Search

Increasingly electron treatment fields are varying widely from small ocular ports to large breast square or long neck's rectangular ports. Their dose distributions are commonly calculated by pencil beam algorithms computing broad dose distributions from summing several elemental electron Pencils. Accurate electron pencil modeling (Pinnacle) requires determining the pencil's critical parameters including its spatial spread (/spl tau///spl rho/) typically one value for each electron beam energy. Spreads (/spl sigma//sub RMS-air/) are deduced from penumbral decrements (i.e., ICRU-35) of the broadest electron beam profiles in air, which algorithms scale to tissue through water-to-air ratios of electron mass-scattering powers (/spl tau///spl rho/). It is possible however, to assess pencil spreads (/spl sigma//sub RMS-water/) directly by ionization scans at depths in water for narrow blocked as well as broad electron fields at different source-to-skin distances (SSD). This paper presents extensive electron pencil spreads in rising field sizes and SSD between air and water, and assesses the media pencil spreads (/spl sigma/(RMS-air) /spl sigma//sub RMS-water/) ratios changing with typical treatment cutouts versus standard open fields. Gaussian pencil spreads (/spl sigma//sub RMS-air/) in air were found in direct constant proportionality (1/[2/spl pi/]/sup 1/2/) to penumbra widths for limited broad fields, but poorly correlated for narrow beams. Water widths and spreads showed limited constant scaling yet their ratios deviated significantly from proportionality for shrinking field ports. Hence for accurate clinical dose commission, an electron pencil that models small as well as large field treatments, may demand RMS values characterized both in air and water, and also scaled empirically with rising electron field diameters or treatment distances.

The concept of equivalent square has been looked at by many authors. The approaches range from simple approximations to summation methods and integration methods. These studies looked at output factors, collimator scatter, TMR, or percent depth dose. In this study, output factor, collimator factor, and TMR data were measured for rectangular and irregular fields for a Siemens MLC. The measurements were made for 6 and 18 MV photon beams. To provide comparison with previous studies, the field shapes described by Xiao et al., Phys. Med. Bio. 44, 1053-1065 (1999) and Yuen et al., Med. Phys., 26, 2385-2389 (1999) were used. It was found that the average field length approximation is accurate to within 3.4% for all parameters and fields studied. The average deviation for output factor was 0.1% for rectangular fields with 28 /30 fields within 1%. The average deviation was -1.1% for the irregular fields with 17/31 fields within 1%. The average deviation for TMR for rectangular fields was 0.3% with ah fields within 1%. For irregular fields the average was -0.6% with 24/31 within 1%. The average deviation was -1.1% for the irregular fields with 17/31 fields within 1%. The average deviation for collimator scatter factor for rectangular fields was 0.1% with 26/30 fields within 1%. For irregular fields the average was -0.8% with 21/31 within 1%. Thus a simple approximation can be used to determine dosimetric parameters for clinical fields with an accuracy better than 2%.

Metal Oxide-Silicon Field Effect Transistor or MOSFET radiation detectors have become excellent in-vivo dosimeters for radiation therapy, but the MOSFET dosimeter's responses to clinical electron beams have yet to be fully evaluated. The authors measured MOSFET dose responses to varying electron beam energies and angles on patients and anthropomorphic phantoms. MOSFET responses in mV per cGy are compared against electron skin doses determined using a thin parallel plate ionization chamber. Results indicate the MOSFET dosimeter is near energy independent to within 1%. The MOSFET responses are dependent on the electron source to patient skin distances (SSD) within +/-5%, indicative of some dependence on dose rate or accelerator pulse rate by the field effect transistor. These SSD responses compare favorably with the response typically observed with diode in vivo dosimeters, except at extremely short SSD (i.e., 95 cm) close to the mouth of the electron cones where both the chamber and the MOSFET or diode readings are difficult to interpret. MOSFET responses correlated within 2% of the doses in different electron field sizes including blocked or small cutout down to 4 cm. The authors concluded MOSFET is dosimeter suitable in electron beams and it necessitates only a nominal dose calibration factor for in-vivo dosimetry treatment field conditions, except when patient treatment distances are considerably greater (20 cm) than accelerator nominal isocenter.

When a shielding jaw or multiple leaf collimator (MLC) traverses across an X-ray field, it modulates the field photon intensity, thereby inducing dose distribution patterns in cross-beam profiles usually termed dynamic (DW) or virtual wedges (VW), or virtual filter for intensity modulated radiation treatment (IMRT). Measured virtual filtration or transmission factors have, unlike those of conventional metals, exhibited rapid drops with modulated field sizes and at times, shown a discontinuity in the transmission when the stepping segmental width is altered amid modulation. We formulated a generalized analytical virtual transmission factor VTF/sub /spl Delta// for a modulated photon beam irrespective of the accelerator (Varian, Siemens, Philips-Elekta) schemes of modulation. WTF/sub /spl Delta// equals a time integral of TR/sub p/(t) plus GR/sub s/(t), and respectively, are the filter modulation primary fraction and the scatter gradient. The modulation duration t or equivalently, is the accumulating counts of partial monitor units (MU). When tested against actual segmental treatment tables (SST Varian), the stated virtual transmission equation reproduced acceptably the measured (DW) factors and predicted their downward trends in the 6 MV photon beams with 60/spl deg/ and 45/spl deg/ DW filters. Further, it is shown that upon taking the primary fraction as the integrated MU-step function /spl delta/, and taking the number (n) of the modulating collimator steps to infinity n/spl rarr//spl infin/, one has the analytical virtual transmission approach the asymptotic value limit VTF/sub /spl Delta//(n/spl rarr//spl infin/)= 1/2 . This corollary theorem is shown supported by (Varian DW) virtual transmission factors as they taper to 0.48 and 0.51 experimental values.