The effects of compensation for scatter, lead X-rays and high-energy contamination on lesion detectability and activity estimation in Ga-67 imaging

Compton scatter, lead X-rays and high-energy contamination are major factors affecting image quality in Ga-67 imaging. Scattered photons detected in one photopeak can originate from photons emitted in the same photopeak, as well as from higher energy photons which interacted in the collimator and crystal and lost energy. Furthermore, lead X-rays can be detected in the main energy photopeak (93 keV). We have previously developed two energy-based methods, based on artificial neural networks (ANN) and on a generalized spectral fitting approach (GS), to compensate for scatter, high-energy contamination and lead X-rays in Ga-67 imaging. The aim of the present study is to evaluate under realistic conditions the impact of these phenomena and their compensation on lesion detection and estimation tasks in Ga-67 imaging. ANN and GS were compared on the basis of performance of a three-channel Hotelling observer (CHO), which incorporated internal noise, in detecting the presence of a sphere of unknown size on an anatomic background, as well as on the basis of estimation of lesion activity. Spherical lesions ranging from 2 to 6 cm in diameter, located at several sites in an anthropomorphic torso phantom, were simulated using a Monte Carlo program that modeled all photon interactions in the patient as well as in the collimator and the detector for all decays between 91 and 888 keV. One hundred noise realizations were generated from cacti very low noise simulated projection. Scatter worsened both the CHO signal-to-noise ratio (SNR) and the estimation accuracy. On average, the presence of scatter led to a 12% reduction in CHO SNR. Correcting for scatter further diminished CHO SNR but to a lesser extent with ANN (5% reduction compared to WIN) than with GS (12%). Both scatter corrections improved performance in activity estimation compared to WIN. ANN yielded better precision (1.8 vs 4%) but greater bias (5.1 vs 3.6%) than did GS.