Experimental validation of absolute SPECT/CT quantification for response monitoring in patients with coronary artery disease

Autor: Petra Dibbets-Schneider, Lioe-Fee de Geus-Oei, Douwe E. Atsma, Cornelis H. Slump, Alina van de Burgt, Arthur J.H.A. Scholte, Floris H. P. van Velden
Přispěvatelé: Digital Society Institute, Robotics and Mechatronics, Biomedical Photonic Imaging
Jazyk: angličtina
Rok vydání: 2021
Předmět:
Zdroj: EJNMMI Physics, Vol 8, Iss 1, Pp 1-13 (2021)
EJNMMI physics, 8(1):48. Springer
EJNMMI Physics, 8(1). SPRINGER
EJNMMI Physics
ISSN: 2197-7364
Popis: Myocardial perfusion scintigraphy (MPS) is based on visual qualitative interpretation and hence interobserver variability might affect clinical decision-making for the patient. Quantification of single photon emission computed tomography (SPECT) imaging enables absolute measurement of treatment efficacy. The aim of this experimental study is to assess quantitative accuracy and precision of iterative image reconstruction (Evolution; Q.Metrix package; GE Healthcare, Little Chalfont, UK), making use of both phantom and patient studies, for potential application of 99mTc-tetrofosmin SPECT/CT in coronary artery disease (CAD) patients. This research consists of a phantom study and an experimental validation study with patient data. An anthropomorphic torso phantom with cardiac insert was filled with technetium (99mTc) and acquired on a SPECT/CT gamma (Discovery 670 Pro; GE Healthcare). Seven sets of 6 acquisitions were evaluated containing a cardiac insert with various defect sizes and configurations. Subsequently, volumes‐of‐interest of the defects were manually drawn on CT to assess the recovery coefficient (RC). Bull’s eye plots were composed to evaluate the uptake per segment. Finally, 99mTc-tetrofosmin SPECT/CT scans were collected retrospectively in ten CAD patients after bone marrow cell treatment and evaluated using Evolution. The phantom study shows that the activity concentration converged after seven iterations (ten subsets) using Evolution. SPECT/CT data were reconstructed with and without Butterworth post-filtering. The average repeatability deviation of all configurations was 2.66% and 2.90% (%SD mean) for the filtered and unfiltered data, respectively. The accuracy after Butterworth post-filtering was lower compared to the unfiltered data with a mean(SD) RC of 0.66 ± 0.06 and 0.71 ± 0.05, respectively (p
Databáze: OpenAIRE