zhang-qiang-github

I also feel very strange. Only some of my data show this in-closed surface, and most of my data is normal. I don't know why it happen~~~

And the following code also show in-closed surface: ``` from vedo import * ellip1 = Circle(pos=[0, 0, 0], c='r') ellip2 = Circle(pos=[0, 0, 3], c='r') ribbon = Ribbon(line1=ellip1, line2=ellip2) show(ellip1,...

@marcomusy Thank you for your kindly reply. I indeed need the boundary hits. Sometime, the mask is very thin. For example:  Thus, the `vol.isosurface(1)` will not return a closed...

Is it possible to cut it by free-hand ROI?

There may be a solution. Firstly, we can create a 3D mask using the ROI. Then, we rotate the mask according to the transform matrix. Last, we apply mask*volume to...

I find it is difficult for me to create such a 3D mask. [Slicer](https://github.com/Slicer/Slicer/blob/master/Modules/Loadable/Segmentations/EditorEffects/qSlicerSegmentEditorScissorsEffect.cxx) has implement such function.

For Brats Dataset, it is OK. But, for another dataset, the channel may be not 4. So, `self.Unet = Unet(in_channels=num_channels, base_channels=16, num_classes=4)` may be more robust.

@daihahaha 应该是df_node[df_node["file"] == img_file]这句话有错吧，我改了一下： df_node[df_node["file"] == img_file[0:-4]] 你可以调试一下，看看df_node里面到底路径是啥，要不要加上mhd之类的

@taigw, I am sorry that the wrong result is caused by wrong input. The input spacing should be [1, 1, 1], but I just input one value. In addition, is...

> Yes: https://vedo.embl.es/autodocs/content/vedo/vedo/pointcloud.html#Points.cut_with_box Thank you very much. I will try it.