Your code
for i=1:file_len
I{i} = dicomread(filename{i});
Q{i}=imresize(I{i},0.4);
end
II = cell2mat(Q);
is creating II (and therefor the later AA) by concatenating the data long the second dimension. I suspect you wanted to concenate along the third dimension. If so then right before the above for loop insert
Q = cell(1,1,file_len);
Also I do not understand why you later squeeze(AA). Which dimension might be singleton?
When I am reading other people's code, I am more comfortable with them using permute() or reshape() rather than squeeze(), at least if it is not very obvious that one dimension has just been made singleton and the other dimensions are definitely not singleton. When the reader of the code does not know ahead of time what the dimensions of the data are to be, the reader of the code must assume that squeeze() could be operating upon an unexpected dimension, such as if it turned out that the data only had a single row. permute() and reshape() remove the possibility of changing any dimension not explicitly intended to be changed.
