Closed frfa1 closed 2 years ago
So to clarify: I'm using the get_prob_heatmap
function in dm_image.py
to get the heatmap of an image. The function output is 3-dimensional with the last dimension being 5: (..., ..., 5). Each of those 5 correnspond to a heatmap for one of the classes. But I'm unsure of the order?
The patch training scripts are set up to train three classes: (background, malignant, benign).
If you could clarify this, I would be so happy, as I'm using the heatmaps in a thesis.
best regards
off the top of my head, the order should be: background, benign calcification, malignant calcification, benign mass, malignant mass.
Do you have any way to confirm this?
A weak link: In the paper the ordering is: background, malignant mass, benign mass, malignant calcification and benign calcification.
This is a different order. So just want to be sure :-))
Do you have any way to confirm this?
A weak link: In the paper the ordering is: background, malignant mass, benign mass, malignant calcification and benign calcification.
This is a different order. So just want to be sure :-))
Yes, I can confirm that. Refer to the Fig. 2 and Fig. 4 of the original paper.
Is it noted in the paper, that the patch classifiers are all trained to predict the following classes: background, malignant mass, benign mass, malignant calcification and benign calcification. It seem consistent with the model prediction outputs of the patch classifiers in the Google Drive, which has a size of 5 output classes.
But how do we check which class each of the patch model outputs are associated with? I.e. what is the ordering of the outputs.