Recent developments of new therapy techniques using small photon beams, such as stereotactic radiotherapy, require suitable detectors to determine the delivered dose with a high accuracy. The dosimeter has to be as close as possible to tissue equivalence and to exhibit a small detection volume compared to the size of the irradiation field, because of the lack of lateral electronic equilibrium in small beam. Characteristics of single crystal diamond (tissue equivalent material Z = 6, high density) make it an ideal candidate to fulfil most of small beam dosimetry requirements. A commercially available Element Six electronic grade synthetic diamond was used to develop a single crystal diamond dosimeter (SCDDo) with a small detection volume (0.165 mm(3)). Long term stability was studied by irradiating the SCDDo in a (60)Co beam over 14 h. A good stability (deviation less than ± 0.1%) was observed. Repeatability, dose linearity, dose rate dependence and energy dependence were studied in a 10 × 10 cm(2) beam produced by a Varian Clinac 2100 C linear accelerator. SCDDo lateral dose profile, depth dose curve and output factor (OF) measurements were performed for small photon beams with a micro multileaf collimator m3 (BrainLab) attached to the linac. This study is focused on the comparison of SCDDo measurements to those obtained with different commercially available active detectors: an unshielded silicon diode (PTW 60017), a shielded silicon diode (Sun Nuclear EDGE), a PinPoint ionization chamber (PTW 31014) and two natural diamond detectors (PTW 60003). SCDDo presents an excellent spatial resolution for dose profile measurements, due to its small detection volume. Low energy dependence (variation of 1.2% between 6 and 18 MV photon beam) and low dose rate dependence of the SCDDo (variation of 1% between 0.53 and 2.64 Gy min(-1)) are obtained, explaining the good agreement between the SCDDo and the efficient unshielded diode (PTW 60017) in depth dose curve measurements. For field sizes ranging from 0.6 × 0.6 to 10 × 10 cm(2), OFs obtained with the SCDDo are between the OFs measured with the PinPoint ionization chamber and the Sun Nuclear EDGE diode that are known to respectively underestimate and overestimate OF values in small beam, due to the large detection volume of the chamber and the non-water equivalence of both detectors.

The realization of the unit of absorbed dose at LNE-LNHB is based on calorimetry with the present GR8 graphite calorimeter. For this reason the calorimetric technique must be maintained, developed and improved in the laboratory. The usual quasi-adiabatic operating mode at LNHB is based on the thermal feedback between the core (sensitive element) and the jacket (adjacent body). When a core-jacket temperature difference is detected, a commercially available analogue PID (Proportional, Integral, Derivative) controller sends to the jacket an amount of electrical power to reduce this difference. Nevertheless, the core and jacket temperatures increase with irradiations and electrical calibrations whereas the surrounding is maintained at a fixed temperature to shield against the room temperature variations. At radiotherapy dose rates, fewer than ten measurements, or electrical calibrations, per day can be performed. This paper describes the new constant-temperature operating mode which has been implemented recently to improve flexibility in use and, to some extent, accuracy. The core and the jacket temperatures are maintained at fixed temperatures. A steady state is achieved without irradiation. Then, under irradiation, the electrical power needed to maintain the assigned temperature in the core is reduced by the amount of heat generated by ionizing radiation. The difference between these electrical powers, without and with irradiation, gives the mean absorbed dose rate to the core. The quality of this electrical power substitution measurement is strongly dependent upon the quality of the core and jacket thermal control. The core temperature is maintained at the set value using a digital PID regulator developed at the laboratory with LabView software on PC for this purpose. This regulator is versatile and particularly well suited for calorimetry purposes. Measurements in a cobalt-60 beam have shown no significant difference (<0.09%) between the two operating modes, with an equivalent reproducibility (1sigma < 0.06%). These results corroborate the negligible difference of heat transfer between steady and irradiation periods when working in quasi-adiabatic mode with thermal feedback between the core and the jacket. The new constant-temperature mode allows numerous and fully automated measurements. The electrical calibration is an integral part of the measurement; no extra runs are needed. It also allows faster thermal equilibrium before starting runs. Moreover the quality of vacuum within the gaps between the bodies is less important.

To extend the dosimetric reference system to field sizes smaller than 2 cm × 2 cm, the LNE-LNHB laboratory is studying an approach based on a new dosimetric quantity named the dose-area product instead of the commonly used absorbed dose at a point. A graphite calorimeter and a plane parallel ion chamber with a sensitive surface of 3 cm diameter were designed and built for measurements in fields of 2, 1 and 0.75 cm diameter. The detector surface being larger than the beam section, most of the issues linked with absolute dose measurements at a point could be avoided. Calibration factors of the plane parallel ionization chamber were established in terms of dose-area product in water for small fields with an uncertainty smaller than 0.9%.