Phantom and Automated Analysis for Advanced Computed Tomography Systems

PHANTOM AND AUTOMATED ANALYSIS FOR ADVANCED COMPUTED TOMOGRAPHY SYSTEMS Abstract This application is a supplement application to U01 EB034205-01, “Multiple x-ray source array (MXA) computed tomography”. The Corgi phantom is an image quality and radiation dosimetry phantom designed specifically for CT scanners and cone beam CT systems and is sold

commercially by the Phantom Laboratory. A key component to this physical phantom is analysis software which will automatically generate rigorous image quality metrics such as the modulation transfer function, the noise power spectrum, linearity, spatial uniformity, and cupping artifact – automatic assessment is performed by uploading the DICOM CT information to the

Phantom Laboratory website. Aim 1 of this proposal is to add an additional module to the existing modular Corgi phantom, which contains signals which allow the assessment of both volume and texture. Volume accuracy is important in cancer imaging (etc.), and texture analysis is important to the growing field of CT radiomics. Aim 2 of this proposal will make use of the

existing Corgi capabilities in addition to the new capabilities to assess the development of the MXA-CT scanner prototype which is the subject of the parent award. Aim 3 of the proposal will be to develop a user group of CT physicists around the country who will use the infrastructure created by the NIST/NIBIB Phantom Lending Library to access both the Corgi phantom and the

evaluation software and evaluate unique CT scanner technologies at their institutions. The user group will be able to assess and perhaps improve the infrastructure at the Phantom Lending Library, in recognition that this phantom includes both hardware and software licensure requirements. Overall, the proposed research will result in an improved Corgi phantom capable

of rigorous and quantitative assessment of CT scanner performance. It will be useful to researchers developing new hardware or reconstruction software for CT, and it will also be very useful to clinical medical physicists as they survey CT scanners at their institutions. The comprehensive set of quantitative metrics, combined with the ease of use and reproducibility of

the assessment software, will improve CT performance evaluation and help to standardize the metrics for CT image quality performance.