Table 3.

Reading acuity.

Viewing condition (v)	Phosphene pattern density	MNREAD acuity (logMAR)	Logistic acuity (logMAR)	Logistic slope (%/logMAR)
1*	n/a	1.00 ± 0.00	n/a	n/a
2	High	1.14 ± 0.05	1.07 ± 0.07	338
3	Medium	1.31 ± 0.06	1.26 ± 0.06	292
4	Low	1.50 ± 0.03	1.48 ± 0.09	305

The first viewing condition, marked with an asterisk (

^*) is the Clear condition where text is presented in the clear as a control, without a simulation of artificial vision, as presented in Table 2. The values for reading acuity computed with the MNREAD technique, are shown for each phosphene pattern, illuminating the relationship between the number of phosphenes and the performance of the simulated artificial vision provided: Increasing the number of phosphenes by an approximate factor of 2 results in an improvement of about 0.2 logMAR in acuity. Acuity for the control condition was limited by the set of font sizes used, and is a very loose upper bound (contrast with the population mean acuity of 0.09 logMAR from the Snellen chart task). Values are population mean and standard deviation of per-subject computed values. MNREAD and logistic-fit acuities for each phosphene viewing condition were not significantly different from normal distributions (Kolmogorov–Smirnov test, p > 0.5 in all cases). The logistic acuity values represent the 50% threshold value (see Figure 6), a more robust metric than the MNREAD computation γ, and are consistently lower than the MNREAD values.