Online Program Home
  My Program

Abstract Details

Activity Number: 244 - Contributed Poster Presentations: Section on Medical Devices and Diagnostics
Type: Contributed
Date/Time: Monday, July 31, 2017 : 2:00 PM to 3:50 PM
Sponsor: Section on Medical Devices and Diagnostics
Abstract #324346
Title: Implications for Gamma Characteristic of Reconstructed PET Data
Author(s): Tian Mou* and Jian Huang and Finbarr O'Sullivan
Companies: University College Cork and University College Cork and University College Cork
Keywords: Image processing ; Gamma distribution ; IRLS ; PET
Abstract:

Improved understanding of the characteristics of PET imaging measurements could lead to improvements in clinical decision making. Iterative statistically based EM reconstructions, so widely used in clinical imaging now, impose positivity constraints and impact the noise properties of reconstructed data. Motivated by analysis of a recently acquired physical phantom dataset from a PET/CT scanner in routine clinical use, we consider both traditional filtered back-projection (FBP) and EM reconstructions of the data. FBP data are quite Gaussian but the EM reconstruction process gives rise to a Gamma-like skewness. The iteratively re-weighted least squares techniques can be adapted for analysis in Gamma parameterization. Model fitting and diagnostics for regions of interest are considered. We highlight use of a Gamma-based probability transform in producing normalized residuals associated with such analyses. The approach is demonstrated for quality assurance analyses associated with physical phantom studies - recovering estimates of local bias and variance characteristics of an operational scanner.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2017 program

 
 
Copyright © American Statistical Association