Compton-scatter tissue densitometry: calculation of single and multiple scatter photon fluences

The accurate measurement of in vivo electron densities by the Compton-scatter method is limited by attenuations and multiple scattering in the patient. Using analytic and Monte Carlo calculation methods, the Clarke tissue density scanner has been modelled for incident monoenergetic photon energies from 300 to 2000 keV and for mean scattering angles of 30 to 130 degrees. For a single detector focussed to a central position in a uniform water phantom (25 x 25 x 25 cm3) it has been demonstrated that: (1) Multiple scatter contamination is an inherent limitation of the Compton-scatter method of densitometry which can be minimised, but not eliminated, by improving the energy resolution of the scattered radiation detector. (2) The choice of the incident photon energy is a compromise between the permissible radiation dose to the patient and the tolerable level of multiple scatter contamination. For a mean scattering angle of 40 degrees, the intrinsic multiple-single scatter ratio decreases from 64 to 35%, and the radiation dose (per measurement) increases from 1.0 to 4.1 rad, as the incident photon energy increases from 300 to 2000 keV. These doses apply to a sampled volume of approximately 0.3 cm3 and an electron density precision of 0.5%. (3) The forward scatter densitometer configuration is optimum, minimising both the dose and the multiple scatter contamination. For an incident photon energy of 1250 keV, the intrinsic multiple-single scatter ratio reduces from 122 to 27%, and the dose reduces from 14.3 to 1.2 rad, as the mean scattering angle decreases from 130 to 30 degrees. These calculations have been confirmed by experimental measurements. (author)