WO2007028059A2

WO2007028059A2 - Multiplexer for high density ultrasound arrays

Info

Publication number: WO2007028059A2
Application number: PCT/US2006/034296
Authority: WO
Inventors: Bhushan Shanti Asuri; Patrick L. Von Behren
Original assignee: Iqonic Corporation
Priority date: 2005-09-02
Filing date: 2006-09-01
Publication date: 2007-03-08
Also published as: WO2007028059A3

Abstract

A multiplexing system enables communication between an array (2 and 6) of transducer elements (23, 24, 25 and 26; and 50, 53, 56 and 59) and a base unit (276) using a small number of wires (40; 70; and 176, 177, 178 and 179) or wireless channels. The transducer elements may be ultrasound transducer elements. For transmissions from the transducer elements to the base unit, one or more bimodal multiplexers (8; 35; 69 and 75; 172, 173, 174 and 175) are utilized. A 'bimodal' multiplexer is one which provides two shifts, with one shift being in the relative timing of signals to enable 'coherent summing' and the other shift being one that enables discrimination of signal energy generated by the different transducer elements. With respect to signaling generated at the base unit for controlling operations at the array, the number of connections is reduced by employing at least one transmit demultiplexer (232, 233, 234, 235, 236, 237 and 238) at the array for receiving a multi-element drive signal for sequencing the trans¬ ducer elements. The sequencing can be configured to achieve desired beam characteristics, such as beamforming.

Description

MULTIPLEXER FOR HIGH DENSITY ULTRASOUND ARRAYS

TECHNICAL FIELD

[0001] This invention relates to front-end electronics in systems used for diagnostic ultrasound imaging.

BACKGROUND ART

[0002] Ultrasound is based on the transmission of sound waves through the human body and recording the pattern of received echoes. The timing of the echoes determines the depth of the object producing the echo and its strength determines the contrast that the echo producing object has with respect to its environment.

[0003] Typical ultrasound systems use a variety of means for scanning the transmitted energy through a volume or area of interest. This extended ultrasound map of the body is designated by various letters to signify the type of scanning: B-mode refers to the method of scanning a slice of the human body; volume scanning refers to the scanning of a whole volume of the body.

[0004] The transmit and receive beams can be scanned by the use of ultrasound phased arrays that contain many piezoelectric elements. During transmit, the relative phase and amplitude of the signal emitted from each element is chosen to form and focus a transmit beam toward the point of interest. A receive beam is formed and focused similarly by delaying the received signal from different elements, scaling them appropriately and summing the received signals. Conventional B-mode scanning employs one-dimensional (1 D) transducer arrays to capture tomographic image slices of areas of the human body. The phased array in this method scans the beam only in the azimuth direction. A slightly advanced form of B-mode scanning is realized with 1.5D arrays, where multiple adjacent rows of 1D array are combined to control the beam width and steering in the elevation direction.

[0005] Volume scanning is desirable because, by forming and collecting volumes rather than slices, it presents more complete structural information and image slices that are derived from those volumes are viewed in a more complete anatomic context than slices derived from B-mode. Image volumes may be acquired by mechanically scanning 1 D arrays to acquire multiple adjacent B-mode image slices and assembling them for display. Mechanical scanning is cumbersome and slow, however, and electronic scanning is preferred. Electronic volumetric scanning requires two-dimensional (2D) arrays to steer the transmit beam in two dimensions, whereas B-mode scanning requires the transmit beam to sweep with 1D arrays in only one dimension.

[0006] The spatial resolution of 2D arrays is determined by the aperture size. Element pitch is determined by the acoustic wavelength of the ultrasound signal. These constraints often lead to upwards of 2000 elements being required for fully populated 2D arrays with resolution in the elevation direction comparable to the azimuth resolution of traditional B-mode scanners.

[0007] The ultrasound transducer is typically composed of a piezoelectric material appropriately machined into individual elements that are connected to a specially constructed cable. This cable, which might contain hundreds of micro-coax wires or ribbon wires, carries signals from the transducer and delivers drive signals from the drive electronics back to the transducer.

[0008] The cables for ultrasound systems demand a flexible mechanical design to prevent ergonomic injuries to ultrasound technicians. For example, the cable that connects the transducer to the system console must be flexible and lightweight so as not to impede the scanning that is performed by hand and not to fatigue or unduly stress or, in the long term, cause injury to the sonographer during the time of the scanning procedure. These demands generally trade-off with the number of micro-coax wires that a cable can support, since adding more wires to a cable bundle makes the cable more stiff and difficult to maneuver. Furthermore, tighter packing of multiple cables greatly impacts the electrical performance of the wire bundle. These factors make the design and manufacture of 1000-wire cables extremely costly and cumbersome.

[0009] Another trend is that the electrical impedance of an individual element rises as its size is reduced. The size reduction is particularly severe for true 2D arrays, where the piezoelectric is diced in both azimuthal and lateral directions to form rows and columns of elements. The higher impedance decreases the strength of the signal transferred from transducer to cable unless special measures are taken. These measures could include using wire bundles with lower capacitance per unit length and/or terminating the cable at the transducer end with an impedance-matching circuit. These choices place additional demands on the conventional system design, but various conventional ultrasound systems employ one or both of them. Being passive, they are relatively straightforward to implement and have limited effect on overall system performance.

[0010] Other measures include the incorporation of active electronics in the transducer scan head, so as to have driver circuits to actively induce currents into the transducer cables to reduce or eliminate losses in those cables.

[0011] Another set of solutions to accommodating the larger numbers of elements in 2D arrays centers around the use of active multiplexing of signals from various transducer elements into a single wire. These methods include the use of time-domain multiplexing, frequency-domain multiplexing, and some method of additively combining the signals prior to transmitting them on the micro coax wire. In practice each of these methods used independently is unlikely to be able to support very large arrays. Using a combination of methods to increase the multiplexing capacity of these systems requires a large amount of space (area) and/or power (heat).

[0012] Accordingly, the use of 2D arrays for volume-scanned ultrasound phased array systems faces various challenges, the most significant being the need to support a large number of array elements and, in turn, a very large cable bundle. While several individual means of multiplexing have been proposed, each of these is severely limited in the number of elements it can support while maintaining reasonable power dissipation and area consump- tion.

SUMMARY OF THE INVENTION

[0013] The invention provides an arrangement that enables communica- tion between an array of transducer elements and a base unit using a relatively small number of wires or wireless channels. The transducer elements may be ultrasound transducer elements in an array. With respect to electrical signals generated by the transducer elements for transmission to the base unit, a multiplexing system and method utilize one or more bimodal multiplexers. A "bimodal" multiplexer is defined herein as a multiplexer which provides two shifts, with one shift being in the relative timing of signals to enable "coherent summing" and the other shift being one that enables discrimination of signal energy generated by the different transducer elements at the base receiver. Each bimodal multiplexer forms an output signal that comprises signal energy from multiple transducer elements for transmission via a single wire or wireless channel. With respect to drive signals generated at the base unit for controlling operations at the array, the number of connections between the array and the base unit is reduced by using at least one transmit demultiplexer which receives a multi-element drive signal via a single wire or wireless channel for sequencing the transducer elements. The sequencing can be configured to achieve desired beam characteristics. For example, beamforming is possible. [0014] The transducer elements that participate in the coherent summation are collectively referred to herein as "sub-arrays." The shift in the relative timing by a bimodal multiplexer may be implemented as phase shifts, while the shift that enables discrimination among sub-arrays may be implemented as frequency shifts. The coherent summation output signal for a first sub-array may be combined with similar such signals from other sub-arrays within a single wire, if the composite signal in the wire is formed such that signals of specific sub-arrays may be identified by the receiver. For example, the different coherent summation output signals may be frequency shifted at the transmitter to enable discrimination among the different signals.

[0015] The invention is particularly advantageous in a context of the use of the bimodal multiplexers in ultrasound systems. Such systems are required to support a large number of channels in order to provide high image quality. One of the upper bounds on the number of channels is determined by the number of wires in a cable that is connected to the ultrasound transducer. The invention provides a technique that processes signals from transducers and enables the signals to be transmitted over significantly fewer wires than is required using convention techniques. However, it is contemplated to use the bimodal multiplexers in other applications.

[0016] As one aspect of the invention, a multiplexing system includes a number of inputs connected to different sources of input signals, such as ultrasound transducer elements. The system includes at least one multiplexer coupled to the inputs to generate a coherent summation output signal in response to the input signals. The multiplexer is configured to provide both shifts in the relative timing of the input signals in order to enable coherent summing and shifts in frequencies of the input signals to enable discrimination of the coherently summed signals from the sub-arrays. The multiplexer includes an output for transmission of the output signal. The signal is transmitted to the "base unit" for further processing which may perform analog-to-digital conversion, image processing, and image display, as well as other possible functions. For a system in which there are a number of multiplexers, each multiplexer may be assigned to a different sub-array. There may be a one-to-one correspondence between the multiplexers and the output lines, such as wires, to a receiver for conducting the coherent summation output signals. Alternatively, some or all of the outputs of the multi- plexers may be combined for transmission over a single line. The outputs from the multiplexers may be varied in frequency to enable discrimination. As an alternative to providing a separate bimodal multiplexer for each of a number of different sub-arrays, the sub-arrays may be connected to a single multiplexer. A "transducer element selection arrangement" may be used to enable and disable signal flow between the multiplexer and the various sub-arrays, so as to provide flexibility in addressing the ultrasound transducer elements.

