Page 54 - Fister jr., Iztok, and Andrej Brodnik (eds.). StuCoSReC. Proceedings of the 2017 4th Student Computer Science Research Conference. Koper: University of Primorska Press, 2017
P. 54
SPECIFICATIONS ness and the statistical significance of the registration pro-
cedure.
The image registration environment is meant to improve and
speed-up the process of implementing new image registration Img[i].roi: Image coordinates of region of interest.
procedures. In order to improve and speed-up the implemen-
tation phase this environment needs to satisfy the following Img[i].O: Image origin in millimeters. It is the point around
requirements. which the global image registration transformations are de-
fined.
The first requirement is a data structure that will store im-
ages and related registration properties, such as image ori- Img[i].T: 3D global geometric transformation defined as
gin, global transformation, region of interest and much more. 4x4 matrix.
This data structure is meant to help implementing different
image registration procedures in a common way. The sec- Img[i].D: Deformation field that enables describing com-
ond requirement is a graphical user interface that simplifies plex, including nonrigid geometric transformations, defined
work with medical images and registration techniques. It as a 3D displacements of individual image voxels.
needs to work with multiple 3D medical images which are
saved in different file formats. The third requirement is the 2.2 Graphical user interface
possibility to select different image registration procedures
and to display their results. The fourth requirement are The second requirement of the image registration environ-
various tools that can be easily used to implement image ment is a graphical user interface that simplifies work with
registration components, such as measuring similarity using medical images and registration techniques. It allows load-
different similarity measures, resampling images and much ing 3D images in different file formats like DICOM [7], Brain-
more. Web [3] and RIRE (Retrospective Image Registration Eval-
uation project) [12]. It provides a slice by slice visualization,
2.1 Data structure image statistics like histograms and comparison of geometric
relations between images. The graphical user interface also
The first requirement of the registration environment is a allows exporting and importing the described data structure
well defined data structure that stores images, their proper- in order to provide the capability to save temporary results
ties, registration results and other registration parameters. that can be reused in future. One of the main features of
The main reason of such data structure is to provide a com- the graphical user interface is to load and execute different
mon way to manipulate images in different implementations image registration procedures. This feature simplifies and
of registration procedures and to display those images and improves the testing phase of the development process.
registration results in the graphical user interface. The de-
scribed data structure unifies data requirements of previous Figure 1: Graphical user interface.
image registration implementations and is organized as an
array of images. It includes a predefined set of properties The graphical user interface provides a set of tools that helps
which can be easily extended to suffice the needs of image researchers to analyze image relations and registration re-
registration procedures. sults. This tools are divided into two groups. The first
group consists of tools that can be executed over a single
Predefined properties of the data structure: image like plotting a histogram as shown in figure 2 or set-
ting a region of interest.
RefIdx: Index of the image in the Img array to be used as
a reference image. The second group consists of tools that helps analyzing ge-
ometric relations between two images. In this scope the
MovIdx: Index of the image in the Img array to be used graphical user interface includes displaying the joint inten-
as a moving image. sity distribution histogram [8], absolute difference and checker-
board image.
Img: Array which stores all medical images and related reg-
istration properties.
Img[i]: Structure defining a single medical image and re-
lated registration properties.
Img[i].name: Image name.
Img[i].path: Absolute path to the image data.
Img[i].voxelSize: Image voxel size in millimeters.
Img[i].data orig: Original image data in case of losing im-
age information in the process of conversion to uint8 format.
Img[i].data: Image data used in registration procedures
which is converted to uint8 format. The conversion is an es-
tablished compromise between the necessary image bright-
StuCoSReC Proceedings of the 2017 4th Student Computer Science Research Conference 54
Ljubljana, Slovenia, 11 October
cedure.
The image registration environment is meant to improve and
speed-up the process of implementing new image registration Img[i].roi: Image coordinates of region of interest.
procedures. In order to improve and speed-up the implemen-
tation phase this environment needs to satisfy the following Img[i].O: Image origin in millimeters. It is the point around
requirements. which the global image registration transformations are de-
fined.
The first requirement is a data structure that will store im-
ages and related registration properties, such as image ori- Img[i].T: 3D global geometric transformation defined as
gin, global transformation, region of interest and much more. 4x4 matrix.
This data structure is meant to help implementing different
image registration procedures in a common way. The sec- Img[i].D: Deformation field that enables describing com-
ond requirement is a graphical user interface that simplifies plex, including nonrigid geometric transformations, defined
work with medical images and registration techniques. It as a 3D displacements of individual image voxels.
needs to work with multiple 3D medical images which are
saved in different file formats. The third requirement is the 2.2 Graphical user interface
possibility to select different image registration procedures
and to display their results. The fourth requirement are The second requirement of the image registration environ-
various tools that can be easily used to implement image ment is a graphical user interface that simplifies work with
registration components, such as measuring similarity using medical images and registration techniques. It allows load-
different similarity measures, resampling images and much ing 3D images in different file formats like DICOM [7], Brain-
more. Web [3] and RIRE (Retrospective Image Registration Eval-
uation project) [12]. It provides a slice by slice visualization,
2.1 Data structure image statistics like histograms and comparison of geometric
relations between images. The graphical user interface also
The first requirement of the registration environment is a allows exporting and importing the described data structure
well defined data structure that stores images, their proper- in order to provide the capability to save temporary results
ties, registration results and other registration parameters. that can be reused in future. One of the main features of
The main reason of such data structure is to provide a com- the graphical user interface is to load and execute different
mon way to manipulate images in different implementations image registration procedures. This feature simplifies and
of registration procedures and to display those images and improves the testing phase of the development process.
registration results in the graphical user interface. The de-
scribed data structure unifies data requirements of previous Figure 1: Graphical user interface.
image registration implementations and is organized as an
array of images. It includes a predefined set of properties The graphical user interface provides a set of tools that helps
which can be easily extended to suffice the needs of image researchers to analyze image relations and registration re-
registration procedures. sults. This tools are divided into two groups. The first
group consists of tools that can be executed over a single
Predefined properties of the data structure: image like plotting a histogram as shown in figure 2 or set-
ting a region of interest.
RefIdx: Index of the image in the Img array to be used as
a reference image. The second group consists of tools that helps analyzing ge-
ometric relations between two images. In this scope the
MovIdx: Index of the image in the Img array to be used graphical user interface includes displaying the joint inten-
as a moving image. sity distribution histogram [8], absolute difference and checker-
board image.
Img: Array which stores all medical images and related reg-
istration properties.
Img[i]: Structure defining a single medical image and re-
lated registration properties.
Img[i].name: Image name.
Img[i].path: Absolute path to the image data.
Img[i].voxelSize: Image voxel size in millimeters.
Img[i].data orig: Original image data in case of losing im-
age information in the process of conversion to uint8 format.
Img[i].data: Image data used in registration procedures
which is converted to uint8 format. The conversion is an es-
tablished compromise between the necessary image bright-
StuCoSReC Proceedings of the 2017 4th Student Computer Science Research Conference 54
Ljubljana, Slovenia, 11 October