User's Guide

PMOD 3D Rendering Tool

(P3D)
Version 3.5

PMOD Technologies

Printed on 16 October, 2013

 i

Contents
PMOD 3D Rendering Tool Introduction 2

P3D Data Processing 8

Preparation Steps ........................................................................................................................................... 11
Image Data Loading ................................................................................................................................ 11
Image Data Segmentation ....................................................................................................................... 11
Object Rendering...................................................................................................................................... 24
Visualization Options ................................................................................................................................... 28
Object Management and Adjustments of Properties .......................................................................... 28
Viewing Options for Surface Rendering (SR) Objects ........................................................................ 29
Viewing Options for Volume Rendering (VR) Objects ....................................................................... 30
Image Plane Objects ................................................................................................................................. 35
Marker Objects ......................................................................................................................................... 38
Cutting away Parts of the VR or SR Information ................................................................................ 39
Rendering of VOIs ................................................................................................................................... 41
Scene Operations ...................................................................................................................................... 42
Protocol Files ............................................................................................................................................ 44

3D Rendering Examples 46
Example 1: Surface Rendering of a Brain Tumor ...................................................................................... 47
Example 2: Animated Texture on Cardiac Surface ................................................................................... 49
Example 3: Fused Volume Rendering for Angio CT ................................................................................ 51

3D Scatter Plots (from PFUS) 53

P3D Configuration 56

Index 58
 PMOD 3D Rendering Tool Introduction 2

PMOD 3D Rendering Tool Introduction

Physicians trained in cross-sectional imaging are able to understand the spatial relationship
between tissue structures by mere exploration of cross-sectional slice images. However, for
communication purposes it is often helpful to generate simulated views of isolated objects of
interest from different viewpoints. 3D image processing allows deriving such objects from
slice images and calculating virtual reality scenes which can be explored interactively.

Surface Rendering (SR)

One way to derive a virtual object is to track (segment) an organ boundary in the slice
images, build a 3D surface from the contours, and shade the surface. The P3D tool supports
such segmentations by applying thresholds and/or region growing, optionally restricted
within volumes-of-interest. As a special option, the object can be colored by projecting the
information of matched images onto its surface (texturing), even animated in time. This
feature allows, for instance, visualizing the concentration changes of the NH 3 perfusion
tracer at the myocardium surface throughout a dynamic acquisition.

Volume Rendering (VR)

A different way to derive a virtual object is to take a certain viewpoint in front of the object,
cast rays through the object and record the ray values when the rays pass a plane behind the
object, thereby producing an image. The ray value recorded in the image plane is a
combination of the values in all the pixels met along the way from the viewpoint to the
image plane, thus the name Volume Rendering. Typically, the combination is just the sum of
the pixel values, each multiplied by a certain weighting function, called opacity. The result
depends heavily on the image values and the opacity function. There are very successful
configurations available for contrast enhanced CT examinations which provide the illusion
of a three-dimensional representation, especially if the viewpoint is interactively changed.

Combination of 3D objects from Matched Series

A unique feature of P3D is the ability to combine and manipulate different types of virtual
reality objects (VRO) in one common scene, even when they stem from different studies.
Once a scene has been created, it can be saved in a VRML format file (SR objects only). These
files can later be loaded into P3D (or an external VRML-Browser) to continue scene
exploration. It is in principle also possible to import VRML-scenes from some dedicated 3D
rendering programs and combine them with P3D objects, but the import may sometimes fail.
So far, AMIRA data has been successfully imported.

CAUTION: When combining multiple studies in a single rendering, please match

beforehand all images series in the PFUS tool and save the resampled images. Otherwise,
unexpected shifts between the objects might occur in the scene.
 PMOD 3D Rendering Tool Introduction 3

Hardware Requirements and Performance

The P3D tool is based on the OpenGL implementation in the Java3D library
(http://java.sun.com/products/java-media/3D/index.jsp). Therefore, using P3D requires a
graphics board which supports OpenGL, preferably with at least 512 MB video RAM to
accommodate high quality texture displays. As the 3D rendering implementation in Java3D
is quite memory demanding (particularly with VR), a 64-Bit operating system and a RAM
size of at least 4GB is highly recommended when working with highly resolved or dynamic
data. We strongly suggest that you test P3D on your platform with your own data before
purchasing this PMOD option.

An automatically Acceptance test is taking place when P3D module is first started. There are
several 3D scenes available for visual inspection and a dialog window opens for each scene.
Please answer the questions and do not perform any other operation until test completion.

The following tests are performed:

1) test of texture 3D support.

2) test of a non power of 2 texture support.
3) test of maximum resolution of Volume rendering (from 128x128x128 up to 512x512x512).
4) test of a full 3D scene loaded from protocol.
5) test of a scene capture.

At the end of all steps the test results will be summarized in a message window. If the test
succeeded the results should be similar like the one below:

Starting the 3D Rendering Tool

The P3D tool is started from the PMOD ToolBox by clicking the dedicated button.

The P3D organizes the functionality on two pages which can be selected by the upper tabs:
3D Rendering and Segmentation respectively.
 PMOD 3D Rendering Tool Introduction 4

The 3D Rendering application window consists of a large display area for showing the 3D
scene and a control area organized by the two tabs Input and View.
 PMOD 3D Rendering Tool Introduction 5

The Segmentation window consists in two main areas: the image viewports with the
dedicated controls on the upper part and the segmentation parameters settings in the lower
part. To make the division between images and parameters more flexible a top-bottom split
is available.
 PMOD 3D Rendering Tool Introduction 6

The taskbar on the right side of the application windows is a detachable taskbar with
shortcuts along the right edge. Please note the tooltips which provide short explanation of
the button functionality.

Load input image data. The arrow below the load button is used for switching
among the available image formats.

If the pin is fixed, the newly selected studies will be appended to the present one(s).
Otherwise the prior studies will be overwritten.

Save results of the last segmentation. The arrow below the save button is used for
switching among the image formats which can be used for saving.

Contains standard rendering definition: the predefined protocols.

Load a P3D protocol file (data and rendering parameters) to restore a prior
processing session.

Save the processing session (data and rendering parameters) as a P3D protocol file.

Allows capturing the 3D scene.

Allows printing a report related to the 3D scene.

Allows hiding/showing image controls

Toggle button for enlarging/shrinking the image viewport in the 3D Rendering

page, Input tab.

View planes locator.

