next up previous
Next: Optimising Scale and Translation Up: Back to Rigid and Previous: Scale


Levels were lowered down from [2 2] to [1 1] and finally [0 0] throughout the three stages. The first stage makes significant improvements, quite unsurprisingly. The subsequent stage does make improvements as well. It's refining the overlap of the brain edges. The third stage makes noticeable differences, but it is hard to see if these are beneficial. It would be interesting to look at the difference in objective function value.