JP4382884B2 - Ultrasonic image processing method and apparatus using harmonics - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
  The present invention relates to an ultrasonic diagnostic method and a body image processing method, and more particularly to a novel method and apparatus for ultrasonic diagnostic image processing using a response frequency different from a transmission frequency, which was filed on September 27, 1996. This is based on a continuation-in-part of US Patent Application Serial Number 08 / 723,483, which is “Ultrasound diagnostic image processing method using contrast medium”.
[0002]
[Prior art]
  Ultrasound diagnostic image processing apparatuses have been used for image processing of the body due to the improvement of image quality using ultrasonic contrast agents. A contrast agent is a material that is biocompatible and has specially selected acoustic properties that return an easily identifiable echo signal in response to sonication. Contrast agents have several properties that allow improved image quality of ultrasound images. One is the non-linear characteristics of many contrast agents. When irradiated with ultrasound of one frequency, the contrast agent is manufactured to exhibit a resonance mode that returns energy at other frequencies, particularly harmonic frequencies. When irradiated at the fundamental frequency, the harmonic contrast agent returns the second, third, fourth and even higher harmonics of that frequency.
[0003]
  It has long been known that tissues and body fluids also inherently have non-linear characteristics. Tissues and body fluids create and send back echo response signals at frequencies that are not their own fundamental frequency, including harmonic signals at the fundamental frequency, even in the absence of contrast agent. Muir and Carstensen studied these properties of water at the beginning of 1980, and Starritt et al. Found these properties in human calf muscle and were clipped. The study was conducted on the cattle liver.
[0004]
[Problems to be solved by the invention]
  Although these non-fundamental frequency echo components of tissue and body fluids are usually not as large as the harmonic components returned by the harmonic contrast agent, they can be beneficially used in ultrasound imaging. Shows a number of features. One of the inventors (M. Averkiou) has conducted extensive research on these properties in the studies described in the doctoral thesis. In this research and other studies, we found that the main lobe of the harmonic beam was narrower than its fundamental frequency lobe, and they found that image processing through a narrow mouth such as the ribs. Is the possibility of scattering reduction. They found that the side lobe level of the harmonic beam was lower than the corresponding side lobe level of the fundamental beam, and found the possibility of reducing off-axis scattering. They also found that the return of harmonics from the near region is relatively smaller than the return energy at the fundamental frequency, and found the possibility of preventing near region scattering. As described below, these characteristics are utilized in the method and configuration examples of the present invention.
[0005]
[Means for Solving the Problems]
  According to the present invention, there is provided an ultrasonic image processing apparatus and method for performing image processing of tissue and body fluid from a response frequency different from the transmission frequency returning from the tissue or body fluid, in particular, a harmonic echo of the fundamental transmission frequency. This image processing apparatus comprises means for transmitting ultrasonic waves of fundamental frequency, means for receiving echoes of harmonic frequencies, and an image processor for creating an ultrasonic image from the harmonic frequency echoes. In a preferred embodiment of the invention, the transmitting and receiving means consists of a single ultrasonic probe. Another feature of the present invention is that the probe utilizes a broadband ultrasonic transducer for both transmission and reception.
[0006]
  A further feature of the present invention is that a partially decorrelated component of the received harmonic echo is created and used to remove a virtual image from the harmonic image to clearly identify tissue boundaries such as endocardial images. It is to provide possible images. In the preferred embodiment, partially uncorrelated components are created from harmonic echoes at different passbands. The method of the present invention includes the use of harmonic echoes to reduce near-field or multipath scatter in an ultrasound image, for example, an image created during image processing through a narrow acoustic window such as a rib. . Additional features of the present invention include noise reduction by blending harmonics and fundamental frequency echoes into a common image, image processing at significant depth, and overcoming depth dependent attenuation effects.
[0007]
  FIG. 1 is a block diagram illustrating an ultrasonic diagnostic image processing apparatus according to the present invention. 2, 3, 4, and 5 illustrate some properties of harmonic echo that can be beneficially applied to ultrasound imaging applications. 6 and 7 are passband characteristic curves used to explain the behavior of the example of FIG. FIG. 8 illustrates the typical fundamental and harmonic frequency passbands of the present example. FIG. 9 illustrates a suitable FIR filter structure for use in the example of FIG. FIG. 10 is a block diagram illustrating a preferred embodiment of the present invention. FIG. 11 illustrates the operation of the normalization stage of the example of FIG. FIG. 12 is a block diagram of one of the multiplier accumulators used in the example filter of FIG. FIG. 13 illustrates the typical fundamental and harmonic frequency passbands of the example of FIG. FIG. 14 illustrates blending of fundamental and harmonic signal components into one ultrasound image. FIG. 15 illustrates the passband of the time varying filter used for blended image formation.
[0008]
BEST MODE FOR CARRYING OUT THE INVENTION
  Referring initially to FIG. 1, an ultrasound diagnostic image processing apparatus constructed in accordance with the present invention is shown in a block diagram format. The central controller 120 instructs the transmission frequency control 117 to transmit a desired transmission frequency band. Transmission frequency band variable, f_trAre sent to the transmission frequency control 117 to cause the transducer 112 of the ultrasonic probe 110 to emit ultrasonic waves in the fundamental frequency band. In the configuration example, a frequency band having a center frequency of approximately 1.67 MHz is transmitted. This is lower than the conventional transmission image processing frequency, which was normally in the range of 2.5 MHz or more. However, the use of typical transmit frequencies of 3 or 5 MHz creates 6 and 10 MHz harmonics. These high frequency harmonics will experience substantial attenuation while they return to the probe, since high frequencies are attenuated more significantly through the body than low frequencies. Harmonic signals are generated as the transmitted wave travels through the tissue, so it does not fade as much as the fundamental frequency signal that experiences attenuation due to the round trip from the transducer is attenuated, but this reduces the penetration depth, The image quality at a larger image processing depth is deteriorated. In order to overcome this problem, the center oscillation frequency in the illustrative example is 5 MHz or less, preferably 2.5 MHz or less, thereby forming lower frequency harmonics that are less susceptible to depth dependent attenuation, Then, harmonic image processing at a deeper location is possible. An outgoing fundamental frequency of 1.67 MHz creates a return signal of the second harmonic of 3.34 MHz in the illustrative example. Of course, it will be understood that any frequency of ultrasound can be used with due consideration of the desired penetration depth, transducer and sensitivity of the ultrasound device.
