SimpleITK With Slicer HansJohnson

SimpleITK Tutorial
Image processing for mere mortals
Insight Software Consortium
Sept 23, 2011
January 10, 2012
(Insight Software Consortium)
SimpleITK - MICCAI 2011
Sept 2011
1 / 142
This presentation is copyrighted by

The Insight Software Consortium
distributed under the
Creative Commons by Attribution License 3.0
http://creativecommons.org/licenses/by/3.0
What this Tutorial is about
Provide working knowledge of the SimpleITK platform
Sept 2011
3 / 142
Program
Virtual Machines Preparation (10min)

Introduction (15min)
Basic Tutorials I (45min)
Short Break (10min)
Basic Tutorials II (45min)
Coee Break (30min)
Intermediate Tutorials (45min)
Short Break (10min)
Advanced Topics (40min)
Wrap-up (10min)
Sept 2011
4 / 142
Tutorial Goals
Gentle introduction to ITK

Introduce SimpleITK
Provide hands-on experience
Problem solving, not direction following
...but please follow directions!
Sept 2011
13 / 142
ITK Overview
How many are familiar with ITK?
Sept 2011
14 / 142
Ever seen code like this?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
// Setup image types .

typedef
float InputPi xelType ;
typedef
float O ut pu t Pi xe lT y pe ;
typedef itk : : Image<InputPixelType , 2> Inpu tImageTy pe ;
typedef itk : : Image<OutputPixelType ,2> O ut pu tI m ag eT yp e ;
// Filter type
typedef itk : : DiscreteGaussianImageFilter<
InputImageType , O ut pu tI m ag eT yp e >
FilterType ;
// Create a filter
FilterType : : Pointer filter = FilterType : : New ( ) ;
// Create the p i p e l i n e
filter >SetInput ( reader >GetOutput ( ) ) ;
filter >SetVariance ( 1 . 0 ) ;
filter >S e t M a x i m u m K e r n e l W i d t h ( 5 ) ;
filter >Update ( ) ;
Ou t pu tI ma g eT yp e : : Pointer blurred = filter >GetOutput ( ) ;
Sept 2011
15 / 142
What if you could write this?
1
2
3
import SimpleITK
input = SimpleITK . ReadImage ( filename )
output = SimpleITK . D i s c r e t e G a u s s i a n F i l t e r ( input , 1 . 0 , 5 )
Sept 2011
16 / 142
What if you could write this?
1
2
3
import SimpleITK
input = SimpleITK . ReadImage ( filename )
output = SimpleITK . D i s c r e t e G a u s s i a n F i l t e r ( input , 1 . 0 , 5 )
We are here to tell you that you can...
Sept 2011
16 / 142
Goals of SimpleITK
Be an on-ramp for ITK

Simplify the use of ITK by
Providing a templateless, typeless layer for ITK in C++
Providing wrappings in scripting languages
Providing access to most ITK algorithms
Sept 2011
17 / 142
SimpleITK Architectural Overview
Conceptually, SimpleITK is an application library built on ITK

All functionality provided by ITK
Components:
Template expansion system
C++ library
Small SWIG definition (more details later)
Glue code for several scripting languages
Some language utilities
Open Source, Apache licensed project

(http://www.opensource.org/licenses/apache2.0.php)
Hosted by GitHub (https://github.com/SimpleITK/SimpleITK)
Sept 2011
18 / 142
Templates in ITK
1
2
3
4
5
6
7
8
9
10
11
typedef unsigned char PixelType ;

enum {Imag eDimensi on = 2 } ;
typedef itk : : Image<PixelType , ImageDimension> ImageType ;
typedef itk : : Vector<float , ImageDimension> VectorType ;
typedef itk : : Image<VectorType , ImageDimension> FieldType ;
typedef itk : : Image<VectorType : : ValueType , ImageDimension> Float ImageTyp e ;
typedef ImageType : : IndexType IndexType ;
typedef ImageType : : SizeType
SizeType ;
typedef ImageType : : RegionType RegionType ;
typedef itk : : M u l t i R e s o l u t i o n P D E D e f o r m a b l e R e g i s t r a t i o n
<ImageType , ImageType , FieldType> R e g i s t r a t i o n T y p e ;
Sept 2011
19 / 142
Template Freedom
1
2
3
4
5
6
7
8
9
10
11
using itk : : simple ;

