-
Hello, Bear with me, I am pretty novice in this field. I am trying to segment out bone in some 2D ultrasound datasets using 3D Slicer and MONAILabel. The images I am working with are 0 to 255 grayscale pngs (although I can convert them to mha files). I would like to manually segment bone using MONAILabel, create labels, and train a model. But I am not sure what model to use or if it is even possible. I have successfully installed both 3D Slicer and MONAILabel and the server does run without errors. I just am not sure where to go from there. Thank you! |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments 6 replies
-
You could manually annotate the target just as you show above, once finished labeling, you could click "submit label" button to save the label. Till there are bunches of labels save, you could click training to train a segmentation model. But before that, you need to choose or create a model, using radiology app. But there are no existing 2D ultrasound models, you might need to create a customized app with a defined network, transforms, etc... |
Beta Was this translation helpful? Give feedback.
-
Hello, I'm having a similar issue as you, I wonder if you trained the model for the ultrasound bone segmentation using MONAI? |
Beta Was this translation helpful? Give feedback.
You could manually annotate the target just as you show above, once finished labeling, you could click "submit label" button to save the label. Till there are bunches of labels save, you could click training to train a segmentation model.
But before that, you need to choose or create a model, using radiology app. But there are no existing 2D ultrasound models, you might need to create a customized app with a defined network, transforms, etc...