TracPETperf-UTU-accuracyFlow2systems

The objective of this study was to: 

The objective of this study was to assess flow quantification accuracy over various injected activities using a flow phantom and two PET/CT systems. QuantifyDCE software was used for flow modelling in the phantom, to extract the time-activity curves and derive the parameters corresponding to phantom flow.

Files contained in the dataset are:

Dynamic DICOM PET series from two PET/CT systems, with folder names corresponding to the manufacturer acronym (D690 or DMI) for a particular PET/CT system. The D690/DMI folders then contain three subfolders with the following names: DICOM, TACs, Results.

Each of the sub-folder names in the DICOM folder (e.g. PT_OSEM_325_MBq) correspond the activity in the phantom at the scan start time, containing dynamic PET DICOM data. The user can extract various information from the DICOM data, including count rates, dead-time and scatter fractions. Moreover, time-activity curves can be extracted from the PET images using either the provided VOI or by VOI drawn by the user.

The files in TACs folder contain VOIs for the phantom input cylinder (bloodVOI) and the exchange cylinder (myoVOI), simulating the blood input and myocardial tissue activity curves. These VOIs are represented as binary files, which can be read e.g. using Vinci as flat images or can be loaded into MATLAB or any other software.

Moreover, the TACs folder contains files (e.g. TACs_PT_OSEM_325MBq) which contain tissue activity curves representing input curve, tissue curve and the simulated tissue curve used to derive the model fitting. These are represented as CSV files, which can be imported to any software. The curves can be used to calculate the AUCs, or the user can fit their own model where the simulated tissue curve from the QuantifyDCE software can be used as a reference for model fitting.

The Results folder contains all of the derived data from the DICOMs, VOIs and TACs. The folder contains all the fitted parameters from the phantom using the model implemented in QuantifyDCE. These parameters can be used as a reference to estimate a custom model fit to the data and can be also compared to the reference values Qref reported in the supplementary material of the linked publication.

Especially important are the K1 and k2 values, which can be used to derive the Qin and Qout flow values form the phantom. K1 and k2 are expressed as 1/min can be converted to Qin and Qout representing mL/min by multiplication of the exchange cylinder volume 161 mL.