Siemens Medical Systems, Inc.

Siemens Medical Solutions Usa, Inc.

Dong-Chyuan Liu

Jin H Kim

Zoran B Banjanin

Christian Deforge

Ja-Il Koo

Steven R Marquis

Jian-Feng Chen

Ming Li

Stephen P Czenszak

Thomas B Lindh

Patrick L Sutcliffe

King-Yuen Wong

Seojoong Kwon

Craig Robinson

Bruce A McDermott

A system and a method for three dimensional (3D) volume editing for medical imaging applications is disclosed. The method and system comprise the steps of projecting vertices of the at least one ROI in one plane and transforming the data within the at least one ROI to allow all of a plurality of slices on the inside of the at least one ROI to be along one axis of a three axis coordinate system. The method and system also includes representing the inside of the at least one ROI as a plurality of line segments, wherein only two coordinates and the length of a line segment are stored. The system and method is particularly useful for removing unwanted structure by defining one or more regions of interest (ROI) in freehand from an arbitrary slice of a 3D data volume. In a preferred embodiment, the system and method comprises a fast algorithm to process a set of compact ROI volumes to remove either the inside or the outside ROI volumes. Because the computational complexity is linear to the boundary of the ROI, the system and method can allow 3D volume editing to be faster and more user interactive.

Method and system for three-dimensional volume editing for medical imaging applications

A 3-D region of interest (ROI) of a body is imaged, preferably using an ultrasound transducer. The intensity of an echo signal from each of a pattern of pixels within the ROI is measured and stored. The pixels are then grouped into pixel blocks, and the intensity values for the pixels in each sequentially selected current block are compiled in a current histogram. Each current histogram is compared with a reference histogram that corresponds to the distribution of echo signal intensities in a known region of speckle. A structure likelihood value is then computed based on the comparison, where the structure likelihood value indicates how likely it is that the current block corresponds to a structure of interest as opposed to noise. The intensity value for any pixel likely to be noise is preferably scaled down. The scaled intensity values are then projected, preferably using a maximum intensity projection, onto a 2-D display plane, a representation of which is then displayed for a user to see. The invention is able to image structures throughout the 3-D ROI even when they would otherwise be obscured by the speckle.

3-D ultrasound imaging system and method

An ultrasound imaging system produces three-dimensional tissue/flow images by first computing a plurality of separate two-dimensional tissue images and two-dimensional flow images. Separate tissue and flow volumes are created by applying the tissue and flow images to a three-dimensional construction algorithm. Each of the separate tissue and flow volumes is analyzed using a three-dimensional rendering algorithm to produce a rendered tissue and flow image. The separately rendered tissue and flow images are combined to produce the combined tissue/flow image. In addition, the present invention provides visual cues that allow a user to create more even scans. The invention also includes a method for correcting for probe movement. The invention also produces an image that is calculated from the partial volume data as it is being created to give a user feedback on the quality of the image they are producing. Finally, the invention compensates for variations in the cardiac cycle when selecting data to be used in creating a flow volume.

Three-dimensional tissue/flow ultrasound imaging system

A method (280) and system (128) for selectively smoothing color flow images in a Doppler ultrasound system (100). A set of original digitized ultrasound signals is received for points (m, m-1, m-2) on a beamline (l-1), each point having associated with it a velocity signal (.phi..sub.j), a magnitude of autocorrelation signal (R.sub.i), and a power signal (P.sub.i). A point (X(1,1)) along a beamline is selected for processing. The set of digitized ultrasound signals is applied to a digital filter (152), and the digital filter computes a set of smoothed digitized ultrasound signals by computing an average velocity signal (.phi..sub.av), an average magnitude of autocorrelation signal (R.sub.av), and an average power signal (P.sub.av) of each point in the selected beamline and points in adjacent beamlines. For the selected point along the selected beamline, a selected velocity signal (.phi.(1,1)) is compared to a velocity threshold (V.sub.th), a selected magnitude of autocorrelation signal (R(1,1)) is compared to a magnitude of autocorrelation threshold (R.sub.th), and a selected power signal (P(1,1)) is compared to a power threshold (P.sub.th). Color flow information is computed from the smoothed digitized ultrasound signals when each of the selected velocity signal, the selected magnitude of autocorrelation signal, and the selected power signal is less than the velocity threshold, the magnitude of autocorrelation threshold, and the power threshold, respectively. The color flow information is computed from the set of original digitized ultrasound signals when any one of the selected velocity signal, the selected magnitude of autocorrelation signal, or the selected power signal is greater than or equal to the velocity threshold, the magnitude of autocorrelation threshold, or the power threshold, respectively. The color flow information is overlaid onto an ultrasound image.