[0009]
  Array transducer 112 of probe 110 emits ultrasonic energy and receives echoes returned in response to the transmission. The response characteristics of the converter show two passbands, one near the fundamental transmission frequency and one near the harmonic frequency in the reception passband. For harmonic image processing, a broadband converter having a passband that includes both the outgoing fundamental frequency and the received harmonic passband is preferred. The transducer is manufactured and tuned to show a response characteristic as shown in FIG. 6, where the lower hump 60 of the response characteristic curve is the outgoing fundamental frequency f._rNear the center and the higher hump 62 is the received harmonic frequency f of the response passband._rThere is a center in the vicinity. However, the transducer response characteristic curve of FIG. 7 is preferred because the single principal characteristic curve 64 allows the probe to be suitable for both harmonic image processing and conventional broadband image processing. The characteristic curve 64 indicates the transmission fundamental frequency f._tAnd the frequency f_LAnd f_cAnd the frequency f_rThe harmonic reception passband centered around is also included. As mentioned above, a fundamental oscillation frequency as low as 1.67 MHz forms a harmonic return echo signal with a frequency of 3.34 MHz. A response characteristic curve 64 of approximately 2 MHz is adequate for these fundamental and harmonic frequencies.
[0010]
  Body tissues and cells change the fundamental frequency signal transmitted while in progress, and the return echo contains harmonic components of the fundamental frequency originally transmitted. In FIG. 1, these echoes are received by converter array 112, transferred through T / R switch 114, and digitized by analog-to-digital converter 115. Sampling frequency f of A / D converter 115_sAre controlled by a central controller. The desired sampling rate indicated by the sampling theory is the highest frequency f in the received passband._cThe frequency in the previous example is at least about 8 MHz. A sampling rate higher than the minimum required is also preferred.
[0011]
  The echo signal samples from the individual transducer elements are delayed and summed by the beamformer 116 to form a coherent echo signal. The in-phase digital echo signal is then filtered by a digital filter 118. In this example, the transmission frequency f_tIs irrelevant to the receiver, and therefore the receiver can freely receive a frequency band different from the transmitted band. The digital filter 118 has a frequency f in FIG._LAnd f_cThe signal in the passband delimited by can be bandpass filtered and also moved to a lower or baseband frequency range. The digital filter in the above example has a 1 MHz passband and can be a filter with a center frequency of 3.34 MHz. A preferred digital filter is a series of multipliers 70-73 and accumulators 80-83 as shown in FIG. This arrangement is controlled by the central controller 120, and performs multiplier weighting and decimation control that controls the characteristic curve of the digital filter. Preferably, this arrangement is controlled to act as a finite impulse response (FIR) filter to perform both filtering and decimation. For example, only the first stage output 1 can be controlled to act as a 4: 1 decimation ratio 4-tap FIR filter. The temporally discontinuous echo sample S is transferred to the first stage multiplier 70. As samples S are transferred, they are multiplied by the weight given by the central controller 120. Each of these products is stored in accumulator 80 until four such products are accumulated (added). An output signal is then created with a first stage output 1. Since the accumulated sum consists of four weighted samples, this output signal is filtered by a 4-tap FIR filter. Since a time for 4 samples is required to create the output signal, a 4: 1 decimation ratio is achieved. One output signal is formed for every four input samples. The accumulator is cleared and this process is repeated. It can be seen that when the decimation ratio is high (the interval between output signals is long), the number of effective taps of the filter increases.
[0012]
  If necessary, the time-divided samples are delayed by a delay element τ, applied to four multipliers 70-73, multiplied, and accumulated in accumulators 80-83. After each accumulator accumulates two products, the four output signals are combined into one output signal. That is, the filter operates as an 8-tap filter having a 2: 1 decimation ratio. Without decimation, this arrangement can operate as a 4-tap FIR filter. The filter can also be operated by transferring the echo signal to all multipliers simultaneously and by selectively ordering the weighting factors. Under the control of the central controller, filter weighting and decimation ratio programming allows a full range of filter characteristic curves. The use of digital filters has the advantage that it can be changed quickly and easily to provide different filter characteristic curves. The digital filter can be programmed to first pass the received fundamental frequency and then pass the harmonic frequency. Thus, the digital filter can be operated to alternately create images or scan lines of the fundamental and harmonic digital signals, and the time-interleaved sequence by simply changing the filter coefficients during signal processing. Can be operated to alternately generate scan lines of different harmonics.
[0013]
  Returning to FIG. 1, in order to perform image processing only on frequencies that are not the fundamental frequency, the digital filter 118 is controlled by the central controller 120, the fundamental frequency is excluded, and the echo signal of the harmonic frequency is passed and processed. Harmonic echo signals from the tissue are detected and processed by either the B-mode processor 37 or the contrast signal detector 128 for display as a two-dimensional ultrasound image on the display 50. The filtered echo signal from the digital filter 118 is also transferred to the Doppler processor 130 for conventional Doppler processing, forming a speed and power Doppler signal. The outputs of these processors are transferred to the 3D image creation processor 162 for 3D image creation, which is stored in the 3D image memory 164. These three-dimensional representations can be performed as described in US patents [Application Serial No. 08 / 638,710] and US Pat. Nos. 5,474,073 and 5,485,842, and the latter two One patent describes three-dimensional power Doppler ultrasound image processing techniques. The contrast signal detector 128, the signals from the processors 37 and 130, and the three-dimensional image signal are transferred to the video processor 140, where the signal is two-dimensional or three-dimensional on the image display 50, according to instructions selected by the user. Selected for display.
[0014]
  It has been found that harmonic imaging of tissue and blood can reduce the scattering of nearby regions in the ultrasound image. It is believed that the harmonic response effect of the tissue depends on the energy level of the transmitted wave. Near the array transducers that are focused deeper, the transmitted wave components are not focused and there is insufficient energy to generate a harmonic response that can be detected in close-range tissue. However, because the outgoing wave continues to invade the body,OutgoingAs the wave component begins to focus, the higher intensity energy creates harmonic effects. While both the near and remote regions return fundamental frequency responses, the scatter from these signals is eliminated by the passband of the digital filter 118 set to the harmonic frequency band. The harmonic response from the tissue is then detected and displayed, while the scatter from the near region fundamental frequency response is eliminated from the displayed image.
