Volume Visualisation

Part 1

1
Farhan bin Mohamed
Mohd Shafry Mohd Rahim
 • Content of Volume
Visualisation, Part 1:

Overview: § Introduction
• About Volume Data
Volume #1 • Overview of Techniques
§ The VolVis Conceptual
Framework

2
Farhan bin Mohamed
Mohd Shafry Mohd Rahim
 Volume Visualisation
• Introduction:
§ VolVis = Visualisation of Volume data
• Picture/image 3D → 2D/3D/4D
• Projection, slicing, surface extraction, volume rendering,
etc..
§ Volume Data =
• 3D x 1D Data
• Scalar data, 3D space, space filling (dense – as opposed to
sparce)
§ User goals =
• To gain insight into 3D Data
• Depends strongly on what user is interested on (focus +
context)

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Volume Data
• Where does the data come from?
§ Medical Application
• Computer Tomography (CT)
• Magnetic Resonance Imaging (MRI)
§ Material testing/control
• Industry-CT
§ Simulation
• Finite Element Method (FEM)
• Computational fluid dynamics (CFD)
§ And other sources
Farhan bin Mohamed
Mohd Shafry Mohd Rahim
 Data Source
• E.g. Different modalities for different electro
magnetic spectrum
Data Acquisition example - Telescope

Hubble Space Telescope view of the "Pillars of

Creation," towering formations of cosmic gas and dust
that lie in the center of the Eagle Nebula.
(Image: © NASA/ESA/Hubble Heritage Team
(STScl/AURA))
Data Acquisition example – CT Scan

Siemens – Somatom Force

 Data Explosion
• Previously :
– 100 slices, 5122 pixels (50MB)
• Then:
– 24000 slices , 5122 pixels (20GB)
• Now:
– > 48000 slices, 10242 pixels (>80GB)
 3D Data Space
• How is the volume data organised?
§ Cartesian, i.e., regular grids:
• CT/MRI: often dx = dy < dz. E.g., 135 slices (z) by 5122 values
in x and y
• Data Enhancement: iso-stack computation = interpolation of
additional slices, so that dx=dy=dz → 5123 Voxel
• Data: Cells (Faces), Corners: Voxel
§ Curvilinear grid i.e., unstructured:
• Data organised as Tetrahedra or Hexahedra
• Often: Conversion to Tetrahedra
Terminology

Tetrahedron. “A 3D primary cell that is a simplex with four triangular faces, six
edges, and four vertices.
Farhan bin Mohamed
Mohd Shafry Mohd Rahim
 • Challenges:
§ Rendering projection,
so much information and
so few pixels
Volume § Large datasets, e.g.,

Visualisation 512 x 512 x 1024 voxels at

16 bit/voxel = 512
Megabytes
(VolVis) – § Computational speed,

Challenges Interaction is very

important, > 10 frames per
second (fps)

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Voxels vs. Cells
• Two ways to view the volume
data:
§ Data as a set of voxels
• Voxel = short for volume element
(recall: pixel = “picture element”)
• Voxel = point sample in 3D
• Not always interpolated
§ Data as a set of cells
• Cell = Cubic primitive (3D)
• Corners: 8 voxel
• Value(s) in cell:
Are always interpolated

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Conceptual VolVis Framework
Sampled Data Analytical Data
(Measurement) (Modelling)

Iso-surfacing
Voxel space Geometric Surface
(discrete) voxelisation (analytic)

(discrete) volume Surface rendering

rendering Pixel space
(discrete)

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Conceptual VolVis Framework
• Example 1:
§ CT measurement
§ iso-stack computation
§ iso-surface computation
(marching cubes)
§ Surface rendering
(openGL/Direct3D)

Whole body contrast-enhanced CT image data set of a rhesus macaque.

