Abstract: A CT scanner non-invasively examines a volumetric region and generates voxel values. A user adjustable 3D axis system defines mutually orthogonal sections and volume reprojections thus defining a cross reference relation between them. An affine transform translates and rotates the axis system from object space to image space whereby each axis defines the orientation of origin intersecting sections in one view port and the viewing direction of volume reprojections in another view port. An operator console selects an angular orientation of the coordinate axis. A cursor position designates coordinates in image space causing the cursor, typically crossed axes, to be displayed on a monitor at a corresponding location in each displayed image. The view port rotation and translation of the projected crossed cursors are reverse affine transformed to rotate and translate the axis system.

Abstract: An x-ray source (12) is mounted to a rotatable gantry (16). The x-ray source irradiates an examination region (14) with penetrating radiation as (i) the x-ray source rotates around the examination region and (ii) a patient coach (30) moves a patient axially through the examination region. Detectors (24) positioned on an opposite side of the examination region converts radiation which has traversed the examination region along spiral paths (80). Data from the detectors are collected and stored (40) in spiral data sets. In the prior art spiral data on either side of a thin slice central plane was interpolated into a thin slice data set, each thin slice data set was reconstructed into a thin slice image, and several thin slice images were combined to make a thick slice image with significant partial volume artifacts. By distinction, an integrating interpolator (42) weights (88) data values in several spiral sets in accordance with a weighting values W.sub.1, W.sub.2, . . .

Abstract: A CT scanner (10) has an x-ray source (12) which transmits a plurality (N) of x-ray fans across an examination region (14) to a plurality of rings of radiation detectors (28). An adjustable septum (80) adjusts a gap between the fan beams and outer collimators (82) adjust the width of the fan beams such that the effective spacing and width are adjusted. The x-ray source rotates around the examination region as a subject support (32) moves longitudinally through the examination region such that the x-ray fans move along interleaved spiral trajectories. Radiation attenuation data from each of the x-ray fans is combined and reconstructed into a plurality of images along parallel slices orthogonal to the longitudinal axis. The spacing between the x-ray beams is adjusted relative to the effective pitch of the spirals such that the leading edges of the x-ray fans intersect a common transverse plane at 180.degree./N angular intervals around the longitudinal axis.

Abstract: X-ray detectors (24) of a CT scanner (10) generate data values which are preprocessed and assembled (26) into binary data lines. The binary data lines are convolved by a convolver (30) which convolves a plurality of data lines concurrently. A backprojector (32) includes a parametric cubic spline preinterpolator (40) which performs a high order interpolation on the convolved data lines and a linear interpolator (42) which performs a linear interpolation on the high order preinterpolated data lines. The interpolated data lines are backprojected into an image memory (34). Data from the image memory (34) is converted (36) to appropriate format for display on a video monitor (38). The parametric cubic spline preinterpolator (40) includes a first adder (52) which adds most adjacent data lines and a second adder (54) which adds next most adjacent data lines. A first barrel shifter (58) shifts the binary sum of the most adjacent data lines by 1-N spaces, where N is an integer, 4 in the preferred embodiment.

Abstract: A subject in an examination region (14) moves axially as an x-ray source (12) rotates therearound. Views corresponding to a spiral path around a volume of interest in the patient are sampled, interpolated (46), and reconstructed (54) into a three-dimensional image representation (56). This process is repeated a plurality of times to generate a plurality of three-dimensional image representations of the same volume at time displaced intervals. The plurality of image representations are temporally interpolated (110) to generate a series of image representations, each one of which represents the same time or time interval, i.e., a four-dimensional image representation that is linear in all four dimensions. Preferably, a contrast agent is injected into the patient such that the data represents the movement of contrast agent through the volume of interest.

