Comparison of two methods for SPECT attenuation correction without transmission measurements

The purpose of this study is to evaluate and compare the ability of two methods of attenuation correction without transmission measurements to achieve absolute or relative quantification. The first method, ConTraSPECT, was developed by Welch et al. and finds the uniform elliptical attenuation distribution most consistent with the measured data. In a second more classical approach, the attenuation map is supposed to be uniform on the contour of the body. For both methods varying attenuation coefficients were tested and the attenuation maps obtained were included in a maximum likelihood expectation maximization (ML-EM) algorithm. Both methods were tested using both simulated and experimentally acquired data. The simulated data were generated for a particular slice of the mathematical cardiac torso (MCAT) phantom with its non uniform attenuation map. Data with additive Poisson-distributed noise were simulated. Scatter was not simulated. The experimental data were acquired using the Radiology Support Device striatal phantom. In conclusion we observed that the contour method worked for the purpose of relative quantification as long as the constant attenuation coefficient was carefully chosen. In contrast, ConTraSPECT seemed even suitable for absolute quantification, since it was more robust with respect to variations in the choice of the attenuation coefficient.