AI-assisted segmentation extension

After I downloaded the plug-in, an error appeared and the interface showed as follows. Could you please tell me how the author can solve this problem.
Traceback (most recent call last):
File “E:/Slicer 5.4.0/”, line 198, in fetchAIAAModels
models = self.logic.list_models()
File “E:/Slicer 5.4.0/”, line 1076, in list_models
return aiaaClient.model_list(label)
File “E:\Slicer 5.4.0\\Extensions-31938\NvidiaAIAssistedAnnotation\lib\Slicer-5.4\qt-scripted-modules\NvidiaAIAAClientAPI\”, line 190, in model_list
status, response = AIAAUtils.http_method(‘GET’, self._server_url, selector)
File “E:\Slicer 5.4.0\\Extensions-31938\NvidiaAIAssistedAnnotation\lib\Slicer-5.4\qt-scripted-modules\NvidiaAIAAClientAPI\”, line 620, in http_method
conn = httplib.HTTPConnection(parsed.hostname, parsed.port)
File “E:\Slicer 5.4.0\lib\Python\Lib\http\”, line 857, in init
(, self.port) = self._get_hostport(host, port)
File “E:\Slicer 5.4.0\lib\Python\Lib\http\”, line 891, in _get_hostport
i = host.rfind(‘:’)
AttributeError: ‘NoneType’ object has no attribute ‘rfind’

The authors of the NVIDIA-AIAA engine has moved on to develop MONAILabel package and Slicer extension.

You can still use the NVIDIA-AIAA extension, but it is not supported and not recommended anymore. Demo server is no longer available so you need to set up your own server and specify its address in the module in Slicer.

It is no longer necessary to set up a server for segmentation and not necessary to even have a GPU. Instead, there are several new extensions that allow segmenting images much more easily. For example, TotalSegmentator and MONAIAuto3DSeg extensions.

Hi, Tamas
I’m looking for the trained model (h5 file) shown in this video ( Tracked ultrasound AI segmentation and 3D reconstruction tutorial ( I’m not able to find it in the github repository. Could you help me with that, please?

Hi, that is a model saved in TensorFlow, in a version that is already a few years old. I wouldn’t recommend using it now. This folder contains the original data and a trained model: 2019_SpineUs-Queens
I recommend training a new model using MONAI rather than relying on the trained model in the folder.


Thank you for your response! I used the model available in the repository. Now, I would like to train a new model as you suggested. I’m very new to AI, so could you please guide me on how to do this using MONAI? Also, can I use the data provided in the repository to train the new model?

Thank you! :slight_smile:

Here are some related posts: