Spaces:

keras-io

/

3D_CNN_Pneumonia

like 5

Runtime error

Currently, the example is a preprocessed 3D image. The inference function too assumes that it gets a preprocessed 3D image of a CT scan. I did this out of simplicity, but I don't think its a good idea to make it this way.

To work around this, I can maybe break down the code of original keras script so that gradio takes ".nii" file as input and then plots 20 slices out of it