Automated identification of neural cells in the multi-photon images using deep-neural networks

摘要

Correlative super-resolution structured illumination microscopy combined with array tomography for high accuracy synapse detection Electron microscopy (EM) has been the gold standard method for synapse detection owing to its ultra-high spatial resolution. Despite three-dimensional tissue image acquisition with high res- olution is available with EM, however, overall procedure to image acquisition and analysis still remains challenging. In addition, mul- tiple labeling method for EM is extremely limited. On the other hand, fluorescence microscopy (FM) has been essential tool for bio- logical research. FM offers large field-of-view (FOV), high contrast, and multiple labeling combination. However, conventional opti- cal microscopy meets challenge, which is the diffraction limit. It makes FM does not provide sufficient spatial resolution for synapse detection.