5 Results and Discussion

Though the information content is maintained across degeneration, animals still experience a decline in their visual perceptions. Thus, this method for information calculation may not necessarily be a good metric for evaluating the retina’s code after disease. Mutual information estimation is is a function of an estimated probability distribution used to describe the stimuli and the responses. If retinas were not exposed to a dynamic range of stimulus, then estimated response distribution may not be accurate. Another explanation for the lack of change in information transmission could be the bias in the probability distribution estimate. It is clear that undersampling of the pattern distribution creates a downward bias in information estimates (Paninski) from an insufficient quantity of trials. Whether that bias is consistent across degeneration is uncertain. Thus, information theory is not necessarily the best way to study a biological system as there are more questions of its reliability compared to more empirical methods.

Some methods that involve the spiking precision are less constrained than applications of information theory. For example, the Fano factor showed that there was greater variability in responses with its spike frequency. Further steps about that study the spike train precision may reveal specific changes to the code that information theory may not. The proximity of spikes and the changes in spike counts as blindness worsens reveal more about how the retina is firing. Understanding those changes are significant because they can contribute to further statistical models for neurons.

Further steps to uncover the changes to the retinal code will be to evaluate the effectiveness of the mutual information estimate and also explore other metrics to evaluate the retinal code. In addition, the question of whether there are improvements to the code in response to genetic treatment needs to be explored.