Abstract: |
A computed tomography (CT) scanner has been installed in a linac room at Morristown. FOCAL (Fusion Of CT and Linac) radiation, with daily treatment localization using the CT-Linac without portal imaging have been delivered since June 2000 for over 100 prostate, lung and liver cancer patients. We have an image management system from Siemens called the "Syngo" that can be used for registration between planning and treatment CT images. This study compares the two image fusion techniques available: Landmark Registration and Visual Alignment. Landmark Registration means rigid matrix transformation based on 3 radio-opaque markers placed on the skin (BBs), deduced by computer. Visual Alignment is the manual dragging of one image until it superimposes on and align with the other image. A Rando phantom was scanned in 6 different positions with the CT simulator and the CT-Linac, and the shift of 24 points were deduced from both methods. BBs were placed on the phantom during CT scans, and shifts were also directly measured from the BBs on the phantom. The isocenter shifts deduced from the fused images generally agreed well with the directly measured BB shifts on the Rando phantom, with average absolute error of 0.92 mm in anterior-posterior (AP) direction, 0.97 mm in right-left (RL) direction, and 2.00 mm in superior-inferior (SI) direction. The uncertainty included those in typical patient set-up. The image fusion algorithms were confirmed to be accurate. Some scans with Landmark Registration gave erroneous AP shifts when the anterior radio-opaque marker (BB) registration was significantly off in the AP direction, since the computer tried to match the 3 BBs which in fact did not match. The manual Visual Alignment method was actually more robust than the automatic Landmark Registration in these clinical situations. Details of the FOCAL treatment, such as dose and fraclionation, beam arrangement and margins, acceptance criteria for dose volume histograms, will also be presented. |