Deformable Groupwise Image Registration using Low-Rank and Sparse Decomposition

Groupwise image registration describes the problem of simultaneously aligning a series more than two images through individual spatial deformations and it is common task in processing medical sequences. Variational methods with data fidelity terms based on robust PCA (RPCA) have proven successful accounting for structural changes intensity stemming, e.g., from uptake contrast agent functional imaging. In this article, we investigate drawbacks most commonly used RPCA term derive an improved replacement that employs explicit constraints instead penalties. We further present multilevel scheme theoretically justified scaling to solve underlying fully deformable model. Our numerical experiments synthetic real-life confirm advanced adaptability RPCA-based showcase accuracy our algorithm when compared related groupwise approaches.