// Read the image file
I ma ge Fi l eR ea de r reader ;
reader . SetFileName ( " / my / fancy / file . nrrd " ) ;
Image image = reader . Execute ( ) ;
// This filters perform a gaussian bluring with sigma in
// ph y s i c a l space . The output image will be of real type .
S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r gaussian ;
gaussian . SetSigma ( 2 . 0 ) ;
Image blurredImage = gaussian . Execute ( image ) ;
Sept 2011
20 / 142
Programming Models
Object oriented
Function oriented
More about this later
Sept 2011
21 / 142
Wrapping
Transformed by SWIG
Parses header and interface files
Automatically creates scripting glue
Wrappings available for:
Python
Java
C#
Tcl, Lua, Ruby
Others as requested. . .
Sept 2011
22 / 142
Image Anatomy
SimpleITK images contain itk::Image

Hides the templates
Adds useful utility functions
Sept 2011
23 / 142
Filter Anatomy
SimpleITK filters contain itk::ImageToImageFilter

Hides the templates
Adds useful utility functions
(Mainly) output = Filter.Execute ( input )
Sept 2011
24 / 142
Basic Tutorial
Basic Tutorial
Sept 2011
25 / 142
Note on the Tutorial
Most examples will be Python

Obvious translation to other languages
C++ usage (generally) obvious
Sept 2011
33 / 142
Image Class
Creation
Number of dimensions, size, origin, spacing
Pixel access
Back to iPython (Examples/BasicTutorial1/Image.py)
Sept 2011
34 / 142
What just happened?
1
2
3
4
5
import SimpleITK as sitk

# Create an image
image = sitk . Image ( 2 5 6 , 2 5 6 , 2 5 6 , sitk . sitkInt16 ) ;
# How about 2 d ?
twoD = sitk . Image ( 6 4 , 6 4 , sitk . sitkFloat32 )
sitk is the module

Image is the constructor for the Image class
Height, width, depth (omit depth for 2D images)
Datatype (more on this later)
Sept 2011
35 / 142
What just happened?
1
2
3
4
# Addressing pixels
image . GetPixel ( 0 , 0 , 0 )
image . SetPixel ( 0 , 0 , 0 , 1 )
image . GetPixel ( 0 , 0 , 0 )
Get the voxel value at [0,0,0]?

Hmm, I dont like it, so set to 1
What is the value at [0,0,0] now?
Sept 2011
36 / 142
What just happened?
1
2
3
4
# Addressing pixels
image [ 0 , 0 , 0 ]
image [ 0 , 0 , 0 ] = 10
image [ 0 , 0 , 0 ]
Without warning, we sprinkled syntatic sugar on you!

image[0,0,0] is shorthand for Image.GetPixel(0,0,0)
image[0,0,0] = 10 is shorthand for Image.SetPixel(0,0,0,10)
Sept 2011
37 / 142
Summary
Images are created using SimpleITK.Image ( w, h, d, Type )

Images can be 2- or 3-dimensional
Images can describe themselves
Images have simple pixel accessors
Questions before we move on?
Sept 2011
38 / 142
Memory Management
Memory Management
Sept 2011
39 / 142
Images
Images...
usually allocated on the stack
are copy-on-write
use internal smart-pointers
Back to iPython (Examples/BasicTutorial1/MemoryManagement.py)
Sept 2011
40 / 142
Image Memory Management
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
image = SimpleITK . Image ( 3 2 , 3 2 , 3 2 , SimpleITK . sitkInt16 )

print image
...
Image ( 0 x94f2d98 )
Reference Count : 1
...
# Clone image
b = SimpleITK . Image ( image )
print image
...
Reference Count : 2
...
print b
...
Sept 2011
41 / 142
Image Memory Management
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
print b
...
...
b [0 ,0 ,0] = 1
print b
...
Image ( 0 x94f4cb0 )
Reference Count : 1
...
print image
...
Reference Count : 1
...
Sept 2011
42 / 142
Filters
Filters...
usually allocated on the stack
tend to clean up after themselves
do not hold on to images
no access to the ITK pipeline
...more on this later...
Sept 2011
43 / 142
Memory Management Strategies
C++. . .
No need for explicit management
Let images clean up after themselves
Let filters clean up after themselves
Wrapped. . .
Utilize language-specific memory management
Automatic in Python, Java, Ruby, C#, Lua
More manual in Tcl
Sept 2011
44 / 142
Basic Tutorial 2
Basic Tutorial 2
Sept 2011
50 / 142
Hands On
Hands On
Examples/BasicTutorial2/Filters.py
Sept 2011
51 / 142
What just happened?
1
2
3
4
5
6
7
# Simple smoothing
smooth = sitk . S m o o t h i n g R e c u r s i v e G a u s s i a n ( image , 2 . 0 )
sitk . Show ( sitk . Subtract ( image , smooth ) )
...
RuntimeError : Exception thrown in SimpleITK Subtract : . . .
sitk : : ERROR : Both images for S u b t r a c t I m a g e F i l t e r don t match type or dimension !
...
The output of SmoothingRecursiveGaussian is of type float

