「Medical Image Analysis」Note on Kid-Net: ConvNets for Kidney Vessels Segmentation

版权声明:本文为博主原创文章,未经博主允许不得转载。 https://blog.csdn.net/dgyuanshaofeng/article/details/83247150

QQ Group: 428014259
Tencent E-mail:[email protected]
http://blog.csdn.net/dgyuanshaofeng/article/details/83247150

作者:Ahmed Taha, Pechin Lo, Junning Li, Tao Zhao
单位:University of Maryland, College Park;Intuitive Surgical, Inc

0 摘要

Kid-Net完成肾脏血管分割,包括动脉、静脉和收集系统(输尿管ureter)。训练方案的特点:handles unbalanced data, reduces false positives and enables high-resolution segmentation with a limited memory budget。具体方法包括:dynamic weighting, random sampling and 3D patch segmentation。分割时间为1-2分钟(图像大小为512的立方)。

1 介绍和相关工作

提及U-Net和V-Net。
提及该分割任务的难点,实际上和大多数的方法差不多,但是分割血管那就更难了。

2 方法

3 实验

[1] Kid-Net Convolution Networks for Kidney Vessels Segmentation from CT-Volumes MICCAI 2018 [paper]

猜你喜欢

转载自blog.csdn.net/dgyuanshaofeng/article/details/83247150