directly comparable to those of conventional acuity tests. Normal limits will depend on the local test conditions but are not required for the intended use, namely, self-monitoring of any changes in performance over time.