Finding the optimal statistical model to describe target motion during radiotherapy delivery--a Bayesian approach.

Early approaches to characterizing errors in target displacement during a fractionated course of radiotherapy assumed that the underlying fraction-to-fraction variability in target displacement, known as the 'treatment error' or 'random error', could be regarded as constant across patients. More recent approaches have modelled target displacement allowing for differences in random error between patients. However, until recently it has not been feasible to compare the goodness of fit of alternate models of random error rigorously. This is because the large volumes of real patient data necessary to distinguish between alternative models have only very recently become available. This work uses real-world displacement data collected from 365 patients undergoing radical radiotherapy for prostate cancer to compare five candidate models for target displacement. The simplest model assumes constant random errors across patients, while other models allow for random errors that vary according to one of several candidate distributions. Bayesian statistics and Markov Chain Monte Carlo simulation of the model parameters are used to compare model goodness of fit. We conclude that modelling the random error as inverse gamma distributed provides a clearly superior fit over all alternatives considered. This finding can facilitate more accurate margin recipes and correction strategies.

Daily online bony correction is required for prostate patients without fiducial markers or soft-tissue imaging.

AIM: To compare online position verification strategies with offline correction protocols for patients undergoing definitive prostate radiotherapy. MATERIALS AND METHODS: We analysed 50 patients with implanted fiducial markers undergoing curative prostate radiation treatment, all of whom underwent daily kilovoltage imaging using an on-board imager. For each treatment, patients were set-up initially with skin tattoos and in-room lasers. Orthogonal on-board imager images were acquired and the couch shift to match both bony anatomy and the fiducial markers recorded. The set-up error using skin tattoos and offline bone correction was compared with online bone correction. The fiducial markers were used as the reference. RESULTS: Data from 1923 fractions were analysed. The systematic error was ≤1 mm for all protocols. The average random error was 2-3mm for online bony correction and 3-5mm for skin tattoos or offline-bone. Online-bone showed a significant improvement compared with offline-bone in the number of patients with >5mm set-up errors for >10% (P<0.001) and >20% (P<0.003) of their fractions. CONCLUSIONS: Online correction to bony anatomy reduces both systematic and random set-up error in patients undergoing prostate radiotherapy, and is superior to offline correction methods for those patients not suitable for fiducial markers or daily soft-tissue imaging.

Implementation of an image-guided radiation therapy program: Lessons learnt and future challenges.

The aim of this paper is to detail the experience obtained in implementing an image-guided radiation therapy program at the Northern Sydney Cancer Centre. This required retrofitting a Varian Clinac 21EX with an on-board imager. The commissioning and quality assurance procedures, organisation of a multidisciplinary image guided radiation therapy group, and the development of clinical protocols for orthogonal kV and cone beam computed tomography implementation are described. Reassessment of the image-guided radiation therapy program has continued as new equipment and software versions were made available in the department.