Method and system for selectively smoothing color flow images in an

A scan conversion method employing a graphical circle generator to calculate a circle having a predetermined radius in the raster domain. The resulting set of (x.sub.i y) points are assigned values from corresponding vector domain points using closest fit and interpolation methods. This process occurs for each value of r in the vector domain. Gaps in the pixel display resulting from quantization errors in the circle generation algorithm are accommodated via a gap interpolation method. In one embodiment, the Bresenham circle generator is employed.

Ultrasound scan conversion method

Ultrasound echo data is analyzed to compress jump noises by recalculating the value of data points whose magnitude exceeds the magnitude of a previous point by more than a predefined threshold. Any data that exceeds the threshold is recalculated according to the value of the slopes of the points that precede and after the point to be replaced. The value by which a data point is compressed is selected so that the DC component of the points after compression is nearly equal to the DC component of the points prior to the jump noise. In addition, to remove saturating clutter in a Doppler display or audio signal, the present invention determines if the power spectrum of the Doppler signals to be displayed exceeds a maximum power display and if so scales a portion of the I and Q signals that comprise the Doppler power spectrum by the ratio of the maximum power spectrum computed and the maximum that can be displayed.

Method and apparatus for noise suppression in a doppler ultrasound

An improved digital wall filter comprises an infinite impulse response (IIR) second order digital filter. The coefficients w(-1) and w(-2) are pre-initialized from a set of m samples of the input signal. The number of samples m is related to the ratio of a constant .gamma., selected between 0 and 1, and the feedback filter coefficients. To ensure maximum performance of the filter, the filter is periodically re-initialized before filtering a new set of input data. To produce a video line, the filtered data created from the input samples used to initialize the filter are averaged with data that was not used to initialize the filter.

Ultrasound doppler wall filter

Ultrasound gray-scale image data and corresponding processed gray-scale image data are color-enhanced to generate an enhanced multi-color image. At one step, a color map look-up table is defined from the processed gray-scale image data to define a color palette of K colors, (e.g., K=4-16). Histogram analysis, nonlinear mapping or segmentation processing is used to map the gray-scale levels to the color palette. The processed gray-scale image data then is transposed into a transposed color image data set. Data interpolation then is performed in which each pixel of the original gray-scale image data is interpolated with a corresponding pixel among the transposed color image data set. Interpolation weighting is adjustable by the user to determine how much detail from the original gray scale image is preserved.

Method for enhancing ultrasound image

The present invention is an ultrasound imaging system that displays images with improved dynamic range. An echo signal is sampled so that each sample contains at least N bits in order to create a digital echo signal having at least 2.sup.N possible values that determine a dynamic range of the digital echo signal. A single-color lookup table is generated having approximately 2.sup.N RGB component entries, each entry defining a pixel having a unique hue, intensity, and saturation value. Each sample of the digital echo signal is mapped to an RGB component entry of the single-color lookup table, and the mapped RGB components are displayed on the display to produce a single-color image having a dynamic range that is substantially equal to the dynamic range of the digital echo signal.

Method and apparatus for displaying an ultrasound image with an improved

A method for enhancing an ultrasound image. The image is assumed to be in the form of an ordered array of pixels. Each pixel corresponding to the intensity of the echo generated by a corresponding voxel in an ultrasound energy field. Axial and lateral directions relative to the ultrasound beam used to generate the sound field are defined within the ordered array. The method of the present invention includes the steps of dividing the image into a plurality of processing blocks, picking blocks that satisfy certain statistical constraints for edge enhancement, enhancing the pixels in the selected blocks, and then displaying the pixels in each block, whether enhanced or not, that are above a display threshold for the block in question. The statistical constraints are determined by second order statistics with reference to the axial and later directions. In one embodiment of the present invention, sum and difference histograms are used to approximate the second order statistics, thereby reducing the computational load.

Edge enhancement system for ultrasound images

An apparatus and method for removing noise from a signal S(x). The noise is removed by filtering S(x) through a low pass filter having an adjustable pass band. A control processor analyzes S(x) prior to S(x) being input to the low pass filter and adjusts the pass band in response to an estimate of the SNR or rate of change of S(x) such that the pass band is reduced when the estimated SNR decreases and the pass band is increased when the SNR increases. In embodiments in which the control processor utilizes the rate of change of S(x) to control the pass band, the pass band is reduced when the rate of change of S(x) increases and increased when the rate of change of S(x) decreases. In embodiments used in ultrasound applications, the estimate of the SNR can be obtained from x, the power in S(x), or the variance in the measured velocity. The estimate of the rate of change of S(x) may be computed from an estimate of the correlation distance in S(x).