 PMOD 3D Rendering Tool Introduction 7

Allows triggering the scene refresh once all VR properties have been adjusted.
 P3D Data Processing 8

P3D Data Processing

Overview

The P3D tool starts with a 3D Rendering empty scene. The user can then create SR or VR
objects by loading images and applying the available segmentation techniques in the
Segmentation application window. The Segmentation page is designed as a process driven
interface: allows applying the available segmentation technique, adding the segmented
object to a list, defining the rendering type and optionally the object properties and finally
creating the 3D object. As soon as an object is created, it is displayed in the scene with the the
predefined properties such as color and transparency. If the appearance of an object is not
optimal, its properties can be changed and the scene refreshed until the user is satisfied.

Finally, the composition of the objects will show a meaningful scene which can be
interactively explored to understand the spatial relationships of the segmented tissue
structures. To this end the scene can be rotated in any direction, zoomed, and objects
obstructing the view to deeper ones can be temporarily hidden or set to a high degree of
transparency. Furthermore, planes showing the original image data can be added, or
volumes-of-interest (VOI). To document meaningful results the current rendering can be
saved as a screen capture, or a movie of a rotating scene can be generated. It is also possible
to save the current configuration into a protocol file, so that it may be reconstructed from the
original data at a later time.

Image Data Requirements

There are situations where objects are derived from a single data set, for example the
boundary of a brain tumor and the outer brain contour. In this case the spatial alignment of
the generated objects will be inherently correct.

However, if objects from different image series are rendered in a single scene, some
requirements must be met to ensure that the objects are correctly positioned in the scene. The
positioning rules in P3D are as follows:

1) The generated objects are positioned relative to the center of the scene volume (i.e. the
center of the image volume is aligned with the center of the scene).
2) For the placement and the size of the objects the pixel size is taken into account.

So if the centers of two image volumes are aligned, objects derived from them can be
combined, even if the pixel size is not the same.

Note: To ensure this requirement it is highly recommended to match the studies to be

combined in the image fusion tool PFUS, save the fused studies, and use them for the 3D
rendering.

When importing virtual reality (VRML) objects generated by other programs this behavior
may not be adequate. In this case it can be changed in the P3D tool configuration.
 P3D Data Processing 9

Control Tabs in the 3D Rendering Page

Along the right border there are two control tabs which group the different functions needed
for data processing and image rendering. While they are stacked in practice, the graphics
below shows them side by side to provide an overview of the user interface elements.

They implement the following functionality.

Input Loading of the different images which can be used for segmentation or
texturing.
Defining the image color tables, window/level settings, and the current
slice/frame of the plane images shown in the scene.
Pre-processing of the images, eg. smoothing, masking, scaling.
Loading a volume-of-interest (VOI) definition for restricting a segmentation
spatially, or for rendering VOIs as 3D objects.
Defining the segmentation process.
Defining the 3D rendering type (SR, VR, VR HD) to apply.
Starting the creation of 3D objects.
 P3D Data Processing 10

View Inspecting the available 3D objects which are organized in a tree structure.
Modifying the properties of each object (color, texture, visibility, etc.) to adjust
the scene.
Adding a marker to the scene which highlights an important location.
Adding plane images to the scene.
Adding light to the scene.
Cutting parts out of the objects.
Rotating or zooming the scene.
Recording a movie of a rotating scene.
Saving the SR objects of a scene as a VRML file.
Saving screen captures of the scene.

The procedures to generate and manipulate 3D scenes are described in the next sections.
 P3D Data Processing 11

Preparation Steps
Image Data Loading
In P3D image data are used for two different purposes:

1) For the derivation of virtual reality objects by segmentation processes (Input images).
2) For the coloring of the objects (Texture images). When using texturing for surface
rendering (SR) objects the color at a surface point is derived from the color at the same
point in the texture images. When using textures for volume rendering (VR) objects the
color is derived by a weighted addition of the texture colors during ray tracing.

The images can be loaded from the Menu with Load Input, from the DB Load page or using
the button in the taskbar to the right.

Note that several images can be loaded at once. It is also possible to incrementally load
additional images later on. To do so, the append pushpin below the button should be
activated to enter the append mode . The format of the loaded images can be changed
with the small button.

Image Data Segmentation

Segmentation is a crucial step of 3D rendering. In P3D this task is recommended to be
performed in the Segmentation page. The Segmentation application window allows
creating SR and VR objects in a step wise fashion procedure:
 P3D Data Processing 12

 selection of the image to segment

 selection among the available segmentation technique
 interactive definition of the segmentation parameters
 generation of the masked segment
 saving of the mask segment
 appending the segment to a Segments list
 definition of the segment property and rendering type
 generate 3D objects.

Alternatively, the same steps can be achieved in the 3D Rendering page in the lower section
of the Input tab. However, the compact design of the Input tab can make the segmentation
process quite tedious and time consuming.

The aim of segmentation is to find the boundary of a tissue structure, which can be
transformed into a virtual 3D object. In general, segmentation is a process which is difficult
to fully automate. The best results are obtained if the structure is clearly separated from the
remainder by a substantial image contrast, and if prior anatomical knowledge is used to
guide the segmentation process.

P3D offers several general segmentation methods for finding the object contours. The
contour finding process can further be restricted by an enclosing volume-of-interest (VOI).
The Smooth box allows enabling a smoothing function during the segmentation. Note,
however, that smoothing will change the volume of the segments and is thus not
recommended for small objects.

P3D incorporates segmentation methods which allow generating multiple segments at once
(e.g. multiple threshold definition). The Color box allows assigning automatically a different
color to each segment.

Each Input study has its own segmentation definition which will be updated when
switching between the studies. Therefore, to begin with a segmentation task, first select the
appropriate Input study, and then adjust the segmentation method.
 P3D Data Processing 13

Segmentation Page

Initially, the Segmentation page appears with the loaded images in the left display area, in
the Input pane. The down arrow below the info button allows selecting between stacked
image series.

When working with dynamic series, it is recommended to average within an appropriate

range in order to obtain an appropriate data set for masking. The range can be specified by
the From and To numbers or using the slider handles. When the Average button is activated,
the average uptake in the specified frame range is calculated and the result image is shown
on the Averaged pane.

Segmentation for Creating a Mask

The next step consists of generating segments which represent tissues of interest. The
segments can then be combined into a single mask. Segmentation can be performed on Input
series but also on the Averaged images, depending on which tab is selected. The Histogram
of the pixel values is updated according to the selected images.

