Premium
Poster — Thur Eve — 18: Differential Dose‐Volume Histogram Modeling Using the Gaussian Error Function
Author(s) -
Chow J,
Markel D,
Jiang R
Publication year - 2010
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.3476123
Subject(s) - fiducial marker , prostate , nuclear medicine , prostate cancer , radiation treatment planning , medicine , histogram , dosimetry , radiation therapy , dose volume histogram , computer science , radiology , artificial intelligence , cancer , image (mathematics)
The Gaussian error function (GEF) was first used to model rectal differential dose‐volume histograms (dDVH) for prostate intensity modulated radiation therapy (IMRT) plans incorporated with the interfraction prostate motion. Seven‐beam IMRT treatment plans were created in three patients with small (40 cm 3 ), medium (53 cm 3 ) and large (87 cm 3 ) prostate volume, selected from a group of 20 patients. The interfraction prostate motions were measured by comparing the digitally‐reconstructed radiographs (anterior and lateral views) from the original treatment plans to the corresponding daily electronic portal images in the treatment unit based on the implanted fiducial gold markers. The ranges of prostate motion were found to be 8 – 2 mm, 4 – 8 mm and 4 – 3 mm along the anterior‐posterior directions for the small, medium and large prostate patient, respectively. Rectal dDVH varying with the interfraction prostate motion were determined by the treatment planning system (TPS), and modeled by the GEF for the three patients. It was found that the rectal dDVH from the prostate plans modeled by the GEF agreed well with those calculated by the TPS. The successful modeling of dDVH results in a significant reduction of the dDVH database, because typically about 12 parameters of the GEF model can be used to substitute about 800 – 1000 dose‐volume bin set for each dDVH. This can greatly reduce the computer memory in the normal tissue complication probability calculation associated with a huge dDVH database.