PO-0772: Patient-specific realtime error detection for VMAT based on transmission detector measurements

Purpose or Objective To investigate a new transmission detector for online dose verification. Error detection ability was examined and the correlation between the changes in detector output signal with γ passing rate and DVH variations was evaluated. Material and Methods The integral quality monitor detector (IQM, iRT Systems GmbH, Germany) consists of a single large area ionization chamber which is positioned between the treatment head and the patient. The ionization chamber has a gradient along the direction of MLC motion and is thus spatially sensitive. The detector provides an output for each single control point (segment-by-segment) and a cumulative output which is compared with a calculated value. Signal stability and error detection sensitivity were investigated. Ten types of errors were induced in clinical VMAT plans for three treatment sites: head and neck (HN), prostate (PC) and breast cancer (MC). Treatment plans were generated with Pinnacle (V.14) for an Elekta synergy linac (MLCi2). Geometric errors included shifts of one or both leaf banks for all control points toward (i) and away (ii) from the central axis of the beam and unidirectional shifts of both leaf banks (iii) by 1 and 2mm, respectively. Dosimetric errors were induced by increasing the number of MUs by 2% and 5%. Deviations in dose distributions between the original and error-induced plans were compared in terms of IQM signal deviation, 2D γ passing rate (2%/2mm and3%/3mm) and DVH metrics (Dmean, D2% and D98% for PTV and OARs). Results For segment-by-segment evaluation, calculated and measured IQM signal differed by 4.7%±5.5%, -2.6%±4.6% and 4.19%±6.56% for MC, PC und HN plans, respectively. Concerning the cumulative evaluation, the deviation was -1.4±0.25%, -6.0±0.3% und -1.47%±0.97%, respectively. Signal stability for ten successive measurements was within 0.5% and 2% for the cumulative and the segment- by-segment analysis. The IQM system is highly sensitive in detecting geometric errors down to 1mm MLC bank displacement and dosimetric errors of 2% if a measured signal is used as reference. Table 1 reports IQM signal deviations for a range of introduced errors. Regarding MLC errors affecting the field size, large deviations from reference were observed in the IQM signal, while unidirectional shifts introduced deviations below detection limit. A similar behavior was observed for 2D γ and DVH parameters. Figure 1 illustrates the correlation of Dmean(PTV) and IQM signal deviation, indicating that clinically relevant errors can be identified. Conclusion The deviation between calculated and measured signal is relatively high, therefore a measurement should be defined as reference. With this limitation, the system is not yet capable of treatment plan verification but is a powerful tool for constancy testing. The detector provides excellent signal stability and is very sensitive regarding error detection. The signal deviation correlates with 2D γ and DVH metric deviations; this information can be used for identifying action limits for the IQM.