Yes, but it would be tedious relative to how super easily DirectRT could do it. In DirectRT, your images can go into independent or paired stimulus lists from which DirectRT can draw using a variety of randomization methods. Consequently, you could design one trial and you'd be done as it would simply repeat for as many pairs as you have to present. There are various ways to handle the randomization depending on what you want to do. If you still prefer to do it in MediaLab let me know and I can suggest a few things.