27. NONLINEAR ESTIMATION OF SPIKE RATES FROM NEURONAL INTRACELLULAR CALCIUM SIGNALS

Department: Bioengineering
Faculty Advisor(s): Gabriel Silva

Primary Student
Name: Christopher L. MacDonald
Email: clmacdon@ucsd.edu
Phone: 858-752-4361
Grad Year: 2011

Student Collaborators
Krystal Chiao, Kchiao@ucsd.edu

Abstract
Being able to measure the spike rate of large groups of neurons in real time is critical to the understanding of large scale neuronal networks and bridging the knowledge gap between single cell kinetics and large scale fMRI studies. Classical methods to record action potential firing in individual neurons at high temporal resolution focus on electrophysiology, which while providing detailed information about one or a few neurons in one experiment, cannot record simultaneously from large networks of neurons. Recent research has focused on the potential of calcium imaging, where bolus loading of cells in a field of view allows real time tracking of the intracellular calcium levels in neurons. While this allows the calcium levels to be measured, there is not a direct correlation between the calcium signal and the spike rate of the neuron. Linear deconvolution of the calcium signal with a kernel has been used to get an estimate of the spike rate of the neuron. This work shows that while this is an accurate estimate for certain stimuli, the relationship between calcium and spike rate is inherently nonlinear due to the kinetics of the voltage gated calcium channels, and the linear method is shown to break down under certain circumstances. Constructing a nonlinear model of the relationship between the voltage and calcium, along with numerical methods to optimize the parameters in the model allows the calculation of the voltage of the neuron from the calcium signal. The method is tested in silico and then compared to experiment by simultaneously recording calcium signals and membrane voltage from the same cell. These results show a novel way of calculating the spike rate for large neuronal populations in real time, which is essential for increasing understanding of large and complex neuronal circuits.

« Back to Posters or Search Results