Validation of post-treatment PET-based dosimetry software for hepatic radioembolization of Yttrium-90 microspheres

Research output: Contribution to journalArticlepeer-review

15 Scopus citations


Purpose: Yttrium-90 (90Y) microsphere radioembolization enables selective internal radiotherapy for hepatic malignancies. Currently, there is no standard postdelivery imaging and dosimetry of the microsphere distribution to verify treatment. Recent studies have reported utilizing the small positron yield of 90Y (32 ppm) with positron emission tomography (PET) to perform treatment verification and dosimetry analysis. In this study, we validated a commercial dosimetry software, MIM SurePlan™ LiverY90 (MIM Software Inc., Cleveland, OH), for clinical use. Methods: A MATLAB-based algorithm for 90Y PET-based dosimetry was developed in-house and validated for the purpose of commissioning the commercial software. The algorithm is based on voxel S values and dosimetry formalism reported in MIRD Pamphlet 17. We validated the in-house algorithm to establish it as the ground truth by comparing results from a digital point phantom and a digital uniform cylinder to manual calculations. Once we validated our in-house MATLAB-based algorithm, we used it to perform acceptance testing and commissioning of the commercial dosimetry software, MIM SurePlan, which uses the same dosimetry formalism. A 0.4 cm/5% gamma test was performed on PET-derived dose maps from each algorithm of uniform digital and nonuniform physical phantoms filled with 90Y chloride solution. Average dose (Davg) and minimum dose to 70% (D70) of a given volume of interest (VOI) were compared for the digital phantom, the physical phantom, and five patient cases (27 tumor VOIs), representing different clinical scenarios. Results: The gamma-pass rates were 97.26% and 97.66% for the digital and physical phantoms, respectively. The differences between Davg and D70 were 0.076% and 0.10% for the digital phantom, respectively, and <5.2% for various VOIs in the physical phantom. In the clinical cases, 96.3% of the VOIs had a difference <5% for Davg, and 88.9% of the VOIs had a difference <5% for D70. Conclusions: Dose calculation results from MIM SurePlan were found to be in good agreement with our in-house algorithm. This indicates that MIM SurePlan performs as it should and, hence, can be deemed accepted and commissioned for clinical use for post-implant PET-based dosimetry of 90Y radioembolization.

Original languageEnglish
Pages (from-to)2394-2402
Number of pages9
JournalMedical physics
Issue number5
StatePublished - May 2019


  • PET
  • Y-90
  • post-implant dosimetry
  • radioembolization


Dive into the research topics of 'Validation of post-treatment PET-based dosimetry software for hepatic radioembolization of Yttrium-90 microspheres'. Together they form a unique fingerprint.

Cite this