2.1 Anatomy and Physiology of the Heart

sploshtribeSoftware and s/w Development

Dec 14, 2013 (3 years and 10 months ago)

66 views

Quantitative Analysis and

Visualisation of Heart

Tissue Structure



Degree:

Computer Science

Module:

Darwin

Stage:

Research Proposal


Date:

18/12/09


Supervisor:

Richard Clayton

Team:

Darwin

Three



The authors submit this report in partial fulfilment

of the requirement for the degree of
Masters of Science in Computer Science.

Signed Declaration


All sentences or passages quoted in this report from other people’s work have been
specifically acknowledged by clear cross
-
referencing to author, work and p
age(s). Any
illustrations, which are not the work of the author of this report, have been used with the
explicit permission of the originator and are specifically acknowledged. I understand that
failure to do this, amounts to plagiarism and will be conside
red grounds for failure in this
project and the degree examination as a whole.


Name: Ryan Jacob Bloor

Signature:


Na
me: Robert Bryan White

Signature:


Name: Samia
Abdalhamid

Signature:


Name:
Jaeseon Lee

Signature:


Name: Amruta Mane

Signature:


Name:
Georgios Kyprianou

Signature:


Name:
Ramya Kotagiri

Signature:



Date: 18/12/09

Abstract



Computer scientists and cardiologists today, still struggle to produce accurate visualisations
of the whole mammalian heart. Previously, work has been carried out
on modelling and
visualising the left ventricular (LV) region of the heart, [1] but other regions of the heart
remain difficult to model due to their complex structure.

In this project a software program
will be written, entitled ‘HeartVis’, which will attempt to simulate a 3D heart using high
resolution
DT
-
MRI data. This data is composed of both animal and human hearts that have
been imaged at the University of Leeds and

John Hopkins University. Half of the hearts
suffered some form of heart disease unknown to us, while the other
s

are fully healthy ones.

Our main

objective is to attempt to quantity the fibre orientations within the whole heart and
locate regions where the
se fibres are disorganised.
It will be interesting to discover what
other useful
quantitative information regarding the structure of the whole heart can be
extracted fro
m the generated visualisations. Also, i
t will be beneficial to find comparisons
between

diseased and healthy hearts regarding the fibre sheet orientation and alignments.








Table of Contents


Abstract

................................
................................
................................
................................
......

3

List of Figures

................................
................................
................................
............................

5

Glossary and Terminology

................................
................................
................................
.........

6

Main Terminology

................................
................................
................................
.................

6

Other Termin
ology
................................
................................
................................
.................

6

Chapter 1 Introduction

................................
................................
................................
.............

7

1.1 Background

................................
................................
................................
....................

7

1.2 Project Overview

................................
................................
................................
...........

7

1.3 Aims and Objectives

................................
................................
................................
......

8

1.4

Report Structure

................................
................................
................................
.............

8

Chapter 2 Background

................................
................................
................................
.............

9

2.1 Anatomy and Physiology of the Heart

................................
................................
...........

9

2.1.1 Heart Introduction

................................
................................
................................
...

9

2.1.2 Heart Structure

................................
................................
................................
........

9

2.1.3 Heart Function

................................
................................
................................
......

12

2.2
Imaging Techniques

................................
................................
................................
.....

13

2.2.1 CT Imaging

................................
................................
................................
...........

13

2.2.2 MRI Imaging

................................
................................
................................
.........

14

2.2.3 DT
-
MRI Imaging

................................
................................
................................
..

15

2.3 Literature Review (Amruta)

................................
................................
.........................

15

2.4 Existing Software

................................
................................
................................
.........

16

2.4.1 OpenGL and JOGL

................................
................................
...............................

16

2.4.2 VTK

................................
................................
................................
......................

17

2.4.3 Java 3D API

................................
................................
................................
..........

17

2.4.4 Mathworks MatLab

................................
................................
...............................

17

Chap
ter 3 Research Programme
................................
................................
.............................

18

3.1 System Architecture

................................
................................
................................
.....

18

Bibliography

................................
................................
................................
............................

19





List of Figures


Figure 1: A labelled diagram of the heart with colour
-
coded regions for oxygenated and
deoxygenated blood (red and blue respectively) [9]

................................
................................
..

9


Figure 2: This shows a cross
-
section of a part of the cardiac muscle and how

the three
separate layers are oriented within it. [10]

................................
................................
...............