[0015]
  2, 3, 4 and 5 illustrate some characteristics of the harmonic return signal that can be beneficially used for ultrasound imaging. These properties, and some of their interactions, are not yet fully and generally understood by scientists and are still considered for research and discussion. FIG. 2 illustrates the spatial response of the fundamental frequency and harmonic signals received by the transducer array 112, particularly the main and side lobes. In this example, the array is instructed to image a body region behind the ribs, such as the heart, and the main lobe appears to spread between ribs 10 and 10 '. Lying on the ribs is a tissue boundary 12 such as a fat layer between the skin and the ribs. The figure shows the main lobe of the basic signal FL1, and there are side lobes FL2FL3 on either side of the main lobe. The figure also shows the harmonic main lobe HL1 of the fundamental frequency and the side lobes HL2 and HL3 of the harmonic main lobe.
[0016]
  In this example, it can be seen that the main lobe of the basic echo is enlarged so as to include the ribs 10, 10 ′. Therefore, the fundamental frequency acoustic energy is reflected back to the transducer 112 as indicated by arrow 9. On the other hand, some of this reflected energy returns to the transducer and is received directly, but in this example, some of the reflected energy is reflected again by the tissue boundary 12, as shown by arrow 9 '. This second energy reflection reaches the other rib 10 'where it reflects again as indicated by arrow 9 "and returns to the transducer 112 for reception.
[0017]
  Since the intent of this image processing technique is to image the heart behind the ribs, these echoes reflected by the ribs are unwanted virtual images that contaminate the ultrasound image. Undesired echoes that have been reflected many times before reaching the transducer, such as echoes following the path of arrows 9, 9′9 ″, are called multipath artifacts. This is called “scattering” and obscure nearby areas, in some cases the entire image. This near area fogging or scattering may obscure the structure of interest near the transducer. In addition, multipath virtual images may be regenerated in deeper images due to the lengthy number of paths that these virtual images follow, causing noise and obscuring deeper areas of interest There is. However, if only the harmonic return signal is used to create the ultrasound image, the scattering from this fundamental frequency is filtered out. Since the main lobe HL1 of the received harmonic echo is narrower than the lobe of the fundamental frequency, in this example, it passes between the ribs 10, 10 'without intersecting them. There is no harmonic return from the ribs and no multipath virtual images from the ribs. Accordingly, the harmonic image in this example is an image that is less scattered and less cloudy than the fundamental frequency image, particularly in the proximity region.
[0018]
  FIG. 3 shows a second example in which the main lobes of both fundamental frequency and harmonic return do not intersect the ribs and the problem discussed in FIG. 2 does not occur. However, in this example, ribs 10, 10 ′ are closer to the skin surface and transducer 112. The main lobe does not intersect the rib, but the fundamental frequency side lobe FL2 reaches the rib and reflects the side lobe energy back to the transducer, as shown in reflection path 9. Again, this causes scattering in the fundamental frequency image. However, the smaller and narrower sidelobe HL2 of the received harmonic energy does not reach the ribs. Again, the harmonic image is displayed with less scatter than the fundamental frequency image.
[0019]
  FIG. 4 illustrates the fundamental frequency and harmonic beam patterns in the perspective view across the lobes of FIGS. 2 and 3, i.e., across the transducer axis. This figure illustrates the relative amplitude response of the fundamental frequency and the second harmonic beam shape. Dynamic response DRF between the main lobe (FL1) and the first side lobe (FL2) of the fundamental component of the acoustic beam, and between the main lobe (HL1) and the first side lobe (HL2) of the second harmonic component The dynamic response DRH is described. If the response due to the main lobe is regarded as the desired signal response and the response due to the side lobe is scattered or noise, the harmonic signal-to-noise ratio is greater than that at the fundamental frequency. That is, the sidelobe scattering is relatively smaller in the harmonic image than in the corresponding fundamental frequency image of the same transmission, or DRH> DRF.
[0020]
  FIG. 5 illustrates another comparison of the characteristics of the fundamental and harmonic signals, which is the energy emanating from deeper locations Z in the body at the fundamental and second harmonic frequencies (in units of acoustic pressure P). Relative amount of Fund. The curve labeled with indicates that the advanced acoustic energy at the fundamental frequency has been built up. It can be seen that the curve reaches the apex at the focus of the array transducer, but it can be seen that there is still a perceivable amount of fundamental energy in the shallow part before the focus area. In comparison, there is relatively much less energy and less energy enhancement at higher harmonic frequencies traveling in these shallow regions. Therefore, because less energy is used for multipath echo and other anomalies, harmonic image processing has less near-field scatter than image processing of the return of the fundamental frequency echo from the same transmission.
[0021]
  FIG. 8 illustrates a typical digital filter and received signal band of the example of FIG. 1 of the present invention in a 4-cycle transmission signal of 1.67 MHz sound waves. Multi-cycle transmission narrows the bandwidth of the transmitted signal; the greater the number of cycles, the narrower the bandwidth. In response to this transmission, the converter 112 receives a fundamental signal with a bandwidth of 90, which is seen to reach the peak at a transmission frequency of 1.67 MHz. As the fundamental frequency band decreases, it can be seen that the harmonic band 92 rises and peaks again at the harmonic frequency of 3.34 MHz. The received signal is transferred to a digital filter with a passband characteristic curve 94, which is found to be centered around a harmonic frequency of 3.34 MHz. As FIG. 8 shows, this passband essentially suppresses the fundamental frequency signal while allowing the harmonic signal to pass through for further processing and imaging. When imaging the heart in this way, the harmonic response of the heart's intracardiac tissue is very substantial, and the heart's harmonic tissue image is seen to show clearly identified boundaries within the heart. It was issued.
[0022]
  Other signal processing techniques other than filtering can be used to separate and extract the harmonic signal from the received echo information, such as separating the fundamental frequency in the wideband signal and leaving only the harmonic frequency. For example, US patent [application SN08 / 728,318] discloses a two-pulse technique whereby each scan line is rapidly and sequentially irradiated with continuous fundamental frequency pulses of opposite phase. When the resulting echo is received from two pulses and combined on a spatial basis, the fundamental frequency is removed and a non-linear or harmonic frequency remains. Thus, the harmonic frequency is separated from the wideband echo signal without the need for a filter circuit.
