The purpose of this study was to estimate the CTV-PTV margin required for prostate and head and neck cancer treatments at the radiotherapy departments of Karolinska University Hospital.    Portal image data from patients treated at the radiotherapy departments during the period of 2009-2011 was used to estimate the set-up displacements for each treatment area. By using the acquired images the magnitude of the systematic, i.e. preparatory, and random, i.e. execution, error was determined in the anterior-posterior (AP), superior-inferior (SI) and right-left (RL) direction. The calculated PTV margin is based on the systematic and random errors of the entire patient populations. A total of 40 patients were used for the analysis of prostate treatments and 47 patients for head and neck treatments. The evaluation of the PTV margin was done for three different matching protocols; no matching (skin marker alignment), five day matching and daily matching.      With no image verification in prostate treatments the calculated PTV margin taking both inter- and intrafractional errors into account was 13.6, 9.2 and 7.9 mm in AP, SI, and RL direction respectively. The corresponding PTV margin in head and neck treatments was found to be 6.7, 5.3 and 4.9 mm. Using a five day matching protocol of the bony anatomy showed no considerable reductions in margins for neither prostate nor head and neck treatments. With daily matching of the bony anatomy in prostate treatments the calculated margins was reduced to 8.1, 7.9 and 2.4 mm in the AP, SI and RL direction respectively.  Measurements of the residual deviations of individual cervical vertebrae after daily image verification and correction in head and neck cancer treatments showed that all matching protocols will require larger margins in the lower vertebrae in order to account for the set-up error in the AP direction. The corresponding margins needed using daily matching of the bony anatomy would be 3.9, 5.4 and 6.0 mm for C1, C4 and C5 respectively in the AP direction.    In the absence of daily imaging the currently used PTV margins might be inadequate for covering to movement of the targets. The deviations in the AP direction of the cervical vertebrae in head and neck cancer treatments should be investigated further in order to ensure that the motion of the target is covered and that no risk organs are subjected to harmful dose levels.

External Beam Radiation Therapy (EBRT) is a common treatment option for patients with prostate cancer. When treating the prostate with EBRT, a geometric volume (PTV margin) is added around the prostate to account for uncertainties in treatment planning and delivery. Current methods for estimating PTV margins rely on the analysis of population-based inter- and intra-fraction motion data. These methods do not consider the patient-to-patient differences in demographic or clinical presentation of patient specific factors (PSFs), such as age, weight, body-mass index, health and performance status, prostate-specific antigen levels, Gleason scores, presence of bowel problems, or other health conditions. The purpose of this thesis is to investigate the feasibility using regression-based predictive algorithms to predict the extent of prostate motion for the purpose of personalizing the PTV margin using PSFs as inputs. Benchmarking simulations of Linear, Ridge, LASSO, SVR, kNN, and MLP algorithms were performed by simulating prostate intra-fraction motion and realistic variations in PSFs. Sample sizes ranged from n=20 to 800, with varying levels of noise into the motion data (0-10mm). Leave-one-out cross validation was used to train and validate algorithm performance. The results suggest that algorithm performance improves significantly within the first 50 – 100 patients, and this rate of improvement is independent of noise in prostate motion. The Ridge regression algorithm predicted intra-fraction motion to the lowest mean absolute error in simulated motion, performing especially well in small datasets. To evaluate the clinical utility of this approach, pre- and post-treatment prostate motion data, treatment time data, and rectal distension data was recorded in 21 patients, along with a variety of PSFs. In the analysis of patient data, the LASSO algorithm out-performed the Ridge algorithm, predicting the mean and standard deviation of an individual prostate cancer patient’s intra-fraction motion to within 0.8mm and 0.4 mm mean absolute error, respectively. However, prostate motion predictions did not correlate with PSFs, possibly due to the small sample size. This work demonstrates the feasibility of using regression-based algorithms for predicting prostate motion, and hence the opportunity to personalize PTV margins in prostate cancer patients. / Graduate / 2018-08-22