基于极大似然法的超分辨荧光显微成像方法研究

Study on fluorescence super-resolution imaging based on maximum likelihood method

由于光的衍射极限限制,人们无法观察精细的细胞结构和获得纳米级的蛋白质以及生物分子信息.本研究采用极大似然算法来实现单分子定位.首先,对荧光分子图像进行解卷积和二值化处理降噪;其次,对荧光显微成像过程建立极大似然模型;最后,用优化工具箱函数进行求解.模拟结果表明:相邻20 nm的2条分子带可以分辨;将此算法应用于实际生物样品的超分辨成像时,原本不能区别的2条微丝可以进行区分,且分辨能力得到了提高;极大似然法可以很好地完成分子定位,进而实现超分辨图像的重构.

Due to the diffraction limit of light,it is impossible for people to observe the fine cell structure and obtain nanoscale protein and the biomolecular information.In this study,maximum likelihood algorithm is used to realize single molecule localization.Firstly,the image of fluorescence molecule is deconvoluted and binarized to reduce noise.Secondly,the maximum likelihood estimation model is established for the fluorescence microscopic imaging process.Finally the optimization toolbox function is applied to solve the problem.The simulation results show that the two molecular bands adjacent to 20 nm can be distinguished.When this method is applied to the super-resolution imaging of real biological samples,the original two micro-filaments can be distinguished and the resolution ability is improved.The maximum likelihood method can well complete the molecular localization and reconstruct the super-resolution image.