Locating abnormalities in brain blood vessels using parallel computing architecture

Abstract

CT and MRI scans are widely used in medical diagnosis procedures, but they only produce 2-D images. However, the human anatomical structure, the abnormalities, tumors, tissues and organs are in 3-D. 2-D images from these devices are difficult to interpret because they only show cross-sectional views of the human structure. Consequently, such circumstances require doctors to use their expert experiences in the interpretation of the possible location, size or shape of the abnormalities, even for large datasets of enormous amount of slices. Previously, the concept of reconstructing 2-D images to 3-D was introduced. However, such reconstruction model requires high performance computation, may either be time-consuming or costly. Furthermore, detecting the internal features of human anatomical structure, such as the imaging of the blood vessels, is still an open topic in the computer-aided diagnosis of disorders and pathologies. This paper proposes a volume visualization framework using Compute Unified Device Architecture (CUDA), augmenting the widely proven ray casting technique in terms of superior qualities of images but with slow speed. Considering the rapid development of technology in the medical community, our framework is implemented on Microsoft.NET environment for easy interoperability with other emerging revolutionary tools. The framework was evaluated with brain datasets from the department of Surgery, University of North Carolina, United States, containing around 109 MRA datasets. Uniquely, at a reasonably cheaper cost, our framework achieves immediate reconstruction and obvious mappings of the internal features of human brain, reliable enough for instantaneous locations of possible blockages in the brain blood vessels.

ConnectViz: Accelerated Approach for Brain Structural Connectivity Using Delaunay Triangulation

“…Presents the networks' interactive operations such as separation of edges and nodes, rotate, zoom-in and zoom-out However, in order to achieve optimum performance, we accelerated the entire processes in this study with Compute Unified Device Architecture, CUDA. The four algorithms proposed for SurLens Visualization System [45] were adopted and utilized, coupled with the designed and the implemented Connectomic Visualization Algorithm. The Connectomic Visualization algorithm and its subroutine, Nectiz, are presented below.…”

“…Computational challenges of medical visualization algorithms have been an open issue to address in the recent years [45].…”

“…However, other possible approaches do exist such as Cg (C for graphics), High Level Shader Language (HLSL), OpenGL Shading Language (GLSL), but they are inherent shading languages in which all computations must only be expressed in graphic terms [49]. Some of the related works that utilized GPU computing for acceleration procedures in visualization are documented in Adeshina et al [45,49].…”

“…SurLens Visualization Architecture, which has been tested and found efficient [45] was utilized in the development of the application. SurLens architecture framework consists of four major phases, the SurLens Data Pre-processing, SurLens Accelerating Hardware, SurLens Graphic Execution and SurLens Volume Rendering phases.…”

“…CUDA has been leveraged in many circumstances in solving the computational overheads in visualization research [45,53]. In the phase 2, the Visualization architecture, the Step 1 handles fragmentation and apparently the fragmented results could be more reliable since there is less human interaction, the entire processes are controlled with Nvidia CUDA Compiler, nvcc.…”