[0017] Each multiplexer may include circuitry that enables variable gain. A multiplexer that includes the variable gain would support a high signal dynamic range.

[0018] In one embodiment of the bimodal multiplexer, the multiplexer includes a number of mixers and a Trans Impedance Amplifier (TIA). The mixers are connected between the signal sources (such as ultrasound transducer elements) and the TIA. Each mixer is associated with one of the sources and with a local oscillator signal selected to enable frequency shifting and phase shifting, which implements the shifts in relative timing of the input signals. The TIA provides a virtual ground for outputs of the mixers, so as to convert voltage signals to current signals. The multiplexer is configured to combine the various current flows to provide the coherent summation output signals. Gain control may be included by means of a variable feedback across the TIA.

[0019] A multiplexer may include a linearized transconductance stage for each input. Each linearized transconductance stage leads to a Weaver architecture, as well as first and second local oscillator signal inputs. Each Weaver architecture is configured such that shifting the phase of a local oscillator signal input implements the shift in relative timing for enabling coherent summation.

[0020] There are advantages to providing filters in the multiplexer prior to transmission to the base unit in accordance with the invention. The filters may be adjustable or may be fixed. The filters may be at the input end, the output end or both ends of a multiplexer. For example, at the output ends, low-pass and band-pass filters may be used to isolate tissue-generated second harmonic energy detected during the use of an ultrasound transducer. The coherent summation may then be adjusted to process the tissue- generated second harmonic energy.

[0021] In addition to the inputs from the signal sources, the multiplexer may include control signal inputs and local oscillator signal inputs. The local oscillator signal inputs may be connected to mixers which enable the two shifts of a bimodal multiplexer. In one implementation, the local oscillator input signals include in-phase and quadrature-phase signals. The multiplexer may also include circuit cells that are specific to providing the mixers with local oscillator signals having target characteristics. For example, the circuit cells may be Gilbert cells.

[0022] As an alternative to multiplexers which receive in-phase and quadrature-phase signals, a multiplexer may include frequency division circuitry connected to receive a single high frequency local oscillator signal, with the circuitry being enabled to generate both in-phase and quadrature- phase signals from the single signal.

[0023] In a method in accordance with this first aspect of the invention as applied to ultrasound transducers, transducer elements are connected to inputs of circuitry enabled to provide the desired multiplexing. The method provides shifting of characteristics for signals generated by individual trans- ducer elements, including providing shifts to enable coherent summing of the signals from a sub-array and shifts to enable discrimination of the signals from the sub-array following the coherent summation. The output signal from one or more sub-arrays is directed via a single wire to a receiver for processing. There are advantages to dividing the transducer array into sub-arrays. The sub-arrays may be assigned to different multiplexers that are enabled to implement the shifting of signal characteristics, providing multiple coherent summation output signals. The signals may be combined for transmission over a single wire or may be forwarded to the receiver via different wires. If the signals are to be combined, the frequencies of the different output signals may be adjusted in order to enable discrimination of the signals at the receiver.

[0024] The method may include steps of enhancing signal integrity with respect to the output signals. As one possibility, the ultrasound transducer elements may be driven at two distinct drive levels. Then, error is identified on the basis of differencing (1) output signals formed while driving the ultrasound transducer elements at the first drive level and (2) the output signals formed while driving the ultrasound transducer elements at the second drive level. A scaled copy of the output signals at the first drive level is subtracted from the output signal at the second drive level. The difference represents the non-linear signal components generated in the tissue and the multiplexer. This measurement may be utilized to estimate non-linear components at other drive levels, thereby improving signal integrity.

[0025] Signal cross-talk is also a concern in ultrasound imaging and related applications. When the method is used in applications in which multiple multiplexers provide output signals transmitting through multiple wires, signal cross-talk among the wires may be controlled by selecting the arrangement of wires such that output signals of wires that are adjacent each other are frequency shifted such that signal energy is in significantly different frequency bands. [0026] Pre-equalization may be performed on the output signal prior to directing a coherent summation output signal to the receiver. The pre- equalization is tailored on the basis of anticipated signal degradation within the wire. Calibration pattern information may be exchanged with the receiver to adaptively adjust the pre-equalization or other characteristic of the multiplexing. Coordination between the receiver and the adaptive circuitry may be provided with respect to the start and the duration of the adaptive adjustment process.

[0027] In a second aspect of the invention, the multiplexing system includes an array of ultrasound transducer elements, a number of filters, and a number of multiplexers. The filters are specific to preferentially passing particular frequency bands. The filters are assigned to functionally equivalent filter sets that divide a signal input to one of the sets and to multiple frequency bands. Each ultrasound transducer element provides an input to one of the sets of filters. Each multiplexer is operatively associated with one of the frequency bands and is connected to the collection of filters which preferentially pass the associated frequency bands. Each multiplexer is configured to provide both first shifts to enable coherent summing of the inputs from the filters connected to that multiplexer and frequency shifts to enable discrimina- tion of the signal energy originating from different ultrasound transducer elements. Thus, similar to the first aspect, a multiplexer generates a coherent summation output signal. However, the output signal is specific to a frequency band.

[0028] The first shifts may be phase shifts that are designed to provide the coherent summation. The multiplexers may be cooperative such that phase shifts for the collection of frequency bands of a signal from a given transducer element may approximate a step-wise quadrature function. As another possibility, the phase shifts may be provided to achieve approximation of a phased array ultrasound system. One or more wires may be used to conduct the outputs to a base unit receiver, which includes an array of circuits for recombining transducer energy from the output signals to reconstruct original signals from the ultrasound transducer elements.

[0029] The focus of the third aspect of the invention is upon signal addressing and conditioning using transmit demultiplexers. The combination of addressing of transducer elements and addressing sub-arrays may be used for providing beamsteering during transmit. In the multiplexing system, a sequence control is configured to define sequencing of the individual transducer elements within each sub-array and is configured to define sequencing of the signaling from the different sub-arrays. The sequence control of the system may include switching circuitry that is responsive to voltage pulses having a pulse pattern for implementing the beamsteering during transmit.

[0030] The signal conditioning may include pre-equalizers specific to providing compensation for signal degradation during transmission through wires. Calibration may be exchanged with the receiver of the output signals from the transducer elements in order to enable adaptive adjustments of equalization and filtering characteristics. The system may include a receiver, which may have equalization for processing the received coherent summation output signals. As possibilities, the equalization may be one of digital multi-tap Finite Impulse Response filters or a multi-tap amplifier in which different paths have different spectral responses.

[0031] As described above, each multiplexer may include mixers and a TIA. Each of the mixers is associated with a particular ultrasound transducer element and with a local oscillator signal selected to enable frequency and phase shifting. The TIA provides a virtual ground for mixer outputs, so as to convert the voltage signals to current signals. In another approach, the multiplexer includes the linearized transconductance stage and Weaver architecture previously described. [0032] In a method in accordance with the invention, the number of wires or wireless channels can be most significantly reduced by combining the use of bimodal multiplexers with the use of transmit demultiplexers.

BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Fig. 1 illustrates a particular embodiment of a 2D array used in ultrasound systems.

[0034] Fig. 2 illustrates one embodiment of a conventional ultrasound system.

[0035] Fig. 3 is an embodiment of a bimodal multiplexer front end.

[0036] Fig. 4 illustrates a mixer that could possibly be used in a fully differential realization of the embodiment of the invention illustrated in Fig. 3.

[0037] Fig. 5 illustrates a differential implementation of the invention in Fig. 3.

[0038] Fig. 6 illustrates the method of using two bimodal multiplexer front ends to drive signals into the output through the wires.

[0039] Fig. 7 represents another embodiment of the bimodal multiplexer based on the concept of a Weaver transmitter.

[0040] Fig. 8 represents one of the challenges that will arise if the bimodal multiplexer is used in presence of tissue-generated second harmonic signals.

[0041] Fig. 9 illustrates a method for correcting the impact of this cross-talk illustrated in Fig. 8. [0042] Fig. 10 represents signaling in the use of the bimodal multiplexer according to the invention for use in harmonic imaging.

[0043] Fig. 11 illustrates a method of enhancing signal integrity for harmonic imaging.

[0044] Figs. 12 and 13 illustrate an aspect of the bimodal multiplexer according to the invention that is related to the mapping of signals to the wires in the cable bundle.

[0045] Fig. 14 illustrates an embodiment of the invention with applicability to high-frequency, high-bandwidth transducer signals.

[0046] Fig. 15 illustrates an embodiment of the invention that enables the use of multiple wires to carry signals from a wide-band transducer.

[0047] Fig. 16 illustrates a system in which the signal from the transducers is amplified by a gain element 201 and is directly launched through non- overlapping filters.

