Spring 2009

VISUALIZATION FRIDAY FORUM

Fridays 12-1pm LSRC D106

Lunch Served


January 16 - Above the Fold: Info Graphics at the New York Times

Casey Alt

Art, Art History, and Visual Studies

I will discuss my experience as an intern and freelance information graphics editor at The New York Times. This 40-minute presentation will cover: the internship selection process, the work environment and culture, the graphics department's mission, NYT style and design standards, the graphics reporting workflow (from assignment to publication), and several project examples from my work.

see a video of the talk - the talk starts about 10 minutes into the video


January 23 - DiVE into Alcohol: Immersive Virtual Reality Science Education

David McMullen and Marcel Yang

Psychology and Neuroscience (David) and Pharmacology (Marcel)

This talk will be a discussion on “DiVE into Alcohol,” an education module that uses immersive virtual reality technology to enhance students' understanding of basic biochemical principles such as oxidation. The discussion will cover design and production of the education module, means of dissemination, and evaluation of student learning.

see a video of the talk


January 30 - Looking at Looking at Looking

Golan Levin

School of Design and School of Computer Science, Carnegie Mellon University

I am interested in the medium of response, and in the conditions that enable people to experience "flow", or sustained creative feedback with reactive systems. In this regard I have found inspiration in the engaging interactive artworks of Myron Krueger and Toshio Iwai, and in the research of cognitive psychologist Mihalyi Csikszentmihalyi. I am drawn to the revelatory potential of information visualization – whether brought to bear on a single participant, the world of data we inhabit, or the formal aspects of mediated communication itself. Here I have drawn from many teachers in the disciplines of conceptual art and information design. And I am fascinated by how abstraction can connect us to a reality beyond language, and the ways in which our gestures and traces, thus abstracted, can reveal the unique signatures of our spirits. My recent projects have explored the gestures of the hand and voice; in my new work, I have turned to the gestures of the eye, with the aim of creating engrossing, uncanny and provocative interactions structured by gaze. This presentation will discuss a wide range of my own works and those of others, with a particular attention to how the use of gestural interfaces, visual abstraction, and information visualization can support new modes of interaction and play.
 

 See two images of Golan's work: image1 image2  

See a video of the talk (talk starts 8 minutes into the video)


February 6 - An Introduction to Quartz Composer: Image Processing made easy?

Steve Feller

ECE

This talk will give an introduction to the Quartz Composer application development suite that is part of the Mac OS-X Leopard installation. QC is a graphical programming environment for real-time image and video processing in the Mac environment. The software makes complicated image manipulation techniques available for novice programmers and abstracts out many of the more tedious details of conventional programming.

See a video of the talk (talk starts about 5 minutes into the video)


February 13 -  Visualizing Statistics on Uncertain Data

Jeff Phillips

Computer Science

The data points for many geospatial data sets have positional uncertainty.  That is, the true location of the data points is not known precisely, but rather is given by a probability distribution.  For instance, the location of atoms in a protein structure, height values in a terrain, or obstacle locations sensed from a robot's laser range finder, may be modeled with a probability distribution based on the error model from the sensing technology.  In this talk I will describe how to visualize answers to shape fitting questions on such uncertain data.  For instance, how do you represent the smallest enclosing ball of an uncertain point set?  Discussion will follow on possible alternatives to various choices made in the visualization as well as how to visualize more complex questions on uncertain data.

Joint work with Maarten Loffler.

 
see a video of the talk (skip the first 8 minutes)
Thursday, February 19th, 9:30am-3:00pm, Room 1411 CIEMAS/Fitzpatrick
 
Autostereoscopic Display Demo
Tridelity Display (http://www.tridelity.de/) makes a nice auto-stereoscopic monitor.  Both a desktop version and larger versions for classrooms.  It can work with any software that supports stereoscopic displays.  They will be on campus to demonstrate their monitor on Thursday (details above).  For those of you who use Amira, you should come see this monitor if you have 10 minutes.  The driver for the monitor is compiled into the Amira software, so it's really easy to use with Amira.

 


February 20 - Changes in the size and shape of the human head contour from infant to adult

Andre Loyd

Biomedical Engineering

This talk will present a novel technique on capturing and analyzing head contour
data.  The talk will present average head contour and show how the head shape
and size changes with age.
 