It is recommended to change the colortable to Gray, and to enable the overlay Ovr box.
Then, select one of the segmentation methods (described below (on page 16)) to specify an
inclusion criterion. The pixels which satisfy the criterion are colored in red in the image
overlay. Note that overlay updating might be slow when changing a segmentation
parameter, depending on the segmentation method. Segmentation performs the actual
segmentation and shows the result in the Current tab to the right. While standard
segmentations create binary images with 0 (background) and 1 (segment) pixel values, there
are clustering approaches which generate multiple segments in a single calculation. These
 P3D Data Processing 14

segments are distinguished by increasing integer pixel values. Each Segmentation button
activation overrides the previous contents in Current.

The F toggle button in front of the Segmentation button is applicable if a dynamic series was
selected as Input. If it is enabled, only the currently shown frame is processed, otherwise all
of the dynamic frames. The purpose of dynamic segmentation is to construct a series of
objects which can be animated over time.

Multiple segments can be combined into a mask. To prepare such a combination, copy
promising segments to the Combined tab using the Add button to the right of Save
Segment. By repeated Segmentation and Add operations a list of segments can be built up
in the Segments pane. Note that when the Combined pane is selected, the Add button is not
available anymore, as illustrated below.

The No entry in the Segments list indicates the number by which a segment is identified in
the Combined image, Name provides some descriptive information, Volume its physical
volume, and Method identifies the applied segmentation method. Multiple segments in the
list can be selected and transformed into a single segment by the Merge button. Hereby, the
initially distinct values are replaced by a common value, and the original list entries deleted.

In certain circumstances, the segmentation methods alone may not be sufficient to separate
an object form other structures. If this happens, the user can defined a VOI which prevents
segmentation from leaving the area of main interest. To do so, the Use VOI(s) box has to be
enabled and the Edit VOI button activated. The VOI tools interface appears and allows
 P3D Data Processing 15

drawing a VOI. Outline the VOI. Quit the VOI tools with the OK button to confirm the VOI
selection. Make sure the overlay Ovr box is enabled. Finally, activate the Segmentation
button to perform the actual segmentation within the VOI. The result is shown in the
Current tab to the right.

Edit Segments

Each segment in the Segments list has its own properties. The properties can be easily
changed double clicking with the left mouse button on it. A dialog window appears and
allows changing the default settings as shown below:

The functionalities of the Edit Segment window are summarized in the table below:

Segment Allows changing the name of the segment

name

Rendering Allows defining whether the data resulting from segmentation are
type subject to surface rendering or volume rendering. Additionally, the
color property of the object can be defined before the rendering process.

 with the surface radio button enabled (default), a SR object will be

generated. Activating the Set segment color icon a new color can be
assigned to the segment, as shown below:

 with the volume radio button enabled, a VR object will be

generated. When the Rendering type is volume, the color selection
is switched automatically to colortable selection, as shown below:
 P3D Data Processing 16

Ok The Edit window is closed while the new properties of the segment are
confirmed.

Cancel The Edit window is closed without changing the initial properties of the
segment.

Saving the Segment

Saving can be performed using the Save Segment pane in any of the supported image
formats.

Segmentation Methods

ITK-based Implementations

Some of the segmentation methods are based on the use of the ITK (Insight Toolkit) libraries.
ITK can be used under the open-source BSD license which allows unrestricted use, including
use in commercial products (see www.itk.org). The ITK-based methods are clearly denoted
with ITK in brackets.

Note: PMOD Technologies cannot be held liable for permanent support of the ITK interface,
nor for the performance of the provided libraries.

Region Growing Methods

"Region growing algorithms have proven to be an effective approach for image

segmentation. The basic approach of a region growing algorithm is to start from a seed
region (typically one or more pixels) that are considered to be inside the object to be
segmented. The pixels neighboring this region are evaluated to determine if they should also
be considered part of the object. If so, they are added to the region and the process continues
as long as new pixels are added to the region. Region growing algorithms vary depending
on the criteria used to decide whether a pixel should be included in the region or not, the
type of connectivity used to determine neighbors, and the strategy used to visit neighboring
pixels." The ITK Software Guide. (http://www.itk.org/ItkSoftwareGuide.pdf)
 P3D Data Processing 17

Segmentation Preview in the 3D Rendering Page

As a help during the definition of the segmentation criterion a preview of the expected result
is shown in the Input display if the Ovr (Overlay) box is checked. The example below
illustrates REGION GROWING after clicking into a mouse kidney. Note how the color of
the segmentation overlay changes depending on the color table used to display the images.

To start with segmentation, navigate to a slice (in any direction) which shows the tissue of
interest and click with the left mouse button onto a central point. Then adjust the criterion
until the overlay indicates a promising segmentation.

Please use the toggle button in the taskbar for enlarging/shrinking the area of the
segmentation preview image. Note that the image display controls are not accessible in the
enlarged mode, so the image presentation should be adjusted beforehand.

For further expanding of the 2D view port activate the Hide segmentation controls
button located under the segmentation settings. Note that upon activation of this button the
Segmentation panel collapse and only the segmentation method selection list is still
available. Therefore, any further settings should be set beforehand.

CAUTION: The preview only shows the segmentation within the current slice. As there may
exist indirect connections through neighboring slices, additional pixels may be contained in
the final segment.
 P3D Data Processing 18

Segmentation Preview in the Segmentation Page

In the Segmentation page the preview of the expected result is shown in the Input display if
the Ovr (Overlay) box is checked. However, upon activation of the Segmentation button, the
result of the segmentation is shown in the Current viewport and thus available for
inspection. The example below illustrates REGION GROWING preview after clicking into a
mouse kidney and the corresponding segmentation result: a mask. The color of the
segmentation overlay changes depending on the color table used to display the images.

REGION GROWING Segmentation

Region growing is a method by which the user defines a starting point (seed) within the
object of interest, and the algorithm tries to find all connected pixels which fulfill a certain
criterion.

After selecting REGION GROWING the following control elements appear.

The slider and the number field serve for defining the pixel value. Together with the
Direction radio button the criterion for pixel inclusion is formed.
 P3D Data Processing 19

 = include all connected pixels with a value threshold ± Deviation;

 >= include all connected pixels above the defined threshold value;
 <= include all connected pixels below the defined threshold value.

