Master of Science (MS)


Physics and Astronomy

Document Type



Introduction: In order to maintain uniform standards in the accuracy of fractionated radiation therapy, quantification of the delivered dose per fraction accuracy is required. The pupose of this study was to investigate the feasibility of a transit dosimetry method using the electronic portal imaging device (EPID) for dose delivery error detection and prevention. Methods and Materials: In the proposed method, 2D predicted transit images were generated for comparison with online images acquired during treatment. Predicted transit images were generated by convolving through-air EPID measurements of each field with pixel-specific kernels selected from a library of pre-calculated Monte Carlo pencil kernels of various radiological thicknesses. The kernel used for each pixel was selected based on the calculated radiological thickness of the patient along the line joining the pixel to the virtual source. The accuracy of the technique was evaluated in homogeneous plastic water phantoms, a heterogeneous cylindrical phantom, and an anthropomorphic head phantom. Gamma analysis was used to quantify the accuracy of the technique for the various cases. Results: In the comparison between the measured and predicted images, an average of 99.4% of the points in passed a 3%/ 3 mm gamma for the homogeneous plastic water phantoms. Points for the heterogeneous cylindrical phantom analysis had a 94.6% passing rate. For the anthropomorphic head phantom, an average of 98.3% and 96.6% of points passed the 5%/3mm and 3%/3mm gamma criteria, respectively for all field sizes. Failures occurred typically at points when object thickness was changing rapidly or at boundaries between materials, and at the edges of large fields. Discussion: The results suggested that the proposed transit dosimetry method is a feasible approach to in vivo dose monitoring. The gamma analysis passing rates are within the accuracy needed for transit dosimetry. Future research efforts should include evaluation of the method for more complex treatment techniques and assessment of the sensitivity to changes in EPID or linac hardware, as well as characterization of any dependency the method may have on image ghosting or lag, gantry angle, or long-term stability.



Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Fontenot, Jonas