Click here to


Are you sure ?

Yes, do it No, cancel

Value Proposition of Online Adaptive Radiotherapy: Plan Fidelity and Deliverability

K Mittauer1*, A Shepard2, L Bayouth2, J Bayouth2, (1) Miami Cancer Institute, Miami, FL, (2) University of Wisconsin-Madison, Madison, WI, (2) University of Wisconsin-Madison, Madison, WI (2) University of Wisconsin-Madison, Madison, WI


(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

Purpose: To assess the risk of IMRT fidelity and deliverability in the value proposition of online adaptive radiotherapy (ART), through retrospectively quantifying the deliverability accuracy of the first one-hundred clinical cases in a clinical MR-guided online adaptive radiotherapy (MRgoART) program.

Methods: One-hundred(N=100) clinical MRgoART cases were prospectively planned and delivered online to the patient, based on their anatomy of day on an MRgRT linac. Post-MRgoART delivery, a measurement-based IMRT QA was delivered to a phantom to quantify plan fidelity. For this study, plan fidelity of the initial simulation plans was compared to MRgoART plans. Correlations in IMRT complexity and plan fidelity were evaluated for MU/prescribed dose (MU/Gy), mean MU/segment (MU/seg), mean aperture area (A), mean aperture perimeter (P), and mean aperture area/perimeter (A/P). All per-segment parameters were also calculated using relative MU/segment weighting as well. Machine deliverability during MRgoART was quantified through dynalog files for MU, MLC and gantry position.

Results: For plan fidelity, average gamma passing rates across respective plans were 98.6±1.3% (initial plans; N=100; minimum 93%) and 97.9±1.4% (MRgoART plans; N=100; minimum 93.0%) at 3%/3mm and 87.6±4.8% at 2%/2mm (MRgoART plans; minimum 76.4%). Two MRgoART and one initial plan had gamma passing rate between 90%-95%. Correlation was found between modulation (A/Pweighted) and gamma passing rate. There was no statistical difference (p>0.67) between plan complexity for initial versus MRgoART plans. Minimal deviation was found when comparing delivered and programed MU/plan, MLC positions, and gantry angles, with a per-segment average error of 0.00±0.04MU, -0.05±0.08cm, 0.00±0.03°, respectively.

Conclusion: Beam model agreement along with highly accurate machine deliverability allowed for safe delivery of online adaptive plans. Reduction of gamma passing correlated with A/Pweighted and can be potentially predict plan failures in MRgoART. Evaluation of dosimetric consequence of online ART is underway to assess the “reward” in the value proposition of online ART.

Funding Support, Disclosures, and Conflict of Interest: Dr Mittauer reports honorarium and travel expenses by ViewRay Inc., and co-founder of MR Guidance, LLC. Dr Bayouth reports honorarium and travel expenses by ViewRay Inc., and co-founder of MR Guidance, LLC.


Clinical Trials, Low-field MRI, Quality Assurance


IM/TH- MRI in Radiation Therapy: MRI/Linear accelerator combined (general)

Contact Email