Open NanYoMy opened 5 years ago
Hello @NanYoMy
If you could give me a sample image (or its representative array) I could give you a better answer. Even if it happens to be the case that this library doesn't work for your images then if I have an example maybe I could update my code to support them.
hi, @AlyShmahell suppose I have two 3-D images i1 = tf.placeholder(dtype=tf.int64, shape=(None,None,None,None)) #batch,width,height,deepth i2 = tf.placeholder(dtype=tf.int64, shape=(None,None,None,None))#batch,width,height,deepth i wanna know if the "mutual_information" works for such 3-D images. centr = mutual_information(i1, i2) ? thanks!!!
Hello @NanYoMy
At the moment it only supports ndims==2, but I'm working on a second version which includes proper gradients, I will try to support higher dimensions within it. It will take me some time as this is a pet project which I will integrate into something else next year.
hi, AlyShamahell I am working on the mutual information of the 3D CT/MR image. I want to know if your code suit for such 3D image data.