Abstract: The present invention, in one form, is a method for correcting for artifacts caused by highly attenuating objects in a CT image data using a correction algorithm. In accordance with one embodiment of the algorithm, the highly attenuating objects are identified in the image data using the CT numbers from the image data. The segmented image data for each highly attenuating material are used to produce separate component images for each material. The component image data for each material is then separately forward projected to generate projection data for each material. The projection data for each material is then adjusted for the attenuation characteristic of the material to generate projection error data for each material. The resulting projection error data are then filtered and backprojected to produce error-only image data. The error-only image data are then scaled and combined with the original image data to remove the highly attenuating object artifacts.

Abstract: The present invention, in one form, is a method for modifying slice thickness during a helical scan without interrupting the scan. The method includes identifying adjacent and different regions within an object to be scanned. A transition region also is identified to include a portion of each of two adjacent regions and the interface therebetween. Slice thickness is modified during the scan so that redundant data is obtained in the transition region. Particularly, in one embodiment, a variable collimator is used to scan a first region with a first slice thickness. The variable collimator is rotated at the interface between the two adjacent regions, without interrupting table translation, to scan the second region with a second slice thickness. When changing the slice thickness, the collimator also is swept so that the x-ray beam with the second slice thickness re-scans a portion of the first region within the transition region.

Abstract: A correction algorithm for substantially eliminating "stair case" type artifacts in dental scans is described. In one specific embodiment, all high density object boundaries in the reformatted images are identified. To identify high density object boundaries, structures in the reformatted image are separated into two classes, namely, structure containing teeth and structure not containing teeth. Then, within each class, fuzzy logic is used to define the membership grade of each pixel. Particularly, linear interpolation is utilized to determine the boundary, and to reduce the probability that spike noise will be erroneously considered as high density objects, the N by N neighbors of the boundary candidate are searched to ensure that the number of pixels that belong to the high density object exceeds a certain predefined threshold. Such searching can be performed by summing the N by N neighbors of the membership function, .xi., and comparing the summation against a pre-defined threshold.

Abstract: A combination of double and triple cell ganging which resolves any incompatibility between the number of detector channels and the lower number of DAS channels without requiring any significant hardware and software changes is described. In one specific embodiment, at least some detector cells on one side of the detector outside the FOV are wired in pairs, i.e., ganged, to form a set of 2 mm channels, and on the other side of the detector outside the FOV, at least some detector cells are wired together, i.e., ganged, to form a set of 3 mm channels. Such ganging of detector cells avoids having to make any significant hardware and software changes to known multislice CT systems.

Abstract: A CT system having an adjustable pre-patient collimator and a partial volume artifact reduction algorithm is described. In one embodiment, a scout scan is performed to collect scout scan data. Utilizing the scout scan data, the algorithm identified the boundaries and radius of a partial volume artifact producing object. The algorithm then determines an average attenuation coefficient and an attenuation index of the object. The object radius, the attenuation coefficient and attenuation index are then graded to identify the object. A collimator aperture index is then determined based upon a variation of the attenuation characteristic along the z axis for the identified object. Utilizing the collimator aperture index, the CT system computer adjusts the collimator aperture for the appropriate size to eliminate, or significantly reduce, partial volume artifacts.

Abstract: A CT Fluoro system having an architecture and algorithms which facilitate increasing the frame rate and providing acceptable image quality is described. Generally, and in one embodiment, the system includes apparatus and algorithms that speed-up image reconstruction and reduce image artifacts that may result from such fast reconstruction. The fast reconstruction is achieved by performing, for example, view compression, channel compression, backprojection with reduced delay, and parallel processing.

Abstract: The present invention, in one form, is a system for generating a high resolution image of an object from projection data acquired during a computed tomography scan. The system includes a gantry having an x-ray source which rotates around the object. The x-ray source emits an x-ray beam which is collimated with a collimator having a collimator aperture to define an x-ray beam width, or slice thickness. The projection data is reconstructed to generate image data for adjacent image slices. A deconvolution algorithm is applied to the image data to generate a deconvolved image having a finer, i.e., smaller, resolution than the collimator aperture.

Abstract: The present invention, in one form, is a system for generating a high resolution image of an object from projection data acquired during a computed tomography scan. The system includes a gantry having an x-ray source which rotates around the object and emits an x-ray beam toward a detector. The system identifies a region of x-ray beam movement and divides the region into subregions. Linear Q-CAL vectors are then generated for each subregion so that each vector is representative of detector gain in one of the subregions. These Q-CAL vectors are then applied to projection data to generate image data.

Abstract: The present invention, in one form, is a system for reducing noise artifacts in three-dimensional image reconstruction using data acquired in a helical scan. More specifically, a standard deviation ratio of reconstructed images with and without helical weighting is identified. Such standard deviation ratio is then used to generate filter coefficients, which are then applied to the data in an adaptive smoothing algorithm.