10


Figure 3: This shows an accurate representation of the sheet structure of the heart. DT
-
MRI
was used to find the eigenvectors of sheet position and this was then plotted. [13]

...............

11


Figure 4: Action potential velocities in the conduction tree. [18]

................................
...........

12


Figure 5: Comparison between different body postures after rigid image
-
based mutual
information registration. Top row: MRI supi
ne (left), CT supine (right). Bottom row: MRI
prone (left), MRI right decubitus (right). [24]

................................
................................
.........

14


Figure 6: Average geometry and fibre tracking on the DT
-
MRI atlas [28]

.............................

15



Glossary and Terminology


The following
describes the terminology used throughout the report.

Main Terminology


The Heart
is a muscular organ located between the lungs within the chest and is protected by
the rib cage. It supplies blood to all parts of the body through a double circulatory syste
m.
Blood is pumped away from the heart through arteries and brought back to the heart through
veins.


The Atria
are the upper
-
most chambers of the heart (the left and right atriums respectively)
and act as storage for blood returning to the heart.


The

Ventricles
are the lower
-
most chambers of the heart (the left and right ventricles)
situated directly below the two atriums and pump blood from the heart around the body.


Heart Valves
prevent a backflow of blood in the heart between the chambers of the
heart.


The Heart Wall
is made up of cardiac muscle allowing the heart to contract, thus, producing
heart beats. It is composed of three layers


epicardium, endocardium and myocardium. This
project will be focusing on the myocardium layer of the heart.


The Myocardium
forms the muscular middle layer of the heart containing cardiac muscle
fibres that allow the heart to contract.


Heart Tissue
is made up of minute rod
-
shaped cells, which are arranged in fibres and sheets.


Muscle fibres

contract along the length of the fibres and sheets and during contraction the
sheets actually slide over one another.


Muscle Fibre Orientation
describes the angle at which the muscle fibres are situated in the
heart wall. The orientation of fibres and sh
eets varies throughout the heart, especially across
the myocardial wall. As the heart grows, this arrangement slowly changes to find an optimum
and more efficient arrangement.

Other Terminology


Arrhythmia
is a clinical term used to define an i
rregular he
artbeat.


Superior Vena Cava
sends blood back from upper part of the body to the Right Atrium. It is
also considered as one of the largest veins.


Inferior Vena Cava r
eturns blood back to the Right Atrium from the lower part of the body.


Pulmonary Arte
ries
take the blood from the Right Ventricle to both of the lungs to be
oxygenated.


Pulmonary Veins

carry the oxygenated blood back to the Left Atrium in the heart.


Chapter 1 Introduction


This chapter will provide a brief introduction to the project, discussing background
knowledge regarding the anatomy
and
of the heart and the fibre sheet architecture. It will also
outline the aims and objectives of the project and provide a quick summary
of what the
remaining chapters will be focused upon.

1.
1 Background



The mammalian heart is the most important organ in the human body and is actually the first
to start growing. The heart supplies blood and oxygen to all parts of the body through a
double circulatory system. It is essentially a biological pumping system, w
hereby blood is
pumped away from the heart through arteries and brought back to the heart through veins.
Naturally, a pumping system requires valves in order to function correctly, which prevent a
backflow of blood in the heart. It is perhaps interesting t
o note that during the lifespan of a
typical human being, the heart will beat more than 2.5 billion times, 35 million times a year
and
approximately
100 thousand times a day. [
2
]


Heart tissue is comprised of minute rod
-
shaped cells, which are arranged in
fibres and sheets.
The orientation of the fibres and sheets varies throughout the heart, but it is widely accepted
that this variability is much greate
r across the myocardial wall. [3
,
4
] This orientation is
highly important because fibres contract along t
heir length and during contraction sheets
actually slide over each other. As the heart grows the arrangement of

these

fibres and sheets
is thought

to change throughout the heart;

as an optimum and more efficient arrangement is
found.


The myocardial fibre
sheet structure influences most of the mechanical and electrical
properties within the heart. It affect
s both mechanical contraction [5] and electrical
propagation, [6
] meaning it is absolutely vital that an accurate estimate of the orientation and
its var
iability concerning the fibre sheet structure is obtained. A technique can be used to
measure the orientation of fibres, known as diffusion tensor magnetic resonance imaging
(DT
-
MRI), which measures the motion of protons within tissue.

