ABSTRACT
This chapter outlines the limits and differential advantages of analytical versus Monte Carlo (MC) codes with increased complexity of physics engines making their way from purely research oriented efforts to fast-paced clinical environments, particularly for proton beams with the advent of GPU-ported MC codes. Regardless of the optimization algorithm, the dose influence matrix determines the overall accuracy of the dose optimization. Consequently, errors and uncertainties that derive from the dose calculation carry over to the dose optimization itself and directly affect the patient treatment. Analytical approaches to the pencil beam algorithm will handle heterogeneity effects on the dose distribution with a varying degree of accuracy. Performance of these models are dependent on location and severity of the heterogeneity. Adaptive radiotherapy is a combination of online dose calculation with fast dose optimization. Similar to online calculation, the machine delivery parameters are fed to the calculation engine that calculates the dose in real time.
