Skewness and kurtosis as measures of range mixing in time resolved diode dosimetry

Technical Note describing studies, in 2011, of time resolved diode dosimetry in an anthropomorphic phantom at the Francis H. Burr Proton Therapy Center, Massachusetts General Hospital, Boston, MA. This technique measures the water equivalent path length (WEPL) to a diode dosimeter with sub-millimeter precision, using very little dose. It may allow one to use the proton stopping point (otherwise too uncertain) to cut between tissue to be treated and tissue to be spared, in cases (such as the prostate) where a diode can be placed in a cavity (the rectum) distal to the target. A potential problem is range mixing, where (because of scattering) protons may arrive at a diode with different energy-loss histories. The WEPL reported by such a diode is meaningless and must be ignored. We therefore use an array of diodes, of which some are likely to be at more favorable locations. In addition to the overall analysis of time resolved data, this Note describes in detail how range mixed diodes can be identified by looking at the skewness and kurtosis of the time dependent signal.