Left-invariant metrics for diffeomorphic image registration with spatially-varying regularisation

Tanya Schmah (1) , Laurent Risser (2) , François-Xavier Vialard (3)

1 Department of Computer Science
2 IMT - Institut de Mathématiques de Toulouse UMR5219
3 CEREMADE - CEntre de REcherches en MAthématiques de la DEcision

We present a new framework for diffeomorphic image registration which supports natural interpretations of spatially-varying metrics. This framework is based on left-invariant diffeomorphic metrics (LIDM) and is closely related to the now standard large deformation diffeomorphic metric mapping (LDDMM). We discuss the relationship between LIDM and LDDMM and introduce a computationally convenient class of spatially-varying metrics appropriate for both frameworks. Finally, we demonstrate the effectiveness of our method on a 2D toy example and on the 40 3D brain images of the LPBA40 dataset.

Domaines: Statistiques [math.ST]; Théorie [stat.TH]; Imagerie médicale

Tanya Schmah, Laurent Risser, François-Xavier Vialard. Left-invariant metrics for diffeomorphic image registration with spatially-varying regularisation. Lecture Notes in Computer Science, 2013, 8149, pp.203-210. ⟨10.1007/978-3-642-40811-3_26⟩. ⟨hal-00869476⟩