Coronary Artery Disease (CAD) is responsible for an unacceptable death rate, compromised life quality and economic burden to the US healthcare system. SPECT MPI is widely utilized for diagnosing, management and predicting outcomes of patients with CAD. Recent studies applying attenuation and spatial resolution correction (ACR) to SPECT have significantly improved these clinical measures over conventional rest-stress protocols and 'stress-only'protocols;the latter having significantly lower patient radiation dose. However, these new methods are applied to less than 10% of patient studies due to transmission imaging hardware and source costs on new systems, and poor incentive for manufacturers to develop solutions for the installed SPECT base. The long-term objective of this research is to develop and commercialize a clinically-validated low-cost software-only ACR solution with significantly lower radiation dose for the un-served SPECT market and the population at-large. The methods will be implemented with rest-stress and 'stress- only'imaging protocols, acquired with 'full-time'or 'half-time/half-dose'protocols proven previously only on the transmission-based systems. The primary objective of the Phase I work is to demonstrate feasibility of a combined computational and methodological approach to estimate the nonuniform attenuation information from components of the emission-spectrum only. It will be shown that the SPECT MPI images obtained in this manner are quantitatively and diagnostically comparable to the 'gold-standard' transmission source methods.
Specific Aims : All investigations will be conducted using the 4D-XCAT/NCAT digital phantom and SIMIND Monte Carlo transport code, and cardiac phantom image data acquired on a model SPECT system.
Aim 1 : Characterize 'gold-standard'transmission imaging and attenuation correction for conventional (full time) 'rest-stress', 'half-time''stress-only'imaging protocols. The results will provide a benchmark for evaluating the feasibility of the new methods.
Aim 2 : Optimize a principal component decomposition (PCD) approach for 1. estimation of attenuation information from Compton scatter components of the emission spectrum, 2. development of a maximum a posteriori reconstruction algorithm for attenuation map estimation from the Compton-based estimates and 3. optimization of tissue boundary detection algorithms for body and lung boundaries from the maximum a posteriori images for a range of body habitus, body truncation and perfusion distributions depicting normal, ischemic and infarct patterns.
Aim 3 : Demonstrate the new ACR SPECT MPI images are quantitatively comparable equivalent to the gold-standard methods, and consistent with qualitative clinical expectations, using image analysis measures, and quantitative measures used clinically for SPECT MPI.
SPECT is a noninvasive imaging test for coronary artery disease. The majority of testing done today is suboptimal and uses excessive radiation exposure. More accurate methods with significantly lower radiation have been proven in large clinical trials, but are generally not obtainable to the installed base. This research develops methods that improve the accuracy and efficiency while dramatically lowering radiation exposure.
