A Hybrid Method for 3D Reconstruction of MR Images

Abstract

Three-dimensional surface reconstruction is a well-known task in medical imaging. In procedures for intervention or radiation treatment planning, the generated models should be accurate and reflect the natural appearance. Traditional methods for this task, such as Marching Cubes, use smoothing post processing to reduce staircase artifacts from mesh generation and exhibit the natural look. However, smoothing algorithms often reduce the quality and degrade the accuracy. Other methods, such as MPU implicits, based on adaptive implicit functions, inherently produce smooth 3D models. However, the integration in the implicit functions of both smoothness and accuracy of the shape approximation may impact the precision of the reconstruction. Having these limitations in mind, we propose a hybrid method for 3D reconstruction of MR images. This method is based on a parallel Marching Cubes algorithm called Flying Edges (FE) and Multi-level Partition of Unity (MPU) implicits. We aim to combine the robustness of the Marching Cubes algorithm with the smooth implicit curve tracking enabled by the use of implicit models in order to provide higher geometry precision. Towards this end, the regions that closely fit to the segmentation data, and thus regions that are not impacted by reconstruction issues, are first extracted from both methods. These regions are then merged and used to reconstruct the final model. Experimental studies were performed on a number of MRI datasets, providing images and error statistics generated from our results. The results obtained show that our method reduces the geometric errors of the reconstructed surfaces when compared to the MPU and FE approaches, producing a more accurate 3D reconstruction.

Editorial for the Special Issue on “Geometry Reconstruction from Images”

“…In [8], the authors propose improving the geometric precision of 3D reconstruction from medical images thanks to a combination of the Marching Cubes algorithm and smooth implicit curve tracking. The proposed approach produces a space partition (without holes) that ensures 3D consistency.…”

UW Deep SLAM-CNN Assisted Underwater SLAM

“…The suggested approach can recognise optimal points for VO and inhibit unstable points, such as those brought on by moving objects. In order to achieve feature extraction in a conventional backend based on ORB-SLAM [28] architecture, LIFT-SLAM [3] proposes a unique hybrid VSLAM technique based on the LIFT [22] network. Authors asserted that, compared to a conventional VSLAM algorithm, LIFT-SLAM is more reliable without sacrificing precision.…”

“…Machine learning has showcased remarkable achievements in various computer vision tasks, encompassing 3D reconstruction [3], image recognition [4], semantic comprehension [5], and image matching [6]. Moreover, numerous approaches leveraging deep learning have shown promising outcomes in tackling SLAM and VO challenges.…”

An Overview on Visual SLAM: From Tradition to Semantic

“…Modern semantic VSLAM systems cannot do without the help of deep learning, and feature attributes and association relations obtained through learning can be used in different tasks [138]. As an important branch of machine learning, deep learning has achieved remarkable results in image recognition [139], semantic understanding [140], image matching [141], 3D reconstruction [142], and other tasks. The application of deep learning in computer vision can greatly ease the problems encountered by traditional methods [143].…”