Encrypted login | home

Program Information

Computational Monitoring Framework of Four-Dimensional Dose Distributions During Treatment Time Based On a 2D/3D Registration with Adaptive Transformation Parameters in Lung Stereotactic Body Radiotherapy

no image available
T Nakamoto

T Nakamoto1, H Arimura1*, K Nakamura1, Y Shioyama2, TA Hirose3, Y Umezu3, Y Nakamura3, H Honda1, H Hirata1, (1) Kyushu University, Fukuoka, Japan, (2)SAGA Heavy Ion Medical Accelerator in Tosu, Tosu, Japan,(3)Kyushu University Hospital, Fukuoka, Japan


WE-AB-303-7 (Wednesday, July 15, 2015) 7:30 AM - 9:30 AM Room: 303

Purpose: The aim of this study was to develop a computational framework for monitoring four-dimensional (4D) dose distributions during treatment time based on a 2D/3D registration with adaptive transformation parameters (ATPs) in lung stereotactic body radiotherapy (SBRT).

Methods: The 4D dose distributions during the SBRT were calculated by applying a pencil beam convolution (PBC) algorithm to simulated 4D-computed tomography (CT) images during treatment time (hereinafter referred to as “treatment” 4D-CT images). The “treatment” 4D-CT images were derived by deforming 3D planning CT images by using transformation parameters so that 2D planning portal dose images (PDIs) can resemble 2D dynamic clinical PDIs in each frame, which were derived from electronic portal imaging device (EPID) images. The transformation parameters were optimized by the Levenberg-Marquardt (LM) algorithm. The optimized transformation parameters in a certain frame were adaptively employed as initial transformation parameters for optimization of the parameters in the consecutive frame. Gamma pass rates (3mm/3%) between dynamic clinical PDIs and dynamic “treatment” PDIs derived from the “treatment” 4D-CT images were measured for comparing the proposed frameworks without and with the ATPs. These proposed frameworks were applied to the EPID dynamic images (40 frames) of two patients with lung cancer, who underwent the SBRT.

Results: Gamma pass rates without and with the ATPs were 83.01±4.42% and 89.31±3.18% on average, respectively (p<0.05). Moreover, percentage errors between prescribed doses and estimated doses at an isocenter by the proposed frameworks without and with the ATPs were 10.67±5.66% and 6.34±0.47% on average, respectively (p<0.05).

Conclusion: The proposed framework with ATPs could be useful to ensure the quality of the SBRT by monitoring 4D dose distributions during treatment time.

Contact Email: