Search Results (2)

Segmenting white matter tracts in diffusion-weighted magnetic resonance imaging (dMRI) is of vital importance for brain health analysis. It remains a challenging task due to the intersection and overlap of tracts (i.e., multiple tracts coexist in one voxel) and the data complexity of dMRI images (e.g., 4D high spatial resolution). Existing methods that demonstrate good performance implement direct volumetric tract segmentation by performing on individual 2D slices. However, this ignores 3D contextual information, requires additional post-processing, and struggles with the boundary handling of 3D volumes. Therefore, in this paper, we propose an efficient 3D direct volumetric segmentation method for segmenting white matter tracts. It has three key innovations. First, we propose to deeply interleave convolutions and transformer blocks into a U-shaped network, which effectively integrates their respective strengths to extract spatial contextual features and global long-distance dependencies for enhanced feature extraction. Second, we propose a novel channel-wise transformer, which integrates depth-wise separable convolution and compressed contextual feature-based channel-wise attention, effectively addressing the memory and computational challenges of 4D computing. Moreover, it helps to model global dependencies of contextual features and ensures each hierarchical layer focuses on complementary features. Third, we propose to train a fully symmetric network with gradually sized volumetric patches, which can solve the challenge of few 3D training samples and further reduce memory and computational costs. Experimental results on the largest publicly available tract-specific tractograms dataset demonstrate the superiority of the proposed method over the current state-of-the-art methods. Full article

There is great interest in the study of brain structural connectivity, as white matter abnormalities have been implicated in many disease states. Diffusion magnetic resonance imaging (MRI) provides a powerful means to characterise structural connectivity non-invasively, by using a fibre-tracking algorithm. The most widely used fibre-tracking strategy is based on the step-wise generation of streamlines. Despite their popularity and widespread use, there are a number of practical considerations that must be taken into account in order to increase the robustness of streamlines tracking results, particularly when these methods are used to study brain structural connectivity, and the connectome. This review article describes what we consider the ‘seven deadly sins’ of mapping structural connections using diffusion MRI streamlines fibre-tracking, with particular emphasis on ‘sins’ that can be practically avoided and they can have an important impact in the results. It is shown that there are important ‘deadly sins’ to be avoided at every step of the pipeline, such as during data acquisition, during data modelling to estimate local fibre architecture, during the fibre-tracking process itself, and during quantification of the tracking results. The recommendations here are intended to inform users on potential important shortcomings of their current tracking protocols, as well as to guide future users on some of the key issues and decisions that must be faced when designing their processing pipelines. Full article