I am interested in applying Deep Learning for fMRI( 4D) and MRI (3D) data for disease classification. I want to use the data as 3D or 4D, but not convert them into 2D images. I have seen some online resources with code, in which they used input as 2D images. I also want to use transfer learning.
Q1.) Would any one provide some resources( probably with code) to build neural networks for fMRI and MRI ?
Q2.) As these images are 3D and 4D and networks used for transfer learning were trained on 2D images, how should I proceed to apply them on MRI and fMRI data.
I am really struggling from past two weeks on these issues. It would be really helpful, if some one provide some pointers.
Hi!
I am also trying to figure out understandable easy ways to load my 3D MRI data into pytorch, here there is an amazing discussion with some code, https://forums.fast.ai/t/custom-data-loader-for-3d-data/27290/9
For keras I wrote a very simple kernel to make dataset from numpy arrays but currently I am changing it to a pandas based method https://www.kaggle.com/meltematay/dataset-maker-for-any-dim-data-mainly-3d-npy
And for handling nifti data (standard format of processed fmris or mris generally)
I know these are very basic…
For Q2, transfer learning approach would not work here since there is no such thing as imagenet in 3D… If someone trained larger datasets with 3D images it would be possible to use transfer learning. But implementing similar networks and trying to train from scratch can also work.