Automated verification of source dwell position in lunar ovoid applicators

Accurate reconstruction of the source dwell positions in high-dose-rate (HDR) brachytherapy is essential for treatment quality. The Venezia™ gynaecological applicator, with its lunar-shaped ovoids, poses challenges because conventional dummy markers do not match the actual source path, introducing systematic uncertainties.

The aim of this work was to develop and demonstrate an automated, reproducible, and open-source software for autoradiographic verification of dwell positions in Venezia lunar ovoids, and to compare measured positions with manufacturer-provided applicator library data. This pipeline established a standardized approach for clinical commissioning and quality assurance (QA) workflows.

A custom 3D-printed holder was designed to standardize autoradiograph acquisition using radiochromic films. An open-source Python pipeline, referred to as SPICA (Source Position Identification in Circular Applicators), was developed to process scanned films, isolate dwell position, and compute their coordinates with associated uncertainties. The code allows direct comparison between measured positions and manufacturer-provided applicator library data. We performed preliminary validation of the software by measuring the trajectory of an Ir-192 source for the 22 and 26 mm ovoids with a step size of 2.5 mm.

Mean discrepancies between measured and reference positions were 1.2 ± 0.5 and 1.5 ± 0.4 mm for the 22 and 26 mm ovoids, respectively, with a maximum deviation of 2.6 mm. Distal positions showed better agreement. These findings confirm that, in the tested applicator-afterloader combination, vendor libraries provide a generally reliable baseline but may exhibit systematic differences.

The proposed methodology offers a practical, reproducible solution to improve source path verification while reducing manual intervention. It supports more robust commissioning and QA of complex applicators and using the coordinates calculated in the treatment planning software can potentially reduce geometric uncertainties in clinical cases. Individualized commissioning remains essential for applicators with complex geometries.