Inter-slice interpolation of anisotropic 3-D images using multi-resolution contour correlation

Jiann Der Lee*, Shu Yen Wan, Cherng Min Ma

*Corresponding author for this work

Research output: Contribution to journalConference articlepeer-review

Abstract

To visualize, manipulate and analyze the geometrical structure of anatomical changes, it is often required to perform three-dimensional (3-D) interpolation of the interested organ shape from a series of cross-sectional images obtained from various imaging modalities, such as ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), etc. In this paper, a novel wavelet-based interpolation scheme, which consists of four algorithms are proposed to 3-D image reconstruction. The multi-resolution characteristics of wavelet transform (WT) is completely used in this approach, which consists of two stages, boundary extraction and contour interpolation. More specifically, a wavelet-based radial search method is first designed to extract the boundary of the target object. Next, the global information of the extracted boundary is analyzed for interpolation using WT with various bases and scales. By using six performance measures to evaluate the effectiveness of the proposed scheme, experimental results show that the performance of all proposed algorithms is superior to traditional contour-based methods, linear interpolation and B-spline interpolation. The satisfactory outcome of the proposed scheme provides its capability, for serving as an essential part of image processing system developed for medical applications.

Original languageEnglish
Pages (from-to)1112-1122
Number of pages11
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume4684 II
DOIs
StatePublished - 2002
EventMedical Imaging 2002: Image Processing - San Diego, CA, United States
Duration: 24 02 200228 02 2002

Keywords

  • Image reconstruction
  • Medical image processing
  • Wavelet transform

Fingerprint

Dive into the research topics of 'Inter-slice interpolation of anisotropic 3-D images using multi-resolution contour correlation'. Together they form a unique fingerprint.

Cite this