Optimal stochastic correction strategies for rigid-body target motion

Area of Science:

• Medical Physics
• Radiotherapy
• Image-Guided Interventions

Background:

• Image-guided radiotherapy (IGRT) enhances targeting accuracy during treatment.
• Future IGRT devices will enable frequent corrections of patient setup errors.
• Developing new correction protocols to minimize systematic and random setup errors is crucial.

Purpose of the Study:

• To derive optimal correction strategies for setup errors, considering measurement uncertainties.
• To analyze the impact of these strategies on treatment margins.
• To develop a general approach for optimal correction strategies using linear systems theory.

Main Methods:

• Utilized linear systems theory to develop optimal correction strategies for setup errors.
• Employed a Kalman filter to estimate patient spatial displacement from noisy measurements.
• Implemented an adaptive Kalman filter for unknown systematic and random error variances.
• Assessed filter performance through simulations on individual patients and a large population.

Main Results:

• Kalman filter corrections were consistently advantageous over using measured values alone across a patient population.
• Reduced residual standard deviation of random setup errors by approximately 28% for over 90% of patients with significant measurement error.
• Measurement uncertainty partially limits the complete mitigation of setup uncertainties.

Conclusions:

• Kalman filter estimates offer an effective method for daily setup corrections in the presence of measurement errors.
• The linear system approach is versatile and adaptable for more complex state variables in radiotherapy.
• Optimized setup corrections improve treatment accuracy and safety in image-guided radiotherapy.