Method and apparatus for adaptive spatial image filtering

An interrogation region, such as a portion of a patient's body, is imaged as an unfiltered pattern of image elements, to each of which is assigned a numerical value such as brightness. One element is selected either automatically or manually as a reference element, which lies in a substantially homogeneous image portion. For ultrasonic imaging, the homogeneous portion preferably corresponds to a region of speckle noise. A reference histogram of the numerical values is then accumulated for all the elements in a region or window about the reference element. All of the remaining elements are sequentially designated as current elements, for each of which a current histogram is accumulated. Each current histogram is compared with the reference histogram using an error function for each corresponding pair of histogram bins. The error function values for all bins, which may be normalized and weighted, are used to generate a similarity value between the current and reference histograms. The more similar to the reference histogram the current histogram is, the more strongly it is assumed that the current element lies in a homogeneous portion of the image, and the greater it is smoothed. All possible error function values may be prestored to increase computation speed. The degree of smoothing may be adjusted by the user. The user preferably also may adjust a comparison factor, which determines how strongly a difference between a current and a reference bin will contribute to the similarity value.

System and method for adaptive filtering of images based on similarity

A spatial-temporal doppler velocity estimation method expands the detectable frequency range by extending the frequency boundary at which aliasing artifacts may occur. In various embodiments the detectable frequency range is extended beyond the Nyquist limit. To do so, multiple velocity estimation functions are defined. One function is the conventional function typically bound by the Nyquist limit. For additional functions, the (n-1)-th echo is spatially shifted with respect to the n-th echo, (i.e., temporal/spatial pulse-pairs are averaged). For example, in one embodiment a shift of +.DELTA. is inserted to define one additional estimation function, while a shift of -.DELTA. is inserted to define another additional estimation function. A power function then is calculated for each velocity estimation function. The estimation function having the highest power is selected for use in deriving the doppler shift frequency of a given sample point. By selecting a one of the K+1 velocity estimation functions with the highest power, the detectable frequency range, f, becomes: -(2.sup.K /2)f.sub.s &lt;f&lt;(2.sup.K /2)f.sub.s, where f.sub.s is the Doppler sampling frequency.

Doppler ultrasound velocity estimation

A method for operating a digital computer to suppress reverberation artifacts in an ultrasound image. The image includes an ordered array of pixels, each the pixel corresponding to the intensity of the echo generated by a corresponding voxel in an ultrasound energy field. The ordered array has axial and lateral directions defined therein, the axial direction corresponding to voxels on the same radial line from a transducer used to generate the ultrasound energy field. The method starts by dividing the image into a plurality of segmentation blocks thereby generating an ordered array of segmentation blocks. The columns of the segmentation block array are chosen such that all segmentation blocks in the same column correspond to the same axial direction in the ultrasound image. In the preferred embodiment of the present invention, at least one segmentation block is classified as having a strong edge. The method of the present invention finds a first segmentation block that is classified as a strong edge, and then finds a second segmentation block which is not classified as a strong edge in the column containing the first segmentation block. A spatial frequency domain transformed block is then generated from a processing block containing the second segmentation block, and a modified transformed block generated from the spatial frequency domain transformed block by reducing the amplitude of selected peaks in the transformed block. Finally, a new second segmentation block is generated by computing the inverse spatial frequency transform of the modified transform block.

Adaptive artifact suppression for ultrasound imaging

Power values of backscattered ultrasonic echoes from an interrogation region of a patient's body are stored in a 2-D envelope table. The region is thereby represented as a 2-D pattern of image elements. In a processing mode, these elements are grouped into processing windows. After row averaging of the power values for each window, a window attenuation coefficient is generated corresponding to a decay parameter of a non-linear decay model of local attenuation of the interrogation region as a non-linear function of the average positional power values for each row in the processing window. The coefficients are then used to scale the power values before scan conversion and display, and thus to compensate for local attenuation. The invention can also test for tissue heterogeneity by first defining convergence and non-convergence criteria. A rationalized gain control processor then iteratively estimates the window attenuation coefficients until either criterion is met. The RGC processor then indicates heterogeneity of tissue in the corresponding processing window of the interrogation when the non-convergence criterion is met.

Growing community of inventors

Mercer Island, WA, United States of America

Dong-Chyuan Liu