How to plot the original image, the ground truth segmentation along with the predicted one as given in the output image below?

[Euphemiasama]

Hello, first thank you for the tutorial.
I would like to know how did you get the resulting plot example of segmentation after training the segmenter. The plot where we have the original image, the ground truth segmentation along with the predicted one.

Also I didn't get the labeling below each resulting image. I know that the original image is given in the top right but what does slice -1 slice +0 and image for slice +0 refer to. If I understood well, three transverse slices (images) containing the largest nodule from each patient scan are extracted and masks are created for segmenting those nodules. So at the end we'll have 3 masked (segmented) images for each patient. So what does the output in the image above give exactly?

[ghost]

Slice 0 is the slice through the center of the nodule, -1 is the one below and +1 is the one above that slice.

The above image shows an overlay of the true mask and the predicted mask. The color code is explained in the tutorial, I think. Is that helpful at all, or did I misunderstand your question?

[Euphemiasama]

Thank you for the reply.
So the original image on the top right is slice 0 and the overlay of the true mask and the predicted one for that slice is the image on the bottom right entitled "image for slice +0" and the two other images on the left represent the overlay of the true mask and predicted one obtained from 2 other slices (one just before the slice 0 and one just after it). Is it so? I am not quite sure and it wasn't clear to me from the tutorial. Sorry for bothering you.

[aszhanghuali]

@jonrmulholland Hi!Did you run classify_nodes.py successfully ?
I run this successfully ,but It's all 0 in dataY.npy. Why is there no 1?
Looking forward to your reply!