構造化照明蛍光顕微鏡の画像再構成アルゴリズムについて(<特集>離散幾何学から提案する新物質創成・物性発現の解明)

概要

論文の詳細を見る
A structured illumination microscopy (SIM) proposed by Gustafsson has resolution better than conventional resolution limit. SIM illuminates a sample with structured light several times by varying the phase, observes the resulting uorescence image from the sample, compute the Fourier transform of the sample's image from those of the observed uorescence images, and then compute the inverse-Fourier transform to construct the super-resolution image of the sample. The observed images are prone to independent image noises. Unless the observed images continuously join to a single image, the computed image will contain an artifact that does not really exist in the sample. We propose to calculate the weighted mean of the Fourier transforms of the observed images, to construct the super-resolution image of the sample. By a computer simulation, we illustrate that this image reconstruction algorithm can indeed enhance the resolution of SIM.
2013-12-24