Jing Tang publishes paper in Physics in Medicine and Biology

“Dr. Jing Tang received her PhD degree in Electrical Engineering from the University of Illinois at Urbana-Champaign and had her postdoctoral training in Radiology at the John Hopkins School of Medicine. She was employed as a Research Associate at the Division of Medical Imaging Physics at Hopkins before joining Philips Healthcare as a Principle Imaging Physicist. Her research has been on the development and application of acquisition, reconstruction, and analysis techniques in PET [positron emission tomography] and SPECT [single photon emission computed tomography] imaging and also in multimodality imaging such as PET/MR.”

Dynamic PET imaging offers the very notable capability to measure changes in the biodistribution of radiopharmaceuticals within the organs(s) of interest over time. This, in turn, offers very useful information about underlying physiological and biochemical processes, as commonly extracted using various kinetic modeling techniques . . . The conventional approach to dynamic PET imaging involves independent reconstruction of individual dynamic frames followed by kinetic parameter estimation as applied to the reconstructed images . . . This approach, however, can result in the generation of very noisy images as the reconstruction of a given dynamic frame does not utilize information from any other frame.

The emerging field of spatiotemporal four-dimensional (4D) PET image reconstruction seeks to move beyond the conventional scheme, and as a result obtains improved noise levels for a given temporal sampling scheme (thus achieving enhanced noise versus temporal resolution trade-off performance) . . . Direct kinetic parameter estimation methods . . . have the additional advantage that they are able to directly model the Poisson noise distribution in the projection space (avoiding the difficult task of estimating image–space noise correlations: in fact, a common shortcut continues to be to merely neglect or oversimplify space-dependent noise variance and inter-voxel correlations in the reconstructed images). In the context of graphical (linearized) modeling methods, the Patlak model (http://en.wikipedia.org/wiki/Patlak_plot) has been employed to develop direct parametric imaging algorithms . . . These approaches however have been applicable to tracers modeled as effectively irreversibly bound (e.g. 18F-FDG, 18F-FDOPA).

Nonetheless, the most active area in brain PET ligand development and imaging continues to involve receptor/transporter studies involving reversible binding. The focus of this work is therefore to develop a direct 4D parametric image estimation scheme applicable to reversibly bound tracers. Furthermore, while previous works in the area of 4D imaging have been primarily limited to plasma input models . . . , we seek to additionally develop a reference tissue model scheme within 4D image reconstruction. Our approach . . . consisted of utilizing the relative equilibrium (RE) graphical analysis formulation . . . , coupled with a proposed generalization of the AB-EM algorithm . . . within the 4D reconstruction context, allowing the modeling of negative intercept parameters in graphical (linearized) analysis of reversible tracers."