Slicer 3D for Volumetric Image Segmentation/labelling - Basic segmentation usage

In this video I show different tools in Slicer 3D to perform volumetric image segmentation.
The labelling is for instance needed to train neural networks to perform this kind of segmentation automatically, as shown in my video series on 2D UNET: https://www.youtube.com/playlist?list=PLPmOaPsuoeluRw6tZQbGyhRUJNfWCU3mG

In the video I first show how to segment multiple bones in a CT image of the forearm.
Then I show how to segment a single tooth in a full CBCT image of the human head. The tooth segmentation both involves: cropping, image smoothening and manual segmentation to be used for the “grow from seeds” method.

It should be noted that the thresholding technique also works well as an initial step for the grow from seeds. You would then just have to remove a lot of the wrongly labelled data.

3D Slicer is available for Windows, Linux and OSX here: https://www.slicer.org/

Timeline:
0:10 Introduction
0:50 Loading forearm CT image
1:40 Otsu threshold segmentation
3:20 Image properties (histogram)
4:47 Manual threshold segmentation
6:32 Split segmentation into parts
7:52 Manual image segmentation editing (draw/erase)
9:25 Save label mask to disk
10:53 Loading CBCT image
11:15 Cropping a tooth
13:37 Image pre-processing (smoothing)
15:20 Grow from seeds segmentation
19:30 Iterate …
21:42 Save label mask to disk
22:06 Checking out

Remember to like the video and subscribe to the channel for more content like this.