February 27 - RNA Secondary Structure Prediction with few Homolog

Olivier Perriquet, Bioinformatics

New University of Lisbon, Center for Artificial Intelligence

Non coding RNAs (ncRNA), unlike messenger RNAs (mRNA), are directly functional in the cell where they are believed to play a role usually assigned to proteins. Since two decades, a series of new ncRNA families are being discovered, usually coming as a set of putative, sometimes poorly conserved, homologs. The function of these ncRNA in the cell, like for proteins, is related to their three-dimensional shape and an ordinary problem is the computational prediction of this shape (ie. the exact atomic positions) given
just the nucleic acid sequence. This hard problem is addressed stepwise by using intermediate levels of description of the molecule.
Secondary structure, which is more or less the graph of the intra-molecular base-pairing interactions, provides such an intermediate, topological description of the molecule that happens to be well suited for an algorithmic approach: given a sequence, the set of possible secondary structures that minimize a free energy function (within a realistic, yet simplified, thermodynamical model) can be computed in reasonable time by dynamic programming. But as there is usually no clue to guess which structure in that set is the correct one, it makes sense to use several homologs, when available, and improve the accuracy of the results by seeking the common structure.
The extension of the previous framework to several sequences turns out to be computationally expensive, although still polynomial in space and time, and calls for heuristic adaptations. I will present our algorithmic results, that prove to apply successfully to that arduous context, usually unaffordable for other methods.

 see a video of the talk (talk starts about 5minutes into the video)


March 6 - Teaching Neuroanatomy in the DiVE with a Virtual-Reality Brain

Scott Huettel

Brain, Imaging, and Analysis Center

The first goal of our FOCUS Neuroscience course is for first-year undergraduate students to learn basic principles of neuroanatomy, such as the gross organization of the brain and how its different parts contribute to mental function. To aid in this instruction, we created a virtual-reality model of the human brain for the DiVE. Summarized briefly, we took a high-resolution and high-contrast T1 MR volume of the human brain and segmented that volume into key anatomical parts using an automated program (FreeSurfer). A student collaborator then adapted the segmentation output into a format suitable for teaching by cleaning boundaries between adjacent regions and then color-coding sets of structures. Students interact with this brain by interactively adjusting the size, clipping plane, point of view, and transparency of anatomical structures, to gain to gain a deeper understanding – in both senses – of the 3D structure of the brain. Our students can look inside the three-dimensional brain, in a manner impossible with normal physical brain models, when learning its anatomy. The VR brain has now been extended to other undergraduate courses and is now a regular part of the neurobiology course for first-year Duke medical students.

 see a video of the talk - skip the first 6 minutes.


March 13 - SPRING BREAK 


March 20 - Seeing Through Time: Historic Maps, Google Earth, and the  Transformation of Durham

Trudi Abel

History

I will discuss the development of the Digital Durham website (http://digitaldurham.duke.edu) a web repository of maps, photographs, printed
works, manuscript letters and census data from post-Civil War Durham There is tremendous power to these documents as they help us teach core themes such as industrialization, urbanization, emancipation and segregation. Right now, I am interested in using new technology to generate new meanings from historic fire insurance maps. Later this spring, I will launch a new collection of maps on Digital Durham which will feature two georectified fire insurance maps from the the 1880s that were created with funding from a CIT Visualization Grant. I think these maps will help users visualize an industrializing city and gain a better understanding of the process of urbanization in this New South community. This talk will focus on some of the challenges that I and my colleagues have encountered in creating the web site and its georectified maps. I would like to explore with the audience how a historian like myself might draw on the tools of computer scientists to make an interactive map of late 19th century Durham that draws data, people's names, occupations, home addresses, and their race-- from OCR scans of city directories.

 see a video of the talk


March 27 - Segmentation and Image Analysis for Multimodality Cancer Treatment and Detection

