MONAI Label: segmentation of organ at risk in head and neck CT images

[GhTara]

Dear MONAI,

My goal is to segment organs at risk from CT images of the brain such as left/right eye, left/right lens, left/right optic nerve, optic chiasm, brain stem,... overall about 17 structures. recently I have got to know about MONAI Label and I tried it with a 3D slicer but I got an error! I don't know exactly why but, I think because of my dataset (they are in Dicom format).
so I have two questions: first, I don't know I am on the right way so, is the way I am trying, will have an effective outcome?
second, for segmenting these 17 structures, does exist a public model or I should train my own model?
I am new to this field and MONAI, please guide me.

Thank you

[Nic-Ma]

Hi @SachidanandAlle ,

I think this is a question about MONAILabel, could you please help move to the MONAILabel repo and share some comments?

Thanks in advance.

[finalelement]

Hi @GhTara

Thanks for raising the question for MONAI Label. Please start it off here at this link

For us to be able to help you, please provide additional more information when starting off the new discussion on MONAI Label:

• What error did you get?
• How many 3D Volumes do you have?
• What is the way that you are trying?
• Given that they are structures of the brain, I do not expect any off the shelf model to work directly.
• Deepgrow 2D might be worth a shot. If not then starting off training of these new structures using the pretrained weights that we already have would be quite beneficial for the model that you train.

[GhTara]

Hi @finalelement

Thank you for your guidance.

great, I will do it.