The input image is signed short
Most SimpleITK filters with 2 inputs require the same type
Lets fix the problem
Sept 2011
52 / 142
Introducing Cast
1
2
3
4
5
# Much better
print " Before : " , smooth . G e t P i x e l I D T y p e A s S t r i n g ( )
smooth = sitk . Cast ( smooth , image . G et Pi xe l ID Va lu e ( ) )
print " After : " , smooth . G e t P i x e l I D T y p e A s S t r i n g ( )
sitk . Show ( sitk . Subtract ( image , smooth ) , " D i f f W i t h G a u s s i a n " )
Back to iPython (Examples/BasicTutorial2/Filters.py)
Sept 2011
53 / 142
Sizes and Indices
1
2
3
4
# Extract
size = [ 6 4 , 6 4 , 1 ]
start = [ 6 4 , 0 , 0 ]
sitk . Show ( sitk . Extract ( image , size , start ) , " Extracted " )
in C++ / ITK code this would use

1
2
3
4
5
6
7
typedef unsigned char PixelType ;

enum {Imag eDimensi on = 2 } ;
typedef itk : : Image<PixelType , ImageDimension> ImageType ;
typedef ImageType : : IndexType IndexType ;
typedef ImageType : : SizeType
SizeType ;
typedef ImageType : : RegionType RegionType ;
SimpleITK uses STL vectors

Wrapping converts to language-specific constructs (tuples/arrays)
Sept 2011
54 / 142
Pixel-wise Operators
1
2
# Use pixel - wise operators

sitk . Show ( 127 image + 127 sitk . BinaryErode ( image ) , " ThinErosion " )
Table: SimpleITK Pixel-wise Operators
Operator
+
/
&
|
Equivalant
Add, AddConstantTo
Subtract, SubtractConstantFrom
Multiply, MultiplyByConstant
Divide, DivideByConstant
And and
Or or
Not not
Pow, PowToConstant
Usage
A + B, s + A, s + B
A B, s A, s B
A B, s A, s B
A/B, s/A, B/s
A&B
A|B
A
A B, A s
A and B are images (2D or 3D), s is a scalar

python only
Sept 2011
58 / 142
Operator Example
1
2
3
4
5
sitk . Hash ( image + 2 )

sitk . Hash ( sitk . AddConstantTo ( image , 2 ) )
sitk . Hash ( image 2 )
sitk . Hash ( sitk . M u l t i p l y B y C o n s t a n t ( image , 2 ) )
Hash calculates the MD5 or SHA1 hash of the image data

Used to tell if two images are exactly identical
Back to iPython (Examples/BasicTutorial2/Morphology.py)
Sept 2011
59 / 142
SimpleITK with Numpy
Numpy is:
Pythons universal numerics foundation
Provides high performance basic linear algebra
numpy interface functions:
1
2
3
4
5
def G e t A r r a y F r o m I m a g e ( image ) :
""" Get a numpy array from a SimpleITK Image . """
def G e t I m a g e F r o m A r r a y ( arr ) :
""" Get a SimpleITK Image from a numpy array . """
Both of these methods do a deep copy of the image

Ensures safety, bypassing error-prone memory issues
Sept 2011
68 / 142
Python Example
To iPython (Examples/InteractiveTutorial/05-02-Numpy.py)
Sept 2011
69 / 142
Image Measurements
Image Measurements
Sept 2011
70 / 142
The Scenerio
You collected 1000 Diusion Weighted Imaging scan sessions

