# DentalSegmentator : error loading results : failt to load the segmentation. 

**URL:** <https://discourse.slicer.org/t/dentalsegmentator-error-loading-results-failt-to-load-the-segmentation/48375>\
**Category:** Support\
**Created:** [October 1, 2026, 2:52pm UTC](https://discourse.slicer.org/t/dentalsegmentator-error-loading-results-failt-to-load-the-segmentation/48375 "2026-10-01T14:52:08Z")\
**Posts on this page:** 8\
**Page:** 1

<div class="post-metadata">

**Author:** ![samzou974](https://avatars.discourse-cdn.com/v4/letter/s/e9a140/32.png) [@samzou974](https://discourse.slicer.org/u/samzou974)\
**Post date:** [October 1, 2026, 2:52pm UTC](https://discourse.slicer.org/t/dentalsegmentator-error-loading-results-failt-to-load-the-segmentation/48375/1 "2026-10-01T14:52:08Z")

</div>

Operating system:Windows 11 PRO  
Slicer version: 5.12.4  
Expected behavior: Segmentation of Cone-Beam using DentalSegmentator extension  
Actual behavior: error

2026/10/01 12:32:47.256 :: nnUNet is already installed (2.8.1) and compatible with requested version (nnunetv2).

2026/10/01 12:32:50.322 :: Transferring volume to nnUNet in C:/Users/Bureau01/AppData/Local/Temp/Slicer-WlePvN

2026/10/01 12:32:59.229 :: Starting nnUNet with the following parameters:

2026/10/01 12:32:59.229 ::

2026/10/01 12:32:59.229 :: C:\ProgramData\slicer.org\3D Slicer 5.12.4\lib\Python\Scripts\nnUNetv2\_predict.exe -i C:/Users/Bureau01/AppData/Local/Temp/Slicer-WlePvN/input -o C:/Users/Bureau01/AppData/Local/Temp/Slicer-WlePvN/output -d Dataset111\_453CT -tr nnUNetTrainer -p nnUNetPlans -c 3d\_fullres -f 0 -npp 1 -nps 1 -step\_size 0.5 -device cuda -chk checkpoint\_final.pth --disable\_tta

2026/10/01 12:32:59.229 ::

2026/10/01 12:32:59.229 :: JSON parameters :

2026/10/01 12:32:59.229 :: {

2026/10/01 12:32:59.229 :: “folds”: “0”,

2026/10/01 12:32:59.229 :: “device”: “cuda”,

2026/10/01 12:32:59.229 :: “stepSize”: 0.5,

2026/10/01 12:32:59.229 :: “disableTta”: true,

2026/10/01 12:32:59.229 :: “nProcessPreprocessing”: 1,

2026/10/01 12:32:59.229 :: “nProcessSegmentationExport”: 1,

2026/10/01 12:32:59.229 :: “checkPointName”: “”,

2026/10/01 12:32:59.229 :: “modelPath”: {

2026/10/01 12:32:59.229 :: “\_path”: “C:\\ProgramData\\slicer.org\\3D Slicer 5.12.4\\slicer.org\\Extensions-34645\\DentalSegmentator\\lib\\Slicer-5.12\\qt-scripted-modules\\Resources\\ML”

2026/10/01 12:32:59.229 :: }

2026/10/01 12:32:59.229 :: }

2026/10/01 12:32:59.248 :: nnUNet preprocessing…

2026/10/01 12:33:02.363 :: Traceback (most recent call last):

2026/10/01 12:33:02.363 :: File “”, line 198, in \_run\_module\_as\_main

2026/10/01 12:33:02.363 :: File “”, line 88, in \_run\_code

2026/10/01 12:33:02.363 :: File “C:\ProgramData\slicer.org\3D Slicer 5.12.4\lib\Python\Scripts\nnUNetv2\_predict.exe\__main_\_.py”, line 2, in

2026/10/01 12:33:02.365 :: File “C:\ProgramData\slicer.org\3D Slicer 5.12.4\lib\Python\Lib\site-packages\nnunetv2\inference\predict\_from\_raw\_data.py”, line 23, in

