Could Deepgrow use dicom or png image to input?

I wanna use clara_train_deepgrow_aiaa_inference_only model to deepgrow , and get a problem. It’s only allow nifti format, I can’t use dicom or png image to input.
Pleas get me some idea,Thx.

Hi
Thank you for your interest in Clara train AIAA.

You can use 3d Slicer to open dicom images then the plugin would do the conversion for you. make sure you enable the session setting in the AIAA plugin for performance reasons.
If you are looking for a native way to send/receive Dicom from/to AIAA. we are working on integrating AIAA with OHIF viewer which would allow you to send dicoms you see using OHIF directly to AIAA and visualize results. Please stay tuned for this feature

HI
Thank you for respond.
I don’t require 3d Slicer.
I used https://github.com/NVIDIA/ai-assisted-annotation-client/tree/master/py_client/test_aiaa_server.py to send deegrow session to aiaa. I saw model’s config and this model required to input nifti instead of png or dicom.

yes… the model which you find in public ngc needs nifti as input (3D volume)…