1.2 Project Overv
iew


A software program will be written entitled ‘HeartVis’, which will attempt to simulate a 3D
heart using high resolution DT
-
MRI data. After which, useful quantitative data will be
extracted regarding the fibre orientations and sheet alignments. The dat
a to be analysed is
composed of both animal and human hearts that have been imaged at the University of Leeds
and John Hopkins University. Half of the hearts suffered some form of heart disease
unknown to us, while the other
s

are fully healthy ones. There
is huge hope that the sacrifice
made in reducing the number of healthy hearts to accommodate the diseased ones, will enable
us to draw significant comparisons between the two. Both heart shape and fibre orientations
are known to change in a diseased heart
, prompting pumping failures and electrical
arrhythmias. DT
-
MRI allows us to measure the motion of individual protons in the heart wall,
meaning comparisons can be made.


Computer scientists and cardiologists today, still struggle to produce accurate visua
lisations
of the whole mammalian heart. Previously, work has been carried out on modelling and
visualising the left ventricu
lar (LV) region of the heart, [1
] but other regions of the heart
remain difficult to model due to their complex structure. There hav
e been numerous attempts
at reconstructing the heart
-
mus
cle fibres, [7, 8] using DT
-
MRI, which
will provide a clear
and concise framework to follow initially, to help avoid major pitfalls. In this project we are
seeking to reconstruct the heart using the o
rientation of heart
-
muscle fibres.

1.
3 Aims and Objectives


The main aim of this project is to develop a software application which, given high resolution
DT
-
MRI data, is able to generate real time 3D visualisations of the heart.

Our secondary
objective
is to attempt to quantity the fibre orientations within the heart and locate regions
where these fibres are disorganised.
It will be interesting to discover what
other useful
quantitative information can be extracted fro
m the generated visualisations. Last
ly, i
t will be
beneficial to find comparisons between diseased and healthy hearts regarding the fibre sheet
orientation and alignments.


1.4 Report Structure


Chapter 2

provides background information on the anatomy of the heart, its structure and
functi
on. It also discusses relevant imaging techniques with application to medical imaging
and examines existing systems, technologies and eventual programming language candidates.


Chapter 3

provides an explanation of the aims and objectives of the project,
detailed above,
in further detail

and states precisely how we intend to achieve them
. The system architecture
will be proposed for the whole system, making important design choices regarding imaging
techniques, programming language and mathematical algorit
hms used. Later on, there will be
a discussion on how to test and evaluate the system once completed, to give a clear idea of
the successes and failures of the project.

Lastly, a

detailed project plan will be outlined

in
addition to
a Gantt chart for monit
oring progress.


Appendix
provides any additional figures, tables and information deemed relevant to the
project as a whole, but is not considered a compulsory requirement in understanding the
report.



Chapter 2 Background


This chapter will provide a brief
overview of background knowledge regarding the anatomy
and physiology of the heart and explore various imaging techniques so that the muscle fibre
orientations

can be obtained. After which, a literature review shall descri
be the previous work
on heart modelling and visualisation. Lastly, various programming languages and toolkits are
contrasted, highlighting the benefits and disadvantages of each.

2.1 Anatomy and Physiology of the Heart

2.1.1 Heart Introduction


The m
ammalian heart is the most important organ in the human body and is actually the first
to start growing. It i
s about the size of a human fist and weights about 250
-
300g.
The heart
supplies blood and oxygen to all parts of the body through a double circulat
ory system. It is
essentially a biological pumping system, whereby blood is pumped away from the heart
through arteries and brought back to the heart through veins.
The heart is situated between the
lungs within the chest, protected by the rib cage and is

surrounded almost entirely by the
pericardium, which acts as a sac storing the heart inside it.

2.1.2 Heart Structure


The heart is divided in two, each of which provides one function of the circulatory system,
mentioned above. These halves are divided

into two chambers with the two atria situated
directly above the ventricles. The left and right atriums act as storage space for blood
returning to the heart from the body, while the left and right ventricles pump blood from the
heart around the body.

Nat
urally, a pumping system requires valves in order to function
correctly, which prevent a backflow of blood in the heart.
Figure 1 shows a labelled diagram
of the heart, whereby red stands for oxygenated and blue for deoxygenated regions.











Figure
1
: A labelled diagram of the heart with colour
-
coded regions for oxygenated and deoxygenated blood
(red and blue respectively)

[9]

2.1.2.1 Heart Layers


It has been explained previously that the heart is a complex struct
ure that regulates the flow
of blood around the body. The efficiency of the pumping action of the heart is based upon the
make
-
up of the cardiac muscle itself, which allows the heart to contract, producing heart
beats. The cardiac muscle is split up into t
hree main layers: epicardium, myocardium and
endocardium. Figure 1 shows how these layers are split up and their ordering.
















