Independent Component Analysis Of Optical Imaging Of Brain Function In Rats

Many techniques and methods are being increasingly applied to the research of brain with the development of the science and technology. Positron emission tomography (PET), functional magnetic resonance brain imaging (fMRI), CT, SPECT and technique of optical imaging are more and more advanced. The optical imaging of intrinsic signals plays an important role in the research because of the virtues of low cost, small volume and easily acquisition of the characteristics of the signals.Independent component analysis (ICA) is a recently developed method with many special advantages. We apply the ICA to the data acquired from the intrinsic optical signals which are caused by the electricity stimulation to sciatic nerve of rat and provide the reasonable results.In this paper we introduce the characteristics and theory of fMRI, PET and optical imaging that are main techniques of brain imaging. Then why to use optical imaging is illuminated by comparing the differences of these techniques. In order to perform analysis, we introduce the basic theory, principle and elementary concepts of ICA. The FastICA and extended information maximum algorithm, which are most used in our object, are emphasized. We demonstrate that the ICA provides a good fit to the analysis of the optical imaging data. How to estimate the ICA model according to the characteristics of the experiment data and use the combination model of the samples to improve the Signal-to-Noise of the optical imaging data are described in detail. At the end of the paper, based on the analysis and comparison, key factors in ICA of optical imaging are concluded and distilled as suggestion to the future research.