Figure 1
Schematic comparison of three methods for X-ray computed tomography of specimens that are larger than the illumination footprint and detector size, as viewed from above for one object slice if the rotation axis is vertical. The bottom figure row shows views of one slice within the object corresponding to one detector row, while the top figure row shows sinograms from one detector row as the object is rotated. In the local tomography acquisition (LTA) or interior tomography Tomosaic approach, the rotation center is placed within a subregion of the object after which a rotation sequence is acquired; when shown in the overall context of the object, one therefore obtains the corresponding sinogram above. In a projection-oriented acquisition (POA) approach as shown in the middle, one acquires a tiled set of projection images while the object is moved transverse to the illumination direction; the object is then rotated slightly about its overall center and the sequence repeated. In the sinogram-oriented acquisition (SOA) Tomosaic approach shown on the right, the rotation center is moved to an offset position relative to the illumination and detector, and the object is rotated to acquire data from a ring-within-a-cylinder region. Both the POA and SOA Tomosaic approaches involve less exposure overlap, reducing radiation dose. Furthermore, SOA generally provides faster acquisition speed because it involves fewer translational motions from the sample stage. |