PDF version of this document
As we attempted to identify the point of convergence for our algorithm, we discovered a clear flaw with the cost we had defined. It turned out that the way in which we evaluate models neglects to account for small artifacts which must be encapsulated in this model. More crucially, these small artifacts which are left-out residuals need to form part of the model cost. In their absence, our objective function was able to drift away, thereby hiding vital structures in images. Not only was the result of registration poorer due to an improper model cost, but also it was impossible to contend that we ever reach one unique solution.