Multi-Steps Registration Protocol for Multimodal MR Images of Hip Skeletal Muscles in a Longitudinal Study

Abstract

Image registration is crucial in multimodal longitudinal skeletal muscle Magnetic Resonance Imaging (MRI) studies to extract reliable parameters that can be used as indicators for physio/pathological characterization of muscle tissue and for assessing the effectiveness of treatments. This paper aims at proposing a reliable registration protocol and evaluating its accuracy in a longitudinal study. The hips of 6 subjects were scanned, in a multimodal protocol, at 2 different time points by a 3 Tesla scanner; the proposed multi-step registration pipeline is based on rigid and elastic transformations implemented in SimpleITK using a multi-resolution technique. The effects of different image pre-processing (muscle masks, isotropic voxels) and different parameters’ values (learning rates and mesh sizes) were quantitatively assessed using standard accuracy indexes. Rigid registration alone does not provide satisfactory accuracy for inter-sessions alignment and a further elastic step is needed. The use of isotropic voxels, combined with the muscle masking, provides the best result in terms of accuracy. Learning rates can be increased to speed up the process without affecting the final results. The protocol described in this paper, complemented by open-source software, can be a useful guide for researchers that approach for the first time the issues related to the muscle MR image registration.

Deformable image registration for automatic muscle segmentation and the generation of augmented imaging datasets

“…Particularly successful approaches within the literature that used 3D deformable image registration to perform muscle segmentation were those based on longitudinal data, such as Le Troter et al [33] and Fontana et al [27], who attained average DSC of 0.90 and 0.85, respectively. Similar to the latter, were the DSC values here found when registering the left to right limb in the same subject.…”

“…Many different automatic segmentation methods have been investigated within the literature in recent years to replace the manual approach [21,24,25,27]. Statistical shape modelling (SSM) entails the generation of an average atlas geometry, which can be scaled or deformed to fit individuals.…”

“…Simplistic applications, such as two-dimensional (2D) deformable image registration between subsequent MR imaging slices within subjects has been used to propagate segmentations of individual slices into partial sections of 3D muscle geometry using only a few manually segmented slices, with encouraging results (DSC ≈ 0.91) [32]. 3D image registration has been used within longitudinal studies to populate MR images with partial segmentations of a small number of muscles to good effect, such as within the studies presented by Le Troter et al [33] and Fontana et al [34]. Though this longitudinal approach provides insight into the change in muscle characteristics over time, multiple MR image sequences are required from individual subjects at two different timepoints and one dataset must be manually segmented.…”

“…These limitations of manual segmentation prevent the utilisation of muscle segmentation as a technique to inform large-scale quantitative investigations into muscle characteristics. Many different automatic segmentation methods have been investigated within the literature in recent years to replace the manual approach [21,24,25,27]. Statistical shape modelling (SSM) entails the generation of an average atlas geometry, which can be scaled or deformed to fit individuals.…”

Deformable image registration based on single or multi-atlas methods for automatic muscle segmentation and the generation of augmented imaging datasets

“…Many different automatic segmentation methods have been investigated within the literature in recent years to replace the manual approach [13,17,20]. Image registration is one that has been explored within the literature to perform muscle segmentation [23][24][25]. Simplistic applications, such as two-dimensional (2D) deformable image registration between subsequent MR imaging slices within subjects has been used to propagate segmentations of individual slices into partial sections of 3D muscle geometry using only a few manually segmented slices, with encouraging results (DSC � 0.91) [23].…”