Validation of a detector-response correction model for in vivo dosimetry in high-dose-rate prostate brachytherapy.

BACKGROUND: High-dose-rate brachytherapy (HDRBT) is an established treatment for localized prostate cancer. Despite its efficacy, HDRBT can exhibit dosimetric discrepancies between planned and delivered doses. In vivo dosimetry (IVD) using metal-oxide-semiconductor field-effect transistors (MOSFETs) enables real-time monitoring of administered doses, though factors such as temperature, source-to-detector distance, and detector orientation can affect accuracy. Appropriate correction models may reduce these uncertainties. PURPOSE: To evaluate a mathematical correction model for microMOSFET response dependencies under in vivo conditions during HDRBT with 192Ir in prostate cancer patients. METHODS: The model, correcting for temperature, source-detector distance, and angular dependence, was applied to IVD data obtained during 29 HDRBT implant procedures. Corrected and uncorrected detector-measured doses were compared with treatment planning system (TPS) calculations at the neurovascular bundle and periurethral zone. RESULTS: The full correction reduced the mean dose difference between measured and TPS doses from -5.9% to 0.7%. Omitting angular correction yielded a -2.1% discrepancy. For patients with two detectors in distinct regions, the average difference was 0.2% (SD 5.2%). Monte Carlo simulations confirmed that combining measurements from multiple positions improves dosimetric precision and provides quantitative support for a ±10% investigation level for detecting treatment delivery errors. CONCLUSIONS: Applying the correction model to microMOSFET-based IVD in prostate HDRBT significantly improved agreement with TPS-calculated doses, particularly when two measurement points were used. Angular dependence was a primary source of residual error. Dual-point Monte Carlo simulations further supported the robustness of the proposed approach.