neheller / kits21 Goto Github PK
View Code? Open in Web Editor NEWThe official repository of the 2021 Kidney and Kidney Tumor Segmentation Challenge
License: MIT License
The official repository of the 2021 Kidney and Kidney Tumor Segmentation Challenge
License: MIT License
Kindly add the dataset description to the repo (e.g., number of cases, demographics, etc.)
Hello, I am having problem with just training the dataset with baseline nnUnet.
Fabian has kindly written the conversion of dataset python file, and I just used it to
convert kits 21 dataset. However following error occurs during preprocessing.
RuntimeError: Exception thrown in SimpleITK ImageFileReader_Execute: /tmp/SimpleITK/Code/IO/src/sitkImageReaderBase.cxx:105:
sitk::ERROR: Unable to determine ImageIO reader for "/home/mchangaf/nnUNetFrame/DATASET/nnUNet_raw/nnUNet_raw_data/
Task135_KiTS2021/imagesTr/case_00038_0000.nii.gz"
On certain files, Exception thrown in SimpleITK ImageFileReader
occurs. Should I just ignore these errors and continue training?
It's not necessary for every saved json file to be stored. The most recent one for each instance should contain all necessary information about the edit lineage. The remainder can be discarded by the annotation/import.py
script.
There is a error occured when i try to download this dataset, log was that ...
...
300 cases to download...
1/300...
check_hostname requires server_hostname
...
thanks,
Hello,
I meet a problem after I run get_imaging.py. In my downloaded data folder, it seems imaging.nii.gz is missing. I would like to ask how can I access the imaging.nii.gz files? Hope for your reply :)
Hello,
I might be doing something silly, but I used the bash option get_imaging.sh and it only downloads the imaging.nii.gz files. How do I get the labels (segmentations)? I tried installing but pip install always result in errors.
Thanks,
Lorena
I got 300 scans at all, and how can I split them into training and testing sets?
Hi Nick, excellent work with KiTS21 :)
I noticed that Instance 1 Cyst and Instance 3 Tumor refer to the same structure in case 00205, is that intended?. Considering that there was no Tumor region in KiTS19 I guess it is supposed to be a cyst.
Best,
Michael
Hi,
I followed Usage and when I run python3 kits21 / starter_code / get_imaging.py in my terminal I got this error: ModuleNotFoundError: No module named 'requests'.
I'm not very experienced so I'm not sure I'm doing it right or understanding all of this.
Any idea ?
Hi and a very good day,
Thank you for organising this challenge and in turn providing the annotated datasets.
I have a question here, if I would want to remove the cyst label from the segmentation mask, eg: the majority voting of aggregated_MAJ_seg.nii.gz, how can I perform such action?
I have tried manually removing the label 3 (assuming that cyst label is represented by label 3) with ITK snap's Label Editor, but however it does not seem to be able to remove it after I reopen the edited mask.
Would you please give some guidance on this
Best regards,
Danny
Hi.
I ran
python kits21/annotation/sample_segmentations.py
python kits21/evaluation/evaluate_predictions.py output
Then it gets me
multiprocessing.pool.RemoteTraceback:
"""
Traceback (most recent call last):
File "/opt/conda/lib/python3.8/multiprocessing/pool.py", line 125, in worker
result = (True, func(*args, **kwds))
File "/opt/conda/lib/python3.8/multiprocessing/pool.py", line 48, in mapstar
return list(map(*args))
File "/workspace/kits21/kits21/evaluation/metrics.py", line 200, in evaluate_predicted_file_on_samples
raise RuntimeError('segmentation_samples folder missing. Please run kits21/annotation/sample_segmentations.py')
RuntimeError: segmentation_samples folder missing. Please run kits21/annotation/sample_segmentations.py
"""
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "kits21/evaluation/evaluate_predictions.py", line 29, in <module>
evaluate_predictions_on_samples(args.folder_with_predictions, args.num_processes)
File "/workspace/kits21/kits21/evaluation/metrics.py", line 168, in evaluate_predictions_on_samples
metrics = p.map(evaluate_predicted_file_on_samples, predicted_segmentation_files)
File "/opt/conda/lib/python3.8/multiprocessing/pool.py", line 364, in map
return self._map_async(func, iterable, mapstar, chunksize).get()
File "/opt/conda/lib/python3.8/multiprocessing/pool.py", line 771, in get
raise self._value
RuntimeError: segmentation_samples folder missing. Please run kits21/annotation/sample_segmentations.py
I read link, and I'm not sure why I should run python kits21/annotation/sample_segmentations.py
. Okay you said, this is about 5-fold CV. And the description in the kits21/evaluation/evaluate_predictions.py
says whatever in the directory output
will give us the result.