2026/10/01 12:33:02.365 :: from nnunetv2.configuration import default\_num\_processes

2026/10/01 12:33:02.365 :: File “C:\ProgramData\slicer.org\3D Slicer 5.12.4\lib\Python\Lib\site-packages\nnunetv2\configuration.py”, line 10, in

2026/10/01 12:33:02.365 :: default\_n\_proc\_DA = get\_allowed\_n\_proc\_DA()

2026/10/01 12:33:02.365 :: ^^^^^^^^^^^^^^^^^^^^^^^

2026/10/01 12:33:02.365 :: File “C:\ProgramData\slicer.org\3D Slicer 5.12.4\lib\Python\Lib\site-packages\nnunetv2\utilities\default\_n\_proc\_DA.py”, line 23, in get\_allowed\_n\_proc\_DA

2026/10/01 12:33:02.365 :: hostname = subprocess.getoutput([‘hostname’])

2026/10/01 12:33:02.365 :: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

2026/10/01 12:33:02.365 :: File “C:\ProgramData\slicer.org\3D Slicer 5.12.4\lib\Python\Lib\subprocess.py”, line 691, in getoutput

2026/10/01 12:33:02.365 :: return getstatusoutput(cmd, encoding=encoding, errors=errors)[1]

2026/10/01 12:33:02.365 :: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

2026/10/01 12:33:02.365 :: File “C:\ProgramData\slicer.org\3D Slicer 5.12.4\lib\Python\Lib\subprocess.py”, line 671, in getstatusoutput

2026/10/01 12:33:02.365 :: data = check\_output(cmd, shell=True, text=True, stderr=STDOUT,

2026/10/01 12:33:02.365 :: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

2026/10/01 12:33:02.365 :: File “C:\ProgramData\slicer.org\3D Slicer 5.12.4\lib\Python\Lib\subprocess.py”, line 466, in check\_output

2026/10/01 12:33:02.365 :: return run(\*popenargs, stdout=PIPE, timeout=timeout, check=True,

2026/10/01 12:33:02.365 :: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

2026/10/01 12:33:02.365 :: File “C:\ProgramData\slicer.org\3D Slicer 5.12.4\lib\Python\Lib\subprocess.py”, line 550, in run

2026/10/01 12:33:02.365 :: stdout, stderr = process.communicate(input, timeout=timeout)

2026/10/01 12:33:02.365 :: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

2026/10/01 12:33:02.365 :: File “C:\ProgramData\slicer.org\3D Slicer 5.12.4\lib\Python\Lib\subprocess.py”, line 1196, in communicate

2026/10/01 12:33:02.366 :: stdout = self.stdout.read()

2026/10/01 12:33:02.366 :: ^^^^^^^^^^^^^^^^^^

2026/10/01 12:33:02.366 :: File “”, line 322, in decode

2026/10/01 12:33:02.366 :: UnicodeDecodeError: ‘utf-8’ codec can’t decode byte 0x82 in position 86: invalid start byte

2026/10/01 12:33:02.881 :: Loading inference results…

2026/10/01 12:33:07.035 :: Error loading results :

2026/10/01 12:33:07.035 :: Failed to load the segmentation.

2026/10/01 12:33:07.035 :: Something went wrong during the nnUNet processing.

2026/10/01 12:33:07.035 :: Please check the logs for potential errors and contact the library maintainers.

Thank you for your help !! Have a nice day !

---

<div class="post-metadata">

**Author:** ![samzou974](https://avatars.discourse-cdn.com/v4/letter/s/e9a140/32.png) [@samzou974](https://discourse.slicer.org/u/samzou974)\
**Post date:** [October 2, 2026, 5:17am UTC](https://discourse.slicer.org/t/dentalsegmentator-error-loading-results-failt-to-load-the-segmentation/48375/2 "2026-10-02T05:17:16Z")

</div>

I tried with the 5.13.0 on an other computer, and it work well… no error.  
So i uninstall everything (3d Slicer 5.12 and extensions) on the 1st computer, then I installed 5.13 version and extension (DentalSegmentator) : still the same error :

