Open Access. Powered by Scholars. Published by Universities.®
Articles 1 - 3 of 3
Full-Text Articles in Computer Sciences
Dynamic Display And Quantitative Analysis Of Three-Dimensional Left Ventricular Pathology, William A. Barrett, Jayaram K. Udupa
Dynamic Display And Quantitative Analysis Of Three-Dimensional Left Ventricular Pathology, William A. Barrett, Jayaram K. Udupa
Faculty Publications
Techniques have been developed for automated extraction and dynamic interactive display of three-dimensional (3D) left ventricular (LV) surface anatomy from Cine CT images using a PC-based image display architecture. Images of both endocardial and myocardial surface anatomy are generated from multiple views at multiple time instances to demonstrate various LV pathologies including apical akinesis, apical and posterior aneurysms, LV Failure, IHSS, and a left atrial myxoma. Surface generation requires interpolation between scans, surface tracking, and rendering. Generation of 60 views corresponding to a single time instance requires approximately 15 minutes. LV dimensions are measured between two or more surface points …
A Parallel-Processing Subsystem For Rapid 3-D Interpolation Of Ct Images, William A. Barrett, Stephen J. Allan, Scott R. Cannon
A Parallel-Processing Subsystem For Rapid 3-D Interpolation Of Ct Images, William A. Barrett, Stephen J. Allan, Scott R. Cannon
Faculty Publications
An inexpensive parallel-processing subsystem for the rapid interpolation of CT image planes is demonstrated with a variety of node topologies. The subsystem is based on a tree network of INMOS T414 Transputer processors and is hosted by an AT-based image workstation. The subsystem accepts a stack of eight arbitrarily-spaced 256 x 256 image planes from the host. Subsystem output to the host consists of a stack of 32 scaled and evenly-spaced image planes (256 x 256 x 32 with cubic voxels). Benchmark execution times ranged from 12.3 seconds for three nodes to 5.8 seconds for eight nodes.
Digital Neural Networks, Tony R. Martinez
Digital Neural Networks, Tony R. Martinez
Faculty Publications
Demands for applications requiring massive parallelism in symbolic environments have given rebirth to research in models labeled as neural networks. These models are made up of many simple nodes which are highly interconnected such that computation takes place as data flows amongst the nodes of the network. To present, most models have proposed nodes based on simple analog functions, where inputs are multiplied by weights and summed, the total then optionally being transformed by an arbitrary function at the node. Learning in these systems is accomplished by adjusting the weights on the input lines. This paper discusses the use of …