Surgery – Diagnostic testing – Detecting nuclear – electromagnetic – or ultrasonic radiation
Reexamination Certificate
1999-03-12
2001-06-12
Jaworski, Francis J. (Department: 3737)
Surgery
Diagnostic testing
Detecting nuclear, electromagnetic, or ultrasonic radiation
Reexamination Certificate
active
06245016
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to ultrasound imaging of the human anatomy for the purpose of medical diagnosis and, more particularly, to methods and apparatus for imaging fluid or tissue in the human body by transmitting ultrasound waves into the fluid or tissue and then detecting ultrasound echoes reflected therefrom.
BACKGROUND OF THE INVENTION
Conventional ultrasound scanners are capable of operating in different imaging modes, such as B mode and color flow mode. In the B mode, two-dimensional images are generated in which the brightness of a display pixel is based on the intensity of the echo return.
In conventional ultrasound imaging operating in B mode, an ultrasound transducer array transmits a series of multi-cycle (typically 2 or 3 cycles) tone bursts which are focused at different transmit focal positions, while dynamic receive delays are applied. These tone bursts are fired at a pulse repetition frequency (PRF) that is typically in the kilohertz range. Each transmit beam propagates through the object being scanned and is reflected by ultrasound scatterers in the object. The returned radio frequency (RF) signals are detected by the elements of the transducer array and then transformed into in-phase and quadrature (I/Q) components and formed into a receive beam. The transformation into I/Q components is accomplished by bandpass filtering either the separate RF signals in the receive channels or the beamsummed RF signal. The I/Q components are then shifted down in frequency and sent to a B-mode processor which incorporates a detector, e.g., an envelope detector, for forming the envelope of the beamsummed receive signal by computing the quantity (I
2
+Q
2
)
½
. The envelope of the signal undergoes some additional B-mode processing, such as logarithmic compression, to form display data that are interpolated by a scan converter into X-Y format for video display. A video processor maps the video data to a gray scale or mapping for video display and sends the gray scale image frames to a video monitor for display.
The safety, versatility and real-time capability of ultrasound imaging equipment has made it an extremely valuable diagnostic tool. However, due to frequency-dependent pulse attenuation, the non-stationary point spread functions of the imaging systems, and the coherent nature of the ultrasound imaging process, is B-mode images have poor resolution and contrast in comparison to other medical imaging modalities, such as computed tomography and magnetic resonance imaging, and contain significant speckle that masks tissue structure at smaller scales. There are two ways to deal with these problems—process more data in generating a single image through techniques such as frequency or spatial compounding, or do more processing at the back-end of the beamformer to take advantage of known tissue characteristics and spatial correlation, in the underlying reflectivity profile. The former can improve ultrasound imagery at the expense of increased system complexity and lower frame rate, whereas the latter can improve it at the expense of increased computational load.
Known post-beamformer signal processing methods for enhancing B-mode imagery can be divided into two groups, depending on whether they process the beamformer signal directly or the envelope-detected imagery.
Representative examples of methods in the latter class include the nonlinear filtering method described in Kotropoulos et al., “Optimum Nonlinear Signal Detection and Estimation in the Presence of Ultrasonic Speckle”, Ultrasonic Imaging, vol. 14, pp. 249-275 (1992), in which a multiplicative noise model is used to derive a local estimator of signal intensity; the adaptive methods in Loupas et al. “An Adaptive Weighted Median Filter for Speckle Suppression in Medical Ultrasonic Images”, IEEE Trans. on Circuits and Systems, Vol. 36, pp. 129-135 (1989), and Koo et al., “Speckle Reduction with Edge Preservation in Medical Ultrasonic Images Using a Homogeneous Region Growing Mean Filter”, Ultrasonic Imaging, Vol. 13, pp. 211-237 (1991), in which the window size of median and averaging filters are set based on similarity of local statistics; the adaptive filtering method in Bamber et al., “Adaptive Filtering For Reduction Of Speckle In Ultrasonic Pulse-Echo Images”, Ultrasonics, Vol. 24, pp. 41-44 (1986), in which the degree of smoothing is dependent on a statistic quantifying the degree of speckle formation; and the related approach in Dutt et al., “Adaptive Speckle Reduction Filter For Log-Compressed B-Scan Images”, IEEE Trans. on Med. Imag., Vol. 15, pp. 802-813 (1996), in which the log-compressed imagery is filtered. While these methods generally result in images in which speckle is significantly reduced, they typically do not provide significant contrast or resolution improvement, and in many cases result in images of substantially reduced resolution.
