search

frankkramer-lab / MIScnn

A framework for Medical Image Segmentation with Convolutional Neural Networks and Deep Learning
GNU General Public License v3.0
402 stars 116 forks source link

Prediction image dimensions. #92

Open lcaronson opened 3 years ago

lcaronson commented 3 years ago

Hello,

Is there any way to adjust this input image dimensions?

dimension_error

Clearly, case_00300 has different dimensions, compared to the rest of the dataset. Thus, I see the following error:

more_dimension_error

Any help would be greatly appreciated!

muellerdo commented 3 years ago

Hi @lcaronson,

there are two ways which pop in my mind:

1) Fixing this beforehand via loading the suspicious cases and switch the axis via NumPy. This is the way, I would highly recommend as it is probably the most cleanest approach to do it in the 'data exploration & preparation' step of your machine learning pipeline. 2) Fixing this via a subfunction by check any case during the preprocessing if the axes are correct and if not, switch the first one with the z-axis.

Cheers, Dominik