Automated Anatomy-Driven Patient-Specific Silicone Bolus Fabrication In Esapi with Phantom-Based Validation
Abstract
Purpose
To develop and validate a fully automated, anatomy-driven workflow for patient-specific silicone bolus fabrication using the Eclipse Scripting API (ESAPI), and to evaluate geometric and dosimetric accuracy using phantom-based post-fabrication CT imaging and dose measurements.
Methods
An ESAPI-based application was developed to generate two-piece, 3D-printable bolus molds directly within the treatment planning system. Mold separation is determined automatically using the spatial relationship between the bolus structure and the external contour, allowing the workflow to adopt to anatomical geometry without user-defined parting planes. This automation is enabled by a Python-based algorithm integrated into the ESAPI environment. The resulting inner and outer mold shells were exported as STL files and fabricated using polylactic acid (PLA) material. Silicone rubber was subsequently cast within the PLA molds to produce the final patient-specific bolus. Mold geometry was generated using Boolean operations between the bolus, external body, and intermediate expanded structures. Geometric accuracy was evaluated using anthropomorphic phantoms. Repeat CT imaging was acquired after the silicone bolus was placed on the phantom. The original bolus structure was registered to the post-fabrication CT. Surface deviation metrics included the mean surface deviation of 0.31 mm, maximum deviation of 0.51 mm, and 95.4% of surface points within 0.5 mm tolerance. Silicone thickness differed from the planned thickness by 0.3 mm on average. Dosimetric validation was performed at the skin surface and at clinically relevant depths across multiple photon and electron energies, demonstrating surface dose difference of ± 3% and depth-dose agreement within 1%
Results
The automated workflow produced anatomically conformal bolus geometry across nine phantom cases without manual adjustment. Improved bolus conformity reduced air gaps and yielded consistent surface and depth dose agreement.
Conclusion
This study demonstrates a fully automated ESAPI workflow for fabricating patient-specific silicone bolus using 3-D printing technology. Improved geometric conformity supports accurate dose delivery.