Point spread function deconvolution in 3D micro-CT angiography for multiscale vascular tree separation

S. T. Witt, C. H. Riedel, M. Goessl, M. S. Chmelik, E. L. Ritman

Research output: Contribution to journalConference articlepeer-review

13 Scopus citations

Abstract

Micro-CT angiography of small laboratory mammal organs visualizes vascular branches on a large range of scales, ranging from root-level branches (∼ 1 mm) to endarteriolar vessels (10-40 μm). Multiscale vascular tree segmentation is facilitated by the ability to set a single grayscale threshold value for vessels of all generation levels. Due to the non-ideal modulation transfer function (MTF) of the imaging system, object contrast varies significantly with scale, and the definition of a grayscale threshold for vessel segmentation becomes a problem. We found that performing a point spread function (PSF) deconvolution on the micro-CT projection images significantly reduces the thresholding problem in terms of restoring the smallest vessels' grayscale and delineation. The increased noise from performing a PSF deconvolution will not have a significant effect on the overall signal-to-noise ratio of the images. The PSF deconvolution was successful only when it accommodated the spatial variation of the PSF.

Original languageEnglish (US)
Pages (from-to)720-727
Number of pages8
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume5030 II
DOIs
StatePublished - 2003
EventMedical Imaging 2003: Physics of Medical Imaging - San Diego, CA, United States
Duration: Feb 16 2003Feb 18 2003

Keywords

  • 3D micro-CT imaging
  • MTF
  • Multiscale vascular tree representation
  • PSF
  • Volume scanning

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Point spread function deconvolution in 3D micro-CT angiography for multiscale vascular tree separation'. Together they form a unique fingerprint.

Cite this