Preprocessing Steps of Data

[ccomella]

Hello, very good job in implementing this CNN transformer, however I have some doubts in the preprocessing steps of the data

I am trying to use this code but with fMRI data. In my case I resample my dataset to the dimension of the images 384 x 256. However, in my case my dimensions are (88x88x64) where 64 are the number of slices. The dimension of the z-axis (slices) remains the same? In my case it is 64, or do I need to resample it also?

Another question is regarding to the patient dictionary: if I understood well, did you create the partition dictionary dividing training, validation and test by slices or patients? I mean, training will be 70% of my patients dataset taking into account all slices? It is a little confusing as I am saving it as s_patientID when s will be the slice number

Thank you in advanced,

Cristina Comella