The epicardium is the outer
-
most layer of the cardiac muscle and its purpose is to provide
protection to the inner layers as well as producing a fluid, pericardial, that allows the heart
sufficient movement wi
thin the chest. The next layer and the thickest by far is the
myocardium which is the main contracting tissue. Finally, the inner most and thinnest layer
of cardiac muscle is called the endocardium which covers the heart valves and assists in the
smooth fl
ow of blood around the heart. It also regulates the ionic composition of blood which
keeps the blood as clean and pure as possible.

2.1.2.2 Fibre Sheet Structure and Orientation


As the previous section touched upon, the main muscular wall of the heart,
called the
myocardium, is the main pumping mechanism which contracts and relaxes due to
electromagnetic pulses. This in turn pushes blood around the heart and into the body. The
structure of this muscle is a complex mesh of interconnecting sheets approxima
tely four cells
thick [
11
]. A single cell of cardiac muscle is called a myocyte. Figure 2 gives a graphical
representation of the interconnecting sheet structure.


These sheet structures are connected by minute radial fibres and there exist cleavage planes

(areas of space) between some sheets. The amount and the structure of these radial fibres
differentiate between species as well as the area of the heart where the sheet is located. For
example, in the left ventricle, where the efficiency of pumping is par
amount, there is a higher
concentration of sheets and fibres then in other areas of the heart, such as the left atrium,
because its job is to act as a te
mporary area of blood storage [12
]. Also, the wall of the right
Figure
2
: This shows a cross
-
section of a part of the cardiac muscle and how the three
separate layers are oriented within it.

[10]

ventricle is thicker than the wall of t
he atria and the wall of the left ventricle is thicker still
than the right ventricle. This main reason for this is difference in wall thickness is that the left
ventricle generates far more pressure than that of the right ventricle due to its handling of
oxygenated blood


even though equal volumes of blood are pumped from the left and right
side of the heart.

















All of this knowledge has been reached through prior testing via various methods. Due to the
anisotropic nature of the sheet and fibre structures of the cardiac muscle, DT
-
MRI is
an ideal
technique to simulate the organisation of this muscle type. Through experimentation, it has
been discovered that the make
-
up of cardiac muscle also differentiates between individuals of
a species. For example, the arrangement or the amount of shee
t structures in one human's left
ventricle maybe, or will probably be, different to someone else's. Furthermore, it has been
discovered that heart disease can be related to irregular arrangement of myocyte
s in a certain
area of muscle [14
].

Fibre Orientati
on during Contraction


It is important to understand how the fibres in cardiac muscle react to the electromagnetic
contractual stimulus when the heart is attempting to pump blood to the various parts of the
body. Previous sections have shown that the fibre

orientation in the epicardium through to the
endocardium differentiates, which means that they can roll over and slide past
each other
during contraction [15
]. Another effect of contraction results on the fibres is that the stress
impounded on each fibre
causes them to shorten and squeeze together. Consequently, the
combined effects of sliding and squeezing cause the muscle as a whole to become shorter,
hence the pumping analogy.


However, through experimentation it has been proven that there is an additio
nal force, torsion
that acts upon the muscle layers and fibres. The electromagnetic stimulus of contraction puts
Figure
3
: This shows an accurate representation of the sheet
structure of the heart. DT
-
MRI was used to
find the eigenvectors
of sheet position and this was then plotted.

[13]

a lot of stress and pressure on each section of cardiac muscle. The outcome of this is that the
layers and fibres actually twist slightly aroun
d their longitudinal axis. The angle of twist
differentiates in each layer of cardiac muscle and it has been proven that it is higher in the
endocardium than the epicardium [
16
].


In summary, the result of contraction on cardiac muscle is a twisting and co
mpressing motion
caused by the interlinking nature of the fibres in each layer and the intense stress on muscle
that the electromagnetism produces. It is, in more simple terms, analogous to the process of
ringing out water from wet material whereby the mat
erial becomes distorted through pressure
of the person performing the action.

