Fall 2009
Objective: My project will explore the neuronal properties of the primary visual cortex (V1) in ferrets using single unit electrophysiology. The neuronal properties measured will be orientation tuning, direction tuning, ocular dominance and spatial frequency along with receptive field size and location. These properties are used to study independent variables like developmental age, experimental manipulation and pharmacological manipulations. Additionally, the effect of masking and the influence of iconic memory on V1 neurons will be explored.
Method: To measure each neuronal property, a specific visual stimulus is presented to the ferret. Each stimulus is assigned a certain bit address. Therefore when a certain stimulus is displayed on a CRT-based monitor using "Visage," a visual stimulus generator, coincidently Visage outputs a high signal to the bit conditions the stimuli stands for. These digital signals are picked up by an analog to digital convertor. The neuronal signals are collected from V1 by placing a carbon fiber electrode in V1. The data collected is recorded extracellularly, i.e. multi-unit recording of neurons at a single site. Recordings are made relative to a ground that is placed at the ferret's scalp. These are amplified, filtered to minimize electrical noise and digitized along with the stimulus bits. A software suite called Spike2 records these signals as channels in reference to time and provides a programming environment for analysis. The neuronal signals will be differentiated at each site using a template-matching algorithm to identify single unit responses as a function of the stimulus presented. This approach will provide very detailed, high temporal resolution data on the properties of single V1 neurons.
Future Directions: In the future, the process for choosing the next experiment will be automated based on the results collected previously using an active learning approach that we hope to develop. This would optimize data collection for temporal efficiency, resolution and spatial sampling. Such an approach will help us to reach accurate conclusions more rapidly, without conducting unnecessary experiments.