Total body nitrogen (TBN) is measured by in vivo prompt gamma neutron
capture analysis. Usually the background under the nitrogen peak is su
btracted before calculating TEN from the ratio of nitrogen counts (at
10.8 MeV) to hydrogen counts (at 2.2 MeV). The hydrogen acts as an int
ernal standard. The background under the hydrogen peak is usually igno
red. The current study was undertaken to measure the variation of the
hydrogen background (HB) in patients of different sizes and to determi
ne whether the accuracy of TBN measurements is improved when an HB cor
rection is incorporated. Heavy-water (D2O) phantoms were used to quant
ify patient HB. D2O simulates patient neutron scattering without contr
ibuting to the hydrogen peak. Equations were developed to predict HB i
n patient measurements. HB was found to vary from 9% of total counts f
or a 99 kg subject to 19% for a 7 kg infant. It was demonstrated that
the accuracy of TBN measurements is improved for child-sized anthropom
etric phantoms of known composition when an HB correction is incorpora
ted.