Before patient studies, we will have conducted experimental studies on a lung phantom.
The purpose of this step is to comprehensively evaluate the system performance and gain confidence prior to its applications to real patient cases. We will operate the phantom to simulate motion patterns with different motion amplitude/frequency. 4DCBCT images will be reconstructed using our method and the conventional FDK algorithm. The validation will focus on two aspects: HU value accuracy and anatomy location accuracy.
Imaging dose in these scans will be also measured in phantom studies to make sure the radiation dose level is acceptable for patient studies. adiation dose level is acceptable for patient studies.
After testing the reconstruction algorithm on phantoms, we will move on to real patient cases. We will conduct patient studies on lung cancer patients treated under Image Guided Radiation Therapy (IGRT) at Department of Radiation Oncology, UT Southwestern Medical Center. CBCT projection data will be collected (4 gantry rotations, each within 1 minute) in only one treatment fraction. 4DCBCT images will be reconstructed from the current standard FDK algorithm using all the projection data, and from our system using data only in the first gantry rotation. Patient 4DCT image will also be used, which have been acquired during treatment planning stage per standard treatment protocol at UTSW.
The image quality of our 4DCBCT will be assessed in two aspects:
1\) geometry accuracy, by comparing selected anatomical landmark locations with those in 4DCBCT reconstructed by the FDK algorithm. 2) HU accuracy, by comparing HU values in selected regions of interest (ROIs) with those in 4DCT.
The 4DCT images will be extracted from radiation therapy information system MOSAIQ.
The CBCT projection data will be exported from a computer that controls the image acquisition system
