Jung, Jae WonBelcher, Wesley2023-06-052024-05-012023-052023-04-26May 2023http://hdl.handle.net/10342/12862Introduction: Lung cancer remains one of the most fatal forms of cancer globally. Lung cancer is often treated by surgical removal of the malignant tissue however, surgery is not often an option for lung cancer. Radiation is the next treatment technique for non-small cell cancer. Radiation treatment of lung cancer is complicated by respiratory motion. As a patient breathes their lung inflates and deflates causing the lung tissue and the tumor to move. There are many techniques to account for the lung motion. A technique to accommodate lung tumor motion is tracking. Lung tumor tracking can be performed in multiple ways, but one way is fiducial marker directed tracking. Tracking helps the tumor to be located and treated continuously in real time. This allows the treatment margins to be smaller and gets better patient outcomes. However, the accuracy of the fiducials' motion matching tumor motion has yet to be depicted. This study aims to show how well fiducial motion matches tumor motion. The machine tracks a fiducial centroid, which is the center of mass of one or more objects. The study will also look at what makes fiducial centroids track the tumor well. The dosimetric effect of switching centroids will be observed. Patterns will be searched for to improve the fiducial marker placement recommendations. The current recommendations for fiducial marker selection are to select the fiducials closest to the tumor. These are not always the best tracking fiducials. The patterns in good fiducial centroids will also be used to improve the fiducial marker selection recommendations. Methods: This was an IRB approved retrospective study that included 27 patients receiving planning 4DCT for SBRT treatment of either primary or metastatic lung cancer. 20 patients had upper lung tumors with 79 fiducials placed. There were 7 lower lung tumor patients and 23 fiducials. It is expected for there to be more upper lung tumor cases than lower lung tumor cases. In total 102 fiducials were placed. The center of mass (COM) of each fiducial and the pre-contoured tumor were found in every phase of the breathing cycle, using a MATLAB[registered] program. The fiducials were then checked to see if they tracked the tumor well or not. If the fiducial kept a consistent distance within 2 mm of the tumor's COM, it was deemed a good tracking fiducial. Every fiducial combination was found and searched to see what centroid kept the most consistent distance from the tumor and which centroid was closest to the tumor's COM. These centroids were compared with the tracked centroid. The planned dose was shifted by the differences in the centroids locations compared to the tracked centroid. The type of bracketing around the tumor was checked in the centroids of interest. This analysis was done again on upper and lower lung patients separately. Results: The distance was not a determining factor on if the fiducial would track the tumor well. The centroid that was closest to the tumor's COM was closer than the tracked centroid by 4.1 [plus-minus] 1.2 mm (p[less-than]0.05) and the centroid that followed the tumor's motion the best kept the distance more consistently than the tracked centroid by 0.23 [plus-minus]0.05 mm (p[less-than]0.05). The shifted dose profiles were worse for tumor coverage than the original dose profiles but some other organs at risk had better dose sparing. Following analyze of the bracketing present in the centroids of interest, it is most important for the tumor to be bracketed in the superior/inferior and anterior/posterior directions. The superior/inferior bracketing is given higher priority for lower lung tumor patients and the anterior/posterior bracketing is given higher priority for upper lung tumor patients.application/pdfenMedical PhysicsRadiation OncologyReal-Time TrackingLung TumorFiducialMotion TrackingTumor markersLungs--CancerTumorsImaging systems in medicineDoctoral Dissertation2023-06-02