Automated Plan Checks in High Dose Rate Prostate Brachytherapy Using Custom Scripts in a Commercially Available Spreadsheet Application

Abstract

In a busy department that handles hundreds of High Dose Rate (HDR) brachytherapy cases a year, the hurried pace creates opportunity for improvements in patient safety checks and workflow. Existing brachytherapy second check software typically involves only a TG-43 based calculation to one or more points. In order to improve the quality assurance process and facilitate robust plan quality analysis without adding significant time, software code was written in a commercially available spreadsheet application. HDR brachytherapy of the prostate involves implantation of needles in a surgical setting, followed by simulation, treatment planning, and treatment delivery in the same day. After plan approval and prior to treatment the physicist typically performs a visual plan check followed by a TG-43 based spot check of one or more discrete dose points. In order to independently and thoroughly check the plan parameters, dose volume results, and calculated treatment time, code was written to extract data from the plan printouts and analyze the planned data against expected results. Parameters checked include prescribed dose, source decay, and number, length, offset, and orientation of catheters. The system then verifies that appropriate target and risk structures have been contoured and utilized for optimization and compares dose-volume data against benchmarks and gives a stoplight grading. Planned time accuracy is checked with two methods: First with a TG-43 point dose calculation, then against a graph of time vs prostate volume, normalized for dose and source strength, built from clinical data. All results are stored for future analysis. At this time 12 patient cases have been run through the automated checking system. One case of a missing risk structure expansion was caught by the new system and led to an new plan optimization. Of the 12 patients there have been 18 treatment deliveries for which dose and source strength normalized treatment time was plotted against prostate volume, with a resulting linear fit R2 = 0.9786. Use of the automated system has decreased time-to-delivery by several minutes per case by automating data collecting that was previously done manually. Implementing this new automated QA process in tandem with a visual second check by the planning physicist reduces the chance of an error which may otherwise result in significant detriment to the patient. Important plan parameters which directly affect accurate dose delivery are now checked automatically, along with robust quality and accuracy checks that go beyond typical commercially available brachytherapy second check software. This process performs an independent second check of the plan parameters which is especially important in the absence of a second physicist.