[0023]
  FIG. 10 is a block diagram showing a part of a preferred embodiment of the present invention, from the beamformer output to the image display device. This example not only creates a harmonic image of the tissue and blood flow, but also overcomes the signal drop-out drawbacks of conventional image processing devices that occur when imaging patients who are difficult to image a medical condition. In addition, in this example, the virtual image of the phased ultrasound image known as speckle is reduced. In FIG. 10, since the illustrative processor is fully digital, all signal and data lines connected to the block of the block diagram represent a multi-conductor digital data path. Scan line echo data from the beamformer 116 is sent in parallel to the two channels 30a, 30b of the processor described in FIG. 10, one of which is the high frequency channel and the other is the low frequency channel. Each channel of the processor has a normalization stage 32, 132, which integrates the scan line data with a scale factor for each sample and provides different gains or attenuations depending on the body depth to which each sample returns. Form. The scale factor for each channel is provided by a normalization factor created or stored by coefficient circuits 32, 132, which in the preferred embodiment are digital memories. Since the integration factor varies along a series of scan line echoes, a depth dependent gain or attenuation is created.
[0024]
  There are two functions in the normalization stage. One is to compensate for the effect of the transducer aperture expanded by the scan depth. As signals from increasing numbers of transducers are used with increasing depth, the magnitude of the summed beamformed signal increases. This increase is offset by a decrease in gain (increase in attenuation) in the normalization phase proportional to the rate at which the circuit is added to the beamforming process, so that the resulting echo sequence is not affected by the changing aperture.
[0025]
  The second function of the normalization stage is the two channels 30a, 30bNominal(nominal) To make the signal amplitudes equal. Of the passband of the two channelsNominalThe signal amplitudes are preferably equal and the original relative signal level is maintained after the passbands are summed to create the total harmonic passband. However, since ultrasonic signals have depth dependent attenuation that varies with frequency, high frequency signals are more attenuated by depth than low frequencies. Considering this depth dependent attenuation, the normalization factor gives a signal gain that increases with depth. Since the two channels use different frequency passbands, the depth dependent gains of the two channels are different from each other. In particular, the rate of gain increase for high frequency passband channels is greater than that of low frequency passband channels. This will be described with reference to FIG. 11. For the sake of explanation, a normalized gain characteristic curve of a high-frequency passband channel divided into two components is shown. The depth dependent characteristic curve 200 cancels out the effects of increasing apertures in the channel, and the depth dependent characteristic curve 202 compensates for the depth dependent signal attenuation. Low frequency passband channels also have depth dependent gain characteristics, but have different characteristic curves 202 due to different attenuation rates at lower frequencies. High frequency passband channels have a similar but more rapidly increasing depth dependent gain characteristic curve due to a faster decay rate due to higher frequencies. Each depth dependent gain characteristic curve 202 is selected to cancel the depth dependent gain effect for the particular frequency passband used by the channel.
[0026]
  In the preferred embodiment, the coefficient of the coefficient circuit applies a gain or attenuation characteristic curve that combines two characteristic curves 200 and 202. Preferably, the coefficient memories 32, 132 store multi-coupled gain curves that change with memory addressing to suit the scan head characteristics or the type of signal being processed (2D or Doppler). The rate of gain change can be controlled by the rate at which the coefficients change for each normalization stage 30,130 multiplier.
[0027]
  The channel's normalized echo signal is forwarded to each channel's quadrature bandpass filter (QBP). The quadrature bandpass filter performs three functions: band limiting of RF scan line data, creation of in-phase and quadrature pairs of scan line data, and reduction of digital sample speed. Each QBP consists of two separate filters, one creating an in-phase sample (I) and the other creating a quadrature sample (Q), each filter performing multiple FIR filters The multiplier-accumulator (MAC). One such MAC is shown in FIG. When an echo sample of scan line data is sent to one input of the digital multiplier 210, the coefficients are sent to another multiplier input. The product of the echo sample and the weighting factor is stored in accumulator 212 and accumulated with the previous product. Other MACs receive echo samples of different phases and accumulate similarly weighted echo samples. Several MAC accumulated outputs are combined, and the final accumulated product consists of filtered echo data. The rate at which the accumulated output is collected sets the decimation ratio of the filter. The filter length is the product of the decimation ratio and the number of MACs used to form the filter, and determines the number of input echo samples used to create the accumulated output signal. The filter characteristic is determined by the coefficient value of multiplication. Different coefficient sets for different filter functions are stored in the coefficient memories 38, 138 and transferred to the MAC multiplier where the selected coefficients are applied. The MAC is a coefficient expressed in sine and cosine, which effectively convolves the received echo signal and creates an output sample in quadrature relationship.
[0028]
  The coefficients for the MAC forming the I filter perform a sine function, while the coefficients for the Q filter perform a cosine function. For bandpass filtering, the coefficients of the active QBPs are further frequency shifted to perform a low bandpass filter function that forms a bandpass filter for quadrature samples with sine (for I) and cosine (for Q) functions. To do. In this example, the QBP of channel 30a₁Produces the first, low frequency passband scanline data I and Q samples, and QBP of channel 30b₂Creates a second, high frequency passband scan line data I and Q sample. Therefore, the spectrum of the initial wideband echo signal is divided into a high frequency band and a low frequency band. To complete the shedding and speckle removal process, the QBP of channel 30a₁The echo data in the passband created by the₁, And the detected signal is transferred to one input of the summer 48. In a preferred embodiment, detection is performed using the following algorithm:
[0029]
(I²+ Q²)^1/2
[0030]
Is performed digitally by calculating. QBP for channel 30b₂Echo data in the supplementary passband generated by₂These detected signals are transferred to the second input of the summer 48. When the two passband signals are combined by the summer 48, the uncorrelated signal dropout and speckle effects in the two passbands are at least partially removed, and the signal dropout and spec in the 2D image created from the signal. Le virtual image is reduced.
[0031]
  Following the detector in each subchannel, the coefficient memory 42₁, 42₂Multiplier 44 that receives weighting coefficients from₁, 44₂There is a gain stage formed by. The purpose of this gain stage is to balance the analog and digital gain in the ultrasound device for best system performance. Some gains in the echo signal path are automatically performed by the ultrasound device, while other manual gain controls and TGC gains can be controlled by the user. The device distributes these gains and adjusts the analog gain before the beamformer ADCs (analog to digital converters) to optimize the dynamic input range of the ADCs. The digital gain is adjusted to optimize the brightness of the image. Both gains perform a gain adjustment change made by the user.
