How to train feedforwardnet neural network with multiple image inputs?

Hi, I'm doing retinal blood vessel segmentation using feedforwardnet. I've trained and tested the network with one image and it works fine. But the problem arises during training with multiple images. I have total 14 features and I tried to keep these features of each image in cell array. Then use this cell array as input in neural network. But it shows error. I've alrady read almost all the solutions given in mathworks, but none of it applies to my case.

Can anyone tell if my cell array approach is alright or not? If alright, how to give input using cell array in neural network. Example will be helpful. If not alright, what appraoch should I follow?

Any help would be appreciated.

Thank you.