Paolo Maccarini

Cancer Biology

In the fight against solid cancers, the use of thermal therapy to enhance drug delivery and radiation effects has shown significant results and started being acepted by more institutions around the world. Quality assurance of the hyperthermia treatment is fundamental to achieve the best synergy with radiation an drug delivery. Recent advances in simulation tools and hardware contributed significantly towards improved quality of treatments and thus enhanced long term survival and local control. The ability to segment real patient data into the simulation tools in a timely matter is critical for a widespread use of the technology. Duke is world leader in the ability to monitor in real time the 3D temperature of patients treated in the MR. We are currently developing software that allow us to calculate in real time 3D dosimetry as well monitor drug release. Both techniques will significantly improve treatment quality as well as push the technology towards widespread clinical use. Amira has been successfully used to accurately design the treatment planning as well as analyze the staining of drug release.  In the early detection the use of segmented patient data using Amira also has significantly contributed towards optimization of two new novel techniques to detect breast cancer safely and in very early stages. The system still under development but are showing promising results.  We will show some examples of our treatment planning, image analysis and segmentation for applicator/detector development, critical steps towards a better treatment of solid tumors.

 see a video of the talk


April 3 - Intensity Modulated Radiation Therapy in Radiation Oncology

Vorakarn Chanyavanich

Medical Physics

Nearly two-thirds of all cancer patients will receive radiation therapy during their illness, with over one million patients treated with radiation therapy annually.  Three cancers - breast cancer, lung cancer and prostate cancer - make up more than half of all patients receiving radiation therapy.  The challenge in treatment planning for radiation therapy is to maximize the delivery of radiation dose to the tumor, while minimizing dose to adjacent normal tissue and structures.  Visualization of the normal anatomy and the disease sites is an important process in the effective planning of a radiation therapy treatment.
 
see a video of the talk
 

April 10 - Vaccinology, the Movie

Tom Kepler

Biostatistics and Bioinformatics

Vaccines induce protective immunity through the complex orchestration of cellular changes and spatial reorganization.  Investigators in the Multiscale Systems Immunology (MSI) project are using computational modeling and experimental studies to reconstruct the processes that make up the immune response to immunization.  I will summarize our work thus far, with an emphasis on visual representations, including 3D simulations, 3D video vital imaging, and statistical image analysis.

 see a video of the talk (talk starts 10 minutes into the video)


April 17 - Visualizing the RNA Sequenome

Erik Schultes

Computer Science

Genomic sequences code for RNAs and proteins have precise structural properties mediating specific biochemical functions. Since the early days of molecular biology, it has been assumed that sequences chosen at random from sequence space (and therefore without having the benefit of natural selection) would be disordered and without biochemical function. Yet the observed sequence diversity in biology and experiments whereby new functional sequences are isolated from synthetic combinatorial libraries suggest that the opposite - sequence space must be densely packed with well-structured and potentially useful sequences. What is the nature of this multi-dimensional encoding of structure and function? I will introduce analytical methods by which the high-dimensional RNA sequence space can be productively visualized and meaningfully sampled using tractable numbers of randomly-generated sequences. The result is a 'map' of RNA sequence space (i.e., the RNA sequenome) that opens the door to new experiments and theories in molecular evolution.

 


April 24 - The advantages of investigating emotional learning and memory processes in a fully-immersive virtual reality environment

Nicole Huff

Cognitive Neuroscience

A unique fully-immersive virtual reality (VR) environment at Duke University, DiVE  (Duke’s Immersive Virtual Environment) was used as a novel investigational tool to explore how VR influences learning and memory of a fearful experience. Three emotional learning phenomena were examined in healthy human participants using paradigms derived from animal models of classical Pavlovian conditioning which are termed: fear extinction, spontaneous recovery, and fear renewal. In our paradigm dynamic images of snakes and spiders were encountered in the DiVE, and at specific times were followed by a negative event, while participants took a guided virtual walk through 3-D VR environments. Skin conductance was used as the dependent measure of fear to the snakes and spiders encountered in the VR worlds. We found that memory for a fearful experience is stronger in the VR environments than in our previous laboratory studies.  Discussion of the findings and the advantages of conducting affective cognitive neuroscience research in this unique VR setting will follow.
 
see a video of the talk

 

For more information, please contact Rachael Brady

Organized by The Visualization Technology Group (VTG),

Sponsored by Information Science + Information Studies (ISIS)

and The Department of Computer Science