Groupwise non-rigid registration on multiparametric abdominal DWI acquisitions for robust ADC estimation: Comparison with pairwise approaches and different multimodal metrics

Abstract

Registration of diffusion weighted datasets remains a challenging task in the process of quantifying diffusion indexes. Respiratory and cardiac motion, as well as echo-planar characteristic geometric distortions, may greatly limit accuracy on parameter estimation, specially for the liver. This work proposes a methodology for the non-rigid registration of multiparametric abdominal diffusion weighted imaging by using different well-known metrics under the groupwise paradigm. A three-stage validation of the methodology is carried out on a computational diffusion phantom, a watery solution phantom and a set of voluntary patients. Diffusion estimation accuracy has been directly calculated on the computational phantom and indirectly by means of a residual analysis on the real data. On the other hand, effectiveness in distortion correction has been measured on the phantom. Results have shown statistical significant improvements compared to pairwise registration being able to cope with elastic deformations.

Robust estimation of the apparent diffusion coefficient invariant to acquisition noise and physiological motion

“…Both real and synthetic datasets have been processed using different groupwise multimodal metrics [45] as well as pairwise registration methodologies. As 275 for the former, apart from the pipeline here proposed, we have tried the Entropy of the distribution of intensities (Entr.)…”

On motion in dynamic magnetic resonance imaging: Applications in cardiac function and abdominal diffusion

“…We have first tackled the problem of motion insensitive ADC estimation by introducing GW nonrigid registration techniques as a preprocessing MC step previous to the ADC estimation. In Sanz-Estébanez et al (2017) we presented an study intended to assess the adequateness of GW and PW paradigms within an elastic transformation model for the alignment of DW sequences, focusing on the liver. The suitability of different well-known multimodal voxel-based metrics, such as Entropy of the Distribution of Intensities (EDI), Variance of the Local Entropy (VLE), Modality Independent Neighbourhood Descriptor (MIND) and Normalized Cross Correlation (NCC) was also tested.…”

“…The WFT provides a representation of the image spectrum in the surroundings of each pixel of the original image, so HARP band pass filtering techniques can be directly applied on the spatially localized spectrum of the image. To adequately retrieve the shape of the spectral peaks, we have resorted to an anisotropic filtering approach combining Gaussian band-pass and all-pass filters as proposed in Sanz-Estébanez et al (2017). Finally, each of the image phase ϕ i (x) (two for each plane) can be extracted in the spatial domain from the IWFT of the aforementioned filtered spectrum.…”

“…The spatial resolution of the reconstructed tensors depends on the width of the HARP band pass filter (Parthasarathy andPrince, 2003, 2004); the HARP method is upper limited by half of the tag spacing (small deformation assumption). However, WHARP methods (Sanz-Estébanez et al, 2017;Cordero-Grande et al, 2016) try to accommodate the band pass filter to the the local frequency of the signal in order to approach to the maximum achievable resolution. Therefore, effective HARP resolution will vary dynamically, allowing large deformations, as those observed at ES, being captured at a maximal scale of 1.5 times half the tag spacing.…”

“…A GD optimization scheme is used (Sanz-Estébanez et al, 2017), where the step size ω i is updated along iterations according to the variation in the registration metric as:…”