Now, compute the fractional anisotropy measures for a set of
anatomical regions
Sept 2011
71 / 142
Python Example
To iPython (Examples/InteractiveTutorial/MeasureRegions.py)
Sept 2011
72 / 142
Feature Detection
Feature Detection
Sept 2011
73 / 142
Edge Detection
CannyEdgeDetection
1
2
3
4
5
Image C a n n y E d g e D e t e c t i o n ( c o n s t Image& ,
double inLowerThreshold = 0.0 ,
double inUpperThreshold = 0.0 ,
s t d : : v e c t o r <d o u b l e> i n V a r i a n c e = s t d : : v e c t o r <d o u b l e >(3 , 0 . 0 ) ,
s t d : : v e c t o r <d o u b l e> inMaximumError = s t d : : v e c t o r <d o u b l e >(3 , 0 . 0 1 ) ) ;
SobelEdgeDetection
1
Image S o b e l E d g e D e t e c t i o n ( c o n s t Image&
);
ZeroCrossingBasedEdge
1
2
3
4
5
Image Z e r o C r o s s i n g B a s e d E d g e D e t e c t i o n ( c o n s t Image& ,
double i n V a r i a n c e = 1 ,
u i n t 8 t i n F o r e g r o u n d V a l u e = 1u ,
u i n t 8 t i n B a c k g r o u n d V a l u e = 0u ,
d o u b l e inMaximumError = 0 . 1 ) ;
GradientMagnitudeRecursiveGaussian
1
2
3
Image G r a d i e n t M a g n i t u d e R e c u r s i v e G a u s s i a n ( c o n s t Image& ,
double inSigma = 1.0 ,
bool inNormalizeAcrossScale = f a l s e ) ;
Sept 2011
75 / 142
Image Derivatives
Derivative
1
2
3
4
Image D e r i v a t i v e ( c o n s t Image& ,
u n s i g n e d i n t i n D i r e c t i o n = 0u ,
u n s i g n e d i n t i n O r d e r = 1u ,
bool inUseImageSpacing = true ) ;
RecursiveGaussian
1
2
3
4
5
Image R e c u r s i v e G a u s s i a n ( c o n s t Image& ,
bool inNormalizeAcrossScale = fal se ,
i t k : : s i m p l e : : R e c u r s i v e G a u s s i a n I m a g e F i l t e r : : OrderEnumType i n O r d e r = i t k : : s i m p l
u n s i g n e d i n t i n D i r e c t i o n = 0u ) ;
Sept 2011
76 / 142
Advanced Topics
Advanced Topics
Sept 2011
81 / 142
Building SimpleITK in 5 Easy Commands
git clone --recursive https://github.com/SimpleITK/SimpleITK

mkdir SimpleITK/SimpleITK-build
cd SimpleITK/SimpleITK-build
cmake ../SuperBuild
make -j 5
Or as part of Slicer
ITK_VERSION_MAJOR=4
Slicer_USE_SimpleITK=ON
Sept 2011
82 / 142
More complete version
Check out the code from GitHub

(https://github.com/SimpleITK/SimpleITK)
Run CMake (http://www.cmake.org/) using
SimpleITK/SuperBuild as the source directory
Build using your favorite compiler
Sept 2011
83 / 142
Supported Platforms
Windows: Visual Studio 10

Windows: Visual Studio 9 (Requires TR1 service pack)
Mac OSX: gcc 4.x (Xcode 4.x)
Linux: gcc 4.x
Sept 2011
84 / 142
SimpleITK Architecture
SimpleITK Architecture
Sept 2011
85 / 142
Filter Anatomy
SimpleITK filters create ITK filters

Templated based on input type
Output type is usually the same as input type
Instantiated for many possible image types
Sept 2011
86 / 142
Image and Filter Types

Dimensions
Vector Types
2 dimensional
3 dimensional
int8 t
uint8 t
int16 t
uint16 t
int32 t
uint32 t
float
double
Scalar types
int8 t
uint8 t
int16 t
uint16 t
int32 t
uint32 t
float
double
std :: complex < float >
std :: complex < double >
Label Types
uint8 t
uint16 t
uint32 t
Sept 2011
87 / 142
Filter Anatomy
Filter interrogates input

Instantiates proper ITK filter
Executes ITK filter
Constructs output from ITK image
Sept 2011
88 / 142
Using Filters
Using Filters
Sept 2011
89 / 142
Object Paradigm (C++)
1 #i n c l u d e <S i m p l e I T K . h>
2 namespace sitk = itk : : simple ;
3 ...
4 // Create a s m o o t h i n g filter
5 sitk : : S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r gaussian ;
6
7 // Set a p a r a m e t e r
8 gaussian . SetSigma ( 2 . 0 ) ;
9
10 // " Execute " the Filter
11 sitk : : Image blurredImage = gaussian . Execute ( image ) ;
Sept 2011
90 / 142
Flexibility
3 ...
4 // Create a s m o o t h i n g filter
6
7 // Set p a r a m e t e r ( s ) , then execute
8 sitk : : Image blurredImage = gaussian
9
. SetSigma ( 2 . 0 )
10
. Execute ( image ) ;
Sept 2011
91 / 142
3 ...
4 blurredImage = sitk : : S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r ( )
5
. SetSigma ( 2 . 0 )
6
. SetRadius ( 5 )
7
. Execute ( image ) ;
One line: create anonymous filter, set parameters, and execute
Sept 2011
92 / 142
Function Paradigm (C++)
3 ...
4 // Call the f u n c t i o n version
5 // NB : Drop the " I m a g e F i l t e r "!
6 // S i g n a t u r e :
7 /*
8
sitk :: Image S m o o t h i n g R e c u r s i v e G a u s s i a n (
9
const Image & ,
10
11
bool i n N o r m a l i z e A c r o s s S c a l e = false );
12 */
13 sitk : : Image blurredImage = sitk : : S m o o t h i n g R e c u r s i v e G a u s s i a n (
14
image ,
15
2.0 ,
16
false ) ;
Sept 2011
93 / 142
Mix & Match (C++)
3 ...
4 // Get our g a u s s i a n ready
6 gaussian . SetSigma ( 2 . 0 ) ;
7
8 // What is the effect on the image
9 sitk : : Image difference = sitk : : Subtract (
10
image ,
11
gaussian . Execute ( image )
12
);
13 sitk : : Image difference2 = sitk : : Subtract (
14
image ,
15
sitk : : S m o o t h i n g R e c u r s i v e G a u s s i a n (
16
image , 2 . 0
17
)
18
);
Sept 2011
94 / 142
Code Philosophy
Code Philosophy
Sept 2011
95 / 142
Filter Class Overview (C++)
1 class S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r :
2
public ImageFilter {
3
typedef S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r Self ;
4
5
/* * Default C o n s t r u c t o r that takes no a r g u m e n t s
6
and i n i t i a l i z e s default p a r a m e t e r s */
7
SmoothingRecursiveGaussianImageFilter ( ) ;
In line 1, we declare a subclass of ImageFilter

