Encrypted login | home

Program Information

Evaluation of Patient CT Dose Reconstruction From 3D Diode Array Measurements Using Anthropomorphic Phantoms

M Huang

M Huang1*, A Faught2 , S Benhabib1 , R Cardan1 , I Brezovich1 , D Followill2 , R Popple1 , (1) The University of Alabama at Birmingham, Birmingham, AL, (2) The University of Texas MD Anderson Cancer Center, Houston, TX


SU-F-BRE-6 Sunday 4:00PM - 6:00PM Room: Ballroom E

Purpose: To compare 3D reconstructed dose of IMRT plans from 3D diode array measurements with measurements in anthropomorphic phantoms.

Methods: Six IMRT plans were created for the IROC Houston (RPC) head and neck (H&N) and lung phantoms following IROC Houston planning protocols. The plans included flattened and unflattened beam energies ranging from 6 MV to 15 MV and both static and dynamic MLC tecH&Niques. Each plan was delivered three times to the respective anthropomorphic phantom, each of which contained thermoluminescent dosimeters (TLDs) and radiochromic films (RCFs). The plans were also delivered to a Delta4 diode array (Scandidos, Uppsala, Sweden). Irradiations were done using a TrueBeam STx (Varian Medical Systems, Palo Alto, California). The dose in the patient was calculated by the Delta4 software, which used the diode measurements to estimate incident energy fluence and a kernel-based pencil beam algorithm to calculate dose. The 3D dose results were compared with the TLD and RCF measurements.

Results: In the lung, the average difference between TLDs and Delta4 calculations was 5% (range 2%-7%). For the H&N, the average differences were 2.4% (range 0%-4.5%) and 1.1% (range 0%-2%) for the high- and low-dose targets, respectively, and 12% (range 10%-13%) for the organ-at-risk simulating the spinal cord. For the RCF and criteria of 7%/4mm, 5%/3mm, and 3%/3mm, the average gamma-index pass rates were 95.4%, 85.7%, and 76.1%, respectively for the H&N and 76.2%, 57.8%, and 49.5% for the lung. The pass-rate in the lung decreased with increasing beam energy, as expected for a pencil beam algorithm.

Conclusion: The H&N phantom dose reconstruction met the IROC Houston acceptance criteria for clinical trials; however, the lung phantom dose did not, most likely due to the inaccuracy of the pencil beam algorithm in the presence of low-density inhomogeneities.

Funding Support, Disclosures, and Conflict of Interest: Work supported by PHS grant CA10953 and CA81647 (NCI, DHHS).

Contact Email: