Page 53 - 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. 53
evelopment environment for medical image
registration procedures
Matjaž Šuber
University of Primorska
Faculty of Mathematics, Natural Sciences
and Information Technologies
Glagoljaska 8, 6000 Koper
Koper, Slovenia
matjaz.suber@gmail.com
ABSTRACT the same subject, acquired by the same (mono-modal image
registration) or by different (multi-modal image registration)
Development of novel image registration methods and pro- imaging modalities or at different time points (serial image
cedures is a tedious task because of the need of large number registration).
of specialized lower level functions and an environment with
the abbility to observe results of each step of the process- In general, the process of image registration involves finding
ing. Although several software libraries and environments the optimal geometric transformation which maximizes the
are available to simplify the development process, they re- correspondences across images. Depending on the problem,
quire exhaustive learning and adaptation. To overcome this geometric transformation can model only rotations, trans-
problem we aimed to implement a development environment lations and scaling of images (rigid transformation) or can
that offers all this with only minimal influence on the user include deformations (nonrigid transformation).
development strategies. We have decided to built it on top
of Matlab environment that is already well established in In order to provide high quality image registrations, re-
this research field. The image registration environment in- searchers need development tools, which simplify develop-
cludes a graphical user interface that simplifies testing and ment and testing of methods by offering easy accesible sup-
application of image registration procedures as well as their porting functions. Such an environment needs to work with
development by incorporating a large set of lower level pro- 3D images, to provide a slice by slice visualization, to com-
cessing functions. The environment was tested by using it pare the geometric relations between images, to obtain im-
for the development of the rigid registration procedure. The age statistics like image histograms, to apply and display
results clearly showed that we obtained a reliable environ- registration results and to provide tools that helps imple-
ment, which has simplified and improved the implementa- menting new registration procedures.
tion.
Keywords Development of image registration methods can aid from dif-
ferent types of supporting software including programs like
Medical image registration, Development environment, Graph- 3DSlicer [4] or ImageJ [1], programer libraries like ITK [2]
ical user interface, Image registration toolbox and VTK [10] or toolboxes for common development envi-
ronments like Matlab [11] or Python [9]. The problem of
1. INTRODUCTION most of these tools is that they require user adaptation and
learning.
Medical image registration is a very important component of
many clinical applications. It can be used for the detection In this paper we describe a new image registration environ-
and diagnosis of diseases, for planning the therapy, for guid- ment for Matlab, that improves the implementation of image
ance of interventions and for the followup and monitoring of registration procedures. This environment includes a graph-
patients. ical user interface that simplifies testing and application of
image registration procedures as well as their development
The primary goal of image registration is to find the corre- by incorporating a large set of lower level processing func-
sponding anatomical or functional locations in two or more tions. Its advantage is that it is straightforward and does
images. Image registration can be applied to images from not need extensive learning. In the next section we will
firstly describe our environment, in section 3 we show how
it can support the development and testing of registration
procedures on an example of developing a rigid registration
procedure. In section 4 we conclude the paper with a dis-
cussion.
StuCoSReC Proceedings of the 2017 4th Student Computer Science Research Conference DOI: https://doi.org/10.26493/978-961-7023-40-4.53-56 53
Ljubljana, Slovenia, 11 October
registration procedures
Matjaž Šuber
University of Primorska
Faculty of Mathematics, Natural Sciences
and Information Technologies
Glagoljaska 8, 6000 Koper
Koper, Slovenia
matjaz.suber@gmail.com
ABSTRACT the same subject, acquired by the same (mono-modal image
registration) or by different (multi-modal image registration)
Development of novel image registration methods and pro- imaging modalities or at different time points (serial image
cedures is a tedious task because of the need of large number registration).
of specialized lower level functions and an environment with
the abbility to observe results of each step of the process- In general, the process of image registration involves finding
ing. Although several software libraries and environments the optimal geometric transformation which maximizes the
are available to simplify the development process, they re- correspondences across images. Depending on the problem,
quire exhaustive learning and adaptation. To overcome this geometric transformation can model only rotations, trans-
problem we aimed to implement a development environment lations and scaling of images (rigid transformation) or can
that offers all this with only minimal influence on the user include deformations (nonrigid transformation).
development strategies. We have decided to built it on top
of Matlab environment that is already well established in In order to provide high quality image registrations, re-
this research field. The image registration environment in- searchers need development tools, which simplify develop-
cludes a graphical user interface that simplifies testing and ment and testing of methods by offering easy accesible sup-
application of image registration procedures as well as their porting functions. Such an environment needs to work with
development by incorporating a large set of lower level pro- 3D images, to provide a slice by slice visualization, to com-
cessing functions. The environment was tested by using it pare the geometric relations between images, to obtain im-
for the development of the rigid registration procedure. The age statistics like image histograms, to apply and display
results clearly showed that we obtained a reliable environ- registration results and to provide tools that helps imple-
ment, which has simplified and improved the implementa- menting new registration procedures.
tion.
Keywords Development of image registration methods can aid from dif-
ferent types of supporting software including programs like
Medical image registration, Development environment, Graph- 3DSlicer [4] or ImageJ [1], programer libraries like ITK [2]
ical user interface, Image registration toolbox and VTK [10] or toolboxes for common development envi-
ronments like Matlab [11] or Python [9]. The problem of
1. INTRODUCTION most of these tools is that they require user adaptation and
learning.
Medical image registration is a very important component of
many clinical applications. It can be used for the detection In this paper we describe a new image registration environ-
and diagnosis of diseases, for planning the therapy, for guid- ment for Matlab, that improves the implementation of image
ance of interventions and for the followup and monitoring of registration procedures. This environment includes a graph-
patients. ical user interface that simplifies testing and application of
image registration procedures as well as their development
The primary goal of image registration is to find the corre- by incorporating a large set of lower level processing func-
sponding anatomical or functional locations in two or more tions. Its advantage is that it is straightforward and does
images. Image registration can be applied to images from not need extensive learning. In the next section we will
firstly describe our environment, in section 3 we show how
it can support the development and testing of registration
procedures on an example of developing a rigid registration
procedure. In section 4 we conclude the paper with a dis-
cussion.
StuCoSReC Proceedings of the 2017 4th Student Computer Science Research Conference DOI: https://doi.org/10.26493/978-961-7023-40-4.53-56 53
Ljubljana, Slovenia, 11 October
