Abstract
In the vertebrate retina, a knowledge of the actual time course of a light-evoked increased in extracellular K/sup +/ concentration is used to provide a rigorous test of a hypothesis regarding the electrical origin of a clinically important component of the electroretinogram (ERG). A deconvolution technique is used to improve the estimation of the actual time course of the light-evoked increase in (K/sup +/)/sub 0/, and it is demonstrated that the ionic change is likely to be as rapid in time course (both latency and time-to-peak) as the b-wave. This result has rejected one of the strong challenges to the K/sup +/ hypothesis, namely, that the light-evoked increase in (K/sup +/)/sub 0/ was too slow to produce the b-wave. The use of the deconvolution technique is a significant improvement on the much less rigorous method in which the ERG was filtered by a low-pass filter. The technique should have widespread applications in many areas of physiology and neuroscience where it is critical to know the exact time course of stimulus-evoked changes in ion concentration.< >
