Responses were recorded in a 1.8-log unit range (+1.4 to +3.2 log phot td · s) of full-field, long-wavelength (red Wratten 29 λ >610 nm) strobe stimuli (Novatron, Dallas, TX) presented on a steady, rod-saturating background (∼ +3 log phot td · s). This approach has been used to isolate a cone response in adults with little rod intrusion.
13 The stimuli were incremented in 0.3-log unit steps. On records such as those shown in
Figure 1 , cone photoresponse parameters were derived from the a-wave. The trough-to-peak amplitude of the b-wave and implicit time of the b-wave were measured and examined as a function of log flash intensity.
The initial portion of the a-wave depends on the photocurrent in the rods and cones.
14 15 16 17 18 19 We took care to restrict fit of the transduction model to the early portion of the a-wave, to minimize postreceptoral contamination.
19 20 21 22 The cone photoresponse parameters were calculated by fit of a modification of the Lamb and Pugh
16 18 model of the activation of phototransduction to the first 11 ms of the a-wave. The modification incorporates a cascaded low pass exponential filter that models the capacitance of the cone membrane
13 23 by numerical convolution of the filter output with the delayed Gaussian function used to model the rod response.
16 18 The equation
13 used was
\[R(i,t){=}({\{}1{-}\mathrm{exp{[}{-}0.5}I\mathrm{\mathbf{S}_{\mathbf{CONE}}}(t{-}t_{\mathrm{d}})^{2}{]}{\}}\mathrm{\mathbf{R}_{\mathbf{CONE}})}{\ast}\mathrm{exp({-}}t/{\tau})\]
where
R CONE is the saturated response amplitude (microvolts),
S CONE the gain parameter (phot td
−1 · s
−3),
t d a brief delay (ms), and τ the time constant of the RC filter (in milliseconds). The symbol * represents the convolution operation. In preliminary studies, according to the procedure of Hood and Birch,
13 the effect of varying
t d (1–5 ms) and
τ (1–5 ms) on the root mean square (RMS) error of model fits was examined. For the final calculation of
S CONE and
R CONE , τ was fixed at 1.8 ms and t
d at 3 ms for both infants and control subjects. These values produced minimum RMS errors and are similar to those used previously.
13