[0032]
  In the preferred embodiment, multiplier 44.₁, 44₂The gain imparted to the scan line signal by is selected to coordinate with the gain of previous normalization stages 34, 134 in the channel. The gain of each normalization step is selected to prevent achieving saturation levels with QBPs, such as occurs when a strong signal from contrast agent or harmonic imaging is received. To avoid saturation levels, the maximum gain of the normalization stage is controlled, and for this controlSectionAny reduction made is followed by a multiplier 44.₁, 44₂Recovered by the gain. The gain function performed by these multipliers can be performed anywhere along the digital signal processing path. It can be done by changing the slope of the compression curve discussed below. It can also be performed, for example, in connection with the gain given in the normalization stage. However, this latter implementation deletes the ability to perform the saturation control described above. The inventors have found that the ability to perform this gain function is canceled when applied after detection, preferably by use of a multiplier after detection.
[0033]
  Gain stage 44₁, 44₂The signal produced by, generally exhibits a larger dynamic range than that provided by the display 50. As a result, the multiplier scan line signal is compressed to an appropriate dynamic range by a look-up table. In general, a logarithmic compression processor 46₁, 46₂As indicated by, the compression is logarithmic compression. The output of each lookup table is proportional to the logarithm of the signal input value. These look-up tables can be programmed to change the compression curve and the brightness and dynamic range of the scan line signal sent to the display.
[0034]
  The use of logarithmic compression for echo signal normalization can adversely affect low level signals near the baseline (black) level of the signal dynamic range due to the degradation of the degree and number of echoes in the black level component, i.e. The present inventors have found that a destructive failure manifests from the speckle effect of coherent ultrasonic energy. When echo signals are displayed, many of them are at the black level and will not be detected or missing from the image. The example of FIG. 10 alleviates this problem by creating echo signals that are separated and partially decorrelated in the two channels 30a, 30b. In this example, as shown in FIG. 13, the echo signal version is partially decorrelated by dividing the echo signal component into two different passbands. The two passbands can be completely separated or may overlap as shown in this example. In this example, the low pass band 300a is centered around the 3.1 MHz frequency, the high pass band 300b is centered around the 3.3 MHz frequency, and the center frequency separation is only 200 kHz. Even this small separation sufficiently decorrelates the signal components of the two passbands so that the black level signal dropout of one passband is often not aligned in frequency with its corresponding component in the other passband It turns out to be enough to do. Thus, when these decorrelated replicas of the same echo signal are combined by the summer 48, signal drop and speckle virtual images will be significantly reduced. This is particularly important when trying to image fine structures in deep parts of the body, such as the endocardium. The harmonic image of the intracardiac curtain is significantly improved by the virtual image removal effect of the example of FIG.
[0035]
  As discussed above, the signal gains of the two passbands 300a, 300b of FIG. 13 can be adjusted to preserve the original signal level after summing. However, in the preferred embodiment, the low frequency passband is processed with a narrower dynamic range than the high frequency passband, as shown in FIG. This has the effect of suppressing the contribution of the fundamental frequency in the low frequency passband (which contains more fundamental frequency components than in the high frequency band). This is because, as a component of different compression characteristics, the logarithmic compression processor 46 after separating the wideband signal into split passbands₁, 46₂Or in other strokes in the channels 30a, 30b.
[0036]
  The processed echo signal at the output of the summer 48 is transferred to the low frequency pass filter 52. This low-frequency pass filter is configured by a multiplier-accumulator combination with variable coefficients arranged to perform an FIR filter, such as QBPs, and controls the filter characteristics. The low frequency pass filter performs two functions. One is to remove the sampling frequency and other unwanted high frequency components from the processed echo signal. The second function is to match the scan line data rate to the vertical line density of the display 50 to prevent aliasing in the displayed image. The FIR filter performs this function by selectively decimating or interpolating scan line data. The filtered echo signal is then stored in the image memory 54. If the scan lines have not yet been scan converted, i.e. they have r, theta coordinates, the scan lines are scan converted to rectilinear coordinates by the scan converter and grayscale mapping processor 56. . If scan conversion has been performed in the previous step, or if scan conversion is not required for the image data, processor 56 simply converts the echo data into the desired grayscale map by a look-up table process. The image data is then stored in a final image memory or sent to a video display driver (not shown) for conversion into a display signal suitable for driving the display 50.
[0037]
  Due to the rapid programmability of the digital filter, the above process utilizes one of the channels 30a, 30b by alternately creating a row of signals for each of the two passbands in a time-interleaved manner. Thus, it will be understood that it can be implemented in an embodiment where echo data from a scan line is processed twice. However, the use of two parallel channels allows the creation of harmonic images at twice the frame rate in real-time, as it has twice the processing speed.
[0038]
  The harmonic image created from the high frequency signal may deteriorate due to depth-dependent attenuation as the part where the echo signal returns becomes deeper in the body. Lower frequency fundamental signals have less attenuation and may therefore exhibit better signal-to-noise ratios at deeper sites. The example of FIG. 14 takes advantage of this property by blending fundamental frequency and harmonic image data in one image. For example, a normal tissue image of the heart can be created from the fundamental frequency, and a harmonic tissue image of the heart can be overlaid on the fundamental frequency tissue image to more clearly identify the intracardiac boundary in this composite image Can do. Two images, one from the fundamental frequency component and the other from the harmonic frequency component, are fundamental and by switching the digital filter 118 between the fundamental and harmonic frequencies alternately. Formed by combining the harmonic images separately or using the two parallel filters of FIG. 10 with two passbands, one passing the fundamental frequency and the other passing the harmonic frequency. be able to. In FIG. 14, the filter of the channel 30 a is set to pass the fundamental signal frequency, and the echo signal that has passed through this channel is stored in the fundamental image memory 182. Correspondingly, the harmonic signal frequency passes through channel 30b and is stored in the harmonic image memory. The fundamental and harmonic images are then blended by proportional coupler 190 under the control of blend control 192. The blend control 192 automatically executes a pre-programmed blend algorithm or executes a user instruction. For example, the proportional coupler 190 uses only echo data from the harmonic image at a shallow depth, then combines the echo data from both images at an intermediate depth, and finally at the deeper part of the base image. Create a mixed image using only echo data. This provides a smooth transition from one data type to the other at intermediate depths, with less scatter of harmonic echo data in shallow regions, more fundamental frequency echoes received from deep regions. It combines a large penetration capability with a signal-to-noise ratio. For example, simply switch from one data type to another at a pre-determined depth, or outline the displayed image area with one data type, and change the remaining image to the other data type. Other combining algorithms are also possible, such as using and displaying.
