- 學位論文
發展快速隨機光學建構顯微鏡
The Development of Fast Stochastic Optical Reconstruction Microscopy (STORM)
若您是本文的作者，可授權文章由華藝線上圖書館中協助推廣。
摘要
傳統顯微鏡的解析度主要受到光學繞射的物理限制。在過去幾年內，快速隨機光學建構顯微鏡(STORM)的高解析能力受到很多研究專家的注意。STORM的原理主要是靠著隨機激發一小部分的單螢光分子且定位，時間持續到所有的分子都被定位完畢。因此，STORM在解析方面超越傳統的顯微鏡，並且突破阿貝解析度(Abbe resolution)。 然而，傳統STORM有處理大量影像很慢的缺點。尤其是定位程序往往是程式的瓶頸。我們建立了單分子的光學模式並且運用質量中心、徑向對稱、最小平方及最大可能性預估方法做定位。同時使用了輝達繪圖處理器(NVIDIA GPU)的平行運算能力來加快分子定位，MEX function也提升了整體影像處理的速度。 結合上述兩項而加快了影像處理速度達百倍。我們也運用STORM觀察了微絲和粒腺體。為了研究蛋白質的共區域化，我們建立了色差校正程序。實驗證明了整體的誤差約在0.25像素，相當於33奈米。三維STORM是我們接下來的發展目標。
並列摘要
Resolutions of conventional microscopy are mainly imposed by the diffraction of light. In the past several years, the emergence of stochastic optical reconstruction microscopy (STORM) has attracted many researchers due to its resolving power. Principles of STORM are based on randomly activating a subset of fluorescent molecules until all of single fluorescent molecules are localized. Hence, STORM has potentials of surpassing conventional microscopy in resolutions, the limit of which is Abbe resolution. The traditional STORM has disadvantages of slowly processing large amounts of images. Especially, localizations are always the bottlenecks in image processing. We built up the optical model of single fluorescent molecule and implemented centroid, radial symmetry, nonlinear least-squares minimization and maximal likelihood estimator methods to localize centers of molecules. NVIDIA GPU has capabilities of parallel processing; thus, it is employed to accelerate localizations. MEX functions also increase the speed of data processing. We have enhanced the processing speed up to hundred folds by utilizing GPU and MEX functions. Cellular actin and mitochondria were visualized by STORM. For the purpose of observing colocalizations between proteins, we setup chromatic correction on two channels. It was demonstrated that mapping error is about 0.25 pixels, equivalent to 33 nm. Three dimensional STROM is our next goal in the future.
參考文獻
1. Robert P J Nieuwenhuizen, Keith A Lidke, Mark Bates, Daniela Leyton Puig, David Grunwald, Sjoerd Stallinga and Bernd Rieger. Measuring image resolution in optical nanoscopy. Nature Methods 10, 557–562 (2013).
2. Joshua C Vaughan, Shu Jia and Xiaowei Zhuang. Ultrabright Photoactivatable Fluorophores Created by Reductive Caging. Nature Methods 9, 1181–1184 (2012).
3. Siddharth Nanguneri, Benjamin Flottmann, Heinz Horstmann, Mike Heilemann and Thomas Kuner. Three-Dimensional, Tomographic Super-Resolution Fluorescence Imaging of Serially Sectioned Thick Samples. PLoS One 7(5), 1-8 (2012).
4. Alexandros Pertsinidis, Yunxiang Zhang and Steven Chu. Subnanometre single-molecule localization, registration and distance measurements. Nature 466, 647–651 (2010).
5. Raghuveer Parthasarathy. Rapid, accurate particle tracking by calculation of radial symmetry centers. Nature Methods 9, 724–726 (2012).