Image analysis – Applications – Biomedical applications
Reexamination Certificate
2000-05-08
2004-03-09
Johns, Andrew W. (Department: 2621)
Image analysis
Applications
Biomedical applications
C128S915000, C708S404000, C342S196000, C324S076190
Reexamination Certificate
active
06704438
ABSTRACT:
FIELD OF INVENTION
This invention relates primarily to the field of coherent imaging and, more specifically, to an apparatus and method for improving the signal to noise ratio in a coherent imaging system using coded waveforms.
BACKGROUND OF INVENTION
In ultrasound imaging a transducer is used first to transmit ultrasound waves in the medium to be examined, for example a region of the human body, nd then to receive the ultrasound echoes from various discontinuities in the medium and to transform them into electrical signals. The received electrical signals are then processed in various ways, e.g. amplified, filtered, beamformed, detected, and eventually transformed into a set of digital values (pixels) that can be displayed on an electronic display such as a cathode ray tube (CRT), or can be used to produce a photograph. One of the major limitations of ultrasound imaging and of other coherent imaging systems is electrical noise. When the signal to noise ratio (SNR) is low, the useful information may be totally or partially covered by noise. Manufacturers make a great effort to design low noise systems, however, a certain level of thermal noise is present in any electronic system. Therefore, once this minimal noise level has been attained, the only way to improve the SNR is to increase the signal, and this is achieved by increasing the amount of transmitted energy.
Increasing the signal's amplitude and/or its duration will increase the signal energy. However, there are limits to how much the amplitude of the signal can be increased. For example, in radar there are practical implementation limits to the maximum amplitude; and in medical ultrasound the limits have to do with the safety of the patient, since high sound amplitude may cause tissue damage. In order to increase the signal energy without increasing its amplitude, a long duration waveform must be transmitted. If this long waveform were a simple sinusoidal burst then the signal bandwidth would decrease compared to the short pulse, causing the axial resolution to decrease proportionally. Therefore, the transmitted signal must be frequency modulated—hence the name ‘coded waveform’—such that the long waveform's bandwidth is kept equal to or larger than that of the conventional short pulse. This is roughly equivalent to transmitting a different instantaneous frequency at each transmission period.
When the transmitted signal is a coded waveform, the received echo signal is processed with a ‘pulse compression filter’. A pulse compression filter applies a different phase delay to each frequency component such that all frequencies are phase aligned to constructively add the frequencies, and thereby produce a short, high-amplitude ‘compressed’ pulse. The method of coded waveforms is well known and described in many communication theory and radar textbooks, for example
Radar Principles
by Peyton Z. Peebles, Jr., John Wiley & Sons, Inc., 1998. Specifics of applying the method to medical ultrasound have also been published, for example by M. O'Donnell in the article ‘Coded excitation systems for improving the penetration of real-time phased-array imaging systems’, IEEE Transactions UFFC, Vol. 39, No. 3, May 1992.
A first difficulty in implementing the coded waveform method is the cost of the pulse compression filter. In the high performance applications of concern, the pulse compression filter is implemented as a digital FIR (finite impulse response) filter, whose cost increases with the number of filter taps (i.e., the number of samples in the filter's impulse response). The number of filter taps is the product of the filter duration and the sampling rate, and may be over 512 taps for medical ultrasound and even larger for radar applications. When a long data stream must be filtered in real time, requiring one multiplier for each filter tap, the filter cost becomes prohibitive.
It is well known that for filtering a long data stream it is more efficient to perform the convolution in the frequency domain: a FFT (fast Fourier transform) is first applied to the data, the transforms of the data and of the filter are then multiplied and the result is inversely transformed (IFFT) to produce the filtered signal. Even though this approach reduces the computational load, the task is still very difficult at the data rates encountered in many applications.
A known method to reduce the number of computations is to perform the FFT on two sets of data in parallel by applying one set of data to the real input of an FFT processor and another set of data to the imaginary input of the FFT processor. After transformation the FFT's of the two signals are separated then multiplied by the filter's FFT, and inversely transformed separately. This method reduces the total amount of computations by a little over 20%, not a significant improvement.
Another method to reduce the number of computations needed for pulse compressing the rf coded waveforms is to demodulate the rf signals to baseband and decimate them before pulse compression, thus reducing the sampling rate. However in ultrasound applications, where the useful bandwidth is over 50% of the center frequency, this method produces little if any savings, and may result in poor axial profiles due to additional difficulties in designing the proper pulse compression filter. It is therefore desirable to provide a method to reduce the number of computations needed for pulse compressing the rf coded waveforms.
A second problem of coded waveform systems is caused by the distortion of the signal due to nonlinearities in the transmit and receive circuits and in the medium, and due to frequency dependent attenuation in the medium. The pulse compression filter is designed for a specific waveform shape and its performance deteriorates when applied to distorted waveforms. It is therefore desirable to provide a simple method to compensate for such waveform distortions.
SUMMARY OF THE INVENTION
In a coherent imaging system capable of performing coded waveform imaging, a pulse compression filter is implemented using frequency domain convolution. The pulse compression filter is applied to the real rf data (before quadrature demodulation). The filter's frequency characteristics are designed to correspond to a real impulse response. This allows two real signals to be packed into one complex signal that can be filtered as such without any intermediate operations, thus reducing the number of computations to half. Packing and unpacking circuits are provided at the input and output of the filter.
In one embodiment, the filtering is applied to conveniently sized segments of the signal, corresponding to different depth regions. The frequency domain filter is provided with buffers for multiple filter frequency responses, each of which can be applied to the appropriate depth segment. The various filter frequency responses may be obtained in a calibration process performed off-line before the actual use of the apparatus.
REFERENCES:
patent: 4063549 (1977-12-01), Beretsky et al.
patent: 4593161 (1986-06-01), Desblache et al.
patent: 4950999 (1990-08-01), Agnello et al.
patent: 5379642 (1995-01-01), Reckwerdt et al.
patent: 5455590 (1995-10-01), Collins et al.
patent: 5502747 (1996-03-01), McGrath
patent: 5790475 (1998-08-01), Marmarelis et al.
patent: 5831934 (1998-11-01), Gill et al.
patent: 5886749 (1999-03-01), Williams et al.
patent: 5930305 (1999-07-01), Leib
patent: 6104670 (2000-08-01), Hossack et al.
patent: 6104822 (2000-08-01), Melanson et al.
patent: 6185594 (2001-02-01), Hilton et al.
patent: 6343510 (2002-02-01), Neeson et al.
patent: 6466277 (2002-10-01), McNeely
“Radar Principles”, Peyton Z. Peebles, Jr., Chapter 7, “Pulse Compression With Radar Signals”.
“Coded Excitation System for Improving the Penetration of Real-Time Phased-Array Imaging Systems”, Matthew O'Donnell, Senior Member, IEEE. “IEEE Transactions On Ultrasonics, Ferroelectrics, And Frequency Control”, vol. 39, No., 3, May 1992.
Aloka Co., Ltd.
Bachman & LaPointe P.C.
Johns Andrew W.
Nakhjavan Shervin
LandOfFree
Apparatus and method for improving the signal to noise ratio... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus and method for improving the signal to noise ratio..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and method for improving the signal to noise ratio... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3202496