A preview of the result is shown in the Input display if the Ovr (Overlay) box is checked, for
example after clicking into a hot spot. To start the segmentation, navigate to a slice (in any
direction) which shows the tissue of interest and click with the left mouse button onto a
central point. Then adjust the criterion until the overlay indicates a promising segmentation.

THRESHOLD Segmentation
The THRESHOLD segmentation is conceptually simple. All pixels above the threshold are
included in the segment. Sometimes it is helpful to segment at several threshold levels at
once. This can easily be realized by setting the number in the Thresholds field accordingly,
select More, and enter the threshold values in the appearing Set dialog.

In the example below, three segments of decreasing volume will be generated at thresholds
306.3, 561.55 and 816.8. During visualization, the user can show or hide each of the segments
separately.

IN RANGE Segmentation
The IN RANGE segmentation is similar to the THRESHOLD method, except that an upper
limit is also defined. This can be helpful to find objects with intermediate intensities. The
complementary pixels which are outside the specified range can be obtained by checking the
Inv box. Similar to the THRESHOLD segmentation, multiple Ranges can be defined and
segmented at once

HOTTEST PIXELS Segmentation

The HOTTEST PIXELS segmentation allows obtaining a 3D object of the (potentially
disconnected) pixels with the highest values. The number of included pixels can be specified
in the Num of pixels field.
 P3D Data Processing 20

Otsu Threshold (ITK)

"Another criterion for classifying pixels is to minimize the error of misclassification. The goal
is to find a threshold that classifies the image into two clusters such that we minimize the
area under the histogram for one cluster that lies on the other cluster’s side of the threshold.
This is equivalent to minimizing the within class variance or equivalently maximizing the
between class variance." The ITK Software Guide.
(http://www.itk.org/ItkSoftwareGuide.pdf)

With the mouse CT, Otsu with the settings above finds appropriate thresholds for the
skeleton object and the soft tissue, as illustrated below.

Connected Threshold (ITK)

Connected Threshold is a region growing method with the criterion that (similar to the IN
RANGE method) the included pixels must have values between a Lower and Upper
threshold.
 P3D Data Processing 21

As usual with region growing methods, the user has to specify a seed point by clicking into
the image. The ITK region growing methods are particular in that multiple seed points can
be specified at once by the use of markers as illustrated below.

Once the markers tab has been activated and the Set button enabled, a marker is created as a
seed point for each clicking into the image. When the segmentation is started, the region
growing with the specified value range is performed from each marker.

Neighborhood Connected (ITK)

Neighborhood Connected is a region growing method with two criteria:

1) The included pixels must have values between a Lower and Upper threshold.
2) All pixels in the neighborhood (within a specified pixel radius) must also be within the
value range.
 P3D Data Processing 22

By the combination of these criteria small structures are less likely to be accepted in the
region.

Multiple seed points can be specified by the use of markers.

and correspondingly multiple structures can be found at once.

Confidence Connected (ITK)

"The criterion used by the Confidence Connected method is based on simple statistics of the
current region. First, the algorithm computes the mean and standard deviation of intensity
values for all the pixels currently included in the region. A user-provided factor is used to
multiply the standard deviation and define a range around the mean. Neighbor pixels
whose intensity values fall inside the range are accepted and included in the region. When
no more neighbor pixels are found that satisfy the criterion, the algorithm is considered to
have finished its first iteration. At that point, the mean and standard deviation of the
intensity levels are recomputed using all the pixels currently included in the region. This
mean and standard deviation defines a new intensity range that is used to visit current
region neighbors and evaluate whether their intensity falls inside the range. This iterative
process is repeated until no more pixels are added or the maximum number of iterations is
reached."
 P3D Data Processing 23

The number of iterations is specified based on the homogeneity of the intensities of the
anatomical structure to be segmented. Highly homogeneous regions may only require a
couple of iterations. Regions with ramp effects, like MRI images with inhomogeneous fields,
may require more iterations. In practice, it seems to be more important to carefully select the
multiplier factor than the number of iterations. However, keep in mind that there is no
reason to assume that this algorithm should converge to a stable region. It is possible that by
letting the algorithm run for more iterations the region will end up engulfing the entire
image.

The initialization of the algorithm requires the user to provide a seed point. It is convenient
to select this point to be placed in a typical region of the anatomical structure to be
segmented. A small neighborhood around the seed point will be used to compute the initial
mean and standard deviation for the inclusion criterion."
The ITK Software Guide. (http://www.itk.org/ItkSoftwareGuide.pdf)

Volume-of-Interest (VOI) Restriction

The segmentation methods alone may not be sufficient to separate an object form other
structures. If this happens, the user can define a VOI which prevents segmentation from
leaving the area of main interest.

If such a VOI already exists, it can be selected using the button

Otherwise, open the VOI tool using the button to the right of the Input image, draw a
VOI, save it, quit the VOI tool, and select the VOI. Note that the VOIs are also rendered and
will be available for display in the object tree.
 P3D Data Processing 24

The example below shows how a box object VOI was used which only encloses the head of
the mouse.

After Surface Rendering, the head is shown together with the VOI which is also represented
as an object and can be hidden.

Object Rendering
The rendering of 3D objects can be started in two ways:

1) in the Segmentation page activating the Go to 3D button. The objects properties can be
defined beforehand in the Segments list as described above (on page 13)
2) in the 3D Rendering page, Input tab, after the segmentation procedure has been
defined. The objects properties can be defined only after the objects are rendered.
 P3D Data Processing 25

Object Rendering in the Segmentation Page

The activation of the Go to 3D red button in the lower right corner initiates the 3D rendering
of the segments available in the segments list.
 P3D Data Processing 26

When the rendering is finished, the program switches automatically to the 3D Rendering
page. The results of the rendering are shown in a scene display as below:

Note that all the segments properties defined beforehand, using the Edit segment facility,
are reflected into the scene. The properties of the rendered objects can be further modified as
explained in the Visualization Options (on page 28) section.

Object Rendering in the 3D Rendering Page

After the segmentation procedure has been defined, the 3D objects can be generated. P3D
allows the rendering of several studies as well as dynamic series simultaneously. Therefore
there are a few user interface elements in the lowest row on the Input tab

which determine the rendering result.