Line 3 creates a special typedef for use later
The default constructor is line 7 (never any parameters)
Sept 2011
96 / 142
Filter Class Overview (C++) Continued
1
2
/* * Define the pixels types s u p p o r t e d by this filter */

typedef B a s i c P i x e l I D T y p e L i s t P ix e lI DT yp e Li st ;
Notice PixelIDTypeList in line 2

Used to instantiate ITK filters
Determines valid input image types
BasicPixelIDTypeList expands to:
int8 t, uint8 t
int16 t, uint16 t
int32 t, uint32 t
float, double
Sept 2011
97 / 142
1
2
3
4
5
6
7
8
Self& SetSigma ( double t ) { . . . return this ; }

double GetSigma ( ) { return this >m_Sigma ; }
Self& S e t N o r m a l i z e A c r o s s S c a l e ( bool t ) { . . . }
Self& N o r m a l i z e A c r o s s S c a l e O n ( ) { . . . }
Self& N o r m a l i z e A c r o s s S c a l e O f f ( ) { . . . }
bool G e t N o r m a l i z e A c r o s s S c a l e ( ) { . . . }
Get/Set parameters
Set methods always return Self & (more later)
Generally, a direct mapping to ITK
Boolean parameters generate On and O methods
Sept 2011
98 / 142
1
2
3
4
5
/* * Name of this class */

std : : string GetName ( ) const { . . . }
/* * Print o u r s e l v e s out */
std : : string ToString ( ) const ;
Return the name and description of the filter
Sept 2011
99 / 142
1
/* * Execute the filter on the input image */
2
Image Execute ( const Image & ) ;
3
4
/* * Execute the filter with p a r a m e t e r s */
5
Image Execute ( const Image &,
6
double inSigma ,
7
bool i n N o r m a l i z e A c r o s s S c a l e ) ;
8 } ; /* End of class S m o o t h i n g R e c u r s i v e G a u s s i a n */
9
10 Image S m o o t h i n g R e c u r s i v e G a u s s i a n ( const Image& ,
11
double inSigma = 1 . 0 ,
12
bool i n N o r m a l i z e A c r o s s S c a l e = false ) ;
Run the filter on an image and return the result

Notice extra function (line 10), adds flexibility
Drop ImageFilter from class name to get function name
Sept 2011
100 / 142
Questions?
Questions?
Sept 2011
101 / 142
Using ITK with SimpleITK
Problem: Use ITK from SimpleITK (or vice versa)

./ToITK input.nii output.nii
Steps:
Load image using SimpleITK
Filter using ITK
Save using OpenCV
Sept 2011
102 / 142
To ITK
Starting code: ToITK/ToITK.cxx

