Premium
Four‐dimensional (4D) image reconstruction strategies in dynamic PET: Beyond conventional independent frame reconstruction
Author(s) -
Rahmim Arman,
Tang Jing,
Zaidi Habib
Publication year - 2009
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.3160108
Subject(s) - iterative reconstruction , smoothing , computer science , wavelet , artificial intelligence , computer vision , positron emission tomography , noise (video) , transformation (genetics) , temporal resolution , frame (networking) , sampling (signal processing) , image (mathematics) , physics , optics , nuclear medicine , filter (signal processing) , medicine , telecommunications , biochemistry , chemistry , gene
In this article, the authors review novel techniques in the emerging field of spatiotemporal four‐dimensional (4D) positron emission tomography (PET) image reconstruction. The conventional approach to dynamic PET imaging, involving independent reconstruction of individual PET frames, can suffer from limited temporal resolution, high noise (especially when higher frame sampling is introduced to better capture fast dynamics), as well as complex reconstructed image noise distributions that can be very difficult and time consuming to model in kinetic parameter estimation tasks. Various approaches that seek to address some or all of these limitations are described, including techniques that utilize (a) iterative temporal smoothing, (b) advanced temporal basis functions, (c) principal components transformation of the dynamic data, (d) wavelet‐based techniques, as well as (e) direct kinetic parameter estimation methods. Future opportunities and challenges with regards to the adoption of 4D and higher dimensional image reconstruction techniques are also outlined.