S This toggle button is important if several input series were loaded. If it is

enabled, only the currently selected input study is processed, otherwise all
loaded studies are processed with their respective definitions.
 P3D Data Processing 27

F This toggle button is applicable if a dynamic series was loaded. If it is

enabled, only the currently shown frame is processed, otherwise all of the
dynamic frames. The purpose of dynamic segmentation is to construct a
series of objects which can be animated over time.

Surface This configuration button defines whether the data resulting from
segmentation are subject to Surface Rendering or Volume Rendering. With
the Surface setting (default), a SR object will be generated.

Volume With this configuration setting a Volume Rendering is performed.

Note: VR is only performed on the current input series, not on several series
concurrently.

Volume With this configuration a High Definition Volume Rendering is performed.

HD
Note: Volume rendering is interpolated to the homogeneous smallest pixel
size. Any other VR object available in the View tree becomes an HD VR
once a VR HD object is appended to the scene.

VOIs With this configuration setting the selected set of VOIs is rendered as a set
[Stripes] of stripes with a width corresponding to the slice thickness. Note that VOIs
can also be independently rendered without loading image data.

VOIs With this configuration setting the selected set of VOIs is interpreted as a
[Surface] volume and surface rendered.

Mode of object creation. If the pushpin is fixed ( ) the created objects are
appended to the current list of objects, otherwise the scene is cleared and
populated with the new objects only.

After the segmentation has been configured, select the rendering button, eg Surface Rend, to
start the segmentation and the subsequent rendering. Note that if the rendering method is
changed, segmentation automatically starts.

We recommend incrementally rendering study by study (S) in the append mode.

 P3D Data Processing 28

Visualization Options
Object Management and Adjustments of Properties
All segmentation operations result in VROs (Virtual Reality Objects) which are arranged as
an object tree in the View tab as illustrated in the example below.

Entries with SR represent surface renderings, VR volume renderings, the Surface Light is
the optional lightning source for SR objects, the Markers are synthetic objects usable for
indicating points of interest, the Planes are sets of image slices, and Note is a text shown in a
corner of the 3D scene. If an object is not needed any more, it can be removed from the tree
by the Remove button.

The Add SR and Add VR buttons are used to create ghost objects which are empty SR or VR
objects with defined visualization attributes. These attributes will be used for rendering the
objects resulting from the next segmentation. The advantage of using ghost objects is
avoiding lengthy rendering operations with inadequate default settings.

To delete a 3D object available in the tree initially select the corresponding object and then
the Remove button. To Collapse all Scene nodes in the View tree activate the green button
available in the lower right corner. Immediately the color button switches to blue. The
details of the View tree can be restore any time activating the blue button Expand all
Scene nodes.

To modify the appearance of a 3D object, the corresponding object must first be selected in
the tree. Its properties are then shown in the lower tab named Surface or Volume etc.
according to the object type. If a sub-tree or several objects in the tree are selected, the
common properties may be manipulated for all contained objects at once.

The meaning and use of the properties are explained in the Viewing Options sections below
for the different types of objects. There are also sections explaining the function of the other
tabs Scene, Info and Cut.
 P3D Data Processing 29

Viewing Options for Surface Rendering (SR) Objects

A SR object only consists of a surface. Its properties define how the surface is represented in
the scene.

The Visible box determines whether or not the object is shown. The texture selection allows
to switch off texturing (No Texture), or to select one of the loaded image series for coloring
the surface. If no texture is active, the surface color can be defined using the Color selection.

There are three Modes how to render a surface: solid, as a wire frame, and by points. The
wire frame and the points modes have the advantage that inner objects are visible. The
examples below illustrate the different modes without and with (lower row) texturing.

Transparency

To also allow viewing into solid objects it has a transparency property. High transparency
settings can be used to make enclosed objects partly visible. Currently there are two ways
how transparency is implemented.

1) The standard way (Screen door, box not checked) is to punch holes into the surface. The
higher the transparency, the bigger the holes.
2) The other way (Blended, box checked), which provides a smoother view, is only available
as long as no VR objects are generated, and if Java3D 1.4 or above is in use.
 P3D Data Processing 30

The check box next to the transparency slider allows switching the two modes.

Note: Property changes of SR objects are immediately reflected in the scene.

Viewing Options for Volume Rendering (VR) Objects

VR Principle

VR objects are computed from all the data, not just from a derived surface. The method is
based on ray tracing from a certain viewpoint. Rays are cast through the image volume, and
the ray values recorded when they pass a plane behind the object, thereby producing an
image.

Each voxel is regarded as a "particle" that emits a certain color and absorbs a fraction of the
passing light from the other voxels along the same ray.
 P3D Data Processing 31

The color transfer function (TF) describes the voxel color as a function of the voxel value. For
CT data, for example, the voxel values range from -1000 to 3000, and color mappings as
illustrated below are often used.

The opacity function describes the voxel opacity as a function of the voxel value. Opacity
values range from 0 to 1, whereby 0 means full transparency, and 1 full opacity. Often, ramp
functions are applied as illustrated below. If lower opacity values are used as in the lower
example, information from the inner of the object may become apparent such as the contrast
agent in the ventricular cavity.
 P3D Data Processing 32

The ray value (the VR image pixel color) is calculated by summing up the contributions
along a ray as illustrated below. Ci denotes the color of a voxeli , and Ti the transparency
(1-opacity) of a voxel. Contributions of voxels far from the surface will be attenuated by
multiple intervening voxels and thus be faint, depending on the opacity function.

VR Implementation in P3D

The properties tab Volume of VR objects shows several elements.

The Visible box serves for hiding the VR object altogether. The Texture selection allows
combining volume rendering with color texturing from a matched study. The lower part of
the tab contains the elements for defining the color and opacity transfer functions. The Auto
refresh button serves for enabling/disabling automatic updating the display after any
settings has been changed.
 P3D Data Processing 33

Opacity Functions

There are different shapes of opacity functions which can be selected using the arrow button
highlighted below.

The purpose of the different shapes is to implement a variety of weighting functions, so that
certain intensity ranges can be emphasized or suppressed. Usually, the ramp function is
applied.

