ahatamiz
commented
2 years ago Hi @BelieferQAQ
I'd assume the labels should have a size of BxH×W×D (if 3D) and contain labels from 0 to 4, and setting args.out_channel to 5 shall make it work.
Open BelieferQAQ opened 2 years ago
ahatamiz
commented
2 years ago Hi @BelieferQAQ
I'd assume the labels should have a size of BxH×W×D (if 3D) and contain labels from 0 to 4, and setting args.out_channel to 5 shall make it work.
BelieferQAQ
commented
2 years ago Hello! Thank you for your answer. There are five fracture types in the ribfrac dataset. With the background, it should be set to six types in the JSON file. In addition, the intensity values of five fracture types are 1, 2, 3, 4, -1 (background is 0). Do I need to change the intensity value of this fracture type from -1 to 5.
dreamwish1998
commented
2 years ago If the number of target classes for the task of CT segmentation is not 14, such as 1 or 2, can I still use your pretrained model? If yes, what parameters need to be modified? Looking forward to your reply, thank you!
SiruLiu645
commented
1 year ago 如果CT分割任务的目标类数量不是14个，比如1个或2个，我还可以使用你们的预训练模型吗？如果是，需要修改哪些参数？期待您的回复，谢谢！ I have the same problem，did you solve it？
Hello! Each CT in the btcv data set used in the code contains all 14 organs. I want to apply unetr to the ribtrac data set. Each CT contains at most 5 types of rib fracture (most CT only contains 2-4 types). How can I change the code? Looking forward to your reply, thank you