Can I use a GPU for the neural network toolbox?

Hi. I am fairly new to using the neural network tool (nftool) but so far have achieved some amazing results. My data set consists of 50,000 "Data-sets" (which I call each individual member of the set a "line") Each line is the 150 voltage measurements from a biological sensor array. Each line contains 150 real numbers, and for each line I have a single real-number value (from a very slow and expensive external chemical lab test process) that indicates amount of biological-contamination. The neural fit tool was able to do a fantastic job (I think) of creating a net (45 neurons first layer, one neuron second layer) that could take in the bio-sensor data and "predict" the chemical test data (the goal is to eliminate the slow lab). The only problem is that with 1,000 samples input, it took about 9 hours to create the model. I need to do this much faster (according to my professor) so I was wondering if I could buy / use a GPU to sped this up? According to the documentation it is possible. If I were to go this route, can I still use the super-easy nntraintool GUI or will I be forced to code in Matlab? I am not a great matlab coder, and the GUI tool has made this very easy for me. I really want to be able to continue to use the GUI, is that possible with the nntraintool?

    Thank you!

     ---Lou