Once an opacity function has been selected, it can be manipulated in several ways as
illustrated below:

1) The opacity value is in the vertical direction. The minimal and the maximal values of the
opacity function can be set numerically or by dragging the horizontal handle lines.
Example: 0.2 (= minimal opacity), 0.1 (= transparency at highest opacity)
2) The pixel value is in the horizontal direction. The min/max range of the transfer function
can be specified as absolute numbers (if the A box is checked), or in relative percent
values (if the A box is not checked as in the example). This is only for the convenience of
the transfer function adjustments and has no impact on the rendering. The range can be
specified by entering numbers directly (example: 5.9, 74.8), or by dragging the vertical
handle lines.
3) Sometimes the transfer function should be defined in a small sub-range of the entire
value range. In this case the x-range can be zoomed by changing the limits of the display
range. The example above displays the whole range from 0% to 100%.
 P3D Data Processing 34

Color Transfer Functions

There are different color transfer functions available as illustrated in the example below.

The upper rendering of the angio CT uses a gray color TF, and the lower rendering the Heart
VR color TF. Note the indicated down arrow button which allows selecting among the
available color TF. After a color TF has been chosen, the mapping of voxel values to colors
can be adjusted by moving the upper/lower color TF limits as indicated by the horizontal
arrows above. The Heart VR table makes lower values look fleshy, and higher values white.

Note: With big data sets refreshing of scenes with VR objects may be quite time-consuming.
In this case it is recommended to switch the Auto refreshing to off. As soon as the VR
properties are changed, the alert icon appears in the taskbar to the right. Once all VR
properties have been adjusted, the user can trigger scene refresh by clicking at .

Increasing the lower value of the opacity function allows cutting pixels with the density of
myocardium, and the result only shows the contrast filled ventricles and the coronary
arteries.
 P3D Data Processing 35

Volume rendering can also be combined with textures to introduce functional information
into an anatomical rendering. The example below shows such a 3D fusion rendering of an
angio CT combined with the texture color from a matched SPECT perfusion scan. The
opacity function is linear and with the same range as above.

Notes:
1. The 3D fusion result also depends on the color and window level settings of the texture
study which can be changed on the Input tab.
2. Voxels in the segment which have values below the lower or above the upper thresholds
of the texture color table obtain the color of the selected color TF (Gray scale in the example
above).

Image Plane Objects

Additional elements for composing a meaningful scene are the slice images from the
different image series. Planes also serve for defining the octants to be cut out of VR or SR
objects.
 P3D Data Processing 36

Plane objects can be created by selecting Planes in the object tree, and then on the appearing
Planes tab activating the Add planes button.

A new planes object appears in the tree, per default named 1. This name can be edited in the
Name field. Initially, all three orthogonal planes are displayed, but they can individually be
switched on/off using the Z, Y and X toggle buttons. The x button switches off all three
planes. The planes show the slice images of the study selected on the Input tab. There the
image coloring can be adjusted.

Note: Beause there might be several planes object in the tree, a plane object has to be selected
when changing the coloring or the planes location.

Navigator Window

The position of the planes can best be changed using the planes navigator, which is started
using the button. The appearing window shows three orthogonal slice images and a list
selection to choose among the loaded series.

The triangulation point can be changed by simply clicking into the navigator images. As an
alternative, the slice locations can be changed by selecting other slices in the Input series. In
 P3D Data Processing 37

case the plane locations of the Planes object should no more be changed, select the Lock
pushpin button. The list selection allows to switch the image source for the planes among
the loaded images.

Planes Styles

The planes can be rendered in different styles which can be selected from the list to the right
of the plane buttons.

The axial slice in the example below is Opaque, while the coronal slice is Transparent.
Transparent slices do not obstruct the view, but this style is not applicable if a VR object has
been rendered.

The Contour style suppresses the image information below a threshold which can be
defined with the Thr slider. Only the biggest connected area will be shown. In contrast,
MultiContour, shows all area above the threshold.

If multiple plane sets have been created, View selected only allows showing only the planes
belonging to the object currently selected in the tree.
 P3D Data Processing 38

Marker Objects
Marker objects can be helpful for labeling points of interest in the scene. A marker object can
be created by selecting Markers in the object tree, and then on the appearing Marker tab
activating the Add marker button.

The marker is initially positioned at the plane intersection point of the Input series. Its
position can be changed by changing the triangulation point there, or by opening the
navigator window with and clicking to the point of interest. Once its final location has
been found, activate Lock Marker's Position to prevent unintentionally moving the marker.
 P3D Data Processing 39

The marker can have the shape of a Cross as illustrated above, or a Sphere. If multiple
markers have been created, Selected only allows showing only the marker belonging to the
object currently selected in the tree.

Cutting away Parts of the VR or SR Information

Sometimes it may be helpful to cut parts of the scene. In P3D it is possible to remove parts
from VR or SR objects, limited by orthogonal planes.

Cutting Procedure
1) Make sure a Planes object is available. If not, then create one.
2) Select the VR or SR object(s) in the tree which should be cut. Multiple objects can be cut
at the same time.
3) Select the Cut tab belonging to this object (NOT the Cut tab of the Planes object).

4) Enable the image planes display with the button.

5) Position the plane intersection point such that the three orthogonal planes enclose the
area to be removed.

6) Enable Box to see the wire-frame box with the colored corners. Define the part to be
removed by selecting color circles in the Select corners area. The color circles represent
octants identified by the colored bullets in the corners of the wire box. After selecting the
button(s), only the plane parts enclosing the parts to be removed are shown, and the Cut
 P3D Data Processing 40

button becomes active.

7) The Cut button starts a process with clears all information of the selected object(s) in the
defined area and refreshes the rendering. Use the x button to remove the planes. There is
now an indication in the object tree which objects have been cut.

8) The Full button undoes cutting and brings back the rendering of the full volume.

To apply the cutting procedure to all objects belonging to an image series the root object can
be selected in the object tree as illustrated below.
 P3D Data Processing 41

Rendering of VOIs

Volumes of Interest

Volumes-of-interest can also be rendered by P3D. One application is the visualization of

VOIs independent of an image study. To this end the VOI file is selected with the button
indicated below on the Input tab.

For rendering VOIs, there are two choices available, VOIs [Stripes] and VOIs [Surface].

The results of both rendering types are illustrated below. Note that each VOI results in an
object which can be selected in the tree and modified.
 P3D Data Processing 42

In case a VOI was used for restricting an image segmentation, it is rendered together with
the segment and will also be available in the object tree.

In P3D there is the possibility of rendering dynamic VOIs. Select the VOI file. When no
image is loaded, the activation of one of the VOIs rendering options opens a dialog window
as shown below:

When the Render all frames radio button is enabled, the entire set of VOIs will be rendered.
Differently, when Render selected frame is enabled, the user can select which frame VOIs
are going to be rendered. The selection can be done using the slider or typing the frame
number in the dedicated textbox. Additionally, the VOIs surface can be smoothed if the
Smooth VOI surface is checked. To confirm the settings close the dialog window with Yes.

