Yes, you can represent. But if you want to feed this to 3D CNN, then image shape should be (height, width, depth, 1).