The CT images were acquired following intravenous injection of contrast agent.
The left and right images are the same image data set rotated approximately 90°.
The major vessels of the cardiovascular system, the heart, liver, spleen, kidney,
and urinary bladder can be seen in these volume-rendered images
(X-Ray Computed Tomography, Science Direct(2006)).
Farhan bin Mohamed
Mohd Shafry Mohd Rahim
 Conceptual VolVis Framework
• Example 2:
§ MR Measurement
§ iso-stack computation
§ MIP (maximum intensity
projection)
§ Output image: Blood
vessels in hand

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Conceptual VolVis Framework
• Example 3:
§ Potential function
ρ (x,y,z)
§ Iso-surface
ρ(x,y,z) = ρ0
§ Surface ray tracing

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Chest CT Scan
• THE PICTURE you going to see is an X-ray computed
tomography (CT) scan of a human lung. Go ahead
and take a few seconds to look at it carefully.

• The study was published in press at Psychological

Science. Trafton Drew and colleagues at the Brigham
and Women’s Hospital showed that when people
focus on searching these images for bright white
cancer nodules…

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Chest CT Scan

https://www.nationalgeographic.com/science/phenomena/2013/01/31/when-experts-go-blind/
Volvis
Techniques -
Overview

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 VolVis Techniques - Overview
Simple Methods:
• Indirect volume surface visualisation:
§ Slicing
§ MPR (multi-planar reconstruction)
§ Surface fitting methods:
• E.g., marching cubes (marching tetrahedra)
• Direct volume visualisation:
§ Ray casting
§ Shear-warp factorisation
§ Splatting
§ 3D texture mapping
§ Fourier volume rendering
Farhan bin Mohamed
Mohd Shafry Mohd Rahim
Marching Cubes Cases
Image-order vs. Object-order

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Image-order approach
• Image-order Approach: Traverse the image
pixel-by-pixel and sample the volume

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Object-order approach
• Object-order approach: Traverse the volume,
and project to the image plane.

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Alias and Anti-aliasing
• Terminology

Aliasing: “A rendering technique that assigns to pixels the

colour of the primitive being rendered, regardless of
whether that primitive covers all or only a portion of the
pixel’s area. This results in jagged edges, or jaggies” [The
OpenGL Programming Guide, 5th Edition (The “Red Book”)]

Anti-aliasing: “A rendering technique that assigns pixel

colours based on the fraction of the pixel’s area that’s
covered by the primitive being rendered. Anti-aliased
rendering reduces or eliminates the jaggies that result
from aliased rendering.” [The OpenGL Programming Guide, 5th
Edition (The “Red Book”)]

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Interpolation
Terminology:
Interpolate: “Estimate a value of a function at a
point, p, given known function values and points
that bracket p.”

f(1)● ←d1→ (p)• ←d2→ ●f(2)

• Nearest neighbour: p = f(0)
• Linear: p = f(1)(D-d1)/D + f(2)d1/D

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Interpolation
• Example:
f(1)● ←d→ (p)• ←d→ ●f(2)
T(1)● ←d→ T(x)• ←d→ ●T(2)
100C● ←2→ T(x)• ←(10-2)→ ●200C

• Nearest neighbour T(x)

Linear:
T(x) = 100C(10-2)/10 + 200C(2/10)
120C = 100C(0.8) + 200C(0.2)

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Interpolation
Nearest Neighbour Trilinear

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Interpolation – Influence
Nearest Neighbour Trilinear

Farhan bin Mohamed

Mohd Shafry Mohd Rahim
 Acknowledgements
• We thank the following people for the composition/materials
in this lecture are based on lectures/materials/text of:
§ Robert S. Laramee
§ Michael Meißner
§ Roger Crawfis
§ Hanspeter Pfister
§ Meister E. Gröller
§ Torsten Möller
§ Dirk Bartz
§ Markus Hadwiger
§ Helwig Hauser
§ Anders Ynermann

Farhan bin Mohamed

Mohd Shafry Mohd Rahim