Initial application of a geometric QA tool for integrated MV and kV imaging systems on three image guided radiotherapy systems

PurposeThe poor quality of megavoltage (MV) images from electronic portal imaging device (EPID) hinders visual verification of tumor targeting accuracy particularly during markerless tumor tracking. The aim of this study was to investigate the effect of a few representative image processing treatments on visual verification and detection capability of tumors under auto tracking.MethodsImages of QC‐3 quality phantom, a single patient's setup image, and cine images of two‐lung cancer patients were acquired. Three image processing methods were individually employed to the same original images. For each deblurring, contrast enhancement, and denoising, a total variation deconvolution, contrast‐limited adaptive histogram equalization (CLAHE), and median filter were adopted, respectively. To study the effect of image enhancement on tumor auto‐detection, a tumor tracking algorithm was adopted in which the tumor position was determined as the minimum point of the mean of the sum of squared pixel differences (MSSD) between two images. The detectability and accuracy were compared.ResultsDeblurring of a quality phantom image yielded sharper edges, while the contrast‐enhanced image was more readable with improved structural differentiation. Meanwhile, the denoising operation resulted in noise reduction, however, at the cost of sharpness. Based on comparison of pixel value profiles, contrast enhancement outperformed others in image perception. During the tracking experiment, only contrast enhancement resulted in tumor detection in all images using our tracking algorithm. Deblurring failed to determine the target position in two frames out of a total of 75 images. For original and denoised set, target location was not determined for the same five images. Meanwhile, deblurred image showed increased detection accuracy compared with the original set. The denoised image resulted in decreased accuracy. In the case of contrast‐improved set, the tracking accuracy was nearly maintained as that of the original image.ConclusionsConsidering the effect of each processing on tumor tracking and the visual perception in a limited time, contrast enhancement would be the first consideration to visually verify the tracking accuracy of tumors on MV EPID without sacrificing tumor detectability and detection accuracy.

Section: Introductionmentioning

confidence: 99%

Enhancement of megavoltage electronic portal images for markerless tumor tracking

Cheong

Yoon

Park

et al. 2018

“…The procedure of IsoCal geometric calibration is described in the handbook (6) and automated procedure has been integrated into treatment console. The gQA calibration methodology has been previous reported by Mao et al 8 , 9 Briefly, a phantom, gQA‐13, is designed and manufactured for geometric calibration. It is a stand‐alone solid cubic phantom containing 13 stainless steel ball bearings (BBs) mounted within a polystyrene block (

170 \times 170 \times 240 {mm}^{3}

).…”

Section: Methodsmentioning

confidence: 99%

Long‐term evaluation and cross‐checking of two geometric calibrations of kV and MV imaging systems for Linacs

Chiu

Yan

Foster

et al. 2015

Self Cite

Geometric or mechanical accuracy of kV and MV imaging systems of two Varian TrueBeam linacs have been monitored by two geomertirc calibration systems, Varian IsoCal geometric calibration system and home‐developed gQA system. Results of both systems are cross‐checked and the long‐term geometric stabilities of linacs are evaluated. Two geometric calibration methodologies have been used to assess kV and MV imaging systems and their coincidence periodically on two TrueBeam linacs for about one year. Both systems analyze kV or MV projection images of special designed phantoms to retrieve geometric parameters of the imaging systems. The isocenters — laser isocenter and centers of rotations of kV imager and EPID — are then calculated, based on results of multiple projections from different angles. Long‐term calibration results from both systems are compared for cross‐checking. There are 24 sessions of side‐by‐side calibrations performed by both systems on two TrueBeam linacs. All the disagreements of isocenters between two calibrations systems are less than 1 mm with ± 0.1 mm SD. Most of the large disagreements occurred in vertical direction (AP direction), with an averaged disagreement of 0.45 mm. The average disagreements of isocenters are 0.09 mm in other directions. Additional to long‐term calibration monitoring, for the accuracy test, special tests were performed by misaligning QA phantoms on purpose (5 mm away from setup isocenter in AP, SI, and lateral directions) to test the liability performance of both systems with the known deviations. The errors are within 0.5 mm. Both geometric calibration systems, IsoCal and gQA, are capable of detecting geometric deviations of kV and MV imaging systems of linacs. The long‐term evaluation also shows that the deviations of geometric parameters and the geometric accuracies of both linacs are small and very consistent during the one‐year study period.PACS number: 87.56.Fc

“…Sub‐mm accuracy has been achieved by systems that integrate phantoms with image‐analysis software. These comprise custom made systems 6 , 7 , 8 , 9 and commercial systems. Commercial examples include the Quasar system (Modus Medical Systems, London, Ontario, Canada) and the ISO Cube (Computerized Imaging Reference Systems, Inc., Norfolk, VA), claiming “

\frac{true}{1} mm

accuracy” and “0.1 mm accuracy,” respectively.…”

Section: Introductionmentioning

confidence: 99%

A novel phantom and procedure providing submillimeter accuracy in daily QA tests of accelerators used for stereotactic radiosurgery^*

Brezovich

Popple

Duan

et al. 2016

Stereotactic radiosurgery (SRS) places great demands on spatial accuracy. Steel BBs used as markers in quality assurance (QA) phantoms are clearly visible in MV and planar kV images, but artifacts compromise cone‐beam CT (CBCT) isocenter localization. The purpose of this work was to develop a QA phantom for measuring with sub‐mm accuracy isocenter congruence of planar kV, MV, and CBCT imaging systems and to design a practical QA procedure that includes daily Winston‐Lutz (WL) tests and does not require computer aid. The salient feature of the phantom (Universal Alignment Ball (UAB)) is a novel marker for precisely localizing isocenters of CBCT, planar kV, and MV beams. It consists of a 25.4 mm diameter sphere of polymethylmetacrylate (PMMA) containing a concentric 6.35 mm diameter tungsten carbide ball. The large density difference between PMMA and the polystyrene foam in which the PMMA sphere is embedded yields a sharp image of the sphere for accurate CBCT registration. The tungsten carbide ball serves in finding isocenter in planar kV and MV images and in doing WL tests. With the aid of the UAB, CBCT isocenter was located within 0.10±0.05 mm of its true positon, and MV isocenter was pinpointed in planar images to within 0.06±0.04 mm. In clinical morning QA tests extending over an 18 months period the UAB consistently yielded measurements with sub‐mm accuracy. The average distance between isocenter defined by orthogonal kV images and CBCT measured 0.16±0.12 mm. In WL tests the central ray of anterior beams defined by a 1.5×1.5 cm2 MLC field agreed with CBCT isocenter within 0.03±0.14 mm in the lateral direction and within 0.10±0.19 mm in the longitudinal direction. Lateral MV beams approached CBCT isocenter within 0.00±0.11 mm in the vertical direction and within ‐0.14±0.15 mm longitudinally. It took therapists about 10 min to do the tests. The novel QA phantom allows pinpointing CBCT and MV isocenter positions to better than 0.2 mm, using visual image registration. Under CBCT guidance, MLC‐defined beams are deliverable with sub‐mm spatial accuracy. The QA procedure is practical for daily tests by therapists.PACS number(s): 87.53.Ly, 87.56.Fc