Note: When the P3D option has been purchased, the 3D rendering of the VOIs can also
directly be initiated from the VOI definition tool.

Scene Operations
The main advantage of 3D renderings is that a user can interactively manipulate the scene to
see it from different viewpoints. The scene can be explored in a mouse-operated way:

 To rotate the scene, click the left mouse button into the scene and drag the mouse.
 To zoom the scene, click the middle mouse button into the scene and drag the mouse.
Note that this mouse-operated zooming changes the rotation center. In order to keep the
rotation center in the center of the object, the zoom slicer on the Scene pane should be
used.

To shift the scene out of the center, for instance to examine a zoomed lesion, click the right
mouse button into the scene and drag. To have a full view of the scene use the button in
the taskbar for hiding the controls to the right. Activating the same button again brings the
controls back.

In the View tab Scene pane there are two sub-panes which group different types of
operation.
 P3D Data Processing 43

Scene Views

The list selection elements bring the viewpoint into a well-defined

View position, eg. frontal, from left, etc. The X button to its right resets scene
positioning to the default.

Zoom Zoom factor for scene rendering.

Background
List selection for changing the background color.
Color

View
Show/Hide the box indicating the anatomical directions
Coordinates

View Box Show the wire-frame box.

Scene Rotation and Cines

The Rotation pane contains multiple elements for creating cines or movies.
 P3D Data Processing 44

The sliders allow well-controlled rotations of the scene.

A cine is configured by the following properties:

1) the initial rotation angles (sliders),

2) the rotation axis (icons left to sliders),
3) the Angle increment in degrees (smaller increments produce smoother cines),
4) the rotation Range in degrees,
5) the behavior at the end of the rotation range (bouncing),
6) the Speed.

The cine can then be started with one of the play buttons, eg . Note that the rotation axis
may be changed while the cine is playing, and scene manipulations by the mouse are still
active.

Creating Movies

To record a cine as a movie first configure the cine loop appropriately and test it. Then
enable the movie button and start. A dialog appears

which allows defining the movie format (QuickTime, or DICOM), quality settings and the
destination path.

After confirming with Start, the scene is stepped through the angles until the defined
number of Rotations are covered, creating a jpeg file at each angle. Finally, the jpeg images
are compiled into the movie, and the jpeg files (optionally) deleted.

Protocol Files
Saving a scene in VRML is limited to SR objects, and may create huge files. A more flexible
alternative is to save a description of the P3D rendering in a protocol file using the save
button in the taskbar to the right.
 P3D Data Processing 45

When such a P3D protocol is loaded again, a dialog window appears which indicates the
data to be loaded.

If the user wants to apply the same rendering to similar data, he can replace the input files
with the buttons. If he wants to apply the procedure to data already loaded, the Run
protocol on current data can be enabled.

Normally the scene will wiped out and replaced by the result of the protocol. To add to the
current scene, the Append the protocol data and renderings to the current scene pushpin
should be fixed. To create high definition VR objects the Use Volume HD option should be
enabled.

The button shows a dialog window containing standard renderings definitions.

Currently, the follwoing predefined protocols are supported:

1) Heart Atlas.
2) MR-AAL.
3) Angio CT renderings are offered with and without multi-modal fusion.
4) Bones (CT)
5) AAL VOI
6) Mouse (CT+NM)

Note that when a predefined protocol is selected, appropriate input file have to be set using
the button. Alternatively, the predefined protocol can be applied to data already loaded
enabling the Run protocol on current data radio button. Optionally, the Use Volume HD
checkbox can be activated to create high definition VR objects.
 3D Rendering Examples 46

3D Rendering Examples
The main visualization capabilities of P3D are illustrated by some examples.
 3D Rendering Examples 47

Example 1: Surface Rendering of a Brain Tumor

This example can be reproduced with data of patient PFUS1 available in the example PMOD
database after installation.

These three matched studies are loaded as Input series. The aim is to localize and visualize
the brain tumor. A simple yet useful approach is the following:

FCH The tumor is well delineated in the FCH data but needs a restriction.
On slice 29, click into the tumor and apply a REGION GROWING
segmentation with criterion >=1.7.
Select the ellipsoid VOI Tumor enclosure.
Start rendering with SR (current study only).
Switch the VOI to invisible.
Select the red surface color in the Surface section of the tumor SR object.

FDG The FDG data is used to obtain the outline shape of the brain.
First smooth the FDG data with a 6mm Gaussian.
Activate the append mode by selecting the pushpin.
Apply a THRESHOLD segmentation to the smoothed FDG study with a value
of 7.9.
Render as a SR, then set yellow color.
Set blended transparency using a value of 0.7.

MRI MRI slices can be included for a better anatomical understanding.

On the Input tab select the MRI study.
In the object tree select Planes and then activate Add planes. The three
orthogonal MRI planes are shown.
 3D Rendering Examples 48

Switch off the y and the x plane for display in the scene. Only axial MRI slice is
shown.
Select Transparent mode (selection right to z-plane button)
Define the location of the plane using the navigator window, pointing at the
location of interest. The scene gets updated accordingly.

The result from a specific viewpoint is is shown below.

 3D Rendering Examples 49

Example 2: Animated Texture on Cardiac Surface

This example illustrates the capability of P3D to project a dynamic texture onto a static
surface using a dynamic NH3 cardiac PET study. It can be reproduced with data of patient
PCARD1 available in the example PMOD database after installation.

Anatomy Load the dynamic NH3, Stress study of patient PCARD1 into the Input
area.
With the external tools, average the frames 8 to 18.

Smooth the generated average image in a similar way by applying an

8mm Gaussian Smooth 3D on the Filter tab.
Define REGION GROWING segmentation with threshold 65.5 for the
smoothed, averaged data set. Click into the myocardium and start the
Surface Rendering.

A smooth shape of the myocardium is shown in the scene, represented

by the SR RG_65.5 object in the tree.

Cine of Load the same dynamic NH3, Stress study of patient PCARD1 into the
Dynamic Texture area.
Uptake Select the SR RG_65.5 object in the tree, and switch the texturing to the
dynamic series.
 3D Rendering Examples 50

On the Input pane, configure the cine player to show a movie in time
(Frames mode), then start. The uptake over time is now a shown as a
dynamic texture on the object surface.

