TY - JOUR T1 - A Viewer for Dynamic Whole Body PET/CT Studies JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 705 LP - 705 VL - 58 IS - supplement 1 AU - Jeffrey Leal AU - Evrim Turkbey AU - Lilja Solnes AU - Steven Rowe AU - Arman Rahmim AU - Martin Lodge Y1 - 2017/05/01 UR - http://jnm.snmjournals.org/content/58/supplement_1/705.abstract N2 - 705Objectives: To provide a system for the review and quantitative analysis of dynamically acquired whole-body PET/CT studies. While single bed position dynamic imaging is standard practice in much of PET-based research, only recently have PET/CT scanners possessed the capability to acquire whole-body dynamic PET images. Even so, few clinical review systems have the ability, or user-interface, to allow the appropriate review and analysis of these new datasets.Methods: A PET/CT viewer was developed in the Java programming language for the display and quantitative analysis of dynamic whole body PET/CT data. The program supports multi-time point DICOM encoded series data as well as the manual selection of multiple time frames of whole-body DICOM series. In addition, the software supports the display of co-registered CT data, allowing full PET/CT display along all three primary display planes (Axial, Sagittal and Coronal) and time points. Separate lookup tables are provided for functional and anatomical display data with a blending control to adjust the relative contribution of each to the visual display. The software includes basic parametric image calculation, including ‘integral’ images, in which summed images from the initial scan through to each time point are calculated and available for display. Similarly, voxel-by-voxel least-squares linear fitting of the source image data provides ‘slope’, ‘intercept’, and ‘correlation’ datasets for display as well. These basic parametric datasets can be displayed and used to drive volume-of-interest (VOI) placement for quantification of the source SUV data. This functionality is a preview of Patlak capability which is forthcoming. Quantification is provided through the placement of resizable VOIs. The user is able to place VOIs upon any of the source or derived datasets. A variety of quantitative metrics, in particular MAX and PEAK values, are presented in real-time as time-activity-curve (TAC) data which is dynamically presented to the user as ‘live’ plots which auto-recalculate as the user adjusts and relocates the individual spherical VOIs. Tabular data can be reviewed and exported for further analysis outside the software. TAC plots, as well as the myriad display images (SUV, ‘Integral’, ‘slope’, ‘intercept’, etc.) can be exported in a variety of formats, both as static images as well as dynamic AVI movie loops.Results: Integrated PET/CT display was provided, allowing user navigation of study time points identically to navigating the data spatially. Figure 1(A) is an example of a dynamic whole-body FDG PET/CT study. Intense FDG uptake in the enlarged thyroid gland due to patient’s known thyroiditis was ignored for the purpose of this demonstration. A VOI placed centrally over the liver for reference measurement, as well as VOIs located over lesions in the liver, can be seen. “Live” TAC plots are shown in Figures 1(B) for MAX voxel and Figure 1(C) for maximum PEAK voxel. Liver lesions were pathologically proven to be metastatic adenocarcinoma. Figures 1(D), 1(E), and 1(F) demonstrate experimental parametric capabilities through dynamic MIP displays of ‘integral’, ‘intercept', and ‘slope' images.Conclusion: Today’s modern PET/CT scanners are faster and more efficient than ever before, enabling acquisitions once not even considered. Dynamic whole body PET/CT has the potential of bringing an entirely new dimension of capabilities to our clinical practice. While PET/CT scanners of today may be capable of acquiring these new studies, the image display and analysis tools which are typically available to our clinicians have not kept pace with these new datasets. The viewer presented here provides many of the required tools to properly display and analyze these new datasets, and does so in an implementation which is easily navigable by the user and executable on the widest variety of computing platforms. Research Support: National Institutes of Health (P30 CA006973) ER -