Leonardo M. Rocha

Clustering by optimum path forest and its application to automatic GM/WM classification in MR-T1 images of the brain

A new approach to identify clusters as trees of an optimum- path forest has been presented. We are extending the method for large datasets with application to automatic GM/WM classification in MR-T1 images of the brain. The method is computed for a few randomly selected voxels, such that GM and WM define two optimum-path trees. The remaining voxels are classified incrementally, by identifying which tree would contain each voxel if it were part of the forest. Our method produces accurate results on phantom and real images, similarly to those obtained by the state-of-the-art, does not rely on templates, and takes less than 1.5 minute on modern PCs.

Tree-Pruning: A New Algorithm and Its Comparative Analysis with the Watershed Transform for Automatic Image Segmentation

Image segmentation using tree pruning (TP) and watershed (WS) has been presented in the framework of the image forest transform (IFT) - a method to reduce image processing problems related to connectivity into an optimum-path forest problem in a graph. Given that both algorithms use the IFT with similar parameters, they usually produce similar segmentation results. However, they rely on different properties of the IFT which make TP more robust than WS for automatic segmentation tasks. We propose and demonstrate an important improvement in the TP algorithm, clarify the differences between TP and WS, and provide their comparative analysis from the theoretical and practical points of view. The experiments involve automatic segmentation of license plates in a database with 990 images

Comparative analysis of shell rendering and shear-warp rendering

In Medical Imaging, shell rendering and shear-warp rendering are two of the most efficient and effective voxel-based techniques for volume visualization. This work presents a comparative analysis of shell rendering and shear-warp rendering in terms of storage, speed, and image quality. We have chosen 10 different objects of various sizes, shapes and topologies and one 1-GHz Pentium-III PC with 512 MB RAM for our experiments. Hard and fuzzy boundaries of up to 2,833 K voxels in size have been created to test both methods in surface and volume rendering, respectively. Hard surface shell rendering and surface shear-warp rendering required less than 0.5 second. In the worst case, volume shell rendering required 1.45 second, while volume shear-warp rendering spent 0.65 second for the same task. Shear-warp rendering uses on average from 3 to 6 times more memory space than shell rendering, but it can be up to 2.79 times faster than shell rendering. On average, shear-warp rendering is as fast as shell rendering for hard boundaries and 1.7 times faster than shell rendering for fuzzy boundaries. We have also observed that both can produce similar high-quality images.

Combined approach of shell and shear-warp rendering for efficient volume visualization

In Medical Imaging, shell rendering (SR) and shear-warp rendering (SWR) are two ultra-fast and effective methods for volume visualization. We have previously shown that, typically, SWR can be on the average 1.38 times faster than SR, but it requires from 2 to 8 times more memory space than SR. In this paper, we propose an extension of the compact shell data structure utilized in SR to allow shear-warp factorization of the viewing matrix in order to obtain speed up gains for SR, without paying the high storage price of SWR. The new approach is called shear-warp shell rendering (SWSR). The paper describes the methods, points out their major differences in the computational aspects, and presents a comparative analysis of them in terms of speed, storage, and image quality. The experiments involve hard and fuzzy boundaries of 10 different objects of various sizes, shapes, and topologies, rendered on a 1GHz Pentium-III PC with 512MB RAM, utilizing surface and volume rendering strategies. The results indicate that SWSR offers the best speed and storage characteristics compromise among these methods. We also show that SWSR improves the rendition quality over SR, and provides renditions similar to those produced by SWR.

Automatic method for assessment of telangiectasia degreeing by mathematical morphology

This paper presents an automatic method to assess telangiectasia degreeing, by estimating the reduction of the varicose area from pre to post-treatment using digital image analysis techniques. First both images are geometrically transformed and color space reduced, in order to make them suitable for processing. Next, the varicose veins are segmented via watershed transform and the varicose areas in both images are detected. Finally, the assessment degreeing is done. The method was applied to twenty patients and the results were very consistent, according to the analysis of six vascular specialists.

E-Shell Rendering

In medical imaging, shell rendering is considered one of the most efficient and effective methods for volume visualization. It requires a compact data structure of voxels and uses the voxel splatting technique for surface and volume rendering. In order to avoid holes in the rendition, the size of the voxels are usually made bigger than the size of the pixels for voxel splatting. In such a case, we have identified a conceptual problem in shell rendering which affects the correctness and quality of the renditions. We discuss this problem and propose a solution, which improves image quality without affecting the speed of the method. The new approach is called E-Shell Rendering (extended-shell rendering). It requires an extension of the shell data structure and a variant of the original algorithm. We illustrate and discuss the results with examples created by surface and volume rendering.