Automated Verification of IGRT-based Patient Positioning

Area of Science:

• Radiotherapy Quality Assurance
• Medical Physics
• Computational Optimization

Background:

• Ensuring accurate patient positioning is critical in radiotherapy.
• Current clinical software has automated registration but may require enhanced quality assurance.
• Therapist registration accuracy directly impacts treatment efficacy and safety.

Purpose of the Study:

• To design and clinically validate an automated quality assurance system for therapist registration in radiotherapy.
• To evaluate the system's performance in complementing existing clinical software.
• To compare the optimization capabilities of evolutionary algorithms versus deterministic methods for registration accuracy.

Main Methods:

• Development of a novel automated quality assurance system for radiotherapy.
• Clinical testing and validation of the designed system.
• Convergence analysis comparing evolutionary algorithms with deterministic optimizers for cost function analysis.

Main Results:

• The automated quality assurance system was successfully designed and tested in clinical practice.
• The system complements existing clinical software, enhancing algorithm configuration and performance.
• Evolutionary algorithms demonstrated superior performance in identifying global optima for the cost function.

Conclusions:

• The developed system provides a practical approach to ensuring safe patient setups in radiotherapy.
• Automated quality assurance is crucial for maintaining high standards in radiation oncology.
• Evolutionary algorithms offer a robust method for optimizing registration accuracy in complex treatment scenarios.