Flutter Discrimination: neural codes, perception, memory and decision making

Frequency discrimination in the sense of flutter is a useful model to understand how neural codes are related to perception, working memory and decision making. The task can be thought of as a series of steps: encoding a stimulus frequency, maintaining it in memory, encoding a second frequency, comparing it to the frequency in memory, and communicating the result of the comparison to the motor system. Neurophysiological studies have explored the brain regions that participate in each of these steps and how these regions interact to solve the task. Encoding. How is the flutter frequency represented in the nervous system? Does this neural code reflect behavioural responses? Changes in neuronal firing rate as a function of stimulus frequency are evident in several areas (particularly in the primary somatosensory cortex) during the flutter discrimination task. Moreover, several lines of evidence (particularly microstimulation experiments) indicate that these rate variations affect behaviour. By contrast, the high periodicity that flutter elicits does not seem to contribute to frequency discrimination. Memory. The clearest neural correlate of working memory during frequency discrimination is found in the prefrontal cortex, which contains neurons that increase their activity in a frequency-dependent manner during the delay period between the two flutter stimuli. This activity does not seem to be related to the impending motor response. The prefrontal cortex might not be the only structure in which such a mnemonic correlate exists, as neurons with similar activity have been found in the secondary somatosensory cortex and in the medial premotor cortex. Comparison process and decision making. The comparison between the two stimulus frequencies can be simply conceptualized as the difference between them. The firing of some neurons in the secondary somatosensory cortex shows dynamic changes as a function of both frequencies, and evolves to encode the difference between them. Moreover, the firing patterns of these neurons are a good indicator of the actual behavioural response, indicating that this neural activity might be involved in the decision-making process. Similar, but significantly different, dynamic changes have been found in the prefrontal and medial premotor cortices. There is a large overlap between sensory-, mnemonic- and decision-related activity during frequency discrimination. As a result, the comparison between stored and ongoing sensory information seems to take place in a distributed manner, and no single area can be identified as the unique site of decision making. We therefore propose that the motor plan that is established to respond after a discrimination trial already contains two possible outcomes, and that sensory information helps in the selection of one of them. Future studies should rigorously explore this possibility.