2.1.3 Heart Function

2.1.3.1 Electromechanical Activity within the Heart


The myocardium is the muscular wall of the heart that contracts and expands due to
electromagnetic p
ulses.
In the heart muscle cell, or myocyte, electrical activation occurs from
the transfer of sodium ions across the cell membrane, similar to that of a nerve cell. [1
7
] The
amplitude of the action potential is approximately 100mV and the rate at which th
e heart
conducts these electrical impulses is
described as

cardiac conduction. The action potential is
generated by specialised cells in the Sino
-
Atrial (SA) node commonly referred to as the
‘pacemaker’ of the heart and spread throughout the atria by cell
-
to
-
cell

conduction in any
direction. [18
]
The conduction velocity in the atrial muscle is about 0.5m/sec. The action
potential then enters the base of the ventricle at the Bundle of His and then follows the left
and right bundle branches along the inter
-
ve
ntricular septum. These specialised fibres
conduct the impulses at a very

rapid velocity, about 2m/sec [18
] which results in ventricular
contraction.










2.1.3.2 Cardiac Cycle

The cardiac cycle is the sequence of events that occur in a single heart beat. There are two
important phases involved in this cycle, namely diastole and systole. Systole is the period of
ventricular contraction and diastole of ve
ntricular relaxation. Below, a brief description of a
single cardiac cycle is discussed in detail with the four main parts of the cycle highlighted.

Figure
4
: Action potential velocities in the conduction tree.

[18]

A single cardiac cycle begins in Atrial Systole, which consists of the superior and inferior
vena cava tra
nsferring deoxygenated blood to the right atrium, while the four pulmonary
veins simultaneously transfer oxygenated blood to the left atrium.
[19]
Blood is now ready to
start flowing through the atrioventricular valves into the ventricles. After which the
SA node
transmits an electrical signal that propagates around the myocardium of the atrial, which
causes them to contract, emptying the atria and marks the completion of ventricular filling.
The cycle now moves into Atrial Diastole after the ventricles hav
e finished filling with blood
and starts to relax the atria, which will continue until the current cycle is complete.


The cardiac cycle now moves into the ventricular phase starting with Ventricular Systole.
Firstly, the atrioventricular (AV) node receive
s the electrical signal sent by the SA node and
then conducts impulses through the bundle of His and on

through the Purkinje fibres. [19
]
The outcome of this is the contraction of the ventricles, which pump blood through the
pulmonary arteries and the aort
a. Lastly, the cycle now continues on into Ventricular Diastole
allowing the ventricles to relax


remember that the atria are still relaxed from the Atrial
Diastole phase earlier. Here, the atria begin to fill up with blood ready for the next cycle and
al
lows sufficient time for the myocardium to recover.

2.2 Imaging Techniques


This section will
discuss various imaging techniques with application to medical imaging
associated mainly with the heart. Each imaging technique has its good and bad points, su
ch as
resolution and length of time required to complete a successful scan. Our focus will be on
DT
-
MRI but it is perhaps beneficial to briefly mention one or two other techniques to
highlight the differences between them
, such as CT and MRI scans
.

2.2.1

CT Imaging


CT scans are
frequently

called CAT scans and are

a non
-
invasive painless medical test that
helps physicians diagnose and treat medical conditions. It is extremely accurate and has the
capability of imaging soft tissue, blood vessels and bone
simultaneously.

[
2
0
]

A CT scanner
uses
multiple
X
-
Rays, taken at slightly different angles, to produce highly detailed images

(slices)

of the inside of the human body.
When these image slices are reassembled later by a
computer, the physician is given a hi
ghly detailed view of the human body.


CT scans are a lot less sensitive to patient movement than MRI scans, relying on the patient
to remain

completely still

for the duration of the scan
. But CT scans can be used on patients
who have implanted medical de
vice
s
, such as a pacemaker, unlike MRI, which would
strongly
interfere with it.
Also,

CT scans

provide

physicians

with
real
-
time images

and are
extremely cost
-
effective for hospitals. However, pregnant women and small children are still
persuaded to avoid
CT scans due to

potential
health risks.

Lastly, MRI scans are somewhat
better for imaging soft tissue in the brain and the heart compared to CT scans.



2.2.2 MRI Imaging


Magnetic Resonance Imaging (MRI) is an imaging technique that offers far better contrast of
soft tissue in the human body, in comparison with CT scans.
It uses a strong magnetic field
and numerous radio waves to create detailed images of inside the body. T
he human body
contains water molecules, which in turn contain hydrogen atoms.

