Hi,

I didn't find where you use the centerlines files to prepare the data. In data_prepare.py file you only read *-rib-seg.nii, in this way your problem is a binary segmentation, the label is rib = 1.

I didn't find where you go from a segmented binary volume (*-rib-seg.nii) to 24 rib labels, without using the information from *-rib-cl.nii files.

Hello, I want to change the ribseg code for the segmentation of five rib fracture types in the ribfrac data set. What are the main changes? thank you

ValueError: Cannot load file containing pickled data when allow_pickle=False

Hi,
I found these labels in train_rigseg.py file:

seg_classes = {'rib':[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24],'Earphone': [44, 45], 'Motorbike': [30, 31, 32, 33, 34, 35], 'Rocket': [41, 42, 43],
'Car': [29], 'Laptop': [28 ], 'Cap': [26], 'Skateboard': [46],
'Mug': [36], 'Guitar': [39, 40], 'Bag': [27], 'Lamp': [25],
'Table': [47], 'Airplane': [48], 'Pistol': [38],
'Chair': [37], 'Knife': [49]}

In each Epoch a have these result:

eval mIoU of Airplane nan
eval mIoU of Bag nan
eval mIoU of Cap nan
eval mIoU of Car nan
eval mIoU of Chair nan
eval mIoU of Earphone nan
eval mIoU of Guitar nan
eval mIoU of Knife nan
eval mIoU of Lamp nan
eval mIoU of Laptop nan
eval mIoU of Motorbike nan
eval mIoU of Mug nan
eval mIoU of Pistol nan
eval mIoU of Rocket nan
eval mIoU of Skateboard nan
eval mIoU of Table nan
eval mIoU of rib nan

Hello, I want to locate the location of my rib fracture (for example, the first one on the left or the fifth one on the right). Therefore, I need a CT file with 24 rib segmentation masks, calculate its bbox coordinates according to the mask marks of each rib, and then match it with the bbox coordinates of the rib fracture I detected, so as to locate. Can the CT of the ribseg dataset achieve my goal?

Excuse me, is there a label file for the test set. I only downloaded the test set image file from ribfrac challenge, but there is no test set label file. Can I provide relevant files, thank you!!

Hi,
Thanks for your contribution, I'm trying to reproduce your results and get a nibabel file to see the 3D segmentation.
However, when I try this code:
dilated = sitk.BinaryDilate(lmage_array, 3, sitk.sitkBall)
It raises a TypeError: in method 'BinaryDilate', argument 2 of type 'std::vector< unsigned int,std::allocator< unsigned int > >', do you know why?
The version of my SimpleITK is 2.0.2, hope to get your reply soon.

Hello, I want to convert the post-processing results into NII files. Here is my conversion program and visualization, but the visualization result is not good. Can you give me some advice? I'm looking forward to your reply

1.运行forward报错
‘RibSeg-main/models/pointnet2_part_seg_msg.py", line 45, in forward
cls_label_one_hot = cls_label.view(B,16,1).repeat(1,1,N)
RuntimeError: shape '[8, 16, 1]' is invalid for input of size 8
cls_label的shape是[8,1,1]

改成 cls_label_one_hot = cls_label.view(B,1,1).repeat(1,16,N)
可运行，但不知是否正确

2.data_prepare中采样固定采样数目为30000
choice = np.random.choice(temp.shape[0], 30000, replace=False)
没不是全部输出所有样本，这样data_prepare生成的样本只有30000个点，dataloader中的重采样也没有意义了

The ribseg labels appear to be missing for many of the ribfrac images.

This is causing data_prepare.py to fail with errors such as:

$ python data_prepare.py 
Traceback (most recent call last):
  File "/home/david/miniconda3/envs/RibSeg/lib/python3.7/site-packages/nibabel/loadsave.py", line 42, in load
    stat_result = os.stat(filename)
FileNotFoundError: [Errno 2] No such file or directory: './data/RibSeg/nii/RibFrac43-rib-seg.nii.gz'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "data_prepare.py", line 102, in <module>
    main()
  File "data_prepare.py", line 20, in main
    label = nib.load('./data/RibSeg/nii/'+data[:-12]+'rib-seg.nii.gz')
  File "/home/david/miniconda3/envs/RibSeg/lib/python3.7/site-packages/nibabel/loadsave.py", line 44, in load
    raise FileNotFoundError("No such file or no access: '%s'" % filename)
FileNotFoundError: No such file or no access: './data/RibSeg/nii/RibFrac43-rib-seg.nii.gz'

Hello!

Please tell me how to number the ribs after they are divided. For example, the first root and the second root on the left, can you give me some suggestions

raise ValueError("Cannot load file containing pickled data "
ValueError: Cannot load file containing pickled data when allow_pickle=False

Traceback (most recent call last):
File "e:/code/RibSeg/train_ribseg.py", line 204, in
main(args)
File "e:/code/RibSeg/train_ribseg.py", line 71, in main
MODEL = importlib.import_module('models.'+args.model)
File "D:\Anaconda\envs\pytorch\lib\importlib_init_.py", line 127, in import_module
return _bootstrap._gcd_import(name[level:], package, level)
File "", line 1006, in _gcd_import
File "", line 983, in _find_and_load
File "", line 965, in _find_and_load_unlocked
ModuleNotFoundError: No module named 'models.CLNet'

Hi,
I found this error: NameError: name 'points' is not defined

I found that you put in comment :

choice = np.random.choice(temp.shape[0], 30000, replace=False)

downsample

points = temp[choice, :]

Why ?

First of all, thank you very much for your excellent work. I would like to ask how do you achieve the indicators of center line evaluation? The only metrics I found in the code were for splitting.