Simultaneous Dual-Isotope SPECT Imaging Using Independent Component Analysis Based on Wavelet Transformation: Wavelet Based Dual-Isotope SPECT Imaging Using ICA Analysis

Simultaneous dual-isotope imaging from two different energy gamma rays such as Tc-99m (140 keV) and I-123 (159 keV) is an unusual yet useful technology. Its advantages are less acquisition time, the same physiological condition, and better spatial registration. However, the disadvantages of dual-isotope technique are the scatter and cross-talk effect because of overlapping between two nearby photopeaks. We proposed a new method to correct cross-talk from primary and scattered photons through wavelet transformation followed by independent component analysis with multiple energy windows (W_eICA). We simulated dual-isotope digital images with five, six, and seven energy windows using SimSET software to generate scattered noise. We compared those images between W_eICA and standalone eICA method. According to our results, we verified the performance of WeICA in separating Tc-99m and I-123 energy spectra generated from the SimSET Monte Carlo simulation. W_eICA method not only decreased the number of energy windows required but also separated dual-isotope photopeaks correctly. The results have demonstrated that the W_eICA method improved the quantitative accuracy and might be an effective tool for simultaneous dual-isotope SPECT imaging.