Biomedical Engineering Reference
In-Depth Information
2D and 3D visuals. This is referred to as the active search mode and proved
invaluable for swift data analysis and helped with the identifi cation of two new
genetic markers for schizophrenia. Another search mode is the already men-
tioned passive mode, for which users can scan through the whole chromosome
in all brain areas by manually dragging a search window. The search window
size is variable and can be fi t to a specifi ed number of genes or SNP locations.
Gene information associated with the browsing window, such as the gene
symbol and
p
-value, is displayed as the user pans through the chromosome
and is further augmented with real-time statistics, such as mean value and
standard deviation. Furthermore, data query is also linked to an integrated
Web interface, allowing searches to be applied against other published materi-
als in standard data banks such as GeneCards or publications such as PubMed.
Thereby, the Web interface provides access to a rich collection of knowledge
as an integrative part of the visual analytics space.
27.4.1.2 Analysis of 2D and 3D Image Data
Imaging results from CT,
MRI, position emission tomography (PET), and traditional X rays can either
be viewed in a slice-by-slice 2D mode or be rendered as volumetric 3D images
with proper registration and visualization algorithms. In the context of the
collaborative visual analytics space, all 2D data are directly shown as texture
maps and 3D data as volumes with optional 2D cross sections freely movable
in sagittal, coronal, and axial directions. Both vertex and fragment shader
programs are applied to change display properties at run time. For example,
different color look-up tables can be applied to fl exibly highlight different core
features, while the user-defi nable threshold can be applied to remove unre-
lated brain areas. Study on group subjects can be performed by arranging
images based on the associated data values. For example, while studying a
group of schizophrenia subjects, the underlying statistical tests associated with
specifi c genes can be used to sort their functional images. This facilitates a
quick detection of the pattern that correlates genes and brain activation areas.
Statistical beta maps can be used to perform quantitative comparisons. Figures
27.8 and 27.9 show the digital equivalent of a light-board with the added
advantage that all data can be interactively and intuitively manipulated.
27.4.1.3 Web Interface
Internet resources, such as genecards.org and
PubMed, contain a vast number of categorized research results. Another
example is the UCSC gene browser [48], a Web-based, interactive database
for genomics. These types of public domain resources can greatly enhance the
expressiveness and effectiveness of visualization tools, and a Web portal engine
was implemented as an extension to the WebVR platform [49]. WebVR pro-
vides a Web browser interface which seamlessly translates queried information
into texture space for streamlined use in visualization environments. More
specifi cally, Web queries are sent to WebVR, which retrieves the correspond-
ing page, converts it into a tagged image, and subsequently returns a texture
map to be used within the visualization environment. This active texture, also
Search WWH ::
Custom Search