Artificial Vision: Vision of a Newcomer - Artificial Sight: Basic Research, Biomedical Engineering, and Clinical Advances - page 37

strip 3mm wide, 4mm long, and 180m thick. The dimensions of each
electrode were height above the surface = 120-130m; diameter = 250m;

=

distance between electrodes 500rmm. Their impedance in saline was

approximately 10 k at 1 kHz.
The bundle of insulated leads from the microarray was also sutured at
the limbus with 5-0 Dacron. A stimulator (SEN-7203, NIHON KOHDEN,
Shinjyuku, JAPAN) was then connected through an isolator (A-395R, World
Precision Instruments INC., Sarasota FL USA) to the microelectrode array. The
recording electrode was a screw in the skull bone above the visual cortex. The
EEPs were elicited by monophasic electrical pulses of 0.5ms duration. The
direction of the current was set for inward-flowing currents (electrode array was
positive and reference electrode was negative). Fifty responses were averaged.
A band-pass filter of 5-1 kHz was used.

Threshold Current by STS in Rabbit

In order to determine the minimum threshold of the EEP, each of the eight
electrodes was stimulated with a current of 500A, and the amplitudes of
EEP waves were compared (Figure 2.3b). The electrode that elicited the largest
EEP was selected to determine the threshold current. The electric current was
decreased in steps, and the minimum electric current that elicited the first or
second positive peaks of the EEP was defined as the threshold current. The
relationship between the second and the third peak amplitude of the EEPs and
the inward current was examined in six rabbits (Figure 2.4). This regression
curve showed that the EEP amplitude increased almost linearly at lower stimulus
currents and tended to be saturated with higher stimulus currents. The mean
threshold current that elicited a small EEP was 550

± A for the six rabbits

102C/cm 2 ).
Humayun et al. reported that a charge density of 892-119C/cm 2 was
required to elicit EEPs from rabbits with epiretinal electrodes [5].In exper-
iments on cats, Dawson and Radtke found a threshold charge density of
305C/cm 2 16. Chow et al. implanted their electrodes subretinally in rabbits,
and reported a threshold charge density of 28nC/cm 2 to elicit EEPs [9].
Although the distance from the electrodes to the retina was farther in our
transcleral electrodes than that of the epi or subretinal electrodes, the threshold
charge density to elicit EEPs was comparable with the other electrode placements.
McCreery et al. [17] found that the threshold for stimulation-induced neural
damages is determined by a synergetic interaction between charge density and
charge per phase, and the limit for safe charge density decreased as charge
density increased. With our electrodes, 50 NC/Phase would give a charge density
of approximately 110C/cm 2 and the work of McCreery et al. suggests that
this combination would not be injurious.
Our electrodes were not in direct contact with the neural tissue. Therefore,
the limit of electrical charge density may be expected to be higher than that
with epiretinal stimulation. Indeed, Nakauchi K. et al. [15] reported that for
100m Pt electrode by STS, retinal tissue was not damaged by a current up

± ±

(275 50nCor56 0