3D tumor localization through real-time volumetric x-ray imaging for lung cancer radiotherapy

• DOI: 10.1118/1.3582693
• Publication Date: 2011-05-09T22:06:58Z
• Authors: Ruijiang Li; John H. Lewis; Xun Jia; Xuejun Gu; Michael Folkerts; Chunhua Men; William Y. Song; Steve B. Jiang
• Keywords: Lung tumor

Recently we have developed an algorithm for reconstructing volumetric images and extracting 3D tumor motion information from a single x-ray projection. We demonstrated its feasibility using digital respiratory phantom with regular breathing patterns. In this work, present detailed description comprehensive evaluation of the improved algorithm. The was by incorporating prediction. accuracy efficiency were then evaluated on 1) phantom, 2) physical 3) five lung cancer patients. These cases include both irregular patterns that are different training dataset. For breathing, average localization error is less than 1 mm. On NVIDIA Tesla C1060 GPU card, computation time each projection ranges between 0.19 0.26 seconds, which about 10% improvement over previously reported results. below mm achieved 0.13 0.16 seconds same respectively. patients, 2 in axial tangential directions. card 0.34 seconds.