Enhancement of Perivascular Spaces in 7 T MR Image using Haar Transform of Non-local Cubes and Block-matching Filtering

Abstract

Perivascular spaces (PVSs) in brain have a close relationship with typical neurological diseases. The quantitative studies of PVSs are meaningful but usually difficult, due to their thin and weak signals and also background noise in the 7 T brain magnetic resonance images (MRI). To clearly distinguish the PVSs in the 7 T MRI, we propose a novel PVS enhancement method based on the Haar transform of non-local cubes. Specifically, we extract a certain number of cubes from a small neighbor to form a cube group, and then perform Haar transform on each cube group. The Haar transform coefficients are processed using a nonlinear function to amplify the weak signals relevant to the PVSs and to suppress the noise. The enhanced image is reconstructed using the inverse Haar transform of the processed coefficients. Finally, we perform a block-matching 4D filtering on the enhanced image to further remove any remaining noise, and thus obtain an enhanced and denoised 7 T MRI for PVS segmentation. We apply two existing methods to complete PVS segmentation, i.e., (1) vesselness-thresholding and (2) random forest classification. The experimental results show that the PVS segmentation performances can be significantly improved by using the enhanced and denoised 7 T MRI.

Image processing approaches to enhance perivascular space visibility and quantification using MRI

“…Our quantitative PVS mapping and previous works showed that PVS can be mapped from an individual MRI contrast (Ballerini et al, 2018;Boespflug et al, 2018;Comulada, 2015;Del C. Valdes Hernandez et al, 2013;Descombes et al, 2004;Hou et al, 2017;Jung et al, 2018;Park et al, 2016;Ramirez et al, 2015;Wang et al, 2016;Zong et al, 2016). However, it should be noted that in the presence of pathology, additional image sequences (e.g.…”

“…EPC also uses a spatial non-local mean filtering technique, which has shown to be effective for mapping PVS (Hou et al, 2017). We noted that even in the high-resolution T2w images of the human connectome project (0.7 mm 3 resolution), it is difficult to detect small PVS (Supplementary Figure 2), while they could be identified with this new technique.…”

“…In order to improve the mapping of the PVS, the MRI contrast of the PVS and the neighboring parenchyma should be increased. Besides the image processing approaches, PVS contrast can be enhanced through MRI technological improvement such as optimizing imaging sequence and employing ultra-high field technology (Barisano et al, 2018;Hou et al, 2017;Park et al, 2016;Zong et al, 2016).…”

“…In the clinical practice, PVS is quantified based on the number of visible PVS on the axial slice of a T2-weighted (T2w) image that has the highest number of PVS in the region of interest . This process can be laborious and error-prone, so efforts to improve efficiency and accuracy have been made by using a wide range of automatic or semi-automatic segmentation techniques, from classical image processing approaches to deep neural network modelling (Ballerini et al, 2018;Boespflug et al, 2018;Comulada, 2015;Del C. Valdes Hernandez et al, 2013;Descombes et al, 2004;Hou et al, 2017;Jung et al, 2018;Park et al, 2016;Ramirez et al, 2015;Wang et al, 2016;Zong et al, 2016). Park et al used auto-context orientational information of the PVS for automatic segmentation .…”

Block‐Extraction and Haar Transform Based Linear Singularity Representation for Image Enhancement

“…Therefore, these methods are called local methods. In recent years, some nonlocal image processing methods have been developed for image denoising, i.e., nonlocalmeans (NL means) [8] and block-matching and 3D filtering (BM3D) [9][10][11][12][13][14][15]. These nonlocal methods attempt to find some of the most similar image blocks, i.e., by implementing block-matching and the weighted means on the blocks [8], or by implementing a 3D transform on a 3D array stacked by similar blocks for enhanced sparsity representation and then for hard-threshold shrinkage on the transformed coefficients 2 Mathematical Problems in Engineering to achieve image denoising [9,10,[16][17][18][19].…”