Accuracy and precision of absorbed dose measurements for neutron therapy.

The occurrence of normal tissue complications and probability of tumor control are steep functions of absorbed dose. Consequently the delivery of the dose to the patient should be performed with a precision better than +/- 2% and an overall uncertainty less than +/- 5%. The sequence of dosimetry procedures to deliver the absorbed dose to the patient is analyzed with emphasis on the physical parameters involved in neutron dosimetry; the results of neutron dosimetry intercomparisons are summarized. The protocols for neutron dosimetry formulated by European and American physicists differ in a number of aspects, including the choice of the phantom material. For the treatment of a specific lesion, e.g., a tumor of the floor of the mouth, different treatment plannings have been suggested. Regarding the determination of total absorbed dose at a reference point in a phantom, the required overall uncertainty can be achieved for neutron energies up to 20 MeV. Because of differences in size, shape and composition between the phantom and the patient, somewhat larger uncertainties are to be anticipated for the actual treatment. Further experimental and theoretical studies are needed to obtain more reliable values for kerma in different elements and neutron sensitivity of the photon dosimeters for neutron energies in excess of 20 MeV.