Improved super-resolution optical fluctuation imaging by multiple sparse Bayesian learning method

By exploiting the statistics of temporal fluorescent fluctuations, super-resolution optical fluctuation imaging (SOFI) can implement a fast super-resolution microscopy imaging, which is suitable for dynamic live-cell imaging. However, the main drawback of SOFI is that the imaging spatial resolution can be surpassed by the localization-based super-resolution microscopy techniques. To address this problem, we propose a new method, which is achieved by using multiple sparse Bayesian learning (M-SBL) method. Since M-SBL method can take into account simultaneously temporal fluctuations and the sparsity priors of emitter, it provides the possibility to obtain an enhancement in spatial resolution compared to standard SOFI (only considering the temporal fluctuations). To measure the performance of our proposed method, we designed three sets of simulation experiments. Firstly, we compared the performance of M-SBL and SOFI in resolving single emitter, and simulation results have demonstrated that the M-SBL method outperforms SOFI. Furthermore, the other simulation data with varying signal to noise and frame number were used to evaluate the performance of M-SBL in resolving fine structures. And the results indicate that when using the proposed M-SBL method, the imaging spatial resolution can be improved compared to the standard SOFI method. Hence, the M-SBL method provides the potential for increasing the temporal resolution of super-resolution microscopy while maintaining a desired spatial resolution.