[2
1
]

MRI uses an extremely
powerful magnet to line up these atoms in the direction equal to that of the magnetic field

given off by the magnet
. Next
,

radio frequency waves are re
peatedly fired at the part of the
body under examination, which briefly alters the magnetic field of the hydrogen atoms

making them

change direction.

[2
1
]
Once the atoms have recovered they slowly start to
return to their normal state, giving off minute ra
dio signals which are then intercepted by a
conveniently placed radio receiver, based in the MRI scanner.

Figure 5

shows four image
comparisons between different body postures after rigid image
-
based mutual information
registration.


MRI scans prevent pat
ients from being exposed to potentially harmful X
-
Ray radiation and as
states above, is far better for imaging soft tissue in areas such as the brain and the heart.
It has
the potential to replace at least four other cardiac tests like echocardiogram, MUGA

scan,
thallium scan and diagnostic cardiac catheterisation. [
22
]
It also provides high accuracy in
identifying structural deformities of the body. However, MRI scans can range from a modest
20 minute
s to an unbearable 90 minutes [23
] and usually require
the patient to remain
completely still, as even minor movements

(or metal present)

during a scan can blur or
misshape images.


















DT
-
MRI in addition with conventional MRI offers several advantages since “diffusion data
contains additional physical
information about the internal structure of the tissue being
scanned”. [
25
]

Figure
5
:
Comparison between different body postures
after rigid image
-
based mutual inf
ormation registration.
Top row: MRI supine (left), CT supine (right). Bottom
row: MRI prone (left), MRI right decubitus (right).

[24]

2.2.3 DT
-
MRI Imaging


In this study, diffusion tensor magnetic resonance imaging (DT
-
MRI) has been
used

to
produce

high quality
datasets.
Th
is

data
is obtained using
animal and h
uman hearts
which
have been imaged at the University of Leeds and John Hopkins University
. It is somewhat
similar to MRI, in that it is non
-
invasive, but allows us to produce three
-
dimensional traversal
cuts/lancinations.


MRI relies on the local micro
-
structural characteristics of water diffusion and captures the
fibre orientation data
from the imaged tissues. It specifies the complete random motion of
hydrogen
atoms within water molecules. [26
] Diffusion is most likely ill
ustrated by using a
tensor model, by calculating it in six non
-
collinear directions. The tensor itself has three
eigenvalues, which describe the magnitude of the diffusion coefficient in three orthogonal
directions and three eigenvectors, which specify tho
se directions. The eigenvector associated
with the largest eigenvalue

coincides with the longitudinal axis of the cell. The local fibre
trajectories can be reconstructed by tracing the direction of the greatest diffusion and then
adding points at regular i
ntervals. (E.g. one imaging voxel width)
[27
]

As it was observed within tissues that contain a large amount of fibres (i.e. cardiac muscle),
water diffuses faster alongside the direction that fibres are travelling to, rather than in the
vertical
directio
n to fibres. In tissues that possess smaller quantities of fibres, water disperses
in a more spherical pattern.

[26
] All of this information can easily be interpreted by DT
-
MRI
and
visualised
using

3D representations
. Figure 6

shows the average geometry an
d fibre
-
tracking on the DT
-
MRI atlas.


















Figure
6
:
Average geometry and fibre tracking on the DT
-
MRI atlas

[28
]

2.3 Literature Review

(Amruta)


This section will discuss prior studies relating to 3D heart reconstruction and
visualisations,
focusing strongly on ones that have

used DT
-
MRI
imaging
. Each study will be heavily
examined, highlighting exactly what they
did and did not
achieve, as well as

signalling any
gaps in their research which our project may or may not fill.

2.4 Existing Software


This section focuses attention on the programming language candidates and relevant toolkits,
some of which will be used to construct the 3D visualisatio
n of the heart and help extract and
to explain any quantitative information obtained. Each of these candidates will be heavily
scrutinized so as to decide which will be used.

2.4.1 OpenGL and JOGL



OpenGL (Open Graphics Library) is a cross platform, cr
oss language library that h
elps
produce 2D and 3D images [29
]. JOGL (Java OpenGL) is the adapted library version of
OpenGL which works within a Java application. The basis of OpenGL programming is to
work with simple polygon and mesh objects, which in turn

can be built up to form more
complex objects and shapes. The simplicity of the internal OpenGL graphics pipeline makes
it relatively simple and quick to render and rasterize objects from a local coordinate system
into the 3D view space on screen. An addit
ional library, GLUT (OpenGL Utility Toolkit),
can be integrated into OpenGL to provide functionality to detect keyboard and mouse input
as well as providing supplementary complex objects such as spheres and teapots.


