Assessment of ventricular function using gated blood pool planar and - SPECT imaging: a phantom study
Loading...
Date
2023
Authors
Pieters, Hané
Journal Title
Journal ISSN
Volume Title
Publisher
University of the Free State
Abstract
In the field of Nuclear Medicine, Gated Blood Pool (GBP) investigations are essential in offering vital insights into cardiac function, specifically the Left Ventricular Ejection Fraction (LVEF). The evaluation of ventricle volume changes during the end-diastolic (ED) and end-systolic (ES) phases plays a critical role in detecting, diagnosing, and managing various cardiac diseases. While the longstanding preference for Gated Blood Pool Planar (GBP-P) methods lies in their validation, non-invasiveness, and straightforward application, challenges such as anatomical overlap and the need for background correction persist.
The theoretical superiority of three-dimensional (3D) analogues, specifically Gated Blood Pool SPECT (GBP-S) studies, promises to overcome GBP-P challenges by offering true volumetric representation without the need for background correction. However, this transition introduces complexity in algorithms and processing software needed for GBP-S studies. Accuracy and precision in determining LVEF is paramount in both GBP-P and GBP-S methods to avoid misdiagnosis, improper treatment, or negligence. Rigorous testing and comparison to known or true values are imperative for validating GBP processing software programs to meet set standards.
A key advancement in validating these software programs is the use of digital hybrid phantoms, notably the advanced 4D-XCAT model. The model, blending voxelised and mathematical elements, mimics human anatomy and physiology. Paired with the Monte Carlo (MC) code, SIMIND, these 4D-XCAT models can be used to simulate clinically realistic GBP images, generating a database for testing, and validating various GBP software packages. Importantly, this approach avoids radiation exposure to patients and researchers and enhances the reliability of outcomes by providing benchmark input parameters for software evaluation.
The primary aim of this investigation was to assess ventricular function using MC-simulated GBP-P and GBP-S images of digital patient phantom studies based on 4D-XCAT models with varying cardiac volumes and functions. The aim was achieved by considering three objectives. Firstly, a Monte Carlo simulated cardiac phantom for planar and SPECT studies was validated. The modelled gamma camera was verified using routine quality control procedures outlined by the National Electrical Manufacturers Association (NEMA). Furthermore, 3D cardiac phantoms were printed and imaged according to GBP-P and GBP-S imaging guidelines. Simulated images of these phantoms were generated using the MC code SIMIND and compared to the gamma camera-acquired images. The successful verification of these simulated images led to the next step, namely verifying the use of the 4D-XCAT model in image simulations. By simulating GBP-P and GBP-S studies of a single 4D-XCAT model, the study demonstrates excellent agreement between known and calculated ventricular parameters, confirming the reliability of the 4D-XCAT model in simulating cardiac imaging.
Building on the successful simulation of the 4D-XCAT model, the second objective involved the creation of a comprehensive database comprising 64 clinically realistic GBP patient models. GBP-P simulated images from these models were utilised to evaluate four commercially available GBP-P processing software programs. The study yielded a strong correlation between calculated LVEF values and known values, thereby affirming the reliability and accuracy of the GBP-P processing software.
Lastly, the research was extended to include clinically realistic GBP-S studies as part of the third objective. The phase focused on assessing the commercially available GBP-S processing software, Quantitative Blood Pool SPECT (QBS), from Cedars-Sinai. GBP-S images of the 4D-XCAT GBP database were simulated according to established imaging guidelines. The study assessed the accuracy of QBS in terms of LV volumes and EF values for two different reconstruction techniques and found a strong correlation between the calculated and known LV volumes and EF values. This paved the way for future multi-centre studies to validate other GBP-S processing software packages.
Description
Dissertation (MMed.Sc.(Medical Physics))--University of the Free State, 2023
Keywords
Nuclear medicine, gated blood pool studies, ejection fraction, ventricular function, end-diastolic volumes, end-systolic volumes, stereolithography printing, validating processing software, Quantitative blood pool SPECT (QBS), Monte Carlo simulations, digital 4D-XCAT phantom