How to do the cross validation for the patients data, and what data structure better suits to send it to classifier?

Hi,

I have 44 patients 3D MRIs as training. The task that I am doing is segmentation by using voxel classification. I have extracted some hand-crafted features and send them to the classifier to be classified.

I need to train the 44 patients data which ranging from almost 300k-400k records for each patient.

I need to train the model based on 4-fold cross validation (10/30 for validation and train) how can I set the fitensemble classifier to consider the cross-validation (`cvpartition`) for this?

I have save patients data into a cell

features{1,i} and labels{1,i}   % i is for each patients ranging from 1 to 40

, each element of cell is one patient data. I do not know how to program this!

Your expert opinion is really appreciated

Thanks