Frequency-spatial domain joint optimization for improving super-resolution images of nonlinear structured illumination microscopy
Frequency-spatial domain joint optimization for improving super-resolution images of nonlinear structured illumination microscopy
复制标题
改善非线性结构照明显微镜超分辨率图像的频空域联合优化
DOI：
10.1364/ol.441160
复制
发表时间：
2021
期刊：
影响因子：
3.6
通讯作者：
Simin Li
中科院分区：
文献类型：
--
作者：
Gang Wen;Linbo Wang;Xiaohu Chen;Yuguo Tang;Simin Li
在结构照明显微镜（SIM）中引入非线性荧光团激发可以进一步扩展其空间分辨率，而不受理论限制。然而，如何恢复微弱的高阶谐波信号并重建高保真的超分辨率图像是一个巨大的挑战。在此，我们提出了一种频率域和空间域的联合优化策略来重建高质量的非线性SIM （NL-SIM）图像。我们证明了我们的方法可以在具有模式激活（PA） NL-SIM的BioSR数据集上以更少的伪影和更高的保真度重建SR图像。该方法可以有效地克服NL-SIM成像长期存在的障碍之一，从而促进其在生物学中的广泛应用。
Introducing nonlinear fluorophore excitation into structured illumination microscopy (SIM) can further extend its spatial resolution without theoretical limitation. However, it is a great challenge to recover the weak higher-order harmonic signal and reconstruct high-fidelity super-resolution (SR) images. Here, we proposed a joint optimization strategy in both the frequency and spatial domain to reconstruct high-quality nonlinear SIM (NL-SIM) images. We demonstrate that our method can reconstruct SR images with fewer artifacts and higher fidelity on the BioSR dataset with patterned-activation (PA) NL-SIM. This method could robustly overcome one of the long-lived obstacles on NL-SIM imaging, thereby promoting its wide application in biology.