In aqua vivo EPID dosimetry

• Published in: Medical physics (Lancaster) 2012-01, Vol.39 (1), p.367-377
• Main Authors: Wendling, Markus, N. McDermott, Leah, Mans, Anton, Olaciregui-Ruiz, Ígor, Pecharromán-Gallego, Raul, Sonke, Jan-Jakob, Stroom, Joep, van Herk, Marcel, J. Mijnheer, Ben
• Format: Article
• Language: English
• Subjects: Algorithms; Cancer; COMPARATIVE EVALUATIONS; Computed tomography; Digital computing or data processing equipment or methods, specially adapted for specific applications; dose verification; DOSIMETRY; Dosimetry/exposure assessment; EPID; HOSPITALS; Humans; Image data processing or generation, in general; Image guided radiation therapy; IMAGE PROCESSING; image reconstruction; IN VIVO; in vivo dosimetry; Intensity modulated radiation therapy; IRRADIATION; KERNELS; lung; lung cancer; Lung Neoplasms - radiotherapy; LUNGS; medical image processing; Medical image reconstruction; Medical imaging; NEOPLASMS; PATIENTS; PHANTOMS; PROSTATE; quality assurance; RADIATION DOSE DISTRIBUTIONS; RADIATION DOSES; RADIATION PROTECTION AND DOSIMETRY; radiation therapy; RADIOLOGY AND NUCLEAR MEDICINE; Radiometry - instrumentation; Radiometry - methods; RADIOTHERAPY; Radiotherapy Dosage; Radiotherapy Planning, Computer-Assisted - methods; Radiotherapy, Computer-Assisted - methods; Reconstruction; Therapeutics; Tissues; Treatment strategy; VERIFICATION; X-Ray Intensifying Screens
• ISSN: 0094-2405; 2473-4209
• DOI: 10.1118/1.3665709

Purpose: At the Netherlands Cancer Institute—–Antoni van Leeuwenhoek Hospitalin vivo dosimetry using an electronic portal imaging device (EPID) has been implemented for almost all high-energy photon treatments of cancer with curative intent. Lung cancer treatments were initially excluded, because the original back-projection dose-reconstruction algorithm uses water-based scatter-correction kernels and therefore does not account for tissue inhomogeneities accurately. The aim of this study was to test a new method, in aqua vivo EPID dosimetry, for fast dose verification of lung cancer irradiations during actual patient treatment. Methods: The key feature of our method is the dose reconstruction in the patient from EPID images, obtained during the actual treatment, whereby the images have been converted to a situation as if the patient consisted entirely of water; hence, the method is termedin aqua vivo. This is done by multiplying the measured in vivo EPID image with the ratio of two digitally reconstructed transmission images for the unit-density and inhomogeneous tissue situation. For dose verification, a comparison is made with the calculated dose distribution with the inhomogeneity correction switched off. IMRT treatment verification is performed for each beam in 2D using a 2D γ evaluation, while for the verification of volumetric-modulated arc therapy (VMAT) treatments in 3D a 3D γ evaluation is applied using the same parameters (3%, 3 mm). The method was tested using two inhomogeneous phantoms simulating a tumor in lung and measuring its sensitivity for patient positioning errors. Subsequently five IMRT and five VMAT clinical lung cancer treatments were investigated, using both the conventional back-projection algorithm and the in aqua vivo method. The verification results of the in aqua vivo method were statistically analyzed for 751 lung cancer patients treated with IMRT and 50 lung cancer patients treated with VMAT. Results: The improvements by applying thein aqua vivo approach are considerable. The percentage of γ values ≤1 increased on average from 66.2% to 93.1% and from 43.6% to 97.5% for the IMRT and VMAT cases, respectively. The corresponding mean γ value decreased from 0.99 to 0.43 for the IMRT cases and from 1.71 to 0.40 for the VMAT cases, which is similar to the accepted clinical values for the verification of IMRT treatments of prostate, rectum, and head-and-neck cancers. The deviation between the reconstructed and planned dose at the iso