Image analysis – Image enhancement or restoration – Artifact removal or suppression
Reexamination Certificate
1999-05-07
2003-01-28
Lee, Thomas D. (Department: 2724)
Image analysis
Image enhancement or restoration
Artifact removal or suppression
C382S128000, C600S437000
Reexamination Certificate
active
06512854
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to a procedure for improving displayed ultrasound images and corresponding audio output and, more particularly, to a method and apparatus for adaptively deciding whether image information is noise or signals containing information and for selectively enhancing the information signals in relation to the noise.
BACKGROUND OF THE INVENTION
Ultrasound (U/S) imaging systems are widely used in the medical profession to diagnose a range of pathologies. Information from within the body is acquired primarily by transmission, backscattering, reception and subsequent processing of an acoustic signal. The acoustic signal contains information relevant to the sonofied region of interest. Sensing and subsequent amplification of the acoustic backscatter introduces an undesirable component in the signal stream referred to hereafter as noise.
The condition where the informational component is appreciably larger than the noise component is referred to as “strong signal”; where the informational component and noise are similar as “weak signal”; and where the noise is significantly larger than the informational component as “noise” or “noise-only”. Further, the signal to noise ratio is the ratio of the power of the informational part of the signal to the power in the noise component.
Often, the relevant U/S information is presented on a display amidst a background of interfering noise. When the informational component of the processed acoustic signal is weak relative to the accompanying noise, i.e., a weak signal, evaluation of the data can be difficult and time consuming. That is, it can be difficult for the user to reliably evaluate or diagnose the patient condition in weak signal situations. Spectral Doppler imaging is an important modality of U/S that provides information about blood flow. For spectral Doppler imaging, U/S system performance is sensitive to the transmit power level, the receive gain, the velocity scale and the selected gray-scale function, to name a few of the relevant system controls. In present U/S systems, the operator must make appropriate adjustments to these controls to avoid compromising system performance.
The common method for presenting the U/S data for evaluation is visual display. Generally, the visual (video) display is produced by encoding the information in a black and white (gray-scale intensity) or color format. The relevant information to be displayed is extracted from the “back-scattered” acoustic wave. Commonly in medical applications it is the back-scattered power alone or power in conjunction with the phase that serves as the measurement, displayed visually or output audibly, from which diagnoses are derived. In the spectral Doppler modality, the U/S data is also presented in an audio format as a stereophonic audible output.
Customarily, the sensed acoustic signal spans a wide range of values in response to backscattering and propagation effects within the body. This range of values of signal power is referred to as the dynamic range of the received signal. Most systems afford the operator an ability to adjust the transmit power level and receive gain to accommodate the back-scattering and propagation effects affecting dynamic range.
The range of intensities achievable by a video display is also referred to as the dynamic range or intensity range of the display. Most systems afford the operator the ability to adjust the compression (i.e., mapping function) of the dynamic range of the received signal power to the dynamic range afforded by the display system. For example, a signal value may be represented by a sixteen bit value ranging from zero to approximately 64,000, whereas the video display may support only a dynamic range of eight bits, i.e., 0 to 255. The compression function maps the 12 bit signal dynamic range to the 8 bit dynamic range of the display. This compression function is generally non-linear and is intended to enhance the presentation of the information relative to the noise.
For example, in a black and white (gray-scale) image, the measured value of the received acoustic data is encoded and displayed, using the higher levels of intensity to display information signals. The lower levels of display intensity are allocated to noise-only values. Non-linear mapping attempts to map the range of noise values into a disproportionately small region at the low (weakly visible) end of the dynamic range of the display, thereby leaving the large remainder of the dynamic range to signal values, ranging from weak to strong.
A gray scale or color map is used to convert a measurement value to be displayed (signal amplitude, for example) into either a black and white (gray-scale intensity) or a color video signal. That is, the color map or gray scale contains the transformation information necessary to map the measurement signal into the video signal that controls display illumination on a pixel by pixel basis.
Among the design objectives of a gray-scale conversion function are: (1) an appropriate mapping of the range of U/S measurements values to the range of intensities achievable by the display monitor, and (2) the mapping of values, containing diagnostic information, to intensities that are distinguishable from values consisting primarily of noise. Presently the user of an U/S system controls video display of the processed acoustic signal by selecting one of a number of stored color maps (i.e., compression curves) in conjunction with the transmit power and receive gain settings. The color maps/gray scales are designed to cover a range of weak signal, strong signal and noise conditions, in conjunction with subjective preferences attributable to the user. However, often the imaging is done with system settings, a transmit power level, receive gain and color map or compression curve which do not employ the dynamic range of the display system effectively and hence, do not achieve the desired quality of presentation. Furthermore, the values of the weak signals are similar to those of the competing noise. Thus, delineation of weak signals from noise is difficult, even when the system settings are adjusted properly.
There is a need for improved U/S presentation of information data in the presence of noise. Further, there it is desired to divorce control of the display from signal intensities that are solely based on individual measurement samples.
SUMMARY OF THE INVENTION
The invention implements a method for adaptively using raw image data (before compression) to produce pixel display values in an ultrasound system in accord with variations in signal-to-noise ratio. Initially, for each subject pixel of an image, a first regional value is derived by taking into account values of a neighborhood of pixels that includes the subject pixel. The first regional value is then compared to a threshold value related to noise and each subject pixel is assigned a classification as a tentative signal pixel if the regional value equals or exceeds the threshold value. This conditional assignment is based on the conclusion that the values within the neighborhood have a higher probability of being a signal population than a noise population. Any pixel value not so classified is considered a tentative noise pixel. A second regional value is then derived for each subject pixel by taking into account first regional classifications of a second neighborhood of pixels that includes the subject pixel. If the number of pixels classified as signal in the second neighborhood is less than a second threshold, the subject pixel is classified as noise. Otherwise the subject pixel is classified as a signal pixel. The actual values of the subject pixels are then adjusted to improve a display of the ultrasound image and the corresponding audio presentation, in the case of Doppler. The method also alters the threshold noise value in accordance with second regional pixel values classified as noise pixels.
REFERENCES:
patent: 5497777 (1996-03-01), Abdel-Malek et al.
patent: 5574212 (1996-11-01), Madsen et al.
patent: 5619998 (1997-04-01), Abdel-Ma
Miele Frank R
Mucci Ronald
Koninklijke Philips Electronics , N.V.
Lee Thomas D.
Vodopia John
LandOfFree
Adaptive control and signal enhancement of an ultrasound... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Adaptive control and signal enhancement of an ultrasound..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Adaptive control and signal enhancement of an ultrasound... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3072169