[0048] Fig. 17 illustrates the concept of spatial diversity for high-bandwidth transducers in which the signal from the different frequency slices is arranged so that no two adjacent wires have substantially similar frequency content.

[0049] Fig. 18 illustrates an embodiment of the invention in which the bimodal multiplexer is used for high-frequency narrow-band signals.

[0050] Fig. 19 illustrates a transmit-receive switch with input, output and a control node.

[0051] Fig. 20 illustrates a three-terminal switch with input 223 that can be connected separately or simultaneously to outputs 226 or 227. [0052] Fig. 21 illustrates an embodiment of the invention that uses the elements in Fig. 19 and Fig. 20 to support a bimodal multiplexer with transmit switching.

[0053] Fig. 22 illustrates the method of using the embodiment of the invention illustrated in Fig. 21 to perform transmit beamforming. Here the signals 250, 251 drive the switch 232.

[0054] Fig. 23 is an illustration of one example of the use of the invention in a practical setting.

[0055] Fig. 24 illustrates an invention where the base unit 276 transmits the different frequencies required for the multiplexing, 277-284.

[0056] Fig. 25 illustrates an alternative aspect of the invention where a plurality of wires 286-289 is used to transmit the same oscillator signal but at different phases from a base unit 276.

[0057] Fig. 26 illustrates an alternative aspect of the invention where a multiple of the local oscillator signal 290 is transmitted down the cable from a base unit.

[0058] Fig. 27 illustrates an alternative embodiment of the invention where the signals from multiple local oscillators, or the local oscillator signals for different bimodal multiplexers, are transmitted on the same cable 294 from the base unit 276.

[0059] Figs. 28-31 show examples of various embodiments of the invention.

[0060] Fig. 32 illustrates an invention that is used to aid digital equalization of the received signal. [0061] Fig. 33 illustrates the use of a combination of feedforward and feedback equalizers in the base unit.

[0062] Fig. 34 illustrates an extension to show a system where multiple uncorrelated calibration patterns are launched through adjacent wires.

[0063] Fig. 35 illustrates the use of MIMO for cross-talk cancellation.

[0064] Fig. 36 illustrates the signal chain in the base unit that comprises the receiver 363, which receives the signal from the cable and converts it into digital data.

[0065] Fig. 37 shows a concept of an ultrasound transducer with a 2D array of piezoelectric elements 367 that receive acoustic signals forming the front of an acoustic stack 368.

DETAILED DESCRIPTION

[0066] Fig. 1 illustrates a particular embodiment of a 2D array 2 used in ultrasound systems. The cone 1 represents the volume where the beam is swept in order to form a 3D image of the body. The 2D transducer array 2 is used to scan the ultrasound beam using phased array techniques. Fig. 1 shows the interconnection 3 of the transducer array to electrical leads which might be through the means of a flexible printed circuit board (PCB) or any other such means. A transducer cable consists of individual micro-coax wires 4 bundled inside a cable jacket 5. Each individual wire is connected to the circuitry in the 2D array by direct soldering or other such methods. The wires carry the transmit, receive, and control signals to and from the 2D array.

[0067] Fig. 2 illustrates one embodiment of a conventional ultrasound system. The transducer array 6 may comprise several individual elements. A micro-coax cable assembly 7 is shown, but in other cases the assembly might be replaced with ribbon cables or other such cables. A high voltage multiplexer 8 is configured to select between different transducer handles. A high voltage transmit receive switch 9 protects the receiver input from the high voltages present on the transmitter. A collection of high voltage amplifiers and associated logic 10 drives the transducers during the transmit cycle. A transmit beamformer unit 11 controls the timing, amplitude and other characteristics required to focus the beam. A beamformer controller 12 interfaces between the rest of the system and the transmit and receive beamformers 11 and 16. The system includes a collection of low noise amplifiers 13 and gain control elements 14. These provide signal amplification, buffering, gain and gain control. A collection of analog-to-digital converters 16 each includes integrated anti-aliasing filters. The receive beamformer 16 adjusts the phase and amplitude of the received signals to enable phased array beamforming. A collection of front end boards 19 is typically included in a system to address all available transducer elements. Blocks 17 and 18 represent controls in the base unit that enable user interfacing and image processing, respectively. Dashed box 20 represents the beamformer board. A collection of these boards might be needed to address all the elements of the transducer array. A tuning section 21 matches the impedance characteristics of the electrical input of the amplifiers 13 to the output of the switch 9. A programmable variable bias control 22 may be used for the low noise amplifiers.

[0068] Fig. 3 is an embodiment of a bimodal multiplexer front end. Depicted in this figure is support for a sub-array composed of transducer elements 23,24,25,26. The invention, however, is not limited by the number of elements in the sub-array and more generally the number or elements in the sub-array may vary from 1 to n, where n is arbitrary and depends on the specific design of the ultrasound system. The signals from the four elements are multiplied by local oscillators 31 ,32,33, 34 in mixers 27,28, 29, 30 respectively. A transimpedance amplifier (TIA) 35 provides a virtual ground at a node 41 and converts the voltage signals from the transducer elements 23, 24, 25, 26 to currents appearing as outputs of the mixing processes. These currents, flowing through the transducer elements, are concurrently modulated by the local oscillators 31 , 32, 33, 34.

[0069] Each mixer 27, 28, 29, 30 converts the frequency spectra of two input signal currents into a composite output current containing the sum and difference frequencies of the inputs. The mixer output signal currents from the sub-array elements are summed into a composite signal current upon flowing into node 41. A phase shift can be added to frequency-converted signals simply by modifying the phase of the local oscillator signal. The current signal at node 41 flows through resistor 36 in the TIA and is presented as a voltage to a low-pass filter 38. The signal is thereafter converted to current in- a voltage-to-current converter 39 and presented at an output 40.

[0070] In another embodiment of the invention an element in the form of a variable series resistor is introduced between the mixer and the transducer element. In another embodiment the TIA has variable resistive feedback, allowing one to modify the gain of this element. Variable gain could also be included in elements 38 and 39 as required in the system. In yet another embodiment of the system, variable gain is introduced by changing the local oscillator drive voltage levels for the signal of interest. In one embodiment of the system, the variable gain introduced at or before the mixer is used in conjunction with the phase-shifting techniques to enable coherent combining along with control of the amplitude in the individual channels.

[0071] Fig. 4 illustrates a mixer that may be used in a fully differential realization of the embodiment of the invention illustrated in Fig. 3. Such a realization has numerous practical advantages. The mixer consists of four switches. In the illustrated case, the switches are realized using PMOS-NMOS pass gates 42, 43, 44,45, although this is just an example and they may be realized by any other solid-state switching arrangement. The pass gates in each pair {43,45} and {42,44} are driven in phase. Each pair is driven out of phase with respect to the other pair, however, such that when the input 46 is connected to the output 49, input 48 is connected to output 47. Similarly, when the input 47 is connected to the output 46, input 49 will be connected to output 48.

[0072] Fig. 5 illustrates a differential implementation of the invention in Fig. 3. Here the transducer elements 50, 53, 56 and 59 generate signals that are combined at the virtual ground nodes 65 and 66. The mixing is done in mixer devices 52, 55, 58 and 61 , which alternately swap the signal from a transducer element and the AC ground signal between the nodes 65 and 66. The elements 51 , 54, 57 and 60 provide a copy of the impedance shown by the transducer elements. By virtue of the swapping, the impedance seen from the ground node to the virtual ground nodes is constant, irrespective of the clock phases driving the mixer elements. This is generally a requirement for maintaining stable operation of the transimpedance amplifier (element 64) and the feedback elements 62 and 63. The output from the differential system is presented at the nodes 67 and 68 for further processing by the low-pass filter, etc.

[0073] Fig. 6 illustrates the method of using two bimodal multiplexer front ends, 69 and 75 to drive signals into the output 70 through the wires 73, 74. The signals from front ends 69 and 75 are separated in frequency and can be discriminated at the receiver with a frequency filter. As was the case illustrated in Fig. 3, the elements 69 and 75 combine signals from sub-arrays 71 and 72 using the bimodal multiplexer design described above. The inven- tion, however, can consist of more than two bimodal multiplexers.

[0074] Furthermore, the various bimodal multiplexers 69 and 75 connected to the output 70 need not be operated simultaneously, but rather could be connected alternately, that is, they may be time-multiplexed. In that case, the various sub-arrays the individual multiplexers address could be used to gather signals from a sub-set of all the elements available. This mode of operation could be useful in synthetic aperture multiplexing methods, where the base unit processing signals arriving from the wires in jacket 5 in Fig. 1 might not have the capacity to simultaneously process all the received channels. A further refinement of the bimodal multiplexer could turn off the mixer devices connected to any element of the sub-array. This could provide the flexibility of addressing individual elements in the sub-array. This further refinement could provide a way to build more sophisticated synthetic aperture-based systems. It could also enable selective modification of the beam pattern and width based on scan requirements.