[0039]
  It is also possible to blend the components prior to image construction using two parallel filters, thereby adding a controllable component of the harmonic echo signal to the fundamental frequency signal to improve the quality of the resulting image. In such an example, the need to store the fundamental and harmonic images in separate memories can be eliminated, and the signal components are processed directly into the blended image memory.
[0040]
  A third technique for creating a blended image is to receive each scan line of the image through a depth-dependent, time-varying filter. Such filters are known to improve the signal to noise ratio of the received echo signal in the presence of depth dependent attenuation, as shown, for example, in US Pat. No. 4,016,750. For the creation of the blended fundamental and harmonic images, the passband 210 of the time-varying filter initially receives the echo signal from a shallow portion, so that the harmonic frequency f is shown in FIG._hIs set to pass. When it becomes desirable to begin supplementing the image with the fundamental signal component at a deeper location, the passband 210 moves to a lower frequency, and finally the fundamental frequency f as shown in the passband 212 of FIG._fMove to. In the digital filter as shown in FIG. 9, the change of the passband frequency is executed by changing the filter coefficient with time. When the filter performs this movement, if desired, the passband passes only a few harmonic frequencies and many fundamental frequencies until the passband finally passes only the fundamental frequency at the deepest image depth. By receiving each scan line through such a time-varying filter, each line of the resulting image has a harmonic frequency in the near region (shallow region), a fundamental frequency in the remote region (deepest region), and the middle. Then, it will consist of a blend of both.
[0041]
  A fourth approach to creating a blended image is to send and receive twice along each scan line. One transmission is at the fundamental frequency, followed by reception of harmonic frequency echoes. The other transmissions are at the fundamental frequency, followed by reception of the fundamental frequency. The two fundamental transmission frequencies are the same, but different fundamental frequencies may be used if desired. The harmonic and fundamental frequency echoes are then combined along the scan line in the desired proportions to form a blended scan line and an image area of such a scan line is created to form a blended image. . It is moving by processing the received harmonic tissue echo signal with a processor as described in US patent [Application 08 / 655,394] entitled “Medical Ultrasonic Power Motion Image Processing Method”. A harmonic tissue image of the tissue can also be formed.
[0042]
  Accordingly, the present invention provides a non-basic frequency response to a non-fundamental, preferably harmonic, frequency tissue, and non-fundamental, non-fundamental, non-fundamental, non-fundamental responses to ultrasound of body tissues and fluids. It includes an ultrasonic image processing apparatus that detects an echo signal and performs image processing by forming an image of tissue and body fluid from the non-basic frequency echo signal. Because the principles of the invention described herein are equally applicable to higher and subharmonic frequencies, the term harmonic used herein is much higher than the second harmonic. The next harmonic frequency also means subharmonic.
[0043]
  The present invention discloses an ultrasonic diagnostic image processing method using a response frequency different from the transmission frequency. The present invention relates to an ultrasonic diagnostic image processing apparatus and method for creating an ultrasonic image from a harmonic echo component of a transmission fundamental frequency. About. Preferably, a programmable digital filter is used to pass the harmonic echo components to remove the fundamental frequency signal and process the image. In the preferred embodiment, the virtual image is removed by creating an uncorrelated replica of the harmonic signal, which is then combined and used for image processing. To create an image in the presence of depth-dependent attenuation of the high frequency echo signal, both the fundamental and harmonic echo signals are processed to create a blended image from both fundamental and harmonic echo signal components Used to.
[0044]
  Aspects of the present invention are as follows.
1. An ultrasonic diagnostic image processing apparatus for image processing a harmonic response of a body structure, wherein the ultrasonic energy is transmitted into the body at a fundamental frequency; in response to the transmitted ultrasound energy, at a harmonic frequency of the fundamental frequency 1. an ultrasonic diagnostic image processing apparatus comprising: means for receiving an ultrasonic echo signal; and means for creating an ultrasonic image from the harmonic echo signal; 2. the ultrasonic diagnostic image processing apparatus according to 1 above, wherein the transmitting means and the receiving means comprise an ultrasonic transducer probe; The ultrasonic diagnostic image processing apparatus according to the second aspect, wherein the ultrasonic transducer probe includes a plurality of transducer elements for transmitting ultrasonic energy having a fundamental frequency and receiving an ultrasonic echo signal having a harmonic having the fundamental frequency. 4. 4. The ultrasonic diagnostic image processing apparatus according to 3, wherein the transducer element exhibits a response characteristic including both the fundamental frequency and harmonics of the fundamental frequency. 5. the ultrasonic diagnostic image processing apparatus according to 1 above, wherein the means for receiving an ultrasonic echo signal of a harmonic of the fundamental frequency comprises a filter having a pass band for identifying the harmonic frequency excluding the fundamental frequency; 6. The ultrasonic diagnostic image processing apparatus according to 5 above, wherein the filter is a programmable digital filter; 7. the ultrasonic diagnostic image processing apparatus according to 1 above, wherein the ultrasonic image generating means includes a B-mode processor; 8. The ultrasonic diagnostic image processing apparatus according to 7 above, wherein the B-mode processor has an amplitude detector for detecting an envelope of the harmonic echo signal. 9. The ultrasonic diagnostic image processing apparatus according to 1 above, wherein the structure is a structure in which a body exists. 9. The ultrasonic diagnostic image processing apparatus according to 9 above, wherein the naturally existing structure includes body tissues and cells; A method for creating an ultrasound image from a harmonic response within a body comprising: transmitting ultrasonic energy into the body at a fundamental frequency; receiving an ultrasound echo signal at a harmonic frequency of the fundamental frequency; and 11. a method of generating an ultrasonic image display signal by processing a wave echo signal to generate an ultrasonic image display signal; and displaying the ultrasonic image display signal. 11. The eleventh method, wherein the transmitting and receiving steps comprise transmitting fundamental frequency ultrasonic energy and receiving harmonic echoes using an ultrasonic probe having a transducer array; 13. The method of claim 12, wherein using the ultrasonic probe comprises transmitting fundamental frequency ultrasonic energy and receiving harmonic echo signals with the same transducer element. The step of receiving the ultrasonic echo signal of the harmonic frequency of the fundamental frequency passes the received ultrasonic echo signal by a filter that excludes the fundamental frequency and passes the signal of the harmonic frequency of the fundamental frequency. 