Investigation of Dosimetric Characteristics and Exploration of Potential Applications of Amorphous Silicon Detector

The ability of the electronic portal imaging device (EPID) to acquire a large two-dimensional array of digitized x-ray data in real time is extremely attractive for dosimetric measurements. To evaluate the potential use of an EPID for portal dose measurement in Wellington Blood and Cancer Centre, some dosimetric characteristics of the Varian's PortalVisionTM aS500 were investigated. PortalVisionTM incorporates an amorphous silicon detector (aSi). Some potential applications of EPID in linac QA were also explored. The EPID's performance for linearity with MU and dose rate was verified and it was found to be proportional over the entire measured range. Short term repeatability was found to be excellent. An investigation of calibration method to improve dosimetric accuracy demonstrated two methods of avoiding detector saturation. Firstly, acquiring flood field with the use of additional buildup and secondly, increasing the source to detector distance for calibration. A study of EPIDs behaviour under conditions of varying dose rate which commonly arise in EDW treatment techniques was carried out. The EPID exhibited a field size dependence in addition to a 8% discrepancy on the `hot edge' of EDW profiles. Further investigation into the field size dependence and the discrepancy at hot edge is required. EPIDs ability to acquire asymmetric field profile was also investigated. The profiles acquired using EPID deviated in shape and magnitude by upto 16% from the ion chamber profiles. Some potential applications of EPID to perform QA of linac beam properties, its ability to perform optical and mechanical linac QA have been explored. The EPID's capability to give constant output, flatness, symmetry, wedge angle and wedge factors with high level of accuracy and reproducibility was demonstrated. EPID was also found to be objective, efficient and feasible for performing optical linac QA. The use of EPID for linac QA could be simplified by improving the available software analysis tools thus making it more efficient