- In summary, it is feasible to construct a neutron source, using the 3H(1H,n)3He reaction and the Dynamitron accelerator, which is suitable for the measurement of aluminium in vivo. It is useful to compare this system with other published data for in vivo aluminium measurement. The important parameters are presented in table 2 where MDL represents the minimum detectable level for the stated local dose. The figures in table 2 are very encouraging since the system described in this paper is at a very early stage of development. Significant improvements in sensitivity (detected 28Al counts per unit dose) should be expected from the combined effects of: (i) Increased NaI detector sizes. (ii) Neutron transport calculations to optimise the irradiation cavity. (iii) Experiments to determine the optimum source energy more precisely. (iv) Better dose delivery (giving the same dose in a shorter time). (v) More sophisticated spectroscopic data reduction. Published data on the amounts of aluminium typically found in hands of patients with renal disease (Ellis and Kelleher 1987) show that aluminium levels are likely to be in the 0.3 to about 15 mg range. Any improvements in system performance could largely be used to reduce the patient dose since the measurement sensitivity reported here is already close to that required for meaningful clinical measurements.