engleski

Optimization of myocardial perfusion imaging through image quality descriptors evaluation

Introduction Evaluation of imaging data is crucial for the optimization of the myocardial perfusion imaging (MPI). Investigation was performed to characterize conventional and advanced MPI systems through a comprehensive evaluation of relevant physical parameters: spatial resolution, contrast-to-noise ratio, self-attenuation characteristics of the system and perfusion detection capacity. Work was performed under the International Atomic Energy Agency project CRO6021. Materials and Methods An anthropomorphic phantom with myocardial insert was imaged using two conventional systems (Siemens Symbia) and one advanced IQ SPECT. Data acquired were reconstructed using filtered back projection (FBP) and iterative reconstruction (IR) algorithms (Symbia), and advanced iterative reconstruction (AIR) algorithm (IQ SPECT). Corrections for scatter (SC) and attenuation (AC) were applied, where available. Two patient conditions were simulated: normal and pathological heart. In normal heart simulations the phantom was filled with 99mTc solutions of different activity concentrations to simulate the biodistribution of the radiopharmaceutical in the body. In pathological heart simulations a perfusion defect (PD) insert filled with water was positioned at four different positions. A comprehensive image quality evaluation was performed using various image quality descriptors: left ventricular (LV) wall thickness, sharpness index, contrast between the LV wall and the inner chamber, contrast of the PD with respect to the LV wall, intrinsic contrast and the net contrast. Overall, image quality for five different systems were evaluated: Symbia-FBP, Symbia-IR, Symbia-IR- SC, IQ-AIR-SC-AC and IQ-AIR. Results The spatial resolution and the contrast-to-noise ratio were found to be superior for IQ-AIR-SC-AC. Analysis of IR and FBP data of conventional systems showed improvement of image quality when IR is used. Application of SC on conventional systems increase spatial resolution and constrast- to-noise ratio. Intrinsic contrast values were minimal for IQ-AIR-SC-AC, resulting in a more uniform distribution of the activity on reconstructed images. Other systems overestimate the signal intensity in the anterior position, and underestimate it in the lateral and posterior positions. The lack of the detection capability of MPI systems in the posterior part of the heart for all systems except IQ-AIR-SC-AC were also observed. Net contrast results showed that IQ-AIR- SC-AC and Symbia-IR-SC systems have overall largest net contrast means and higher perfusion detection capability. Conclusions Characterization of MPI systems via multiple image quality descriptors obtained using static phantom is prerequisite for successful optimization of the MPI procedures. This can be additionally enhanced through the use of non-standard dynamic phantom, which will be the next step in this investigation.

myocardial perfusion imaging, SPECT, optimization, image quality

nije evidentirano