Unencrypted login | home

Program Information

A New Method & Schema for Real-Time Prostate Tracking During VMAT Delivery


L Zhuang

L Zhuang1*, D Yan1, J Li2, (1) William Beaumont Hospital, Royal Oak, MI, (2) Oakland University, rochester, MI

TH-C-BRA-1 Thursday 10:30:00 AM - 12:30:00 PM Room: Ballroom A

Purpose: To create a method and schema for intra-treatment prostate motion tracking using simultaneous kV imaging during the VMAT treatment delivery.

Methods: Intra-treatment prostate motion is monitored using the visicoils' position manifested on continuous kV projection images. Instead of using intensity based image registration, we extract the points' representation of the visicoils for alignment. Real-time prostate displacement is determined by minimizing the distance between the visicoil points' positions on the planning CT and their corresponding positions extracted from a sub-arc of projection images. In the clinical operation, kV imaging keeps on continuously during the VMAT delivery, and the prostate instantaneous motion, as well as the corresponding systematic and random components within a fixed time window, is determined in real-time. Following the delivery, the projection images will be reconstructed to a Cone-Beam (CB) CT. The visicoils' position in the CBCT is compared to the average position of the visicoils determined from the kV projections for verification purpose. The performance of our method was evaluated using a computer simulated phantom under four typical prostate motion patterns. The imaging and detection schema was tested during hypo-fractionated (20 fractions) VMAT treatment of prostate cancer patient.

Results: Based on the simulated experiments, the root mean square error between the detected trajectories and the ground truth is (0.1, 0.4, 0.6) mm. The reconstructed mean position is (0.0±0.1, 0.2±0.1, 0.2±0.2) mm from actual value. Based on the patient experiment, the average visicoils' position determined from the real-time projection image is (-0.6±0.4, 0.6±0.8, -0.4±0.5) mm away from the position detected in the CBCT.

Conclusions: A practical detection method and imaging schema were developed and implemented in clinic for monitoring target motion during VMAT delivery of prostate cancer treatment. Phantom and clinical evaluation showed its feasibility for clinical use.

Contact Email