Abstract: The present invention, in one form, is a method for detecting a partial volume artifact in scan data acquired during a tomographic scan. In accordance with one embodiment of the algorithm, an object of interest is scanned to generate scan data. Sampling pairs of the scan data are identified and used to generate an inconsistency map. The inconsistency map is weighted and filtered to generate a partial volume signature which identifies partial volume artifacts.

Abstract: The present invention, in one form, is a method for improving grey-white matter differentiation between regions of an image to be reconstructed from data obtained by a CT scan. More particularly, in accordance with one form of the present invention, a re-mapping function is utilized to generate CT numbers. In accordance with such function, CT numbers that are outside the grey-white matter region are not "stretched". The CT numbers within such region are "stretched" with the larger "stretch" centered on the grey-white matter region and tapering off at the boundary. Using such a function, grey-white matter differentiation is improved without adversely affecting quality and accuracy.

Abstract: The present invention, in one form, is a system for obtaining data measurement signals for producing a tomographic image of an object in a multislice scan. More specifically, detector cells in at least one channel of a detector array are electrically coupled, or ganged, so that the ganged channel provides one data measurement signal to be transmitted through the gantry slip ring. The signal distribution in the z-direction for the ganged channel is then determined using signals obtained by adjacent non-ganged detector cells.

Abstract: The present invention, in one form, is a method for improving image quality in Computed Tomography systems by using a partial volume artifact estimation algorithm. In accordance with one embodiment of the algorithm, an object of interest is scanned to generate image data. The image data is segmented into low attenuation data and high attenuation data. A gradient image is generated for two adjacent slices of the image data. The gradient image is then forward projected and squared. The squared gradient image is an estimation of partial volume artifacts in the image data, and therefore is removed from the image data.

Abstract: The present invention, in one form, is a system for performing image reconstruction from projection data acquired in a helical scan. More specifically, the system implements an incremental reconstruction algorithm for helical scan projection data which does not require filtering, weighting and backprojecting such projection data for generating each image. Particularly, an overscan weighting algorithm generates weighting factors to be applied to projection data to generate base image projection data. An update weighting algorithm generates update weighting factors to be applied to the base image projection data to generate subsequent image projection data.

Abstract: The present invention, in one form, is a system for modulating x-ray tube current as a function of gantry angle and slice location in a computed tomography system. In one embodiment, a desired noise level for a final image is selected, and a desired minimum x-ray photon reading and a desired average x-ray photon reading are identified to produce an image in accordance with the desired noise level. During scanning, actual x-ray photon readings are used with the desired average x-ray photon reading and the desired minimum x-ray photon reading to generate an x-ray modulating factor. This modulating factor is then used to modulate the x-ray tube current.

Abstract: Methods and apparatus for scanning an object in a computed tomography system during an interventional procedure are described. The computed tomography system includes an x-ray source, a detector, and a display. The detector detects x-rays projected from the x-ray source and attenuated by an object. A processor is coupled to the detector and coupled to the display for generating images of the object on the display. A helical scan is executed to generate an image slice of the object corresponding to each gantry rotation. At least one image slice and one three-dimensional image are simultaneously displayed on the display.

Abstract: Apparatus and methods for generating z-axis profiles for a CT system detector are described. In one form, the method includes the steps of directing x-ray beams having different z-axis centroids and slice thicknesses at the detector and collecting detector signals for each beam. The detector signal for a first beam is then subtracted from the detector signal for a second beam to obtain a differential, or composite, detector signal which corresponds to a third beam having yet another z-axis centroid and slice thickness. The full z-axis profile of the detector is generated from measured detector signals and composite detector signals.

Abstract: Methods and apparatus for performing a computed tomography scan are described. In one embodiment of the apparatus, a twin beam computed tomography scanner includes a beam splitter, an x-ray source for generating an x-ray to be projected generally towards, and at least partially through, an object, and a detector array comprising a plurality of detector cells arranged to form at least two cell rows. The beam splitter is positioned so that the x-ray projected from the x-ray source is substantially split to form at least two beams prior to being projected at least partially through the object.

Abstract: The present invention, in one form, is a method for enhancing image sharpness in images generated from CT scan data by using enhancement masks. The enhancement masks are generated, in one embodiment, by generating difference image data from the original age data and low pass filtered image data. The original image data CT numbers are assigned to image regions, e.g., bone, air, and soft tissue, and based on such CT number classifications, certain data in the difference image is fully or partially suppressed. Subsequent to suppressing some difference image data, the difference image data set, which is sometimes referred to as an enhancement mask, is then combined With the original image data to increase image sharpness.

Abstract: Precise alignment of the focal spot position on an x-ray CT system is achieved using a deflection coil that produces a magnetic field which acts on the electron beam path in the x-ray tube. A variable current power supply drives the deflection coil and is controlled by input signals to align the focal spot at a static reference position, to correct for focal spot drift between scans, and to wobble the focal spot position during a scan or between scans.