Accounting for TLD response to CBCT protocols in external beam radiotherapy dose monitoring

Abstract

Purpose: LiF thermoluminescent dosimeters (TLDs) are commonly used in radiation therapy to verify the delivered dose. Examples include dose verification for complicated treatment setups or cardiovascular implantable electronic devices (CIEDs). TLDs may be present for both the pre-treatment imaging kilovoltage (kV) beam and the megavoltage (MV) treatment beams. For low energy x-ray beams, where the photoelectric effect dominates, TLDs respond differently than tissue or water. An overresponse of up to 40% has been previously reported for lower-energy (kV) x-rays when calibrated to higher-energy (MV) beams. In this work, the response of TLDs to various cone beam computed tomography (CBCT) protocols with calibrations in clinical therapy beams (MV) is quantified. Methods: Three Varian OBI (Head, Thorax, and Pelvis) and three Elekta XVI (Fast Head and Neck, Fast Chest, and Fast Pelvis) CBCT protocols were investigated. For each protocol, TLDs were positioned on a tissue equivalent phantom at various distances extending out from the center of the imaging field. The response was determined by calibrating TLDs to a 6 MV photon beam. Results: The maximum in-field TLD response ranged from 0.82 to 4.80 and 0.06 to 2.69 cGy for Varian OBI and Elekta XVI protocols, respectively. The out-of-field CBCT response dropped exponentially from the field edge. Calculated uncertainties were generally less than 3% (k = 1), with exceptions along the edge of the CBCT field (6%) and at the most distal TLD positions (34%). Conclusions: Using the measured TLD responses to CBCT protocols with an MV calibration, the therapeutic dose can be isolated. The therapeutic dose can then be compared to predictions from the treatment planning system (TPS), allowing for more accurate dose verification for complex treatment setups and patients with CIEDs. The CBCT response can change the reported therapeutic dose by up to 2.5%.