RPC Experience with TLD for Output and Energy Monitoring of Radiation Therapy Beams

The Radiological Physics Center (RPC) has been utilizing TLD to verify photon-beam output, and electron-beam output and energy for many years. The RPC currently monitors over 1200 institutions, monitoring 4000 photon and 3500 electron beams per year. Control TLD, irradiated on one specific Co-60 machine, are used for a performance check of our TLD system. Analysis of these data over 5 years, reveals high precision (SD =0.9%) in beam output verification. Analysis of all TLD results, since 1990, for remotely monitored photon beams (27,900) and electron beams (23,000), shows a spread (SD) of 1.9% and 2.2% respectively in beam output. The increased spread arises from the variability in beam energies, makes/models of machines, and institutional performance. In view of these variabilities, the results are extremely encouraging. Institutional performance includes uncertainties in beam-output calibration, setup errors, and beam drifts. The Spread (SD) of individual beams varies widely from beam to beam and Institution to institution. The spread of individual beams has been used to identify “good” beams (SD < 2%) which are to receive TLD less frequently than others. This work is supported in part by PHS grant CA10953 awarded by the NCI, DHHS. Introduction Thermoluminescent Dosimetry is a well-established technique whose scientific basis and its use for irradiation dosimetry is well documented in the scientific literature. It is widely used for in-vivo and experimental dosimetry for complex geometries. The Section of Outreach Physics at MDACC has been using TLD as a remote-monitoring tool for more than 30 years. TLDs are available in the form of rods, chips and powder. The characteristics are basically the same irrespective of the form or shape of the TLD material. The technique for LiF TLD rods or chips has been detailed in other publications. Kirby et. al. (1986) have discussed the use of powder for a mailable TLD system for monitoring calibrations of photon beam energies for Cobalt 60 to 25M and electron energies from 6 to 20 MeV. An uncertainty analysis by Kirby et. al. (1992) for absorbed dose calculations, showed that a ±2% dose uncertainty and ±5% acceptance criteria for TLD dosimetry was possible. This work presents analysis of TLD results from both, the RPC's internal quality assurance and from the users over past 8 years. The former shows a precision of 1% under irradiations with high-precision set up. This implies that the RPC can implement a tighter quality-assurance criterion for beam output. The analysis of history of TLD results for individual beams of users has led to a promising tool to identify two types of potential problems. First, a systematic departure of the G:\USERS\COMMON\amanda\RTailorAAPMPoster.doc multiple TLD results from the expected value indicates a potential problem in beam-output calculation or electron depth dose data. Second, a significant variability of multiple TLD results indicates potentially poor quality-assurance window for beam. The action levels suggested by the data are presented. Materials and Methods • TLD Powder, about 20 mg per sample, disposed of after one use. • TLD readers: Harshaw 3500 and Teledyne 7300; Dry Nitrogen Flushes • TLD are massed with Metler Balances (accuracy of 0.1 mg) • The signal (T) is therefore: TLD reading per unit mass (volt/mg) • TLD irradiation system: Lucite "miniphantom" for photons, "full" phantom for electrons.* • Photons, Dose at dmax only; Electrons, Dose at dmax and energy check • Standards, irradiated on Cobalt 60, are used for system calibration of every session. • Controls, irradiated on another Cobalt unit, are used to monitor stability of the reader throughout the session. Controls also provide a redundant calibration check. • Each Batch of TLD (approximately 100,000 samples) is commissioned prior to use: ü Reproducibility ü Linearity ü Fading ü Energy/Phantom dependence • Quality Assurance Procedures: ü Response of Controls and Standards throughout the session. (Drift is flagged automatically) ü Balance is monitored throughout the session. (standard mass before and after session) ü Glow curve shape is monitored. ü Control response is tracked over time. ü Other tests include: dark current, standard light, record of faults. ü Outliers are evaluated. Concepts and Definitions Standards: TLD are irradiated on a specific Cobalt 60 Irradiator, one set at a time, in the Cobalt “miniphantom”, in the center of a 10cm x 10cm field. One set of Standards is read at the start of a session and one at the end of the session (approximately 18 samples per session). The responses of the two sets are compared to determine reader drift. Controls: TLD are irradiated on a second Cobalt 60 irradiator, 60 samples at a time, in a rotating jig. 4 sets of control TLD are read spread throughout the reading session. Each control TLD is compared with the initial control and outliers are flagged by the computer so additional controls can be read if necessary. Controls monitor for reader drift, and serve as a redundant check of calibration. System Sensitivity (S): Dose per unit TL signal corrected for fading and linearity, calculated from the average response of the standards. (See Kirby et.al.)