Li et al (2023)


A high-performance deep-learning-based pipeline for whole-brain vasculature segmentation at the capillary resolution

Authors: Yuxin Li*, Xuhua Liu, Xueyan Jia, Tao Jiang, Jianghao Wu, Qianlong Zhang, Junhuai Li, Xiangning Li, Anan Li*

* Correspondence: liyuxin@xaut.edu.cn, aali@hust.edu.cn

Original data (whole brain, 1x1x1 µm, registered to CCFv3)
Coronal (100µm / 200µm)

Segmented data (whole brain, 2x2x2 µm, registered to CCFv3 )
Original coronal (100µm)
Sagittal (100µm)
Horizontal (100µm)

Test raw data:
Test data1 : ( 3D TIFF, 0.325x0.325x1 µm, 1000x1000x300 pixels, unregistered)
Test data2 : ( 3D TIFF, 0.325x0.325x1 µm, 1000x1000x300 pixels, unregistered)
Test data3 : ( 3D TIFF, 1x1x1 µm, 800x800x800 pixels, unregistered)
Test data4 : ( 3D TIFF, 1x1x1 µm, 800x800x800 pixels, unregistered)

In additional, the fMOST whole brain image dataset is about 20 TB, which is very huge that cannot be shared via internet. But, it is available from the corresponding author on reasonable request.
Source Code:https://github.com/visionlyx/HP-VSP

ęCopyRight Huazhong Univ. of Sci. & Tech.