| 20110206178 | On-line cone beam CT reconstruction - An in-line 4D cone beam CT reconstruction algorithm, i.e. one that works in parallel with image acquisition, comprises an imaging system for a object exhibiting internal periodic motion comprising a source of penetrating radiation and a two-dimensional detector for the radiation, the source and the detector being rotatable around an axis lying on the beam path from the source to the detector, a storage means for images obtained from the detector, a control means for initiating rotation of the source and the detector and for obtaining images from the detector at a plurality of rotation angles over time, a processing means for (i) condensing the images in a direction transverse to an axis to produce a one-dimensional image, (ii) collating the one-dimensional images obtained up to that point side-by-side into a two dimensional image, (iii); analysing the two-dimensional image thus obtained to identify periodic patterns, (iv) allocating phase information to the images in the storage means on the basis of that analysis, (v) selecting images in the storage means having like phase information, and (vi) backprojecting the selected images, the control means being adapted to invoke the processing means after a plurality of images have been placed in the storage means, and then place further images in the storage means and further invoke the processing means. Thus, we queue a limited number of projection images such that the phase determination algorithm can look-ahead. At regular intervals, the queue is scanned and those images which have enough look-ahead to obtain phase information are filtered and back-projected. The algorithm thus keeps up with the image acquisition speed and produces a 4D reconstruction within a few seconds of the end of scanning. A local rigid registration algorithm is then used to match the tumor region defined in the mid-ventilation frame of our 4D planning CT with each of the phases of the 4D CBCT. An animation technique provides rapid visual verification; the mean position of the tumor is computed and used for correction, while the amplitude is reviewed to validate the margin. | 08-25-2011 |
