This work describes the design characteristics of a medical imaging centre which uses positron emission tomography, with a cyclotron for fluorine-18 and nitrogen-13 production, and which has provided experimental information on operational data recorded by area dosimetry since 1995. Doses to radiopharmacy and medical staff have been measured both in normal work and in some handling incidents. Data on radiation levels in the installation have also been obtained and related to design details and shielding. Area dosimetry was carried out using a five-stationary detector network, with a sampling rate of 2 min(-1), and by thermoluminescent dosimetry (TLD). Staff were also monitored by TLD, using extra chips for finger dosimetry and to duplicate individual whole-body dosimetry in order to measure doses in certain single operations. For normal work, average whole-body doses to radiopharmacy staff were between 0.03 and 0.28 mSv/month, wrist doses were between 0.42 and 2.67 mSv/month, and finger doses were between 1.4 and 7.7 mSv/day for the left hand and 0.8 and 2.4 mSv/day for the right hand; such variation reflects the differing expertise of staff and the role played by optimisation. Finger doses between 16 and 131 mSv were measured in handling incidents, and finger doses of 20.2 and 20.7 mSv for the left hand and 22.0 and 22.3 mSv for the right hand were measured during handling of a syringe without shielding, containing 3 GBq. For medical staff, contributions to the whole-body dose of 2.0 and 1.9 microSv/procedure were measured for injection and placing the patient on the examination couch, respectively. Dose measurement on the middle finger of the right hand gives an average of 70 microSv during the injection. The provisions regarding the shielding design have proved to be adequate and effective during a 3-year operational period. Operational doses to medical staff are comparatively low, while radiopharmacy staff are the most exposed. The finger doses in these professionals may exceed the annual limit, unless operational restrictions in daily practice are adopted. On-line area dosimetry records based on dose rate probes have proved to be effective both for monitoring radiation levels during the operation and for detecting changes in the behaviour of the facility in the irradiation process.

To develop a practical means of auditing the procedures and to optimize the administered radionuclide activity, patient surface radiation doses following the administration of 18FDG for PET imaging have been measured at breast and gonad locations at two PET facilities. Patient dosimetry was performed using LiF TLD-100 chips placed near the breasts and gonads for 2 h following tracer injection. Standard uptake values from the image-reconstruction algorithms of one PET camera were investigated in regions of interest in tomograms of the myocardium and liver with the aim of validating dosimetry at breasts. Mean doses measured on the patient's skin ranged between 3.3 and 6.1 microGy MBq(-1) at the gonads and between 3.9 and 6.4 microGy MBq(-1) at the breasts, noticeably lower than the calculations reported in the literature. These values show good concordance with the injected activities, although they are not proportional. The proportion of injected activity actually contributing to image production seems to decrease gradually as the injected activity increases. Conversely, for a given injected activity, breast and gonadal doses were found to be lower than the values expected from the numerical calculations reported in the literature, showing increasing discrepancies when the injected activity increased. Doses measured at the right breast were consistently higher than for the left, which is indicative of greater radiotracer absorption by the liver compared to the average absorption in the body.

No abstract