Implementation and Evaluation of Phantomless Intensity Modulated Radiotherapy Delivery Verification Using...

The objective was to implement the DynaLog file-based automated FractionCHECKTM software which verifies the delivery accuracy of all patient treatments without any measuring devices and evaluate this tool for dynamic IMRT delivery verification of 40 cases. The present work was aimed to find the generalized results with user-defined tolerances of all treatment deliveries and asses the delivery results in terms of Pass/Warn/fail analysis. In addition to this, an in-depth analysis of mean RMS error of each BANK of the MLC's and mean 95th Percentile error of all the leaves and percentage of pixels passing Gamma between the planned and delivered fluences for all 40 dynamic IMRT cases.

The DynaLog files were analyzed from all 40 IMRT deliveries which include ten plans each of head & neck, brain, cervix and oesophagus patients treated using 6 MV X-ray beam at DHX linear accelerator equipped with Millennium120 multileaf collimator (MLC). Before processing actual patient data, we have validated the FractionCHECKTM software using routine DMLC QA test fields such as X-wedge, Y-Wedge, Garden fence test and Pyramid Test. The mean RMS error and 95th percentile of each leaf was studied in four different sites of dynamic IMRT and fluence delivery accuracy at 1%, 1mm gamma criterion was determined over the entire course of 40 patients.

We have analyzed a total of 1134 fractions of dynamic IMRT cases of 40 patients in this study. Out of 25,166 log files evaluated for the entire course of these patients, 23,906 files have met pass criteria (Green), 1022 log files resulted in warning (Yellow), and 238 files have resulted in failure mode (RED) in the analysis. The overall mean RMS error for all 40 IMRT cases was found to be 0.34±0.08 mm, and the mean 95th percentile error of all leaves was 0.66 ± 0.18 mm, which was well below the recommended by TG142 as well with other published works.

In a busy department where a large number of IMRT cases are being treated, the DynaLog analysis-based FractionCHECKTM software can be used as one of the efficient methods in verifying IMRT treatment deliveries. It can be implemented much quicker without long delay and can also be used for routine DMLC QA tests. We conclude that FractionCHECKTM is a fast, simple and powerful tool for routine QA of the IMRT fields and that based on our test field and patient data results, FractionCHECKTM could be used for treatment delivery assessment accuracy.