Room: Exhibit Hall | Forum 7
Purpose: Ranking the biological effectiveness of alternate treatment plans is essential for ongoing efforts to improve radiation therapy treatment outcomes. In this study, an automated algorithm and biological metrics are used to compare and rank the potential effectiveness of treatment plans delivered using the same or different numbers of fractions using MV x-rays, protons, fast neutrons or other particles.
Methods: We have developed a script to automate the conversion of organ at risk (OAR) and tumor dose volume histograms (DVH) into estimates of the equivalent uniform dose (EUD) with corrections for fraction size and particle relative biological effectiveness (RBE). Estimates of the EUD are then converted into an estimate of the equivalent dose in 2 Gy fractions (EQD2) so that alternate and competing plans can be ranked, regardless of the number of fractions. Plans with a larger tumor to OAR ratio of EQD2 values are ranked higher than plans with a lower tumor to OAR ratio. The automated plan ranking algorithm is tested using representative MV x-ray, proton and neutron therapy treatment plans.
Results: Sample plans for neutron, proton, and photon plans are evaluated using the algorithm. Treatment planning DICOM files are processed to compute metrics. Maximum tumor to OAR EQD2 ratio were identified.
Conclusion: Evaluating DVH is presently the standard for evaluating planning quality; however, assessments of OAR and tumor DVH requires subjective judgement that often translate to a simple â€œsatisfyâ€? or â€œdoes not satisfyâ€? plan constraints, which hamper the automated ranking of alternate and competing treatment plans. The presented study shows the developments of an objective plan evaluation algorithm.