Fluorescence laminar optical tomography for brain imaging: system implementation and performance evaluation.

We present our effort in implementing a fluorescence laminar optical tomography scanner which is specifically designed for noninvasive three-dimensional imaging of fluorescence proteins in the brains of small rodents. A laser beam, after passing through a cylindrical lens, scans the brain tissue from the surface while the emission signal is captured by the epi-fluorescence optics and is recorded using an electron multiplication CCD sensor. Image reconstruction algorithms are developed based on Monte Carlo simulation to model light–tissue interaction and generate the sensitivity matrices. To solve the inverse problem, we used the iterative simultaneous algebraic reconstruction technique. The performance of the developed system was evaluated by imaging microfabricated silicon microchannels embedded inside a substrate with optical properties close to the brain as a tissue phantom and ultimately by scanning brain tissue in vivo. Details of the hardware design and reconstruction algorithms are discussed and several experimental results are presented. The developed system can specifically facilitate neuroscience experiments where fluorescence imaging and molecular genetic methods are used to study the dynamics of the brain circuitries.