[0075] An important concern with practical bimodal multiplexers is the power dissipation in the assembly. According to one embodiment of the invention, the power dissipation of the individual elements in the chain is controlled as a function of the imaging depth. Thus, the signals from the deepest tissue (which will arrive last in time) would be processed by a signal chain running at full power and hence lowest noise. Signals arriving from the skin or other shallow scatterers would be processed by the same signal chain, but with different current consumption for different elements. This is just one example; the invention provides a bimodal multiplexer whose operating current is changed with respect to time such that it maps to the depth of the signal returning to the array. One advantage of this dynamic adjustment of current consumption is that it reduces the amount of power dissipation.

[0076] A further refinement of the bimodal multiplexer is the provision of adjustable low-pass filters and mixer frequencies. These adjustable elements could be used to accommodate a variety of transducer center frequencies. They could also be used in the field to track changes in transducer characteristics with age, temperature, etc. They could also be optimized in the field for different imaging modes and imaging performance.

[0077] Fig. 7 represents another embodiment of the bimodal multiplexer based on the concept of a Weaver transmitter. Here, wires 108-111 carry the signals from four transducers. The four circuit elements 76-79 convert the voltage signals into proportional currents and may have variable gain. One method for voltage-to-current conversion is to use transconductance elements. In one embodiment, circuit elements 76-79 are transconductance elements composed of appropriately sized bipolar or CMOS transistors - the proper size may be determined by skilled engineers using known design methods. The transconductance with variable gain would be attained by using these devices in the common emitter or common source configuration with variable degeneration.

[0078] Single Weaver transmitters, 103-106 each consists of two parallel paths. Each path contains a pair of mixers 96-97 (only one pair is labeled, for simplicity). The mixer pair is followed by a pair of low-pass filters 98-99, another pair of mixers 100-101 , and finally a summer 102, which combines the signals from mixers 100-101 in anti-phase. The Weaver transmitter as such is known in the prior art. The present invention, however, proposes to use a multiplicity of these transmitters 103-106 to perform bimodal multiplexing. One method of doing this is to drive the first pair of mixers

(96-97) in the Weaver transmitter with local oscillator (LO) signals 80, 81 , 82, 83 that are all in phase with each other and are in quadrature phase with LO signals 84, 85, 86, 87. In this embodiment, the phase shifts required to perform coherent combination of the signals from the transducers are introduced in the second pair of mixers. Herein, the LO signals 88, 89, 90, 91 would have phases as required by the need to perform coherent mixing and the LO signals 92, 93, 94, 95 would be in quadrature with 88-91 , respectively. The output from the summer 102 is assumed to be available as current that can be summed at a node 107 by simply connecting all the requisite wires together.

[0079] The particular embodiment of the invention illustrated in Fig. 7 is only one example of a bimodal multiplexer according to the invention - the invention encompasses any other arrangement wherein the signal from a transducer is simultaneously shifted in frequency and modified in phase by modifying the frequency and phase of the local oscillator signal in a mixer. The signals from the transducers in a sub-array may also be shifted by the same frequency, but might be differently phase-shifted and might be differently amplitude-weighted and subsequently combined coherently.

[0080] The coherently summed output of one sub-array can be co-propagated in a wire with outputs of other sub-arrays. The receiver in the base unit distinguishes signals from different sub-arrays based upon the frequency shift they are assigned in the bimodal multiplexer.

[0081] Fig. 8 represents one of the challenges that will arise if the bimodal multiplexer is used in presence of tissue-generated second harmonic signals. Here, reference numbers 114 and 115 represent the positive and negative frequency components, respectively, of the fundamental signal. The signals 113 and 116 represent the second harmonic generated in the tissue and during transmit from the transmitters that are now picked up by the receiver.

[0082] The signals 114, 113, 115, 116 are shifted up in frequency by mixing to become signals 118, 117, 119, 121 respectively. Similarly, the signals 114, 113, 115, 116 are shifted down in frequency by mixing to become signals 122, 120, 123, 124, respectively. The overlap between signals 121 and 122, and between signals 119 and 120, constitutes cross-talk between the fundamental and second harmonic signals. This cross-talk is more severe than that encountered in systems that do not perform any mixing. A low-pass filter 125 is preferably included to remove components of the signal from the output.

[0083] Fig. 9 illustrates a method for correcting the impact of this crosstalk. As illustrated, method step 126 constitutes the step of transmitting a weak signal a1 , which does not generate significant second harmonic components in the corresponding received signal r1. The received signal r1 is recorded in step 127. In step 128, a much stronger signal is transmitted with strength a2. The received signal is recorded as r2 in step 131. Since second harmonic components grow as the square of the signal strength while the fundamental grows linearly with drive signal strength, the invention, in step 130, evaluates the residual second harmonic signal in r2 - rdouble - as being equal to r2 less r2 scaled by the ratio of a2 to a1. Finally, in step 129, the signal rdouble is subtracted from r2 and the residue rclean is the second harmonic cross-talk free signal available to the system.

[0084] Fig. 9 illustrates only one embodiment of a method according to the invention for correcting for second harmonic-induced cross-talk in the fundamental band in bimodal multiplexing systems. Others are Phase Inversion (Siemens) and Pulse Inversion (Philips). Similar correction methods have been employed in the prior art to remove the fundamental energy in imaging the second harmonic. This invention, however, provides for the first time a method to solve a problem unique to bimodal multiplexing, namely, corruption of the fundamental component by frequency shifted versions of the second harmonic energy.

[0085] Fig. 9 also illustrates the method according to the invention for correcting the impact of shifted second harmonic energy falling within the band of the fundamental for the same bimodal front end. The same method can be used to correct for second harmonic energy from one bimodal front end that is available as cross-talk in the fundamental band of the signal from another bimodal front end.

[0086] Fig. 10 illustrates the use of the bimodal multiplexer according to the invention for use in harmonic imaging. Here, the fundamental compo- nents 133, 134 of the signal are frequency-shifted by mixing to occupy the bands 138, 137, 142, 141. The second or other harmonic components of interest 135, 132 of a signal are frequency-shifted by mixing to occupy the bands 143,140,139,136 of the signal, respectively. A low-pass filter 144 removes components of the signal from the output. This differs from methods that do not use bimodal multiplexing, where the second harmonic frequencies are separated from the output signal with high-pass filters or band-pass filters. Furthermore, the second harmonic energy in these conventional systems is always available at higher frequencies than the fundamental. Owing to performance degradations within the system at higher frequencies, this is not an optimum method of capturing second harmonic signals.

[0087] Fig. 10 also illustrates a method according to the invention for translating second harmonic energy to lower frequencies. In the illustrated embodiment, a low-pass filter 144, which is already being utilized for bimodal multiplexing, is used to separate the second harmonic from the fundamental signal. Further, by using a filter 144 with an adjustable bandwidth, the invention is given a tuning capability at the channel level. Tunable filters 144 can also mitigate the amount of front end noise that is propagated to the analog-to-digital converter (ADC). The filter constrains the noise to low-pass frequencies to reduce noise aliasing at the ADC.

[0088] In its use of low-pass filters to isolate the second harmonic signal, the bimodal multiplexer according to the invention is unique. A further elimination of second harmonic energy can be used by signal processing techniques at the system level. One such technique is illustrated in Fig. 11.

[0089] In Fig. 11 , step 145 involves transmitting a drive signal of strength a1. In step 146, the received signal r1 is recorded. In step 147 a signal of strength a2 is transmitted. Signal a1 is at a level where it causes minimal second harmonic generation in the system. Signal a2 is much stronger and it causes significant second harmonic generation owing to the fact that second harmonic artifacts grow as square of signal strength. The received signal is recorded as r2 in step 149. Since second harmonic components grow as the square of the signal strength while the fundamental grows linearly with drive signal strength, the invention, in step 148, evaluates the residual second harmonic signal in r2 -- rdouble - as being equal to r2 less r2 scaled by the ratio of a2 to a1.

[0090] Fig. 12 illustrates an aspect of the bimodal multiplexer according to the invention that is related to the mapping of signals to the wires in the cable bundle. This figure shows the cross section of the cable bundle with the individual cables 150-156, The figure shows a cable bundle in which adjacent wires are arranged in a hexagonal close packing arrangement.

[0091] Consider a bimodal multiplexer as shown in Fig. 3, which multi- plexes signals from two adjacent sub-arrays. By way of illustration, assume a system in which seven pairs of such bimodal multiplexers use seven pairs of oscillators. In this system, the seven pairs of oscillators (fourteen oscillators in all) have dissimilar frequencies. The seven pairs of bimodal multiplexers are then mapped to the wires 150-156. This set of seven wires is in turn placed in a hexagonal close packing arrangement with other sets of seven wires. This is shown in Fig. 13, where the set of wires 157 is arranged in proximity to the set of wires 158 such that no two adjacent wires have the signals at the same frequency. This spatial diversity mode of the bimodal multiplexer enables support of systems requiring a high degree of channel-to- channel isolation. This is especially true of systems like narrow-band Doppler signals, where bandwidth is not a constraining factor but large dynamic range requirements demand excellent inter-channel cross-talk performance.

