Real-Time image interpolation for medical image display

Context: I'm about to start making a GUI in which a 3D medical image data-set will be displayed. Essentially the user will be able to look at one image slice at a time and will scroll to see subsequent slices. Often the size (pixels) of the image being displayed is smaller than the size (pixels) of the image object's parent axis (for example a 512x512 image being displayed on 1024x1024 screen pixels). Thus MATLAB must upsample the image data in order to display the image. I think by default MATLAB just uses nearest-neighbor interpolation. This results in a pixelated appearance if you have many screen pixels compared to image pixels.

Question: Is it possible to change the interpolation MATLAb uses to display images on the screen (linear, cubic, lanczos, etc...)? I know I could use the imresize() function and upsample the image to match the size of the image object's parent axis. I'm worried this method will be too slow if the user is quickly scrolling through many slices. I'm looking for a lower-level solution. Any ideas?