On the other hand, methods based on processing the beamformer signal directly, typically referred to as deconvolution methods, can result in resolution improvement if adequate prior system information can be obtained. The methods are generally based on a one-dimensional or two-dimensional convolutional model of the beamformer signal acquisition process, e.g.,
y
(
t
)=
h
(
t
)*
w
(
t
)+
v
(
t
) (1)
where y(t) is the beamformer output, h(t) is the pulse response of the imaging system (including the transducer pulse along with the axial and/or lateral impulse response of the transducer itself), w(t) is the underlying tissue reflectance (usually assumed to be a white noise process), and v(t) is additive white noise. Convolutional models can be justified by assuming weak scattering (no bones or air pockets), single scattering (no reverberations or multipath reflections) and no absorption (see Jensen, “A Model For The Propagation And Scattering Of Ultrasound In Tissue”, J. Acoust. Soc. Am., Vol. 89, pp. 182-190 (1990), for discussion and references). The improved resolution is obtained because additional system information—including the intensity of the additive noise v(t) and reflection sequence w(t) and the pulse response h(t)—is exploited. While the resulting algorithms are typically not highly sensitive to parameter mismatch, reasonable parameter estimates are required to obtain significant increases in resolution and, just as important from a clinician's perspective, to avoid the introduction of spurious artifacts, especially at tissue boundaries.
Deconvolution filters generally have one or more of the following limitations:
(1) Such filters are depth invariant or piecewise invariant and as a result do not take into account the depth-varying structure of the transducer impulse response resulting from frequency-dependent pulse attenuation and transducer dynamic focusing and aperture effects. Wiener filtering approaches such as that in Liu et al., “Digital Processing For Improvement Of Ultrasonic Abdominal Images”, IEEE Trans. Med. Imag., Vol. 2, pp. 66-75 (1983), and Hoess et al., “Adaptive Wiener Filtering For B-Mode Image Improvement”, IEEE Ultrasonics Symposium, pp. 1219-1222 (1992), are explicitly based on a depth-invariant pulse response, and are typically applied by first segmenting the data and then applying the filter to each segment separately. The Kalman filtering methods proposed in Jensen, “Deconvolution Of Ultrasound Images”, Ultrasonic Imaging, Vol. 14, pp. 1-15 (1992) (see also Kuc, “Application Of Kalman Filtering Techniques To Diagnostic Ultrasound”, Ultrasonic Imaging, Vol. 1, pp. 105-120 (1979)), allow for continuously depth-varying pulse responses; however the authors do not take advantage of this because they estimate their pulses directly from the data using windows of fixed size and spacing.
(2) Deconvolution filters do not take into account abrupt changes in attenuation and scattering characteristics across tissue boundaries. In the Jensen and Hoess papers cited above, variations in scatterer density and strength and pulse attenuation are accounted for by estimating filter parameters
Daft Christopher Mark William
Luettgen Mark Robert
Tenney Robert Ronald
General Electric Company
Jaworski Francis J.
Snyder Marvin
Stoner Douglas E.
LandOfFree
Ultrasound imaging system having post-beamformer signal... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Ultrasound imaging system having post-beamformer signal..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Ultrasound imaging system having post-beamformer signal... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2543490