Surgery – Diagnostic testing – Detecting nuclear – electromagnetic – or ultrasonic radiation
Reexamination Certificate
1998-05-13
2001-03-13
Jaworski, Francis J. (Department: 3737)
Surgery
Diagnostic testing
Detecting nuclear, electromagnetic, or ultrasonic radiation
Reexamination Certificate
active
06200267
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to apparatus and methods for improving an image produced by a medical ultrasound system. More specifically, the present invention uses an optical correlator to provide an optimization metric which is used to improve the image by adjusting the imaging parameters of the ultrasound system, or by processing the image data using the metric produced by the optical correlator.
BACKGROUND OF THE INVENTION
Ultrasound imaging is widely used for diagnosis in numerous medical fields. When properly used and adjusted, an ultrasound imaging system can non-invasively provide a cross-sectional view within soft tissue being imaged, such as tissue of a breast, brain, heart, kidney, liver, lung, eye, abdomen, or pregnant uterus.
A typical ultrasound imaging device operates by directing short ultrasonic pulses, typically having a frequency in the range of 1-30 MHZ, into the tissue being examined. The device then detects echoes of the ultrasonic pulses caused by acoustic impedance discontinuities or reflecting surfaces within the tissue.
A typical scanhead for an ultrasound imaging system has a linear array of ultrasonic transducers, which transmit ultrasonic pulses, and detect returning echoes. The array of transducers provides simultaneous views of the tissue at positions roughly corresponding to the positions of the transducers. The delay time between transmitting a pulse and receiving an echo is indicative of the depth of the discontinuity or surface which caused the echo. The magnitude of the echo is plotted against the position and depth (or time) information to produce a cross-sectional view of the tissue in a plane perpendicular to the face of the scanhead.
There are many sources of distortion which may effect the quality of an ultrasound image. For example, the velocity of the ultrasonic pulses through tissue vary according to the tissue type. The propagation velocity of ultrasound ranges from approximately 1470 m/s in fat tissue, to more than 1600 m/s in muscle tissue, and as much as 3700 m/s in bone. Since the timing of an echo pulse is used to estimate the depth of a tissue feature, this variation in velocity of the pulse can lead to distortions along the depth axis of the image. This source of distortion is especially severe in tissue which is highly inhomogeneous, such as breast tissue.
Inhomogeneous tissues can also cause other distortions in the image produced by an ultrasound system, due to refraction effects. Additional sources of distortion may include scattering of the ultrasonic pulses, and interference between pulses from different transducers.
The distortions and noise present in an ultrasound image may be reduced by adjusting a number of parameters associated with the transmission of the pulses, the reception of the echoes, and the processing of the received echo data. For example, it is possible to apply differing time delays to the echo signals received from each of the transducers, to attempt to better “focus” the image produced.
There are a large number of parameters which must be adjusted to produce a good image from an ultrasound imaging system. These parameters may include the number of transducers which will be used for transmitting ultrasonic signals into the tissue, the wave shape to be transmitted by each transmitting transducer, the amplitude of the wave to be transmitted by each transmitting transducer, the transmit time delay to be used by each transmitting transducer, the number of transducers that will receive echoes, the gain of each receiving transducer, the receive time delay of each receiving transducer, and the filters to be applied to the incoming echo signals. There are complex relationships between these parameters, so that they cannot be optimized independently. Furthermore, the optimal parameter settings depend on the tissue being imaged, and vary with the individual patient and tissue type being examined.
In the past, these parameters have typically been adjusted by human operators, or have been selected from parameter sets which have been clinically determined to be acceptable for specific tissue types for a majority of patients. The obvious disadvantage of using a human operator is the time and skill required to make the adjustments. A disadvantage of using parameters which are selected from a file of known “acceptable” parameters is that the parameters are likely to produce better images in some patients than others.
Systems have been developed which attempt to automatically adjust various ultrasound imaging parameters to remove distortions and improve image quality. Such systems must provide a means of measuring the distortions, or of determining the image quality, so that the distortions may be corrected, or the image quality improved. Making these determinations may be computationally expensive, making it difficult for a system to correct the images in real-time. The speed at which such computationally expensive operations are carried out can be improved by using extremely high speed processors, parallel processing, or specialized hardware to perform the computations, all of which may greatly increase the cost of an ultrasound system.
The automatic ultrasound focussing systems described in U.S. Pat. Nos. 4,989,143 and 4,835,689, to O'Donnell, and U.S. Pat. No. 4,817,614, to Hassler attempt to use cross-correlation of the echo signals received by two or more of the transducers to derive information about the phase distortion of the image. That information is then used to adjust the delays associated with the transducers to improve the image. The cross correlation operations used by these methods are costly to compute and require high-performance equipment to achieve real-time performance.
An alternative approach, discussed in U.S. Pat. No. 4,852,577, to Smith et al., and U.S. Pat. No. 5,331,964, to Trahey et al., both of which are incorporated herein by reference, uses image processing techniques to derive an optimization metric for the image. The time delay parameters for each transducer are then altered, and the optimization metric is used to compare the quality of an image generated using the new parameters with the quality of an image generated using the old parameters. If the quality is improved, the system will continue adjusting the time delays in the same direction until a maximum image quality is obtained. Otherwise, the system will start adjusting the time delays in the opposite direction, and continue its adjustments until a maximum image quality is obtained. This maximizing process must be repeated for each of the transducers which are being adjusted.
The techniques described by Smith et al. and Trahey et al. require a considerable amount of computation. For each of the transducers being adjusted, it is necessary to apply the optimization metric numerous times to find the delay parameters which maximize the image quality. Each application of the optimization metric requires computation for each pixel in the image. If each image has N×M pixels, the maximizing procedure takes an average of K steps, and there are L transducers to adjust, the average number of computation steps required to optimize the image quality will be proportional to K×L×M×N. Even for small values of K, L, M, and N, this product can be very large.
Both Smith et al. and Trahey et al. attempt to keep the amount of computation under control by limiting the number of transducers which are adjusted, limiting the possible range of adjustments, and limiting the area of the image which is examined for the optimization metric to a small region of interest (ROI). For example, Smith et al. suggests using only a few transducers, and examining only five lines out of the ROI to compute the optimization metric. Smith et al. and Trahey et al. also attempt to keep the amount of computation to a minimum by choosing image brightness of the ROI, which is very easy to compute, as the optimization metric, and by using specialized hardware to compute the image brightness.
Using image brightness as an opt
Fish & Neave
Jaworski Francis J.
Pisano Nicola A.
LandOfFree
High-speed ultrasound image improvement using an optical... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with High-speed ultrasound image improvement using an optical..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and High-speed ultrasound image improvement using an optical... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2544303