Automated Treatment Plan Integrity Verification for Proton Therapy: Enhancing Error Detection and Workflow...

Area of Science:

• Medical Physics
• Radiation Oncology
• Quality Assurance

Background:

• Automatic error detection in radiation therapy predominantly focuses on photon therapy, neglecting proton therapy.
• Proton therapy offers advantages but requires robust quality assurance for its complex delivery techniques like pencil-beam scanning.

Purpose of the Study:

• To develop and validate an automated framework, Auto-IniCheck, for initial proton therapy plan review.
• To improve error detection capabilities and workflow efficiency in pencil-beam scanning proton therapy.

Main Methods:

• Designed Auto-IniCheck following AAPM TG-275 guidelines for systematic physics plan integrity verification.
• Integrated site- and modality-specific checklists into a process map with color-coded risk levels.
• Utilized failure modes and effects analysis to determine risk levels and assessed efficiency via risk priority number reduction.

Main Results:

• Auto-IniCheck automated 60-65% of checklist items, reducing plan review time to under 10 seconds.
• Achieved an average 65% risk priority number reduction, reclassifying 15 high-risk items to low risk.
• Demonstrated greatest risk reduction (75%) in robustness optimization, couch confirmation, and target validation.

Conclusions:

• Auto-IniCheck streamlines proton therapy plan reviews, enhancing both efficiency and safety.
• The automated system proactively addresses proton-specific failure modes for practical initial plan review.
• Offers a scalable solution for maintaining rigorous quality assurance as proton therapy programs expand.