In the field of medical imaging, OpenG
L is integrated into some real world applications. 3D
-
DOCTOR is a vector based 3D imaging, modelling and measurement software

[30
]. It is used
to simulate the results of CT or MRI scans in a 3D form. The software can also be employed
to create visual model
s of results from neurological examinations. It requires high end
machines to be effective, due to the vast amount of data needed to be processed. Another
example of OpenGL in medical imaging is MITK (Medical Imaging Toolkit) which can be
used to model dif
ferent structures within the body, such as the teeth and the interior of
arterie
s and valves within the heart [31
]. 3DView is a similar software package to 3D
-
DOCTOR, in the fact that it utilises OpenGL to produce 3D vis
ualisations of CT and MRI
data [32
].



The advantages of OpenGL is that it is available across a wide variety of platforms and can
be integrated into most mainstream programming languages, which means virtually anyone
can download and use of it free of charge. In some of these languages, suc
h as Python, there
is no need to recompile code once something OpenGL related is changed, which means
programs can be produced quicker. A further advantage is that the OpenGL commands are
pretty logical and easy to implement without further assistance from

APIs. However, the use
of OpenGL does have disadvantages also. On machines with poor RAM and graphics cards,
performance can suffer, especially for programs trying to implement complex scenes. In
some languages, such as Python and C++, it can also be diff
icult to debug mistakes made in
the OpenGL code. A final disadvantage is that OpenGL is in direct competition with
Microsoft’s Direct3D package, which monopolises the software market.

2.4.2 VTK


<WILL BE DONE FOR 01 DECEMBER MEETING


DOING TODAY AFTER N
AP>


2.4.
3

Java 3D API


<AMRUTA>

2.4.4

Mathworks MatLab


<GEORGE>



Chapter 3 Research Programme


This chapter will provide a brief introduction to the project, discussing background
knowledge regarding the anatomy of the heart and the fibre sheet architecture. It will also
outline the aims and objectives of the project and provide a quick summary of w
hat the
remaining chapters will be focused upon.

3.1 System Architecture



Bibliography


[1]
Gilbert, S. H., & Benson, A. P. (2007). Regional localisation of left ventricular sheet
structure: integration with current models of cardiac fibre, sheet and b
and structure.
European
Journal of cardio
-
Thoracic Surgey (Volume 32, Issue 2)
, (pp. 231
-
249).


[2]

Heart Facts
. (n.d.). Retrieved November 15, 2009, from
http://www.pbs.org/wgbh/nova/heart
/heartfacts.html



[3]
Hsu, E., & Henriquez, C. (2001). Myocardial fiber orientation mapping using reduced
encoding diffusion tensor imaging.
Cardiovascular Magnetic Resonance 3

, 325
-
333.


[4] McCulloch, A., & Guccione, J. (1991). Theory of Heart: Biomech
anics, biophysics, and
Non
-
linear Dynamics of Cardiac Function.
Springer Verlag

, 121
-
144.


[5]
Waldman, L. K., Nosan, D., Villarreal, F., &

Covell, J. W. (1988). Relation between
transmural deformation and local myofiber direction in canine left ventricle.
Circulation
Research

, 63(3):550
-
562.


[6]
Chen, P. S., Cha, Y. M., Peters, B. B., & Chen, L. S. (1993). Effects of myocardial fiber
orien
tation on the electrical induction of ventricular fibrillation.
AM J Physiology

, 264(6 part
2):J160
-
173.


[7]
Zhukov, L., & Barr, A. H. (2003). Heart
-
muscle fiber reconstruction from diffusion tensor
MRI.
Proceedings of the conference on Visualisation
, (p
p. 597
-
602).


[8]
Zhukov, L., & Barr, A. H. (2002). Orientated tensor reconstruction: tracing neural
pathways from diffusion tensor MRI.
Proceedings of the Conference on Visualisation
, (pp.
387
-
394).


[9]

Cardiovascular Biomaterials
. (n.d.). Retrieved November 15, 2009, from Society for
Biomaterials Special Interest Group (SIG) :
http://www.biomaterials.org/SIGS/Cardiovascular/Heart.htm



[10]

O. ThinkQuest
.
(n.d.). Retrieved November 20, 2009, from Walls of the heart:
http://library.thinkquest.org/C003758/Structure/walls.htm


[11]

Hooks, D. A., Trew, M. L., Caldwell, B. J., Sands, G. B.
, LeGrice, I. J., & Smaill, B. H.
(2007). Laminar arrangement of ventricular myocytes influences electrical behavior of the
heart.