Abstract: A source (A) of images, such as a CT scanner (10), a magnetic resonance imaging apparatus (12), and the like produces a plurality of basis images (I.sub.0, I.sub.1, I.sub.2, I.sub.3 . . . ). Two of the basis images are subtracted and divided (70, 72) by a number of interpolation increments (L.sub.1) to form a first differential image (I.sub..DELTA.1). The first and the third basis images are subtracted and divided (76, 78) by a number of available interpolation increments (L.sub.2) to form a second differential image (I.sub..DELTA.2). Four differential images are selectively combined and divided by a product of the first and second available increments (82, 84) to form a second order differential image (I.sup.2.sub..DELTA.12). An array of adders (D) selectively adds the first differential image to a currently displayed image stored in an image memory E each time a track ball (104) moves a cursor one increment in a horizontal position.

Abstract: A scanner (A), such as a CT or MR scanner, non-invasively examines a region of interest of subject and generates a plurality of views indicative thereof which are reconstructible into an image representation. A filter control (14) generates a bandwidth scaling factor (BWF.sub..theta.) and a bandwidth offset (BWO) for each view. Preferably, the selected bandwidth scaling factor is the maximum of a plurality of bandwidth scaling factors based on different criteria including noise, missing views, material surrounding the region of interest, or noise texture. The bandwidth scaling factor is then used to address a filter curve look up table (18) to select from a plurality (N.sub.T) of filter values. The digital filter values are interpolated (20) or extrapolated as necessary to match the number of values (N.sub.S) in each sampled view.

Abstract: The radiation source (22) rotates around a patient on a patient couch (30) as the couch is advanced through an examination region (14). Radiation detectors (24) detect radiation that has passed through the patient along rays of a plurality of fan shaped views. Each view is identifiable by its angular position around the examination region and its longitudinal position in the spiral and each ray is identifiable by its angular position in the fan. An interpolator (46) interpolates views collected over more than two revolutions of spiral path using an interpolation function (FIGS. 1A-1D). A set of interpolated views is reconstructed (54) into a series of image representations representing parallel planar slices through the imaged volume. In some reconstructions, particularly reconstructions in which a 180.degree. based reconstruction algorithm is used or the energy of the x-ray beam is varied, the filter function is varied from ray to ray within each view.

Abstract: A plurality of segmented radiation sensitive arrays (30) receive radiation from a radiation source (16) which has transversed an examination region (14) of a CT scanner (10). Each array includes a plurality of rows (A, B, C) of radiation sensitive cells, e.g. photodiodes, which produce an electrical signal indicative of the intensity of radiation received. Within each of the rows, there are larger cells and smaller cells. The electrical signals from each of the plurality of cells within each row are serialized (34) and amplified (36A, 36B, 36C). Selected combinations of the electrical signals from the various rows and larger and smaller elements within each row are combined (46) and reconstructed (50) into an image representation (52) for display on a video monitor or the like.

Abstract: A method of minimizing a streaking effect found in reconstructed images obtained from high resolution CT scanning of a small scan circle that corresponds to a limited region of interest, such as portions of the spinal cord of a patient, particularly in scans that include concentrated high density material, such as bone, situated outside the small scan circle. The approximate range of view angles that will project bone from outside the limited region of interest is determined. Detectors corresponding to the determined region are selected out. The projection data acquired by the selected detectors is filtered. The filtering is accomplished by convolving the projection data with a preselected filter function or in machine implemented form by passing the electrical signal representing the projection data through a low pass filter.

Abstract: A method of preprocessing incomplete or truncated projections obtained from high resolution CT scanning of a small scan circle corresponding to a limited region of interest within the scanned object. An air calibration is performed to obtain a set of air values. The intensity values obtained during the partial area scan are subtracted from the air values to convert the intensity values to attenuation values. The approximate slope at either end of the truncated projection is then calculated. Based on the attenuation values and the slope the necessary extrapolation of the truncated projection is performed to complete the projection. No measurements, ionizing or otherwise, are required from the region outside of the small scan circle.The completed projection is ready for conventional reconstruction techniques involving convolution and back projection.