> 2026/10/02 09:06:53.981 :: PyTorch Python package is required. Installing… (it may take several minutes)2026/10/02 09:09:04.957 :: nnUNet installation completed successfully.
> 
> 2026/10/02 09:09:04.961 :: Downloading model weights…
> 
> 2026/10/02 09:09:29.570 :: Transferring volume to nnUNet in C:/Users/Bureau01/AppData/Local/Temp/Slicer-DKvOSY
> 
> 2026/10/02 09:09:41.372 :: Starting nnUNet with the following parameters:
> 
> 2026/10/02 09:09:41.372 ::
> 
> 2026/10/02 09:09:41.372 :: C:\ProgramData\slicer.org\3D Slicer 5.13.0-2026-09-30\lib\Python\Scripts\nnUNetv2\_predict.exe -i C:/Users/Bureau01/AppData/Local/Temp/Slicer-DKvOSY/input -o C:/Users/Bureau01/AppData/Local/Temp/Slicer-DKvOSY/output -d Dataset111\_453CT -tr nnUNetTrainer -p nnUNetPlans -c 3d\_fullres -f 0 -npp 1 -nps 1 -step\_size 0.5 -device cuda -chk checkpoint\_final.pth --disable\_tta
> 
> 2026/10/02 09:09:41.372 ::
> 
> 2026/10/02 09:09:41.372 :: JSON parameters :
> 
> 2026/10/02 09:09:41.372 :: {
> 
> 2026/10/02 09:09:41.372 :: “folds”: “0”,
> 
> 2026/10/02 09:09:41.372 :: “device”: “cuda”,
> 
> 2026/10/02 09:09:41.372 :: “stepSize”: 0.5,
> 
> 2026/10/02 09:09:41.372 :: “disableTta”: true,
> 
> 2026/10/02 09:09:41.372 :: “nProcessPreprocessing”: 1,
> 
> 2026/10/02 09:09:41.372 :: “nProcessSegmentationExport”: 1,
> 
> 2026/10/02 09:09:41.372 :: “checkPointName”: “”,
> 
> 2026/10/02 09:09:41.372 :: “modelPath”: {
> 
> 2026/10/02 09:09:41.372 :: “\_path”: “C:\\ProgramData\\slicer.org\\3D Slicer 5.13.0-2026-09-30\\slicer.org\\Extensions-35317\\DentalSegmentator\\lib\\Slicer-5.13\\qt-scripted-modules\\Resources\\ML”
> 
> 2026/10/02 09:09:41.372 :: }
> 
> 2026/10/02 09:09:41.372 :: }
> 
> 2026/10/02 09:09:41.398 :: nnUNet preprocessing…
> 
> 2026/10/02 09:09:44.911 :: Traceback (most recent call last):
> 
> 2026/10/02 09:09:44.911 :: File “”, line 198, in \_run\_module\_as\_main
> 
> 2026/10/02 09:09:44.911 :: File “”, line 88, in \_run\_code
> 
> 2026/10/02 09:09:44.911 :: File “C:\ProgramData\slicer.org\3D Slicer 5.13.0-2026-09-30\lib\Python\Scripts\nnUNetv2\_predict.exe\__main_\_.py”, line 2, in
> 
> 2026/10/02 09:09:44.919 :: File “C:\ProgramData\slicer.org\3D Slicer 5.13.0-2026-09-30\lib\Python\Lib\site-packages\nnunetv2\inference\predict\_from\_raw\_data.py”, line 23, in
> 
> 2026/10/02 09:09:44.919 :: from nnunetv2.configuration import default\_num\_processes
> 
> 2026/10/02 09:09:44.919 :: File “C:\ProgramData\slicer.org\3D Slicer 5.13.0-2026-09-30\lib\Python\Lib\site-packages\nnunetv2\configuration.py”, line 10, in
> 
> 2026/10/02 09:09:44.919 :: default\_n\_proc\_DA = get\_allowed\_n\_proc\_DA()