[0092] The embodiment shown in Fig. 12 is not limited to the number of wires or the geometrical arrangement of the wires in the cable. Nor does it necessarily require all frequencies in adjacent wires to be different from each other. Further, this aspect of the invention does not necessarily have to employ bimodal multiplexing. The invention encompasses any arrangement and method where wires anticipated to be adjacent or approximately so (close enough to lead to significant cross-talk) are driven with signals that could be substantially different in frequency to provide spatial diversity. Another aspect of the invention in Fig. 12, which applies to systems employing time-domain multiplexing (TDM), is the use of different clock phases to drive adjacent or nearly adjacent wires. Here, the shape of the TDM pulses in adjacent chan- nels and their phase can be chosen to minimize the impact of cross-talk between channels in different wires. The invention thus also encompasses any method for driving adjacent wires in a TDM system with signals differing substantially in phase so that the cross-talk between adjacent or nearly adjacent wires is minimized.

[0093] Fig. 14 illustrates an embodiment of the invention with applicability to high-frequency, high-bandwidth transducer signals. This aspect of the invention is particularly useful for systems in which the cable bandwidth is substantially smaller than the signal bandwidth. Systems with sufficient bandwidth but significant cross-talk at higher frequencies are also expected to benefit. The signals from high-bandwidth, high-frequency transducers 159, 160 are buffered in pre-amplifiers 161 , 162. The output of each pre-amplifier drives low-pass filters 163,164 and band-pass filters 165, 167, 168, 169, 170, 171. The filters are designed such that each filter in each pair 163,164; 165,167; 168,169; and 170,171 has the same frequency characteristic as the other. Further, at the frequency where filter 163 has significant transmission, 165,168 and 170 attenuate the signal significantly. Thus, the filters

163,165,168 and 170 taken together cover the entire bandwidth of interest, while each filter passes a set of frequencies that is mutually exclusive of the other. The outputs from the pairs 163,164; 165,167; 168,169; and 170,171 are then combined using bimodal multiplexers 172, 173, 174 and 175, respectively. In this embodiment, the signals from 163 and 164 can be coherently combined by mixing with the same oscillator frequency. It is well recognized that a phase shift is an approximation of a time delay for a narrowband signal. True phased array therefore generally approximates a time-shift of the signal. A true time shift is equivalent to a phase shift that varies as the square of the frequency.

[0094] The signals arriving through each filter 163, 164, 165, 167, 168, 169, 170 and 171 in Fig. 14 have a relatively narrow bandwidth. By using the phase shift in the oscillator of the bimodal multiplexer 172, 173, 174 and 175, the invention essentially approximates a time delay. However, the phase shift could be different for signals from different filters. In this specific case, if the frequencies at which filter 163, 165, 168 and 170 have substantial transmissions that are monotonically increasing, then the phase shift applied to the local oscillator could be arranged to be the optimum approximation of a quadratic function of frequency. The bandwidth of the filters 163, 165, 168 and 170 as well as their center frequencies could be optimized in conjunction with the sequence of phase shifts associated with the oscillators driving the corresponding bimodal multiplexers 172-175. The output from the bimodal multiplexers drives the wires 176-179, which enters the cable bundle 166.

[0095] The aspect of the invention shown in Fig. 14 is not by any means restricted to two transducer elements or four band-pass filters 163, 164, 165, 167, 168, 169, 170 and 171. Furthermore, the illustrated order of the preamplifier 161 and 162 and the filter bank is not necessary to the invention. The invention encompasses any arrangement and method of operation by which the signal from a broad-band transducer is sub-divided into slices with substantially differing frequency content. The slices are then combined with signals from adjacent transducer elements by a bimodal multiplexer. The local oscillator signals for slices from the same transducer elements have a different phase shift for each slice. The phase shifts are chosen to provide the optimum approximation to a quadratic phase shift across the band of interest.

[0096] In addition to the approximation of the true time delay mentioned above, the invention also encompasses the method of coherently combining signals from different transducers into the same wire, where not all the signals are from the same frequency slice. In this latter case, the signals from different frequency slices are mapped to substantially different frequencies in the wire.

[0097] Fig. 15 illustrates an embodiment of the invention that enables the use of multiple wires to carry signals from a wide-band transducer. Here, the transducers 180 and 181 drive pre-amplifiers 182 and 183. The output from the pre-amplifiers is carved into slices with substantially non-overlapping frequency content by filters 184, 185, 186 and 187 and by filters 188, 189, 190 and 191. The filters 184-187 are substantially similar to the filters 188-191. The outputs from the filters are mixed with local oscillators in mixers 192-195 and 196-199. The mixers 192-195 may or may not be driven by local oscillators with different frequencies. The goal of the mixing is to shift the frequencies of the transducer signals so that they can be transmitted over the cable with better performance. The mixers may or may not be followed by buffers to drive the cable.

[0098] Fig. 16 illustrates a system in which the signal from the transducers 200 (only one of which is shown for the sake of simplicity) is amplified by a gain element 201 and is directly launched through non-overlapping filters 202, 203, 204, 205, 206, 207 and 215. The signals from these filters may or may not be further buffered before driving cables 208, 209, 210, 211, 212, 213, and 214 respectively. This illustrates the mapping of the different frequency slices to wires that are adjacent in a hexagonal close packing arrangement of wires.

[0099] Fig. 17 illustrates the concept of spatial diversity for high-bandwidth transducers, where the signal from the different frequency slices is arranged so that no two adjacent wires have substantially similar frequency content. The invention is not limited to the number of slices or wires. It is also not limited by the arrangement of wires. The invention encompasses any arrangement and method of slicing the signal from transducers into substantially non-overlapping frequency regions and the mapping of these slices into a set of wires such that the adjacent wires have signals with substantially different frequencies. This enables low cross-talk cable assemblies to support high-bandwidth transducers.

[0100] The aspect of the invention shown in Fig. 17 is not restricted for use to the aspect of the invention shown in Fig. 16. In one embodiment, the mixers in Fig. 15 are arranged to map the frequency slices of the signal to a set of frequencies that can then be mapped to the spatial arrangement of the wires. This would then reduce the amount of cross-talk between adjacent wires. [0101] In another embodiment, the invention of Fig. 14 could use different frequencies or sets of frequencies such that it could enable mapping of outputs to wires in the cable with minimum cross-talk. Here again, the invention would ensure substantially different frequency content in adjacent wires.

[0102] The invention is very general in that it encompasses any arrangement and method of slicing the frequency content from a high- bandwidth transducer, processing it, and mapping it into different frequencies. The signals mapped into different frequencies are then assigned to wires in the cable by a method that reduces the amount of cross-talk by ensuring that signals in adjacent wires have substantially differing frequency content. The invention also applies to any other physical or electrical elements whose spatial proximity might cause significant incursion of one signal into another that is on a different wire or connector. The invention thus encompasses any arrangement and method of mapping different frequencies to different physical or electrical elements with a view to impacting the incursion of the signal.

[0103] Fig. 18 illustrates an embodiment of the invention where the bimodal multiplexer is used for high-frequency, narrow-band signals. Here, signal spectrum 216 is mixed with a local oscillator signal 215A. Signals from adjacent elements of an array can be combined in a bimodal multiplexer. This would simultaneously shift the frequency of these signals and add them coherently. The result of the bimodal multiplexing is the low-pass filtered version of the shifted signal bands 217 and 218. The low-pass filter 219 might or might not be explicitly present. Fig. 16 is only an example of one embodiment of the invention in which the bimodal multiplexer allows shifting of high- frequency content into frequency bands that can be efficiently transmitted over the cable.

[0104] The invention is not restricted to systems where the output of the bimodal multiplexer is one frequency band. In general the system might consist of many transducers that are grouped as sub-arrays. The elements of the sub-arrays could be coherently combined into the same frequency band, while a plurality of sub-arrays could be multiplexed into the same wire by shifting them to different frequencies.

[0105] Fig. 19 illustrates a transm it-receive switch with input 220, output 222 and a control node at 221. Such a switch is well suited for use in a system based on the invention.

[0106] Fig. 20 illustrates a three-terminal switch with input 223 that can be connected separately or simultaneously to outputs 226 or 227. Again, such a switch is well suited for use in systems described herein.

[0107] Fig. 21 illustrates an embodiment of the invention that uses the elements in Fig. 19 and Fig. 20 to support a bimodal multiplexer with transmit switching. In Fig. 19, the transducer elements are represented by 228 and 241-247. The signal from the elements passes through a transmit-receive switch 229 before being amplified 230 and combined through the bimodal multiplexer. The output of the bimodal multiplexer connects to the cable through another transmit-receive switch 239. The cable 240 is connected to the transducer elements through a transmit demultiplexer consisting of three sets of terminal switches 232, 233, 234, 235, 236, 237 and 238. Here, the terminal switches can be used to drive any one of the transducer elements with the signal in wire 240. During the transmit mode, the switches 229 and 239 disconnect the input and output, respectively, of the bimodal multiplexer to protect it from high transmit voltages.

[0108] Fig. 21 is only an illustration. In general, the invention encompasses protection against high voltage for the receive circuit in a bimodal multiplexer during the transmit mode and a transmit circuit that can connect one wire to any combination of elements in a sub-array. [0109] Fig. 22 illustrates the method of using the embodiment of the invention illustrated in Fig. 21 to perform transmit beamforming. Here, the signals 250, 251 drive the switch 232. Similarly, signals 252, 253 drive the switch 233; signals 254, 255 drive the switch 236 and signals 256, 257 drive the switch 235. Signals 258, 259 drive switch 234. Signals 260, 261 drive switch 237. Signals 262, 263 drive switch 238.