11. The eleventh method comprising 15. The eleventh method, wherein the processing step comprises B-mode processing of the harmonic echo signal; 16. The fifteen method as described above, wherein the step of the B-mode processing includes the step of detecting the amplitude of the harmonic echo signal. An ultrasonic diagnostic image processing apparatus for generating an ultrasonic image of a harmonic response of a body structure with reduced virtual images: means for transmitting ultrasonic energy into the body at a fundamental frequency; Means for receiving an ultrasonic echo signal at a harmonic frequency of the fundamental frequency; processing the harmonic ultrasonic echo signal to form an at least partially uncorrelated replica of the echo signal Means for combining the decorrelated replicas to create a harmonic echo signal with reduced virtual images; and generating an ultrasound image using the harmonic echo signals with reduced virtual images An ultrasonic diagnostic image processing apparatus comprising: 18. The seventeen ultrasonic diagnostic image processing apparatus, wherein the virtual image is a dropped virtual image; 18. The ultrasonic diagnostic image processing apparatus according to the above 18, wherein the virtual image further includes a speckle virtual image; 17. The ultrasonic diagnostic image processing apparatus according to 17, wherein the processing means includes a band-pass filter that divides a component of the harmonic ultrasonic echo signal into two pass bands having different center frequencies. 20. The ultrasonic diagnostic image processing apparatus according to 20, wherein the means for processing further includes a detector for detecting a harmonic ultrasonic echo signal in each of the passbands. 21. The ultrasonic diagnostic image processing apparatus according to 21, wherein the means for processing further comprises a logarithmic compression processor for logarithmically compressing the detected harmonic ultrasonic echo signal. The means for processing has two parallel channels each having an input connected to receive a harmonic ultrasonic echo signal and an output connected to the coupling means, wherein each of the channels is connected to the other 17. The ultrasonic diagnostic image processing apparatus according to the above 17, having a band-pass filter having a filter characteristic different from the filter characteristic of the channel; 24. the ultrasonic diagnostic image processing apparatus according to 23, wherein the filter characteristic is a peak response frequency of the filter; 23. the ultrasonic diagnostic image processing apparatus according to 23, wherein the filter characteristic is a center frequency of the filter; The ultrasonic diagnostic image processing apparatus according to 23, wherein each of the channels further includes a detector; 26. The ultrasonic diagnostic image processing device of claim 26, wherein each of the channels further includes a logarithmic compression processor; 17. The seventeen ultrasonic diagnostic image processing apparatus, wherein the processing means comprises means for dividing the harmonic ultrasonic echo into two pass bands having unequal dynamic ranges; The 28 ultrasonic diagnostic images, wherein the two passbands are composed of a low-frequency passband and a high-frequency passband, and the dynamic range of the low-frequency passband is smaller than the dynamic range of the high-frequency passband. Processing device, 30. 29. The ultrasonic diagnostic image processing apparatus according to 29, wherein the processing means further includes means for mapping different dynamic ranges of the passbands. A method for creating a harmonic ultrasound image with reduced virtual images comprising: transmitting ultrasonic energy at a fundamental frequency; receiving an ultrasound echo signal at a harmonic frequency of the fundamental frequency; and decorrelating the signal Processing the harmonic echo signal to create a replica; combining the decorrelated replicas to create a harmonic echo signal with reduced virtual images; and a harmonic echo signal with reduced virtual images 32. A method for creating a harmonic ultrasound image comprising the above steps, wherein an ultrasound image is created using 32. The method of 31, wherein the step of processing the harmonic echo signal comprises dividing the components of the harmonic echo signal into two different passbands. 32. The thirty-second method, wherein the step of processing the harmonic echo signal comprises dividing the harmonic echo signal component into two different passbands of two unequal dynamic ranges; Processing the harmonic echo signal divides the harmonic echo signal component into a high frequency passband of a given dynamic range and a low frequency passband of a dynamic range smaller than the given dynamic range. 33. The method according to 33, comprising An ultrasonic diagnostic image processing apparatus for creating a harmonic ultrasonic image of a body structure, comprising: means for transmitting ultrasonic energy of a fundamental frequency into the body; responding to the transmitted ultrasonic energy; the fundamental frequency; and Means for receiving an ultrasonic echo signal at a harmonic frequency of the fundamental frequency; and processing the received fundamental frequency echo signal and the harmonic frequency echo signal, and formed from both components of the fundamental frequency and the harmonic frequency echo signal 35. an ultrasonic diagnostic image processing apparatus including an image processor for generating an ultrasonic image; 35. The ultrasonic diagnostic image processing apparatus according to 35, wherein the receiving means includes means for generating divided fundamental and harmonic frequency echo signals; The means for receiving includes a filter for creating a fundamental frequency echo signal excluding at least a part of the harmonic frequency echo signal and creating a harmonic frequency echo signal excluding at least a part of the fundamental frequency echo signal. Ultrasonic diagnostic image processing apparatus of 38. 35. The ultrasonic diagnostic image processing apparatus according to 35, wherein the image processor comprises means for using more harmonic echo signals in the near region of the image and using more fundamental frequency echo signals in the remote region of the image; An ultrasonic diagnostic image processing apparatus for imaging a harmonic response of a body structure exhibiting a depth-dependent attenuation of ultrasonic energy: means for transmitting ultrasonic energy of a fundamental frequency of 5 MHz or lower into the body; Ultrasound comprising: means for receiving an ultrasonic echo signal of a harmonic of the fundamental frequency of 10 MHz or less in response to ultrasonic energy; and means for creating an ultrasonic image from the harmonic echo signal Diagnostic image processing apparatus, 40. The transmitting means transmits ultrasonic energy into the body at a fundamental frequency of 2.5 MHz or less; and the receiving means receives an ultrasonic echo signal having a harmonic frequency of the fundamental frequency of 5 MHz or less. 41. the ultrasonic diagnostic image processing apparatus according to 39; The 39 ultrasonic waves, wherein the transmitting means transmits ultrasonic energy having a fundamental frequency of 2 MHz or less into the body; and the receiving means receives an ultrasonic echo signal having a harmonic frequency of the fundamental frequency of 4 MHz or less. Diagnostic image processing apparatus, 42. 39. The ultrasonic diagnostic image processing apparatus according to 39, wherein the reception means includes a programmable digital filter programmed to pass the harmonic echo signal excluding the fundamental frequency; An ultrasonic diagnostic image processing apparatus for image processing of a nonlinear response of a tissue: a transmitter for transmitting ultrasonic energy of a fundamental frequency into the body; in response to an echo returned from the tissue after transmitting the ultrasonic energy, 44. an ultrasonic diagnostic image processing apparatus comprising: a receiver that separates a signal that represents an ultrasonic non-linear response; and an image processor that generates an ultrasonic image from the non-linear response signal; 45. The ultrasonic diagnostic image processing apparatus according to 43, wherein the receiver includes a filter circuit that separates a signal that represents the nonlinear response of the tissue as an ultrasonic wave; In response to receiving multiple echoes from the same spatial location in the tissue, the receiver includes a signal processor that combines the multiple echo signals and separates a signal representative of the non-linear response of the tissue in ultrasound. 45. The ultrasonic diagnostic image processing apparatus according to 43, 47. The ultrasonic diagnostic image processing apparatus according to 43, wherein the non-linear ultrasonic response of the tissue includes second-order or higher-order harmonics of the fundamental frequency. 43. The ultrasonic diagnostic image processing apparatus according to 43, wherein the non-linear ultrasonic response of the tissue has a subharmonic of the fundamental frequency.
