3D Reconstruction From Multiple Imaging Planes: A
Pilot Study of Bone Tumor MR Images
Weerayuth Chanapai
Department of Biomedical
Engineering
Faculty of Engineering, Mahidol
University
Nakhonpathom, Thailand
weerayuth.cha@student.mahidol.edu
Praman Fuangfa, MD
Department of Radiology,
Ramathibodi Hospital,
Mahidol University, Bangkok,
Thailand
time_today@hotmail.com
Suphaneewan Jaovisidha, MD
Department of Radiology,
Ramathibodi Hospital,
Mahidol University, Bangkok,
Thailand
rasjv@yahoo.com
Adisak Nartthanarung, MD
Department of Orthopaedics,
Ramathibodi Hospital,
Mahidol University, Bangkok, Thailand
adisak.nar@mahidol.ac.th
Panrasee Ritthipravat *
Department of Biomedical Engineering
Faculty of Engineering, Mahidol University
Nakhonpathom, Thailand
panrasee.rit@mahidol.ac.th
(*corresponding author)
Abstract—MRI inter-slice gap is required during the image
acquisition process in order to prevent leakage protons which
cause blurred images. The gap is one of the major obstacles in
image analysis, particularly in various modern computer aided
diagnosis applications because the data in those inter-slice gaps
disappear. In this paper, an inter-slice interpolation algorithm
based on the projection on convex curve set (POCS) is proposed
to reconstruct MRI voxels and fill inter-slice gaps by using joint
information provided by different MRI viewing planes including
axial, coronal and sagittal planes. Difference of pixel spacing of
these sequences has been transformed and matched to the
reference frame. Pixels from different planes are then located
together in a unit voxel space matrix. They are then interpolated
to create gapless MRI data. The proposed algorithm has been
tested with five MRI datasets of bone tumor. The interpolated
results are then compared with that of frequently used
interpolation techniques including nearest neighbor, linear, and
cubic interpolation. Interpolation results from these algorithms
have been measured in terms of signal to noise ratio which
represents image similarity to the original one. The comparison
results showed that the proposed algorithm is superior to the
others in which average similarity between the interpolated
images and the original MRI images is highest.
Keywords—3D reconstruction; multiple imaging planes; MR
Images; bone tumor
I. INTRODUCTION
Accurate diagnosis within the shortest time is required for
efficient bone tumor management [1]. In clinics, Magnetic
Resonance Imaging (MRI) is a standard, non-invasive
diagnostic tool for examining bone-related lesions. Though it
can present soft tissues better than several image modalities,
such as X-ray, CT etc. [2], bone lesion is not clearly observed
in these MR images. Therefore, differentiation between bone
lesion and normal tissues remains difficult. Currently,
Computer-Aided Diagnosis (CAD) has been developed to
support the decision making process of medical doctors. Useful
information is extracted from the images in order to support
bone lesion detection. Differentiation between normal and
tumor tissues has also been studied. This tool is intended to use
as a second opinion for facilitating physicians in bone tumor
diagnosis and treatment planning. With this tool, physicians
can accurately identify tumor location, size, and invasion better
than in a general examination process. In addition, the
improvement from pre- to post- treatment can be easily
observed.
To provide accurate and necessary information of bone
image data, highly detailed MR images are required. However
during the spin echo MRI imaging, spaces between images are
created in order to prevent the “crosstalk” effect from leaking
of excited protons which may result in noisy images [3]. The
gaps are used as the insulator for partitioning interfering of
excited areas. These gaps are called the “inter-slice gaps” and
they are left as blank spaces between image data. For
applications that require highly detailed images, such as brain
tissue diagnosis, inter-slice gaps can be kept at minimum as
much as possible. Modern MRI scanners may be used for
controlling crosstalk effects. However, it is very costly and
requires bigger data storage and is time consuming. There
exists an alternative technique that can create gapless data. It
uses the gradient echo sequence which is different from the
spin echo sequence in which it does not require 180 degree
proton rephrasing pulse in the imaging process [4]. It thus can
scan faster with minimal excitation of protons which results in
fewer crosstalk effects.
2015 IEEE International Symposium on Signal Processing and Information Technology (ISSPIT)
354