UC San Diego

Fluorescence imaging in biology has become a fundamental tool to understanding structure and function. In addition, the past two decades has seen the first implementation, and subsequent widespread use, of two-photon scanning laser microscopy as a tool to allow fluorescent data to be acquired in new modalities, such as imaging at depth, complete three-dimensional reconstructions, rapid acquisition of cellular activity, and accurate assessment of cerebral blood flow at the level of individual vessels. This thesis examines the use of two-photon microscopy to study neurovascular coupling, the process by which computational mechanism in the brain actively reroute blood flow to regions of increased neural activity, where it is needed the most. Understanding the mechanisms through which this occurs is an active area of research, and represent a "sweet spot'' in terms of complexity: the function of the system can be stated simply, and many of the mechanisms involved are understood in isolation. However, putting everything together into a complete picture has remained elusive, and understanding how the brain accomplishes this seemingly simple task should provide insight into the general computational mechanisms of the brain. Central to this research was the development of scanning hardware and software techniques that allow the rapid acquisition of the biologically relevant variables with high signal-to-noise ratios. This lead to the development of computer control software for creating user-defined scan lines, which can interface with machine learning programs for quickly and accurately identifying relevant regions in the field of view, and photon-counting hardware to make the most out of the limited amount of light that can be collected per unit time. The photon- counting hardware and firmware, developed at UCSD, let to some unexpected and previously unanalyzed findings about what happens to the signal under high light conditions, and the resulting analysis and corrections are generally useful to two-photon microscopy. It is hoped that the data acquisition and analysis tools discussed here, and currently being used in the laboratory, will help advance the field of neuroscience, and provide some insight to long standing questions in brain function