Hi,
I have tried the pretrained MONAI1.5.0 model with the CLI tool on my data, the result looks weried.
By using the web app follow the video and one HCP-YA acpc restore image as input, the segmentation also wrong for both of the 3 models.
It seems the model always infer the brain segmentation from a space that different from the input. I'm not sure there are still some preprocessing should be performed ahead?
Hi,
I have tried the pretrained MONAI1.5.0 model with the CLI tool on my data, the result looks weried.
By using the web app follow the video and one HCP-YA acpc restore image as input, the segmentation also wrong for both of the 3 models.
It seems the model always infer the brain segmentation from a space that different from the input. I'm not sure there are still some preprocessing should be performed ahead?