Hello,
I have successfully run the get_Imaging.py file and got the image.nii.gz under the corresponding case folder. But the each image.nii.gz looks strange. Their grayscale values are almost all zero, with some very small negative numbers. That means when I opend the image.nii.gz by ITK-SNAP, I can not see the any useful image information.
Is it possible to get another implementation of get_imaging in Julia?
Hello, I am working on data collection for my research and I need CT scan images of patients with chromophore renal cell carcinoma and benign renal oncocytoma. Can I get data for these conditions, please? On your site I saw images but where is the type of cancer?? is there any excel sheet??
Hi, thanks for your codes;
how train and test on some samples?
is there any trained model?
hi!recently i have been working on no-contrast CTs(more than 300) kidney segmentation,while 90% of them get very bad performance on KiTS.
it confused me and I have checked the two datasets ,both KiTS and mine ,to ensure they remain highly consistent expect for differences in kidney HU intensity.
I do have some other CTs with contrast (just like the KiTS) and the fact is that when do the inference on kidney filled with medium, KiTS model demonstrats outstanding performance but it fails to segment a whole kidney in no-contrast CTs.
is there any chance that KiTS21/19 model rely on high HU intensity too much to make judgments using contrast,while hardly capturing other features such as position and shape so that when medium disappears the model cannot recognize kidneys?
Hi, I've noticed that my results with Baseline Model (Described here) are inconsistent with those described in this repository.
I used the unchanged KiTS21 Dataset and ran predictions with the 3d_fullres Baseline model without changes in weights. However, after running sample_segmentations
and evaluate_predictions
, I got different results that are described in this repository and paper. All source code I left unchanged (that counts for seeds for randomization).
Do you know any particular reason why this can happen?
Dice_kidney | Dice_masses | Dice_tumor | Dice_average | SurfDice_kidney | SurfDice_masses | SurfDice_tumor | SurfDice_average | ||
---|---|---|---|---|---|---|---|---|---|
3d_fullres | 0.9666 | 0.8618 | 0.8493 | 0.8926 | 0.9336 | 0.7532 | 0.7371 | 0.8080 | |
my 3d_fullres | 0.9732 | 0.9149 | 0.9188 | 0,9356 | 0.9454 | 0.8450 | 0.8485 | 0.8796 |
Can the annotation be used as segmentation classes?
you have 3 segmentation clases (Kidneys, tumor,cyst) but you have more annotation in the RAW folder (Arteries,veins,unretires) how can I make use of theses?
Can someone help me with this. I cloned the repo, ran the get_imaging.py file. initially I want to check the dataset so that is why I only ran for 10 cases. But every time there is this error below-
HTTPSConnectionPool(host='kits19.sfo2.digitaloceanspaces.com', port=443): Max retries exceeded with url: /master_00000.nii.gz (Caused by NewConnectionError('<urllib3.connection.HTTPSConnection object at 0x000001E915A80430>: Failed to establish a new connection: [WinError 10060] A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond'))
Would appreciate if I can get an immediate help in this regard. TIA.
Hello,
recently I tried to add some clinical CTs from hosipitals into KiTS. However I noticed KiTS nii fomat shape be like (num_slices, height, width) , which can be described as (z, x ,y) and origin be like(0,0,0). But my CTs get (x, y ,z) shape and have none-zero origins. One way to solve this problem is that I transposed the dims, pixdims. But how to deal with the qform/sform , origin and srow_x/y/z in the metadata still confuse me.
Could you provide aggregated segmentation, with ureter,artery,vein etc.. ,besides kidney and tumor?
Hello,
I am trying to understand how to compare (or merge) the datasets of kits19 and kits21. The description of the classes in the kits21 webpage (https://kits21.kits-challenge.org/) says that the kidney class should be equivalent for both, however I thought, reading the kits19 data description paper (https://arxiv.org/pdf/1904.00445.pdf, page 4, bullet point 2), that cysts were included in the kidney class for kits19, in which case it would be equivalent to kidney + cysts in kits21?
Thanks!
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.