Clustering algorithms. Three clustering algorithms for segmentation are added: expectation-maximization on Gaussian mixture, DBSCAN, and k-means. Users can use these methods in the "Component" window.
Improved tracking algorithms. Tracking accuracy has been improved. We improved the algorithm for generating the track map and incorporated clustering algorithms to automatically segment during tracking. Users can also adjust a series of parameters in the "Tracking" window to fine tune the tracking.
A new 4D script for tracking sparse particles. The 4D script allows users to track selected features in a time sequence. No initial segmentation is required for it to work. It can be used to track sparse and small features conveniently.
A density setting for component generation. The density setting has been added in the basic component generation. Its concept is based on the clustering algorithm DBSCAN. However, its implementation is based on the synthetic brainbows algorithm, which uses GPU to compute segmentation of dense data sets.
Render view output enlargement. The option has been added when the render view is captured and saved as an image. Users can set an output image size larger than the render view size.
4D script list. The list of built-in 4D scripts has been added to the "Record/Export" panel. Users can easily select and switch 4D scripts without browsing to the actual files.
Alex Lex, Bei Wang Phillips, and Chris Johnson present at the Future in Review Conference
Alex Lex, Bei Wang Phillips, and Chris Johnson were chosen to present at the Future in Review Conference held at the Stein Eriksen Lodge in Park City, Utah.
What is Future in Review? Hosted by Mark Anderson, founder and publisher of the Strategic News Service™, the Future in Review® (FiRe) conference exposes world experts and participants to new ideas in a manner that produces an accurate portrait of the future in technology, including the global economy, cloud computing, biology and medical diagnostics, policy, netbooks, space travel, sustainability, and other fields that contribute to technology outcomes. (Strategic News Service and the SNS weekly global report have been accurately predicting the technology industry's future since 1995.)
This year's theme was titled "The Power of Flows." Photo courtesy of Kris Krüg
Visualizing the Universe
Utah engineers co-developing space simulation software for planetariums and home computers.
Sept. 7, 2016 – If space is the final frontier, OpenSpace could become the final frontier in space simulation software.
Computer scientists from the University of Utah will be working with researchers from New York University's Tandon School of Engineering and the American Museum of Natural History (AMNH) to develop OpenSpace, an open-source 3-D software for visualizing NASA astrophysics, heliophysics, planetary science and Earth science missions for planetariums and other immersive environments. The software also will be developed for use in schools and on home computers.
In this release, we mainly improved the usability of a series of functions for 4D analysis, including the component analyzer, 4D scripts, paint brush tools, and format supports. We have also made a series of video tutorials and published them on YouTube.
Bei Wang Joins the Scientific Computing and Imaging Institute as an Assistant Professor of Computer Science
Bei Wang has joined the University of Utah's Scientific Computing and Imaging (SCI) Institute as an Assistant Professor of Computer Science. The SCI Institute focuses on solving important problems in biomedicine, science, and engineering using computation and is an international research leader in the areas of scientific computing, visualization, and image analysis.
Dr. Wang received her Ph.D. in Computer Science from Duke University in 2010. There, she also earned a certificate in Computational Biology and Bioinformatics. She was a postdoctoral fellow from 2010 to 2011, and a research scientist from 2011 to 2016, both at the SCI Institute, University of Utah.
From Amazon: You have a mound of data sitting in front of you and a suite of computation tools at your disposal. And yet, you're stumped as to how to turn that data into insight. Which part of that data actually matters, and where is this insight hidden?
If you're a data scientist who struggles to navigate the murky space between data and insight, this book will help you think about and reshape data for visual data exploration. It's ideal for relatively new data scientists, who may be computer-knowledgeable and data-knowledgeable, but do not yet know how to create effective, explorable representations of data.
NeuroImage Journal Cover Features SCIRun Renderings
The publication and cover image is the result of a close collaboration between the University of Freiburg (Lukas Fiederer, Tonio Ball and others) in Germany and the NIH-funded Center for Integrated Biomedical Computing (CIBC, Moritz Dannhauer) at the SCI institute (Johannes Vorwerk). The cover of NeuroImage' March (128) issue illustrates different tissues in a model of the human head that are known to have distinct electrical properties. In this study, we simulated the electrical effect of blood vessels on current flow originating from active brain regions as monitored by scalp electrodes (encephalography, EEG). Since the understanding of EEG measurements matters in many clinical applications (e.g., epilepsy), SCIRun software offers a set of tools to investigate their underlying electrical generators. Recently, in the new version 5.0 of SCIRun additional capabilities (BrainStimulator) have been added to simulate the current flow from external stimulation devices targeting brain regions of potential EEG generators.