Share Email Print
cover

Proceedings Paper

Removing image artifacts in optoacoustic tomography using virtual transducer restoration
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

We propose an approach allowing significant reduction or even complete removal of artifacts that can appear in optoacoustic images acquired with limited number of transducers (missing detectors) due to incomplete data. In optoacoustic tomography the image is reconstructed from a set of acoustic transducers located on the surface of tissue irradiated by a laser. The rigorous solution of the tomographic problem requires covering of the entire surface of the illuminated volume by an array of transducers. However, in practice, only portion of the surface is available. As a result of data incompleteness, artifacts (usually looking like arc-shaped shadows extending from the bright objects) can appear. These artifacts limit the spatial resolution, degrade the image contrast and distort shapes of the reconstructed objects. The results of the numerical simulation, presented in this work, show that the intensity and the shape of the “arc-shadow” artifacts depend on the surface area of uncovered by the acoustic detectors. The cause of the artifacts appearance is the violation of the absorbed energy conservation by the image reconstruction algorithm. Such explanation of this fact represents a key for removal of these artifacts. As presented in the paper, the intensity of the artifacts could be reduced by partial restoration of the missed transducers. In case of sufficient a priori information about number of objects, the proposed algorithm can be considered as the interpolation/extrapolation of the data or substitution of the missed signal by averaged real signal taking into account energy conservation. In a common case, the signals of virtual transducers are restored from the distorted image using the solution of the wave equation. Then the cleaned image is reconstructed from the complete set of signals combining real and virtual transducers. These operations can be repeated iteratively until artifacts become weak. The accuracy of the image reconstruction depends on the number of absent transducers, i.e. portion of the surface area uncovered by the detectors.

Paper Details

Date Published: 12 July 2004
PDF: 8 pages
Proc. SPIE 5320, Photons Plus Ultrasound: Imaging and Sensing, (12 July 2004); doi: 10.1117/12.527917
Show Author Affiliations
Igor Patrickeyev, Institute of Continuous Media Mechanics (Russia)
Alexander A. Oraevsky, Fairway Medical Technologies, Inc. (United States)


Published in SPIE Proceedings Vol. 5320:
Photons Plus Ultrasound: Imaging and Sensing
Alexander A. Oraevsky; Lihong V. Wang, Editor(s)

© SPIE. Terms of Use
Back to Top