A proof-of-concept automated method for accurate skin dosimetry: correcting overestimated surface dose measurements

Abstract

Objective. Accurate surface dosimetry is recommended for detecting treatment error and managing skin toxicity. However, measurements using detectors such as optically stimulated luminescence dosimeters (OSLDs) may overestimate surface dose in the buildup region due to detector thickness. We developed a new automated framework to overcome these challenges and enhance dose correction accuracy. Approach. The proposed framework includes: (i) determining the detector's position and orientation on the patient's curved surface; (ii) calculating the beam incidence angle for each detector across all radiation beams using treatment planning data and the orientation of the detector; (iii) assessing detector placement relative to the radiation field, and the alignment within irregular subfield areas; (iv) computing the cumulative field contributions to the detector response across all fields; and (v) applying the detector correction factor, defined as the ratio of surface film measurements to detector readings, while considering beam incidence angle and field size. To demonstrate the proof-of-concept of the developed framework, it was evaluated across four treatment modalities using two OSLD types, nanoDot and myOSLchip. Performance was assessed by comparing the OSLD-measured surface dose with the framework-corrected surface dose. Main results. The framework effectively corrected OSLD response to reference skin dose, thereby enhancing the accuracy of skin dose estimation. In the anthropomorphic phantom study, the deviation of OSLD response from skin dose varied up to 21.48% for nanoDot and 46.63% for myOSLchip. With the proposed framework, the deviation was reduced to within 5% across all treatment modalities for both OSLD types, demonstrating a substantial improvement in measurement accuracy. Significance. This novel framework automates the correction of complex factors affecting surface dosimetry, mitigating measurement variability and improving accuracy and consistency. Its feasibility was demonstrated in a phantom-based, preclinical evaluation using breast radiotherapy treatment plans.