Meetings » 7th CTUG Meeting » Automated analysis of Catphan 500 and 600 images from single and multi-slice CT scanners

Automated analysis of Catphan 500 and 600 images from single and multi-slice CT scanners

Andrew J Reilly1 and Nick Weir2

1 Oncology Physics Dept, Western General Hospital, Edinburgh
2 Medical Physics Dept, Royal Infirmary of Edinburgh

Abstract

Over the past four years a single Catphan 500 phantom has been regularly utilised for performance testing of the various CT scanners within our health board. In particular, it is a key component of the radiotherapy CT-Sim quality assurance programme. A Catphan 600 phantom has recently been purchased, with the main differences between this and the Catphan 500 being:

  • an additional module consisting of bead ramps for measuring narrow slice-thicknesses for multi-slice and spiral protocols;
  • extra reference materials for verifying Hounsfield Unit calibration, including a water reservoir, in a slightly different geometrical configuration from the Catphan 500;
  • two small impulse beads for assessing both in-slice modulation transfer function (MTF) and slice sensitivity profiles, one embedded in the spatial resolution module and the other offset longitudinally from this.

A comparison study was performed to determine whether the two phantoms could be utilised interchangeably. The additional features of the Catphan 600 were also investigated. The phantoms were scanned on single and multi-slice scanners and both axial and spiral protocols were considered.

Software was written to automatically analyse the Catphan images and examine the following parameters:

  • Slice thickness, measured by plate and bead ramps;
  • Hounsfield Unit calibration;
  • Slice sensitivity profiles, determined from the impulse beads;
  • Spatial resolution, in terms of the point spread function (PSF) and MTF, calculated for small and large field of view images of the impulse beads;
  • Uniformity, in terms of index of uniformity, noise power spectrum (NPS) and the plotting of profiles through the uniformity module.

Although it was concluded that the two phantoms could effectively be used interchangeably some interesting differences were found between the two. These will be discussed at the meeting.

Back to meeting timetable