Directory:
SimpleITK-MICCAI-2011-Tutorial/Examples/AdvancedTutorial
1
2
3
4
5
6
7
8
9
10
11
12
namespace s i t k = i t k : : s i m p l e ;
...
// Load t h e image v i a S i m p l e I T K
s i t k : : Image s i t k I m a g e = s i t k : : ReadImage ( i n p u t F i l e n a m e ) ;
// C o n s t r u c t t h e ITK P i p e l i n e
// L i n k p i p e l i n e t o S i m p l e I T K
// Update p i p e l i n e
// C r e a t e o u t p u t S i m p l e I T K image
// Save image v i a S i m p l e I T K
s i t k : : W r i t e I m a g e ( sOutput , o u t p u t F i l e n a m e ) ;
r e t u r n EXIT SUCCESS ;
Sept 2011
103 / 142
ToITK Step 1: Construct the ITK Pipeline
1
2
3
4
5
6
7
8
9
10
11
12
13
// C o n s t r u c t t h e ITK P i p e l i n e
t y p e d e f i t k : : Image<f l o a t ,3> ImageType ;
t y p e d e f i t k : : M i r r o r P a d I m a g e F i l t e r <ImageType , ImageType> P a d F i l t e r T y p e ;
P a d F i l t e r T y p e : : S i z e T y p e upperBound , l o w e r B o u n d ;
P a d F i l t e r T y p e : : P o i n t e r pad = P a d F i l t e r T y p e : : New ( ) ;
f o r ( u n s i g n e d i n t i = 0 ; i < 3 ; i++ )
{
upperBound [ i ] = s i t k I m a g e . G e t S i z e ( ) [ i ] ;
lowerBound [ i ] = s i t k I m a g e . G e t S i z e ( ) [ i ] ;
}
pad >SetPadUpperBound ( upperBound ) ;
pad >SetPadLowerBound ( l o w e r B o u n d ) ;
Sept 2011
104 / 142
ToITK Step 2: Link pipeline to SimpleITK
1
2
3
// L i n k p i p e l i n e t o S i m p l e I T K
ImageType : : P o i n t e r i n p u t I m a g e = ( ImageType ) s i t k I m a g e . G e t I m a g e B a s e ( ) ;
pad >S e t I n p u t ( i n p u t I m a g e ) ;
Sept 2011
105 / 142
ToITK Step 3: Update ITK Pipeline
1
2
// Update p i p e l i n e
pad >Update ( ) ;
Sept 2011
106 / 142
ToITK Step 4: Create the SimpleITK output image
1
2
// C r e a t e o u t p u t S i m p l e I T K image
s i t k : : Image s O u t p u t ( pad >GetOutput ( ) ) ;
Sept 2011
107 / 142
ToITK (Optional) Step 5: Show
1
2
// ( O p t i o n a l ) Show t h e r e s u l t s
s i t k : : Show ( s O u t p u t ) ;
Sept 2011
108 / 142
To ITK Solution
~/Source/AdvancedTutorial-build/ToITK/ToITKSolution \
~/Source/SimpleITK/Testing/Data/Input/RA-Float.nrrd \
/tmp/foo.nii
Sept 2011
109 / 142
Quick Examples from Other Languages
Simple Gaussian in 8
languages
AdvancedTutorial/SimpleGaussian/SimpleGaussian.*
Sept 2011
130 / 142
C++
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
# include < SimpleITK .h >

