1: \begin{abstract}

2: 

3: The class of non-rigid registration methods proposed in the framework of PDE-constrained 

4: Large Deformation Diffeomorphic Metric Mapping is a particularly interesting family of physically 

5: meaningful diffeomorphic registration methods. 

6: % PDE-constrained LDDMM methods are formulated as constrained variational problems,

7: % where the different physical models are imposed using the associated partial differential equations 

8: % as hard constraints.

9: Inexact Newton-Krylov optimization has shown an excellent numerical accuracy and an extraordinarily 

10: fast convergence rate in this framework.

11: However, the Galerkin representation of the non-stationary velocity fields does not provide proper geodesic paths.

12: In this work, we propose a method for PDE-constrained LDDMM parameterized in the space of initial velocity fields

13: under the EPDiff equation.

14: The derivation of the gradient and the Hessian-vector products are performed on the final velocity field and

15: transported backward using the adjoint and the incremental adjoint Jacobi equations.

16: This way, we avoid the complex dependence on the initial velocity field in the derivations and the 

17: computation of the adjoint equation and its incremental counterpart.

18: The proposed method provides geodesics in the framework of PDE-constrained LDDMM, and 

19: it shows performance competitive to benchmark PDE-constrained LDDMM and EPDiff-LDDMM methods.

20: 

21: \end{abstract}

22: