Background:
Patient specific quality assurance measurements remain a crucial part of modern intensity modulated static and rotational radiotherapy (IMRT/VMAT) due to the high complexity of the treatment plans.
Recalculating patient plans on a single phantom geometry and comparing the planned dose with the delivery using detector arrays is the gold standard of the patient specific QA procedure.
Evaluation of patient specific QA includes gamma-passing rate, which takes into account spatial and dose accuracy at the same time. This inevitably depends on the geometrical design of the detector array including the grid resolution and the detectors position.
Goal:
In this master thesis proposal the core initiative is to evaluate correspondence between three different detector array (“X”, “+” and “-“ shapes) and evaluate various gamma conditions for the clinical applications (normo- or hypo-fractionation, large or small stereotactic treatment volumes).
Depending on the candidate, this subject can elaborated further into one of the following three orientations:
- Predictive analysis on plan complexity – (Analytical, statistical approach)
Using cohort based analysis, determine the correlation between plan complexity indexes and various gamma-passing conditions. Perform a multivariate analysis (optimally machine learning), create and validate a predictive nomogram.
- Predictive analysis on intensity distribution – (Phantom performance evaluation)
Perform an analysis of large number of VMAT plans based on the dominant dose/intensity distribution (Antero-posterior, medio-lateral, oblique) and evaluate the sensitivity of the three arrays with regards to the different groups.
- Predictive analysis on geometrical challenges – (Clinical, geometrical analysis)
Evaluate the plan based on the target volume’s shape and the most proximity organ at risks (OARs), and determine the differences between detector arrays for measuring the steepest dose gradient around the target.
Promoters:
Akos Gulyban akos.gulyban@bordet.be
Nick Reynaert nick.reynaert@bordet.be
Olivier Debeir odebeir@ulb.ac.be
Nicolas Pauly Nicolas.Pauly@ulb.be