namespace sitk = itk :: simple ;
int main ( int argc , char * argv [] ) {
// Read the image file
sitk :: I ma g eF il eR e ad er reader ;
reader . SetFileName ( std :: string ( argv [1] ) );
sitk :: Image image = reader . Execute ();
// This filters perform a gauss ian bluring with sigma in physical
// space . The output image will be of real type .
sitk :: S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r gaussian ;
gaussian . SetSigma ( atof ( argv [2] ) );
sitk :: Image blurredImage = gaussian . Execute ( image );
// Covert the real output image back to the original pixel type , to
// make writing easier , as many file formats don t support real
// pixels .
sitk :: C as tI m ag eF il t er caster ;
caster . S e t O u t p u t P i x e l T y p e ( image . G et Pi x el ID Va l ue () );
sitk :: Image outputImage = caster . Execute ( blurredImage );
// write the image
sitk :: I ma ge F il eW ri t er writer ;
writer . SetFileName ( std :: string ( argv [3] ) );
writer . Execute ( outputImage );
return 0;
}
Sept 2011
131 / 142
Ruby
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
require simpleitk
if ARGV . length != 3 then
puts " Usage : Simpl eGaussia n < input > < sigma > < output > " ;
exit ( 1 )
end
reader = Simpleitk : : I ma ge Fi l eR ea de r . new
reader . set_file_name ( ARGV [ 0 ] )
image = reader . execute
in putPixe lType = image . g e t _ p i x e l _ i d v a l u e
gaussian = Simpleitk : : S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r . new
gaussian . set_sigma ARGV [ 1 ] . to_f
image = gaussian . execute image ;
caster = Simpleitk : : C as tI ma g eF il te r . new
caster . s e t _ o u t p u t _ p i x e l _ t y p e inputP ixelType
image = caster . execute image
writer = Simpleitk : : I ma ge Fi l eW ri te r . new
writer . set_file_name ARGV [ 2 ]
writer . execute image
Sept 2011
132 / 142
R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
# Run with :
#
# Rscript -- vanilla S i m p l e G a u s s i a n . R
#
input sigma output
library ( SimpleITK )
args <- commandArgs ( TRUE )
myreader <- I ma ge Fi l eR ea de r ( )
myreader <- I ma ge Fi l eR ea de r _ SetFileName ( myreader , args [ [ 1 ] ] )
myimage <- Im ag e Fi le Re a de r _ Execute ( myreader )
pixeltype <- Image _ G e tP ix el I DV al ue ( myimage )
myfilter <- S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r ( )
myfilter <- S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r _ SetSigma ( myfilter , as . real ( args [ 2 ] ) )
smoothedimage <- S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r _ Execute ( myfilter , myimage )
mycaster <- C as tI ma g eF il te r ( )
mycaster <- C as tI ma g eF il te r _ S e t O u t p u t P i x e l T y p e ( mycaster , pixeltype )
castedimage <- Ca st Im a ge Fi lt e r _ Execute ( mycaster , soothedimage )
mywriter <- I ma ge Fi l eW ri te r ( )
mywriter <- I ma ge Fi l eW ri te r _ SetFileName ( mywriter , args [ [ 3 ] ] )
mywriter <- I ma ge Fi l eW ri te r _ Execute ( mywriter , castedimage )
Sept 2011
133 / 142
C#
1 u s i n g System ;
2 using itk . simple ;
3
4 namespace itk . simple . examples {
5
class Sim pleGauss ian {
6
static void Main ( string [ ] args ) {
7
try {
8
if ( args . Length < 3 ) {
9
Console . WriteLine ( " Usage : Simp leGauss ian < input > < sigma > < output > " ) ;
10
return ;
11
}
12
// Read input image
13
I ma ge Fi l eR ea de r reader = new I ma ge F il eR ea d er ( ) ;
14
reader . SetFileName ( args [ 0 ] ) ;
15
Image image = reader . Execute ( ) ;
16
17
// Execute Gaussian s m o o t h i n g filter
18
S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r gaussian = new S m o o t h i n g R e c u r s i v e G a
19
gaussian . SetSigma ( Double . Parse ( args [ 1 ] ) ) ;
20
image = gaussian . Execute ( image ) ;
21
22
// Write output image
23
I ma ge Fi l eW ri te r writer = new I ma ge F il eW ri t er ( ) ;
24
writer . SetFileName ( args [ 2 ] ) ;
25
writer . Execute ( image ) ;
26
27
} catch ( Exception ex ) {
28
Console . WriteLine ( ex ) ;
29
}
30
}
31
}
32 }
Sept 2011
134 / 142
Java
1 import org . itk . simple . ;
2
3 class Simp leGaussi an {
4
5
public static void main ( String argv [ ] ) {
6
7
if ( argv . length < 3 ) {
8
System . out . println ( " Usage : java S impleGau ssian < input > < sigma > < output > " ) ;
9
return ;
10
}
11
12
org . itk . simple . I ma ge Fi l eR ea de r reader = new org . itk . simple . I ma ge Fi l eR ea de r ( ) ;
13
reader . setFileName ( argv [ 0 ] ) ;
14
Image img = reader . execute ( ) ;
15
16
S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r filter =
17
new S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r ( ) ;
18
filter . setSigma ( Double . valueOf ( argv [ 1 ] ) . doubleValue ( ) ) ;
19
Image blurredImg = filter . execute ( img ) ;
20
21
Ca s tI ma ge F il te r caster = new C as tI ma g eF il te r ( ) ;
22
caster . s e t O u t p u t P i x e l T y p e ( img . g et Pi xe l ID Va lu e ( ) ) ;
23
Image castImg = caster . execute ( blurredImg ) ;
24
25
Im a ge Fi le W ri te r writer = new I ma ge Fi l eW ri te r ( ) ;
26
writer . setFileName ( argv [ 2 ] ) ;
27
writer . execute ( castImg ) ;
28
29
}
30
31 }
Sept 2011
135 / 142
Lua
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
if #arg < 3 then

print ( Usage : Simpl eGaussia n <input> <sigma> <output> )
os . exit ( 1 )
end
reader = SimpleITK . I ma ge Fi l eR ea de r ( )
-- Re m e m b e r that Lua arrays are 1 - based ,
-- and that arg does not contain the a p p l i c a t i o n name !
reader : SetFileName ( arg [ 1 ] )
image = reader : Execute ( ) ;
in putPixe lType = image : G et Pi xe l ID Va lu e ( )
gaussian = SimpleITK . S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r ( )
gaussian : SetSigma ( arg [ 2 ] )
image = gaussian : Execute ( image ) ;
caster = SimpleITK . C as tI ma g eF il te r ( ) ;
caster : S e t O u t p u t P i x e l T y p e ( input PixelTy pe ) ;
image = caster : Execute ( image )
writer = SimpleITK . I ma ge Fi l eW ri te r ( )
writer : SetFileName ( arg [ 3 ] )
writer : Execute ( image ) ;
Sept 2011
136 / 142
Python
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
import SimpleITK as sitk