> 
> 2026/10/02 09:09:44.919 :: ^^^^^^^^^^^^^^^^^^^^^^^
> 
> 2026/10/02 09:09:44.919 :: File “C:\ProgramData\slicer.org\3D Slicer 5.13.0-2026-09-30\lib\Python\Lib\site-packages\nnunetv2\utilities\default\_n\_proc\_DA.py”, line 23, in get\_allowed\_n\_proc\_DA
> 
> 2026/10/02 09:09:44.919 :: hostname = subprocess.getoutput([‘hostname’])
> 
> 2026/10/02 09:09:44.919 :: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 
> 2026/10/02 09:09:44.919 :: File “C:\ProgramData\slicer.org\3D Slicer 5.13.0-2026-09-30\lib\Python\Lib\subprocess.py”, line 691, in getoutput
> 
> 2026/10/02 09:09:44.919 :: return getstatusoutput(cmd, encoding=encoding, errors=errors)[1]
> 
> 2026/10/02 09:09:44.919 :: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 
> 2026/10/02 09:09:44.919 :: File “C:\ProgramData\slicer.org\3D Slicer 5.13.0-2026-09-30\lib\Python\Lib\subprocess.py”, line 671, in getstatusoutput
> 
> 2026/10/02 09:09:44.919 :: data = check\_output(cmd, shell=True, text=True, stderr=STDOUT,
> 
> 2026/10/02 09:09:44.919 :: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 
> 2026/10/02 09:09:44.919 :: File “C:\ProgramData\slicer.org\3D Slicer 5.13.0-2026-09-30\lib\Python\Lib\subprocess.py”, line 466, in check\_output
> 
> 2026/10/02 09:09:44.919 :: return run(\*popenargs, stdout=PIPE, timeout=timeout, check=True,
> 
> 2026/10/02 09:09:44.919 :: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 
> 2026/10/02 09:09:44.919 :: File “C:\ProgramData\slicer.org\3D Slicer 5.13.0-2026-09-30\lib\Python\Lib\subprocess.py”, line 550, in run
> 
> 2026/10/02 09:09:44.919 :: stdout, stderr = process.communicate(input, timeout=timeout)
> 
> 2026/10/02 09:09:44.919 :: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> 
> 2026/10/02 09:09:44.919 :: File “C:\ProgramData\slicer.org\3D Slicer 5.13.0-2026-09-30\lib\Python\Lib\subprocess.py”, line 1196, in communicate
> 
> 2026/10/02 09:09:44.919 :: stdout = self.stdout.read()
> 
> 2026/10/02 09:09:44.919 :: ^^^^^^^^^^^^^^^^^^
> 
> 2026/10/02 09:09:44.919 :: File “”, line 322, in decode
> 
> 2026/10/02 09:09:44.919 :: UnicodeDecodeError: ‘utf-8’ codec can’t decode byte 0x82 in position 86: invalid start byte
> 
> 2026/10/02 09:09:45.449 :: Loading inference results…
> 
> 2026/10/02 09:09:50.156 :: Error loading results :
> 
> 2026/10/02 09:09:50.156 :: Failed to load the segmentation.
> 
> 2026/10/02 09:09:50.156 :: Something went wrong during the nnUNet processing.
> 
> 2026/10/02 09:09:50.156 :: Please check the logs for potential errors and contact the library maintainers.