[12]

Anderson, R., Smerup, M., Sanchez
-
Quintana, D., Loukas, M., & Lunkenheimer, P.
(2009). The three
-
dimenional arrangement

of myocytes in the ventricular walls.


[13]

Pope, A. J., Sands, G. B., Smaill, B. H., & LeGrice, I. J. (2008). Three
-
dimensional
transmural organization of perimysial collagen in the heart.


[14]

Weber, K. (1989). Cardiac interstitium in health and diseas
e: the fibrillar collagen
network.

[15]

Streeter, D. D., Vaishnav, R. N., Patel, D. J., Spotnitz, H. M., Ross, J., & Sonnenblick,
E. H. (1970). Stress Distribution in the Canine Left Ventricle during Diastole and Systole.


[16]

Azari, H., Buchalter, M., Si
deman, S., Shapiro, E., & Beyar, R. (1992). A conical model
to describe the non uniformity of the left ventricular twisting motion.


[17]

Malmivuo, J., & Plonsey, R. (1995).
Bioelectromagnetism
-

Principles and Applications
of Bioelectric and Biomagnetic
Fields (Ch 6).

New York: Oxford University Press.


[18]

Hurmusiadis, V. (2005). Virtual Heart: Cardiac Simulation for Surgical Training &
Education.


[19]

Cardiac Cycle
. (n.d.). Retrieved 11 25, 2009, from About.com:
http://biology.about.com/library/organs/heart/blcardiaccycle.htm



[20]

CT Scans
. (n.d.). Retrieved 11 29, 2009, from Radiology Info:
http://www.ra
diologyinfo.org/en/info.cfm?pg=bodyct



[21]

MRI scans
. (n.d.). Retrieved 11 29, 2009, from NHS:
http://www.nhs.uk/Conditions/MRI
-
scan/Pages/How
-
does
-
it
-
work.aspx



[22]

Fogo
ros, R. (n.d.).
Cardiac MRI
. Retrieved 11 23, 2009, from About.com:
http://heartdisease.about.com/cs/otherhearttests/a/cardiacMRI_3.htm



[23]

Gould, T. (n.d.).
How MRI Wor
ks
. Retrieved 11 23, 2009, from HowStuffWorks.com:
http://health.howstuffworks.com/mri10.htm



[24]

Estepal, &

al., e. (2009). A MRI study of relative organ position in the upper abdomen
for changes in patient orientation: application to image
-
registered endoscopic ultrasound.
International Journal of Computer Assisted Radiology and Surgery

.


[25]

Zhukov, & al.,
e. (2009). Level Set Modelling and Segmentation of DT
-
MRI Brain Data.
California Institute of Technology

.


[26]
Bock, R. (2000).
Diffusion Tensor Magnetic Imaging
. Retrieved 11 16, 2009, from
National Institute of Health:
http://www.nih.gov/news/pr/dec2000/nichd
-
04.htm



[27]
Landsdown, D. (2007). Quantitative Diffusion Tensor MRI
-
Based Fibre Tracking of
Human Skeletal Muscle.
Vanderbilt University, Nashville TN USA
.


[28]

Peyrat, & al., e. (
n.d.).
A computational framework for the statistical analysis of cardiac
diffusion tensors
. Retrieved 11 16, 2009, from
http://ralyx.inrialpes.fr/2007/Raweb/asclepios/uid102.html



[29]
OpenGL Overview
. (n.d.). Retrieved 11 29, 2009, from OpenGL:
http://www.opengl.org/about/overview/



[30]
3D
-
DOCTOR Medical Modelling, 3D Imaging Article
. (n.d.). Retrieved 11 29, 2009,
from 3D
-
DOCTOR:
http://www.3d
-
doctor.com/aioct.html



[31]
Wolf, I., Vetter, M., Bottger, T., Nolden, M., Schobinger, M., Hastenteufel, M., et al.
(2005). Medical Imaging
Toolkit.


[32]
RMR Systems
-

Medical Imaging
. (n.d.). Retrieved 11 29, 2009, from RMR Systems
Limited:
http://www.rmrsystems.co.uk/volume_rendering.htm


[33]


[34]


[35]