[0110] For purposes of explanation, assume that the transmitter needs to tilt the beam at an angle. To accomplish this, one would generally have to drive transducer elements 228 and 241-247 with the same signal with a delay between successive transducer elements. Once again for purposes of explanation, consider a signal that consists of a single pulse. This pulse could be generated at each transducer element by controlling switch 232 such that, in accordance with drive signals 250 and 251, it first supplies a pulse to element 228 and then the same pulse is applied to 241 with a delay determined by the delay between the signals 250 and 251. The nature of the embodiment of invention shown in Fig. 20 enables the invention to have an overlap between drive signals 250 and 251 and consequently the pulses driving 228 and 241. The pulse is generated by connecting a DC voltage to the input of the switch element 238. In general, a sequence of more complex signals can be used to drive the transducer by creating multiple sets of demultiplexers like the ones constructed from elements 232-238. A multilevel signal could be created by connecting the inputs of these demultiplexers to different voltages and driving the constituent switches in the sequence required to generate an appropriate drive signal.

[0111] In conjunction with switches 232, 233, 235 and 236, the switches 234 and 237 could be controlled to direct the drive signal from one part of the array to another.

[0112] The implementation of this embodiment of the invention would require the distribution of a clock to each of the demultiplexer elements, along with a digital control signal that determines the clock instant whe,n each switch would activate. The digital control signal could be modified from one transmit pulse to another by programming from the base unit, or the digital control signal could change from one transmit pulse to another in a pre-set sequence. The sequential activation of the elements would be triggered at a well defined instant by means of some synchronizing signal. The control signals would also determine the sequence of activation of subsequent sub-arrays (not shown in Fig. 21).

[0113] The invention is independent of the means used to control the instant when the switch should activate and is independent of the complexity of waveforms generated. It encompasses the inclusion of circuit elements in the bimodal multiplexer that can drive the transducer elements in a sequence with the appropriate signal waveform to generate the desired ultrasound wavefront.

[0114] Fig. 23 is an illustration of one example of the use of the invention in a practical setting. The system includes an acoustic lens 270 in an ultrasound probe, a transducer array 271 , boards with electronics 272 and 273, and an interconnect array 274. The interconnect wires in the cable bundle 275 connect the probe to a base unit 276, which may comprise any known monitor or processing system, local or remote (such as those that receive signals over a network for remote diagnosis), storage device (for example, for later processing), etc.

[0115] Fig. 24 illustrates the invention where the base unit 276 transmits the different frequencies required for the multiplexing via wires 277, 278, 279, 280, 281 , 282, 283 and 284. This invention reduces the power dissipation in the probe head by moving the problem of generating low-noise oscillators for mixing to the base unit 276, where typically more power can be dissipated. In this Figure, both in-phase (277-280) and quadrature phase (281-284) local oscillators provide signals at four different oscillator frequencies for transmission to the probe. Here, pairs of wires 277,281 ; 278,282; 279,283 and 280,284 carry signals at the same frequency. The probe electronics contain a bank of Gilbert cells 285. These will take the in-phase and quadrature oscillator signals and combine them to provide an oscillator signal at the phase required by the mixers in the bimodal multiplexers. In general, there will be more than two Gilbert cells associated with each oscillator signal. In one embodiment, several Gilbert cells with fixed combining ratios could be used to generate a discrete set of phases and an analog multiplexer could be used to connect the mixers with the Gilbert cell output with the right frequency and phase.

[0116] Fig. 25 illustrates an alternative aspect of the invention, where a plurality of wires 286, 287, 288 and 289 is used to transmit the same oscillator signal, but at different phases from a base unit 276. This reduces even further the burden of generating local oscillator signals in the probe.

[0117] Fig. 26 illustrates an alternative aspect of the invention, where a multiple of the local oscillator signal 290 is transmitted down the cable from the base unit 276. In the probe head, this is divided down to generate the local oscillator signal 292. In this embodiment, the signal from the base unit is thus divided by a succession of "divide by two" circuits 291. If implemented using a D Flip flop, such division produces both in-phase and quadrature signals that can be combined using a bank of Gilbert cell mixers 293. This can be used as in Fig. 24 to generate a continuum of phases. Another possibility is for the base unit to transmit signals that are mixed together or with a fixed frequency in the probe head to obtain the multiples of the local oscillator signal. In this method, of the two signals being mixed, one could be transmitted with a multiplicity of phases on different wires such that the electronics in the probe head could obtain a signal at any desired phase by multiplexing the correct wire.

[0118] Fig. 27 illustrates an alternative embodiment of the invention, where the signals from multiple local oscillators, or the local oscillator signals for different bimodal multiplexers, are transmitted on the same cable 294 from the base unit 276. The signal from the base unit is separated in the probe using frequency-domain filters 295, 296, 297 and 298. The signals from the filters are processed as shown in Fig. 24 and 25 to generate a variety of oscillator phases.

[0119] Figs. 28-31 show only examples: The invention encompasses any arrangement and method of generating a substantial local oscillator signal in the base unit and transmitting it to the probe head. The invention is therefore very general and extends to other systems that employ, for example frequency-domain multiplexing (FDM), TDM or a switched capacitor-based phased array in the probe head. The invention also extends to the method of generating either multiples of the local frequency or different phases of one or more clocks in the base unit and transmitting the resulting signals to the probe head either over a cable or in any other way, including wirelessly. The invention also encompasses the method of processing these signals in the probe head to generate the local oscillator signal or different phases of the local oscillator.

[0120] Fig. 28 illustrates a receiver in the base unit that supports a bimodal multiplexer in the probe head. The figure illustrates one of the wires 315 carrying signal from the probe head 316. The receiver in Fig. 28 preferably includes a low-pass filter 299 and several band-pass filters 300, 301 and 302 that are tuned to the same set of frequencies as the local oscillator signals. The outputs from the low-pass filters and band-pass filters are mixed with another local oscillator in the mixers 303, 304, 305 and 306 such that the output of the mixer is at a fixed frequency. The mixer output is low-pass filtered in 307, 308, 309 and 310 and the array of fixed frequency signals is captured by an array of analog-to-digital converters 311 , 312, 313 and 314. The digitized signals are then sent to a beamformer, which is a well known component in ultrasound systems.

[0121] Fig. 28 shows only one embodiment of the receiver in the base unit. In another embodiment, the receiver in the base unit directly converts the receive signal into digital form using an analog-to-digital converter. The digitized signals are then filtered digitally and shifted to the desired frequency by digitally mixing.

[0122] These are simply examples. The invention encompasses any base unit receiver for the bimodal multiplexer that discriminates the received signals on the basis of their frequency and further provides a digital representation of such discriminated signals to the beamformer.

[0123] Fig. 29 illustrates a base unit receiver that is designed to reconstruct the signals generated by the arrangement shown in Fig. 14. Here, the signal from a cable bundle 317 arrives on a plurality of wires. Each of the wires carries a slice of the signal from the transducer, shifted to some frequency. The signal carried by each of the wires is first mixed in mixers 318, 319, 320 and 321. The mixers might optionally be preceded by a pre- amplifier. The mixers are driven by local oscillator frequencies such that the signal is shifted from each wire to the frequency of the slice that it represented at the transducer. This process thus moves each slice back to its original position. These slices are available after the band-pass filters 322, 323, 324 and 325. The output of these band-pass filters is then summed in an adder 326 before being converted in an analog-to-digital converter 327.

[0124] As an alternative, the signals arriving on different cables could be directly digitized using an array of analog-to-digital converters. The signals could then be shifted to the right frequency and combined in the digital domain.

[0125] These are only illustrations - in general, the invention encompasses any arrangement or method to receive signals in the base unit from a high bandwidth transducer through different wires, where each wire carries a slice of the signal spectrum. The received signal is reconstructed by shifting each slice back to its original spectral location and adding the resulting properly located slices. The reconstructed output is made available to a digital beamformer. [0126] Fig. 30 illustrates a base unit receiver that performs equalization on the received signal to counteract signal degradation resulting during propagation through the cable. The received signal 328 is amplified in an equalizing receiver 329. The receiver output 330 is spectrally analyzed by a low-pass filter 331 and a band-pass filter 332. Rectifiers 333 and 334 create a low-pass filtered version of the envelope of the signals at the frequency bands determined by the filters 331 and 332. The outputs of the rectifiers are converted using linear-to-log converters 335 and 336. The logarithms of the spectral envelope at the different frequencies are subtracted using a suitable component 337. The difference of the logarithms (which is the same as the ratio of the rectifier output signals) is compared with a reference signal 338. The difference is used by the element 339 to drive the feedback to the equalized receiver 329. In equilibrium, the signal at the output of the receiver 329 has a fixed ratio of energy in two spectral bands, that is, the signal spectrum approaches a pre-defined standard.

