A neuronal imaging dataset for deep learning in the reconstruction of single-neuron axons

The project encompasses three types of data: three-dimensional images of neuronal processes, the skeletons of these processes, and the masks generated from the skeletons. The image dataset is collected from fluorescence micro-optical sectioning tomography and consists of 852 carefully selected volume images, each sized at 192×192×192 voxels.本项目涵盖三类数据：神经元突起的三维图像、此类突起的骨架，以及由该骨架生成的掩码。该图像数据集采集自荧光显微光学切片断层成像（fluorescence micro-optical sectioning tomography），包含852张经精心筛选的体域图像，单张图像的尺寸为192×192×192体素（voxel）。