[0045]
【The invention's effect】
  Methods and apparatus are provided for effectively using harmonic frequency echo components that are not fundamental frequencies in ultrasound imaging. The use of harmonic beams reduces scatter when processing through narrow mouths such as ribs, taking advantage of the lower sidelobe level of the harmonic beam being lower than the corresponding sidelobe level of the fundamental beam. Reducing off-axis scattering. Moreover, since the return energy of the harmonics from the proximity region is relatively smaller than the return energy of the fundamental frequency, the proximity region scattering can be prevented.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating an image processing apparatus according to the present invention.
FIG. 2 is a diagram illustrating a main lobe and a side lobe of a harmonic echo signal.
FIG. 3 is an explanation under different conditions for a main lobe and a side lobe of a harmonic echo signal.
FIG. 4 is an explanation of a beam pattern of harmonic echoes.
FIG. 5 is an explanation of another beam pattern of harmonic echoes.
FIG. 6 is a passband characteristic curve used to explain the behavior of the example of FIG.
FIG. 7 is another passband characteristic curve used in FIG. 1;
FIG. 8 is an illustration of typical fundamental and harmonic frequency passbands.
FIG. 9 is an illustration of a FIR filter structure suitable for use in the example of FIG.
FIG. 10 is a block diagram illustrating a part of a preferred embodiment of the present invention.
FIG. 11 explains the operation of the normalization stage of the example of FIG.
12 is a block diagram of a multiplier accumulator used in FIG.
13 is an illustration of the basic and harmonic frequency passbands of the example of FIG.
FIG. 14 illustrates blending of fundamental frequency signals and harmonic signal components.
FIG. 15 illustrates a pass band of a time variation filter for forming a blend image.
[Explanation of symbols]
37 ... B-mode processor, 50 ... Display device, 110 ... Ultrasonic probe, 112 ... Converter, 114 ... T / R switch, 115 ... Analog to digital converter, 116 ... Beamformer, 117 ... Transmission frequency control, 118 ... Digital filter, 70-73 ... Multiplier, 80-83 ... Accumulator, 120 ... Central controller, 128 ... Contrast signal detector, 130 ... Doppler processor, 140 ... Video processor, 162 ... 3D image creation processor, 164 ... 3D image memory.

Claims (14)

An ultrasonic diagnostic image processing apparatus for generating an ultrasonic image of a harmonic response of a tissue in a body with reduced multipath scattering in the absence of an ultrasonic contrast agent , the apparatus comprising:
An array transducer that emits ultrasonic energy into the body at a fundamental frequency and generates a tissue harmonic signal in the proximity region that is less intense than the fundamental signal in the proximity region;
A beamformer that processes echo signals from transducer elements of the array transducer to form in-phase echo signals;
A circuit that attenuates the fundamental frequency signal to a greater extent than the tissue harmonic signal;
An image processor for generating an ultrasonic tissue harmonic image so that scattering by the fundamental frequency signal is reduced in response to the tissue harmonic echo signal passed by the circuit;
Have
The apparatus further includes means for processing the harmonic ultrasound echo signal received by the array transducer to form at least partially decorrelated replicas of the echo signals; combining the decorrelated replicas An ultrasonic diagnostic image processing apparatus comprising: means for creating a harmonic echo signal with a reduced virtual image; and means for creating an ultrasonic image using the harmonic echo signal with the reduced virtual image. The ultrasonic diagnostic image processing apparatus according to claim 1 , wherein the virtual image is a dropped virtual image. The ultrasonic diagnostic image processing apparatus according to claim 2 , wherein the virtual image further includes a speckle virtual image. 2. The ultrasonic diagnostic image processing apparatus according to claim 1 , wherein said processing means comprises a band-pass filter that divides a component of the harmonic ultrasonic echo signal into two pass bands having different center frequencies. The ultrasonic diagnostic image processing apparatus according to claim 4 , wherein the processing means further includes a detector for detecting a harmonic ultrasonic echo signal in each of the pass bands. 6. The ultrasonic diagnostic image processing apparatus according to claim 5 , wherein said means for processing further comprises a logarithmic compression processor for logarithmically compressing the detected harmonic ultrasonic echo signal. The means for processing has two parallel channels each having an input connected to receive a harmonic ultrasonic echo signal and an output connected to the coupling means, wherein each of the channels is connected to the other The ultrasonic diagnostic image processing apparatus according to claim 1 , further comprising a band pass filter having a filter characteristic different from a filter characteristic of the channel. The ultrasonic diagnostic image processing apparatus according to claim 7 , wherein the filter characteristic is a peak response frequency of the filter. The ultrasonic diagnostic image processing apparatus according to claim 7 , wherein the filter characteristic is a center frequency of the filter. The ultrasonic diagnostic image processing apparatus of claim 7 , wherein each of the channels further includes a detector. The ultrasonic diagnostic image processing apparatus of claim 10 , wherein each of the channels further comprises a logarithmic compression processor. 2. The ultrasonic diagnostic image processing apparatus according to claim 1 , wherein the processing means comprises means for dividing the harmonic ultrasonic echo into two pass bands having unequal dynamic ranges. The ultrasound diagnostic of Claim 12 , wherein the two passbands comprise a low frequency passband and a high frequency passband, wherein the dynamic range of the low frequency passband is smaller than the dynamic range of the high frequency passband. Image processing device. 14. The ultrasonic diagnostic image processing apparatus according to claim 13 , wherein the processing means further comprises means for mapping different dynamic ranges of the passbands.

Images