9405146 Frostig One of the most informative ways to study the brain is to record from the brains of intact animals in a non-invasive way the electrical signals that brain cells use to communicate with each other. A new technique to do this is called optical imaging of intrinsic signals, in which signals from the illuminated surface of the brain are recorded with a digital CCD camera. One major impediment to getting good signals from the nerve cells is that the blood vessels on the surface of the brain disrupt the signals from the underlying brain tissue. Another obstacle in the development of the technique comes from the very large data set generated for each measurement (typically 120 images of a grid of 144 x 192 pixels, each pixels with 256 levels of light). This research focuses on ways to improve the digital signals recorded by the camera in studies of the part of the rat cortex that processes the tactile signals coming in from the rat's whiskers. It is a collaborative effort between biologists and mathematicians to develop new mathematical tools to separate the signals from the main blood vessels from those from the brain, so that the optical imaging of the intrinsic signals will have more faithful spatial resolution. A second goal of the project is to develop a more efficient way for the computer to handle the large data set so that the images can have better temporal resolution. The results of this study will have impact not only on neuroscience research, but also development of improved diagnostic tools in clinical neurology as well as impact on other types of image analysis. This multidisciplinary proposal was submitted to the Federal Interagency Coordinating Committee on the Human Brain Project and is being supported by three Programs within the National Science Foundation - Statistics and Probability, Computational Neuroscience, and Computational Biology.