[0127] Fig. 31 shows one embodiment of an analog filter that has a peaked gain path 341 and a flat band path 342. The signal from the cable traverses both these paths and is thereafter combined in a Gilbert cell mixer 343. The relative strengths of the peaked gain and flat band paths are determined by the currents through the sources 344 and 345 that determine the transconductance. The relatively weighted outputs of the two paths are available at the output of the Gilbert cell 343.

[0128] Figs. 31 and 32 are only illustrations of the method of equalization according to the invention, where the received signal is equalized by observing certain characteristics of the received signals. While examples of analog equalization are provided here, the invention extends to any technique that applies signal equalization techniques in the base unit of an ultrasound receiver such that the received signal characteristics are matched with some pre-set requirements. These equalization techniques might or might not be solely implemented in the base unit. These techniques could further encompass changes in the filter characteristics used in the bimodal multiplexer and/or shift in the local oscillator frequencies.

[0129] Fig. 32 illustrates an invention that is used to aid digital equalization of the received signal. Here, the base unit 276 transmits a calibration pattern (e.g. pseudo-random bit sequence - PRBS) through a specially dedicated wire 346. A switch 347 connects the receiver at the probe head to a wire 346. The signal in the wire 346 has noise-like characteristics with substantially the same frequency content as the signal expected from the transducer element 348 . The wire 346 is assumed to have little degradation due to cable impairments. The signal from the wire 346 is amplified and processed in front end circuitry 349, which optionally includes a bimodal multiplexer. The output of the circuitry 349 is carried back to the base unit on a wire 350. During this process, the array element 348 is disconnected from the signal path by switch 347. The base unit compares the received calibration pattern with the transmitted calibration pattern and adaptively adjusts the receive digital filter that will correct for cable degradation. The receive filter could be implemented as a Feedforward Filter (FFE), Distributed Feedback filter (DFE) or a Maximum Likelihood Sequence Estimator (MLSE). The invention is not restricted by the implementation of the filter, but rather encompasses the general use of filtering in the base unit to equalize for signal degradation due to the cable. This adaptive adjustment could be based in one embodiment on the Least Mean Square (LMS) method of adaptive adjustment of digital filters. Moreover, the base unit could transmit a multiplicity of un-correlated signals through multiple receiver channels to adapt a cross-talk correction filter as well.

[0130] Fig. 33 illustrates the use of a combination of feedforward and feedback equalizers in the base unit. Here the output from a feedforward equalizer 351 is combined in a digital adder 353 with that from a decision feedback equalizer 352. The output of the digital adder is sliced in a slicer 354. In the case of an ultrasound system, the slicer will typically be replaced by a multi-level detector. [0131] Fig. 32 is merely an illustration of an embodiment in which a calibration signal from the base unit 276 is propagated to the probe. It is returned from the probe traversing the signal path. This loop-back is used to adjust adaptive filters in the base unit. The filter settings are used during signal reception to mitigate the impact of cable degradation and electrical cross-talk. The loop-back calibration is repeated with sufficient frequency to maintain signal quality without interfering in the diagnostic imaging tasks.

[0132] Fig. 34 illustrates an embodiment of the invention in which a calibration signal known by prior design to a base unit 276 is generated in the probe head in a circuit 355. This can be done for example by having a PRBS pattern generator which drives a filter with a well-defined impulse response. The calibration signal is connected to the signal path through a switch 356. It traverses the signal path comprising the amplifiers and multiplexers in the element 357 and the cable 358. During this process, the array element 348 is disconnected from the signal path by a switch 356. It arrives at the base unit 276, where it is converted with an ADC and the received signal is used to adapt a digital filter that equalizes for cable degradation and corrects for inter- channel cross-talk. In a different embodiment of this system, an LMS adaptive algorithm could be used in the receiver to perform blind equalization - i.e. equalization without a priori knowledge of the transmit pattern.

[0133] Fig. 34 can be extended to show a system in which multiple uncorrelated calibration patterns are launched through adjacent wires. This provides the base unit 276 with an array of received signals which it could utilize to build a far more comprehensive model of the cable 358. This model then guides an adaptive arrangement for correcting inter-channel cross-talk and cable cross-talk constraints. The cross-talk correction can be accomplished by blind cross-talk elimination or by using multiple-input, multiple-output (MIMO) techniques. The invention is not restricted by the method of cross-talk cancellation and instead generally encompasses the use of cross-talk cancellation to improve imaging performance. [0134] Fig. 35 illustrates the use of MIMO for cross-talk cancellation. Here, a MIMO processing block 359 builds a model of the channel comprising ultrasound cables using the known transmit pattern through transmitters 360 and the received signals in wires 361. Once the calibration for cross-talk is completed, the MIMO block is used to correct received signals for cross-talk without any knowledge of the transmitter signals. In this illustration, the received signal with correction is provided along lines 362.

[0135] Fig. 36 illustrates the signal chain in the base unit that comprises the receiver 363, which receives the signal from the cable and converts it into digital data. The MIMO block 364 performs cross-talk cancellation and the equalizer 365 equalizes the signal. The multi-level detector 366 then provides data decisions. The output from the multi-level detector is propagated through the beamformer to the rest of the signal chain.

[0136] Fig. 36 is only an illustration. The invention generally encompasses any arrangement and method whereby a calibration signal originating in the probe head traverses all or part of the signal chain to the base unit. The base unit exploits the received signal and, with or without prior knowledge of the calibration signal, adapts analog or digital circuits that will correct for cable- induced signal degradation, including but not limited to cross-talk and limited signal bandwidth. This adaptation could be repeated as often as required to maintain signal quality without impacting the diagnostic process.

[0137] Another consideration is the collection of heat generated by the bimodal multiplexer electronic circuitry and associated electronic circuitry and by acoustic vibrations in the ultrasound transducer. The invention may include a heat sink for removing the heat to a location away from the transducer. A hollow plate made of metal or other thermally conductive material with a fluid or gas such as water or air pumped through it will be in thermal contact through a thermally conductive path with the electronic and acoustic heat generating components. The fluid or gas becomes warmer and collects the heat, carrying it away from the transducer in pipes exiting the transducer to a remote location. The remote location may be in the ultrasound system console or another location sufficiently distant from the transducer.

[0138] Fig. 37 shows a concept of an ultrasound transducer with a 2D array of piezoelectric elements 367 that receive acoustic signals, forming the front of an acoustic stack 368. The electrical output of the acoustic stack is input into an assembly of electronic circuits 369 that process and deliver the ultrasound signals through interconnections 370 that carry the signals through wires 371 to the an ultrasound system console (not shown). Heat sinks 372 made of copper or other significantly thermally conductive material collect heat generated by the acoustical and electrical power dissipation in the transducer. The heat reaches the heat sink through thermally conductive pathways 373 that also connect to the electronic circuits and the acoustic stack. A liquid and/or gas may be pumped into and out of input and output channels 374 and 375 to remove and carry heat to a location outside the transducer. The pump and heat sink for the fluid heat transfer may be located in the ultrasound system console or at some other location outside the transducer. One possible arrangement is to place the cooling plates behind the acoustic stack and into a slot 376 in the electronics assembly, so that the plate may transfer heat by conduction away from both the acoustic stack and electronics assembly.

[0139] A system with or without the cooling described above may include circuitry for converting signals from an array of ultrasound transducers into digital form with analog-to-digital converters in significant proximity to the transducers. The digital data might be further processed with beamforming or compression. The processed data might be reduced to a form compatible with a network protocol like IOGbase Ethernet or such other protocol. The data so reduced could be transmitted to a base unit by wire or cable or even using a laser connected to an optical fiber. It might also be transmitted to the base unit wirelessly. The system could contain receive high-voltage protection switches and transmit demultiplexers that enable transmit beam steering.

Claims

WHAT IS CLAIMED IS:

1. A multiplexing system comprising: a plurality of inputs connected to sources of input signals; a multiplexer coupled to said inputs to generate an output signal in response to said input signals, said multiplexer being configured to provide both shifts in the relative timing of said input signals for coherent summing of said input signals and shifts in frequencies of said input signals to enable discrimination of individual said input signals within said output signal in which said input signals are coherently summed; and an output coupled to said multiplexer for transmission of said output signal.

2. The multiplexing system of claim 1 wherein said inputs are connected to a plurality of ultrasound transducer elements in an array of ultrasound transducer elements.

3. The multiplexing system of claim 1 wherein said inputs comprise first inputs connected to a sub-array of a total number of said sources, said multiplexer having at least one set of second inputs in which each set is connected to a different sub-array of said sources, said multiplexer being configured to separately provide both of said shifts for said different sub-arrays so as to form a plurality of sub-array output signals, each said sub-array output signal being a coherent summation of input signals from the respective sub-array of said sources.

4. The multiplexing system of claim 3 wherein said multiplexer includes an output summer connected to receive and combine said sub-array output signals for transmission as a composite output signal, said multiplexer being configured to vary at least one signal characteristic of individual said sub-array output signals so as to enable discrimination of said individual sub-array output signals.

5. The multiplexing system of claim 4 wherein said multiplexer varies frequencies of said sub-array output signals to enable said discrimination, said multiplexer including frequency-specific filters connected to regulate frequency bands of said individual sub-array output signals so as to control overlap of said frequency bands.