Let me know if you can help me ! Thank you !

---

<div class="post-metadata">

**Author:** ![Thibault\_Pelletier](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/thibault_pelletier/32/4767_2.png) [@Thibault\_Pelletier](https://discourse.slicer.org/u/Thibault_Pelletier)\
**Post date:** [October 2, 2026, 7:08am UTC](https://discourse.slicer.org/t/dentalsegmentator-error-loading-results-failt-to-load-the-segmentation/48375/3 "2026-10-02T07:08:30Z")

</div>

Hi @samzou974,

Thanks for reporting this error.  
I’m investigating where the problem may come from and I’ll get back to you if I have a solution to test.

---

<div class="post-metadata">

**Author:** ![Thibault\_Pelletier](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/thibault_pelletier/32/4767_2.png) [@Thibault\_Pelletier](https://discourse.slicer.org/u/Thibault_Pelletier)\
**Post date:** [October 2, 2026, 7:48am UTC](https://discourse.slicer.org/t/dentalsegmentator-error-loading-results-failt-to-load-the-segmentation/48375/4 "2026-10-02T07:48:34Z")

</div>

Hi @samzou974,

Looking at the nnUNet code, it seems to be accessing the hostname at some given time which will yield problems depending on the host local langage.

Can you try defining the following env variables in your console and running the inference:

```python
import os

cpu_count = os.cpu_count() or 1
os.environ["nnUNet_n_proc_DA"] = str(min(12, cpu_count))
os.environ["nnUNet_def_n_proc"] = str(min(8, cpu_count))

```

---

<div class="post-metadata">

**Author:** ![samzou974](https://avatars.discourse-cdn.com/v4/letter/s/e9a140/32.png) [@samzou974](https://discourse.slicer.org/u/samzou974)\
**Post date:** [October 2, 2026, 12:42pm UTC](https://discourse.slicer.org/t/dentalsegmentator-error-loading-results-failt-to-load-the-segmentation/48375/5 "2026-10-02T12:42:20Z")

</div>

It works well !! Thank you very much !  
Do I will have to define the env fariables each time i use 3dSlicer or Dentalsegmentator ?

---

<div class="post-metadata">

**Author:** ![Thibault\_Pelletier](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/thibault_pelletier/32/4767_2.png) [@Thibault\_Pelletier](https://discourse.slicer.org/u/Thibault_Pelletier)\
**Post date:** [October 2, 2026, 12:50pm UTC](https://discourse.slicer.org/t/dentalsegmentator-error-loading-results-failt-to-load-the-segmentation/48375/6 "2026-10-02T12:50:40Z")

</div>

Happy that it works.

For now, you can add it to your `.slicerrc.py` file and it will be applied automatically.  
I have pushed a [fix in the SlicerNNUNet repo](https://github.com/KitwareMedical/SlicerNNUnet/pull/30) that will apply this logic internally before starting the subprocess. It should be available in future Slicer versions.

---

<div class="post-metadata">

**Author:** ![samzou974](https://avatars.discourse-cdn.com/v4/letter/s/e9a140/32.png) [@samzou974](https://discourse.slicer.org/u/samzou974)\
**Post date:** [October 5, 2026, 4:05am UTC](https://discourse.slicer.org/t/dentalsegmentator-error-loading-results-failt-to-load-the-segmentation/48375/7 "2026-10-05T04:05:08Z")

</div>

Je viens de capter que tu était français ! Merci pour ton aide et bravo pour Dentalsegmentator et ta collaboration avec la FFO !

J’en profite pour te poser une dernière question : sais-tu si il y aurait un projet de segmentation automatique unitaire des dents avec Dentalsegmentator ?

Ceci afin de proposer une solution open-source en remplacement des fonctions payantes de RELU ou Diagnocat ?

Merci encore, bonne journée !

Samuel JEU

---

<div class="post-metadata">

**Author:** ![Thibault\_Pelletier](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/thibault_pelletier/32/4767_2.png) [@Thibault\_Pelletier](https://discourse.slicer.org/u/Thibault_Pelletier)\
**Post date:** [October 5, 2026, 6:58am UTC](https://discourse.slicer.org/t/dentalsegmentator-error-loading-results-failt-to-load-the-segmentation/48375/8 "2026-10-05T06:58:39Z")

</div>

Merci @samzou974,

Pour le discourse, c’est mieux de garder les posts en anglais pour qu’ils profitent au plus grand nombre 🙂

For English speakers:

> @samzou974 is asking if there is a similar project to DentalSegmentator with individual tooth segmentation to have an opensource alternative to RELU and Diagnocat.

[Total Segmentator](https://openaccess.thecvf.com/content/CVPR2025/html/Bolelli_Segmenting_Maxillofacial_Structures_in_CBCT_Volumes_CVPR_2025_paper.html) now has support for individual tooth segmentation (which should be available from the Total Segmentator extension).

To my knowledge, it was trained on the tooth fairy challenge dataset.