A robust Hough transform algorithm for determining the radiation centers of circular and rectangular fields with subpixel accuracy

Weiliang Du, James Yang

Research output: Contribution to journalArticlepeer-review

28 Scopus citations

Abstract

Uncertainty in localizing the radiation field center is among the major components that contribute to the overall positional error and thus must be minimized. In this study, we developed a Hough transform (HT)-based computer algorithm to localize the radiation center of a circular or rectangular field with subpixel accuracy. We found that the HT method detected the centers of the test circular fields with an absolute error of 0.037 0.019 pixels. On a typical electronic portal imager with 0.5 mm image resolution, this mean detection error was translated to 0.02 mm, which was much finer than the image resolution. It is worth noting that the subpixel accuracy described here does not include experimental uncertainties such as linac mechanical instability or room laser inaccuracy. The HT method was more accurate and more robust to image noise and artifacts than the traditional center-of-mass method. Application of the HT method in Winston-Lutz tests was demonstrated to measure the ball-radiation center alignment with subpixel accuracy. Finally, the method was applied to quantitative evaluation of the radiation center wobble during collimator rotation.

Original languageEnglish (US)
Pages (from-to)555-567
Number of pages13
JournalPhysics in medicine and biology
Volume54
Issue number3
DOIs
StatePublished - 2009

ASJC Scopus subject areas

  • Radiological and Ultrasound Technology
  • Radiology Nuclear Medicine and imaging

Fingerprint

Dive into the research topics of 'A robust Hough transform algorithm for determining the radiation centers of circular and rectangular fields with subpixel accuracy'. Together they form a unique fingerprint.

Cite this