6. The multiplexing system of claim 1 wherein said multiplexer is one of M different multiplexers, where M is greater than one, said multiplexers having inputs connected to different sub-arrays of a total number of said sources such that each multiplexer is uniquely associated with a particular sub-array, each said multiplexer being configured to provide both said shifts to generate an output signal in which input signals from said uniquely associated sub-array are coherently summed.

7. The multiplexing system of claim 6 further comprising frequency shifters connected to said input and said output signals and to generate M signals that are distinguished with respect to frequency bands, said M signals being received by an output summer configured to combine said M signals for transmission via a single wire or wireless channel, said sources being ultrasound transducer elements of a transducer array.

8. The multiplexing system of claim 1 further comprising a transducer element selection arrangement for enabling and disabling signal flow between said multiplexer and specific said sources, said sources being ultrasound transducer elements, said transducer element selection arrangement thereby enabling flexibility in addressing said ultrasound transducer elements.

9. The multiplexing system of claim 1 wherein said multiplexer includes variable gain so as to support high signal dynamic range.

10. The multiplexing system of claim 1 wherein said multiplexer includes a plurality of mixers and a Trans Impedance Amplifier (TIA), said mixers being connected between said sources of said input signals and said TIA, each said mixer being associated with one of said sources and with a local oscillator signal selected to enable frequency shifting and phase shifting of said input signals from said sources, said phase shifting implementing said shifts in relative timing of said input signals, said TIA providing a virtual ground for outputs of said mixers so as to convert voltage signals to current signals, said multiplexer being configured to combine said current flows to provide said output signal in which said input signals from said sources are coherently summed.

11. The multiplexing system of claim 10 wherein said multiplexer further includes variable feedback across said TIA to enable gain control.

12. The multiplexing system of claim 10 having a plurality of said multiplexers connected to sub-arrays of said sources, wherein each said multiplexer generates a coherent summation output signal, said coherent summation output signals being passed through different frequency-specific filters to enable signal discrimination and being input to a summer for combining said coherent summation output signals.

13. The multiplexing system of claim 1 wherein said multiplexer includes a linearized transconductance stage for each said input, said multiplexer further including a Weaver architecture for each said input, each said Weaver architecture being connected to receive an output of an associated said linearized transconductance stage and including first and second local oscillator signal inputs, each said Weaver architecture being configured such that shifting the phase of one of said first and second local oscillator signal inputs implements said shift in relative timing for said coherent summing.

14. The multiplexing system of claim 1 wherein said multiplexer includes at least one filter with means for selectively adjusting filter characteristics.

15. The multiplexing system of claim 1 further comprising circuitry in proximity to said sources, said sources being ultrasound transducer elements, said circuitry including a filter that is adjustable with respect to frequency and phase characteristics.

16. The multiplexing system of claim 1 wherein said sources are ultrasound transducer elements, said multiplexer including low-pass and band-pass filters specific to isolating tissue-generated second harmonic energy, wherein second harmonic signals are multiplexed to generate said output signal.

17. The multiplexing system of claim 1 wherein said multiplexer includes control signal inputs and local oscillator signal inputs, said local oscillator signal inputs being connected to mixers for enabling said shifts.

18. The multiplexing system of claim 17 wherein said mixers are connected to be responsive to said local oscillator signal inputs, which include in-phase and quadrature-phase signals, said multiplexer further including a plurality of circuit cells connected to provide said mixers with local oscillator signals having target characteristics.

19. The multiplexing system of claim 18 wherein said circuit cells are Gilbert cells.

20. The multiplexing system of claim 17 wherein said multiplexer includes frequency division circuitry connected to receive a high frequency local oscillator signal enabled to generate both in-phase and quadrature-phase signals from said high frequency local oscillator signal.

21. The multiplexing system of claim 17 further comprising a plurality of mixers connected to said local oscillator inputs and configured to provide mixing for generating local oscillator frequencies at different frequencies and phases.

22. The multiplexing system of claim 1 further comprising a plurality of said multiplexers, with each said multiplexer being dedicated to a different sub-array of said sources, said sources being ultrasound transducer elements, said multiplexers having outputs in communication with a receiver via at least one wire, wherein the number of said wires for propagating signals from said outputs is less than the number of said multiplexers, said receiver being enabled to discriminate among signals from different said multiplexers.

23. A method of utilizing multiplexing comprising: connecting a plurality of ultrasound transducer elements to inputs of circuitry enabled to provide multiplexing; shifting signal characteristics of signals generated by individual said ultrasound transducer elements, including providing first shifts to enable coherent summing of said signals and providing second shifts to enable discrimination of said signals; summing said shifted signals to form an output signal; and directing said output signal via a single wire or wireless channel to a base unit receiver for processing.

24. The method of claim 23 further comprising assigning sub-arrays of said ultrasound transducer elements to different multiplexers enabled to implement said shifting of signal characteristics, including providing said summing for each said multiplexer.

25. The method of claim 24 further comprising adjusting frequencies of said output signals in which said summing of signals is achieved, said adjusting being implemented to enable discrimination of individual said output signals by said receiver, said method further comprising summing said output signals and directing said summed output signals via said single wire or wireless channel to said receiver.

26. The method of claim 24 further comprising enhancing signal integrity with respect to said output signals by:

(a) driving said ultrasound transducer elements at two distinct drive levels; (b) determining error on a basis of differencing output signals formed while driving said ultrasound transducer elements at a first drive level and a scaled copy of said output signals formed while driving said ultrasound transducer elements at a second drive level.

27. The method of claim 26 wherein said enhancing signal integrity includes isolating second harmonic signal energy generated by non-linear effects in imaged tissue and circuit components for achieving said multiplexing, including estimating said non-linear effects while driving said ultrasound transducer elements at the lower of said first and second drive levels and including removing estimates of said non-linear effects from said output signals, thereby correcting for said non-linear effects.

28. The method of claim 24 further comprising directing said output signals from said multiplexers via a plurality of wires, said output signals of wires that are adjacent each other being frequency shifted to have signal energy in different frequency bands, thereby controlling signal cross-talk.

29. The method of claim 23 further comprising providing protection to the multiplexer from high voltage in signal exchanges between said multiplexer and said base unit receiver.

30. The method of claim 23 further comprising repeatedly reselecting which one of a plurality of sub-arrays of said ultrasound transducer elements for input to said multiplexer, such that said output signal is a coherent summing for different sub-arrays at different times.

31. The method of claim 24 further comprising providing beamforming during direction of said output signals to said receiver, including:

(a) controlling sequencing of said ultrasound transducer elements within said sub-arrays; and (b) controlling sequencing of multiplexers in directing said output signals to said receiver.

32. The method of claim 23 further comprising performing pre-equalization on said output signal from said multiplexing circuitry prior to directing said output signal to said base unit receiver, said pre-equalization being tailored on the basis of anticipated signal degradation in said wire.

33. The method of claim 23 further comprising using at least one variable oscillator frequency for achieving said shifting of signal characteristics and adjusting said variable oscillator frequency on the basis of enhancing performance.

34. The method of claim 23 further comprising exchanging calibration pattern information between said base unit receiver and processing circuitry prox- imate to said multiplexing circuitry for said ultrasound transducer elements, including:

(a) using said single wire as one path for said exchanging; and

(b) adaptively adjusting said processing circuitry on the basis of said exchanging.

35. The method of claim 34 further comprising exchanging calibration pattern information to said base unit receiver via said single wire or wireless channel for adaptively adjusting processing characteristics.

36. The method of claim 34 wherein said exchanging further includes

(c) coordinating said receiver and said processing circuitry with respect to a start and a duration of said adaptive adjusting.

37. A multiplexing system comprising: a probe having an array of ultrasound transducer elements for generating ultrasound energy in response to electrical drive signals; a probe input/output enabled for exchanging signals with a base unit; and a transmit demultiplexer connected to drive said ultrasound transducer elements in response to a multi-element drive signal received from said base unit via a single wire or wireless channel of said probe input/output, said transmit demultiplexer being configured to drive said ultrasound transducer elements in an addressing sequence dictated by said base unit on a basis of achieving desired imaging characteristics.

38. The multiplexing system of claim 37 wherein said transmit demultiplexer is cooperative with said base unit to drive said ultrasound transducer elements by sequencing sub-arrays, thereby enabling sequencing of said sub-arrays of said array and sequencing of said ultrasound transducer elements of individual said sub-arrays so as to achieve said desired imaging characteristics.

39. The multiplexer of claim 37 wherein said transmit demultiplexer includes switching circuitry which is responsive to said multi-element drive signal, said multi-element drive signal being received as a pulse patterning for sequencing the driving of said ultrasound transducer elements.

40. The multiplexing system of claim 39 wherein said switching circuitry is responsive to said multi-element drive signal to achieve beamsteering while said array of ultrasound transducer elements is in a beam transmit mode of operation.

41. The multiplexing system of claim 37 wherein said probe input/output is a wire cable and said multi-element drive signal is transmitted to said probe along a single wire of said wire cable.