import sys
if len ( sys . argv ) < 4 :
print " Usage : Sim pleGauss ian < input > < sigma > < output > " ;
sys . exit ( 1 )
reader = sitk . I ma ge Fi l eR ea de r ( )
reader . SetFileName ( sys . argv [ 1 ] )
image = reader . Execute ( )
pixelID = image . G et Pi xe l ID Va lu e ( )
gaussian = sitk . S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r ( )
gaussian . SetSigma ( float ( sys . argv [ 2 ] ) )
image = gaussian . Execute ( image )
caster = sitk . C as tI ma g eF il te r ( )
caster . S e t O u t p u t P i x e l T y p e ( pixelID )
image = caster . Execute ( image )
writer = sitk . I ma ge Fi l eW ri te r ( )
writer . SetFileName ( sys . argv [ 3 ] )
writer . Execute ( image ) ;
Sept 2011
137 / 142
Tcl
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
if { $argc < 3 } {
puts " Usage: Simpl eGaussia n < input > < sigma > < output > "
exit 1
}
I ma ge Fi l eR ea de r reader
reader SetFileName [ lindex $argv 0 ]
set image [ reader Execute ]
set pixelID [ $image G et P ix el ID V al ue ]
S m o o t h i n g R e c u r s i v e G a u s s i a n I m a g e F i l t e r gaussian
gaussian SetSigma [ lindex $argv 1 ]
set image [ gaussian Execute $image ]
Ca s tI ma ge F il te r caster
caster S e t O u t p u t P i x e l T y p e $pixelID
set image [ caster Execute $image ]
Im a ge Fi le W ri te r writer
writer SetFileName [ lindex $argv 2 ]
writer Execute $image
# Tcl r e q u i r e s e x p l i c i t cleanup Cleanup
reader -delete
gaussian -delete
caster -delete
$image -delete
writer -delete
Sept 2011
138 / 142
Conclusion
Conclusion
Sept 2011
139 / 142
Wrap-up
Gentle introduction to ITK

Introduce SimpleITK
Provide hands-on experience
Problem solving, not direction following
Sept 2011
140 / 142
Where to go next
Some resources for using and extending SimpleITK
Home Page
http://www.simpleitk.org
Documentation
http://www.itk.org/SimpleITKDoxygen/html/
Conventions
http:
//www.itk.org/SimpleITKDoxygen/html/Conventions.html
Contributions
http://www.itk.org/SimpleITKDoxygen/html/Developer.html
Sept 2011
141 / 142

SimpleITK With Slicer HansJohnson

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SimpleITK With Slicer HansJohnson

Uploaded by

Copyright:

Available Formats

SimpleITK Tutorial

Image processing for mere mortals

Insight Software Consortium

Sept 23, 2011

January 10, 2012

(Insight Software Consortium)

SimpleITK - MICCAI 2011

This presentation is copyrighted by

What this Tutorial is about

Provide working knowledge of the SimpleITK platform

(Insight Software Consortium)

SimpleITK - MICCAI 2011

Virtual Machines Preparation (10min)

(Insight Software Consortium)

SimpleITK - MICCAI 2011

Gentle introduction to ITK

(Insight Software Consortium)

SimpleITK - MICCAI 2011

How many are familiar with ITK?

(Insight Software Consortium)

SimpleITK - MICCAI 2011

Ever seen code like this?

// Setup image types .

(Insight Software Consortium)

SimpleITK - MICCAI 2011

What if you could write this?

(Insight Software Consortium)

SimpleITK - MICCAI 2011

What if you could write this?

We are here to tell you that you can...

(Insight Software Consortium)

SimpleITK - MICCAI 2011

Be an on-ramp for ITK

(Insight Software Consortium)

SimpleITK - MICCAI 2011

SimpleITK Architectural Overview

Conceptually, SimpleITK is an application library built on ITK

Open Source, Apache licensed project

(Insight Software Consortium)

SimpleITK - MICCAI 2011

typedef unsigned char PixelType ;

(Insight Software Consortium)

SimpleITK - MICCAI 2011

using itk : : simple ;

(Insight Software Consortium)

SimpleITK - MICCAI 2011

(Insight Software Consortium)

SimpleITK - MICCAI 2011

(Insight Software Consortium)

SimpleITK - MICCAI 2011

SimpleITK images contain itk::Image

SimpleITK - MICCAI 2011

SimpleITK filters contain itk::ImageToImageFilter

SimpleITK - MICCAI 2011

(Insight Software Consortium)

SimpleITK - MICCAI 2011

Note on the Tutorial

Most examples will be Python

(Insight Software Consortium)

SimpleITK - MICCAI 2011

(Insight Software Consortium)

SimpleITK - MICCAI 2011

What just happened?

import SimpleITK as sitk