A two step algorithm to predict portal dose images in arbitrary detector systems has been developed recently. The current work provides a validation of this algorithm on a clinically available, amorphous silicon flat panel imager. The high-atomic number, indirect amorphous silicon detector incorporates a gadolinium oxysulfide phosphor scintillating screen to convert deposited radiation energy to optical photons which form the portal image. A water equivalent solid slab phantom and an anthropomorphic phantom were examined at beam energies of 6 and 18 MV and over a range of air gaps (approximately 20-50 cm). In the many examples presented here, portal dose images in the phosphor were predicted to within 5% in low-dose gradient regions, and to within 5 mm (isodose line shift) in high-dose gradient regions. Other basic dosimetric characteristics of the amorphous silicon detector were investigated, such as linearity with dose rate (+/- 0.5%), repeatability (+/- 2%), and response with variations in gantry rotation and source to detector distance. The latter investigation revealed a significant contribution to the image from optical photon spread in the phosphor layer of the detector. This phenomenon is generally known as "glare," and has been characterized and modeled here as a radially symmetric blurring kernel. This kernel is applied to the calculated dose images as a convolution, and is successfully demonstrated to account for the optical photon spread. This work demonstrates the flexibility and accuracy of the two step algorithm for a high-atomic number detector. The algorithm may be applied to improve performance of dosimetric treatment verification applications, such as direct image comparison, backprojected patient dose calculation, and scatter correction in megavoltage computed tomography. The algorithm allows for dosimetric applications of the new, flat panel portal imager technology in the indirect configuration, taking advantage of a greater than tenfold increase in detector sensitivity over a direct configuration.

The relationship between the pixel value and exit dose was investigated for a new commercially available amorphous silicon electronic portal imaging device. The pixel to dose mapping function was established to be linear for detector distances between 116.5 cm to 150 cm from the source, radiation field sizes from 5 x 5 cm2 to 20 x 20 cm2 and beam energies of 6 to 18 MV. Coefficients in the mapping function were found to be dependent on beam energy and field size. Open and wedged field profiles measured with the device showed agreement to a maximum of 5% and 8%, respectively, as compared to film. A comparison of relative transmission measurements between the EPID and ion chamber indicate a maximum deviation of 6% and 2% at 6 and 18 MV, respectively, for an attenuator thickness of 21 cm and SDD > or = 130 cm. It was found that accuracies of better than 1% could be obtained if detector position and field size specific fitting parameters were used to generate unique mapping functions for each configuration.

A quality control (QC) test suitable for routine daily use has been developed for video based electronic portal imaging devices. It provides an objective and quantitative test for acceptable image quality on the basis of the high contrast spatial resolution and the contrast-to-noise ratio (CNR). The test uses a phantom consisting of five sets of high-contrast rectangular bar patterns with spatial frequencies of 0.1, 0.2, 0.25, 0.4, and 0.75 lp/mm. Data obtained during a one month calibration period were used to determine a critical frequency fc for the relative square wave modulation transfer function and a critical contrast-to-noise ratio (CNRc). Subsequent measurements indicating significant deviations from these critical values result in warning messages to the operator indicating potential problems in system performance. Measurements over a period of two years show that the QC test provides a sensitive indication of imaging performance.

This study involves a fractionated course of external radiation therapy for a 42 year old female weighing 150 kg, diagnosed with stage IIb cancer of the cervix. The patient could not be simulated in the conventional sense due to weight restrictions on the simulator couch, and body casts or molds were impractical. Using an on-line portal imaging device, treatment fields were established during the first session, and intratreatment verification was used before every subsequent treatment to measure and, when necessary, to correct the patient setup. Two courses of treatment were prescribed with a total dose of 60 Gy delivered by a four field box technique (A/P, P/A, and two lateral fields). Out of a total of 108 treatment fields monitored, 12 anterior fields and 1 posterior field were corrected (exclusive of the first, or simulation fraction). Without corrections, 10% of the initial setup displacements would have had displacements greater than 10 mm, 21% greater than 7 mm, and 41% greater than 5 mm. With the application of intratreatment corrections, only 2% of the displacements were greater than 10 mm, 11% were greater than 7 mm, and 32% were greater than 5 mm. It was also found that the second field treated in a parallel opposed pair (i.e., anterior/posterior or left/right lateral) had lower setup displacements and did not require verification or correction.

Projected light fields are used on treatment simulators and teletherapy treatment units to delineate the size and position of the radiation beam. Any discrepancy between these fields will lead to a systematic field placement error, with possibly serious implications with regard to the accuracy of the delivered dose distribution in the patient. Conventionally, film has been used for regular quality control tests of light and radiation field congruence, but this is a time consuming method and is not suitable for daily checks. A new method is described that uses a specially designed test phantom, a video-based electronic portal imaging device and a personal computer to test for radiation and light field congruence on treatment accelerators. This method consists of aligning the test phantom in the light field of a treatment linac and acquiring an electronic portal image. A computer program then automatically analyzes the image and determines the degree of congruence between the two fields. The final result of the test is a go, warning, or no go decision depending on the extent of misalignment between the light and radiation fields. Two algorithms were tested for reproducibility (<0.4 mm), sensitivity to noise (<0.2 mm), and positional accuracy (<0.4 mm) and are shown to give results comparable to the conventional film method. Daily testing of field congruence over a period of 84 days demonstrated differences in the results determined by the two algorithms of less than 0.1 +/- 0.2 mm (standard deviation) at 6 MV and 0.22 +/- 0.13 mm at 23 MV. Routine testing is possible as the effort and time required are minimal, and the test can be performed during daily routine start-up procedures.