Commissioning and clinical implementation of a mega-voltage cone beam CT system for treatment localization
Projection of the 3D voxel in the construction volume onto the 2D plane of the flat panel detector. A projection matrix is used to correlate the two positions for every gantry angle as the x-ray source and the detector move around the reconstruction volume.
(a) Geometry calibration phantom. There are 108 tungsten ball bearings of different sizes embedded on this Plexiglas cylinder to form a single helix. Their known coordinates are used to calculate the transformation matrices projecting the three-dimensional voxel in the reconstruction volume to a two-dimensional pixel on the flat panel. (b) Image quality phantom. The phantom is divided into four sections: a thick solid water section is used to check image uniformity and noise, and three thick sections containing plugs of different electron density are used to evaluate low contrast and spatial resolution. The 3 o’clock position of the beads on the top, middle, and bottom slices is also shown.
(a) Low contrast section of the image quality phantom. Each section has inserts of four different materials: (1) 1% SIG, (2) 3% SIG, (3) brain, and (4) liver. (b) High contrast section of the image quality phantom: (5) inner bone, (6) acrylic, (7) air, and (8) CB2 (bone–50% mineral). The electron density for each of these materials is given in Table I. For each material, there are five inserts of diameters 2, 1, 0.7, 0.5, and . (c) Spatial resolution section of the image quality phantom. Eleven bar groups with different numbers of line pairs per millimeter are inserted in the section. (d,e,f) Slices of the low contrast, high contrast, and spatial resolution sections, acquired with a protocol with “smoothing head and neck” filtering.
MV-CBCT image of a prostate patient with gold markers, taken with the 15 MU protocol. (a) axial, slice reconstruction, MPR view , resolution; (b) axial, slice reconstruction, MPR view , resolution; (c) sagital, slice reconstruction, MPR view , resolution; (d) sagital, slice reconstruction, MPR view , resolution; and (e) axial, slice reconstruction, MPR view , resolution; and (f) axial, slice reconstruction, MPR view , resolution. We use the parameters from (a) in all our imaging protocols.
Measurement of noise and uniformity of the image in Sec. I of the image quality phantom. The means and standard deviations in pixel values of the five outlined regions are within our specifications.
Contrast to noise ratio versus longitudinal field size, for three different electron density materials. The CNR decreases by about when the field size increases from 5 to .
CBCT image of a CT phantom using a 15 MU protocol: (a) head-and-neck case using the inner portion only and (b) pelvis case using both the inner and outer portions, thus extending beyond the CB field of view.
Relative lateral, vertical, and longitudinal measured positions of the 12 beads embedded in the image quality phantom, with respect to their nominal values. There are four beads in the middle section of the phantom and four beads in sections from the middle section in each direction (labeled “top” and “bottom”). The beads on each section are labeled by their azimuthal positions, at 12, 3, 6, and 9 o’clock. The horizontal lines correspond to the tolerance on the position in each direction.
Physical density and electron density relative to background solid water of the inserts in Secs. II and IV of the image quality phantom. The size of the smallest visible insert for each material is also reported.
Geometric arrangement of bars in the spatial resolution section of the image quality phantom.
Typical imaging protocols used in our clinic for different anatomical sites, along with their frequency. For daily localization, the CB dose was accounted for in the treatment plan.
Daily QA, monthly QA, and biannual maintenance program. The daily tasks are performed by the therapists and checked by a physicist, the monthly tasks are performed by a physicist, and the biannual maintenance is performed by the manufacturer and checked by a physicist.
Article metrics loading...
Full text loading...