Note: When a dynamic study is segmented and the F button (single frame) is not selected,
one object per time will be created. In this case a cine of the Input images can be run to show
an animation of the object shape over time. An obvious application for this feature is the
rendering of a beating heart from a gated acquisition.
 3D Rendering Examples 51

Example 3: Fused Volume Rendering for Angio CT

This example illustrates the use of the Predefined protocol for an angio-CT/SPECT fusion. It
can be reproduced with data of patient P3D2 available in the example PMOD database after
installation.

First the CT data was prepared. A cubic sub-volume containing the heart is extracted from
the data using the Resize function available in the external tools. The remaining thoracic
structures are interactively removed by drawing VOIs and setting the values outside the
VOI to -1024.

Next the SPECT data was matched to the CT in the PFUS tool, and the resliced images saved.

In P3D, load the prepared CT study Angio CT Reduced series of patient P3D2 and the
matched SPECT. Then, select Heart (CT+ PET) from the Predefined protocols. Select Run
protocol on current data and than Run Protocol button.
 3D Rendering Examples 52

The protocol performs a segmentation with two RANGE thresholds (-1029-927 HU, 50-400
HU), and performs a VR rendering for each of the resulting segments.

The first segment is textured using the SPECT colors, and results in a rendering as illustrated
below. To see this segment only remove the Visible check of the second and VR object
activate Refresh. After adjusting the SPECT colors a bit the result looks like illustrated below
 3D Scatter Plots (from PFUS) 53

Note that the coronary arteries are colored because the SPECT study is not sharply bonded.
Therefore, the second segment is used for highlighting the coronary artery structure.

When both VR renderings are combined in a single scene (checking both Visible boxes), the
following result is finally obtained.

This scene could be extended by additional objects. For instance, if an important coronary

3D Scatter Plots (from PFUS)

artery could be individually segmented, it could be added as a clearly visible SR object.
 3D Scatter Plots (from PFUS) 54

The 3D tool has a dedicated application for showing a scatter plot of the pixel values
belonging to VOIs of three matched studies in PFUS. The example below shows the matched
images of a brain tumor patient studied with fluorocholine, fluoroethyl tyrosine, and FDG
PET. Two VOIs have been outlined, tumor and gray matter.
 3D Scatter Plots (from PFUS) 55

When Scatter Plot 3D is activated, the following rendering is generated. Each pixel is shown
as a 3D symbol, with the VOI pixels grouped by the symbol color. In the 3D scatter mode
there are some new P3D properties for adjusting the rendering as well as the annotations.

Note that the numeric scatter data can be saved as an ASCII text file for statistical analyses
using the Save button on the IO sub-panel.
 P3D Configuration 56

P3D Configuration
There are a few specific configurations for the P3D tool. They can be accessed by the Menu
entry Settings/Modify, or directly using the button.

The View Box Initially check is for enabling the wire frame box, and View Coordinates
Initially the orientation cube at start-up.

Some graphics boards may also support volume rendering textures in hardware. To exploit
this acceleration feature, check the Enable use of 3D Texture hardware support box, and/or
the Enable use of Non Power OF Two Texture hardware support box. Maximum
resolution of Volume Rendering is identified during the Acceptance test. However it can
be decrease selecting a smaller resolution available on the list.

Objects generated in P3D are positioned relative to the center of the scene volume (i.e. the
center of the image volume is aligned with the center of the scene, Center of matrix). When
importing virtual reality (VRML) objects generated by other programs this behavior may not
be adequate. It can be modified by setting Objects position relative to to Begin of matrix,
Center of matrix or Data origins.

For Exporting the scene, two formats are available: VRML (Virtual Reality Modeling
Language, with or without sensors), or SLT (Stereolitography), a format which is
understood by many CAD programs.

Copyright © 1996-2013 PMOD Technologies Ltd.

All rights reserved.

The PMOD software contains proprietary information of PMOD Technologies Ltd; it is

provided under a license agreement containing restrictions on use and disclosure and is also
protected by copyright law. Reverse engineering of the software is prohibited.

Due to continued product development the program may change and no longer exactly
correspond to this document. The information and intellectual property contained herein is
confidential between PMOD Technologies Ltd and the client and remains the exclusive
property of PMOD Technologies Ltd. If you find any problems in the document, please
 P3D Configuration 57

report them to us in writing. PMOD Technologies Ltd does not warrant that this document
is error-free.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in

any form or by any means, electronic, mechanical, photocopying, recording or otherwise
without the prior written permission of PMOD Technologies Ltd.

PMOD Technologies Ltd

Sumatrastrasse 25
8006 Zürich
Switzerland
+41 (44) 350 46 00
[email protected]
http://www.pmod.com
58 PMOD 3D Rendering Tool (P3D) User's Guide

Index
PMOD 3D Rendering Tool Introduction • 2
3 Preparation Steps • 11
3D Rendering Examples • 46 Protocol Files • 44
3D Scatter Plots (from PFUS) • 53
R
C REGION GROWING Segmentation • 18
Confidence Connected (ITK) • 22 Rendering of VOIs • 41
Connected Threshold (ITK) • 20
S
Cutting away Parts of the VR or SR
Information • 39 Scene Operations • 42
Scene Rotation and Cines • 43
E Scene Views • 43
Example 1 Segmentation Methods • 13, 16
Surface Rendering of a Brain Tumor • 47 Segmentation Page • 13, 24
Example 2
T
Animated Texture on Cardiac Surface • 49
Example 3 THRESHOLD Segmentation • 19
Fused Volume Rendering for Angio CT •
51 V
Viewing Options for Surface Rendering (SR)
H
Objects • 29
HOTTEST PIXELS Segmentation • 19 Viewing Options for Volume Rendering (VR)
Objects • 30
I Visualization Options • 26, 28
Image Data Loading • 11 Volume-of-Interest (VOI) Restriction • 23
Image Data Segmentation • 11
Image Plane Objects • 35
IN RANGE Segmentation • 19
M
Marker Objects • 38
N
Neighborhood Connected (ITK) • 21
O
Object Management and Adjustments of
Properties • 28
Object Rendering • 24
Object Rendering in the 3D Rendering Page •
26
Object Rendering in the Segmentation Page •
25
Otsu Threshold (ITK) • 20
P
P3D Configuration • 56
P3D Data Processing • 8