Abstract
Magnetic Resonance Imaging (MRI) is a technology for non-invasive imaging ofanatomical features in detail. It can help in functional analysis of organs ofa specimen but it is very costly. In this work, methods for (i) virtualthree-dimensional (3D) reconstruction from a single sequence of two-dimensional(2D) slices of MR images of a human spine and brain along a single axis, and(ii) generation of missing inter-slice data are proposed. Our approach helps inpreserving the edges, shape, size, as well as the internal tissue structures ofthe object being captured. The sequence of original 2D slices along a singleaxis is divided into smaller equal sub-parts which are then reconstructed usingedge preserved kriging interpolation to predict the missing slice information.In order to speed up the process of interpolation, we have used multiprocessingby carrying out the initial interpolation on parallel cores. From the 3D matrixthus formed, shearlet transform is applied to estimate the edges consideringthe 2D blocks along the $Z$ axis, and to minimize the blurring effect using aproposed mean-median logic. Finally, for visualization, the sub-matrices aremerged into a final 3D matrix. Next, the newly formed 3D matrix is split upinto voxels and marching cubes method is applied to get the approximate 3Dimage for viewing. To the best of our knowledge it is a first of its kindapproach based on kriging interpolation and multiprocessing for 3Dreconstruction from 2D slices, and approximately 98.89\% accuracy is achievedwith respect to similarity metrics for image comparison. The time required forreconstruction has also been reduced by approximately 70\% with multiprocessingeven for a large input data set compared to that with single core processing.