EP2268419A1

EP2268419A1 - Multiple frequency band acoustic transducer arrays

Info

Publication number: EP2268419A1
Application number: EP09700502A
Authority: EP
Inventors: Rune Hansen; Svein-Erik MÅSØY; Tonni F. Johansen; Sven Peter NÄSHOLM; Bjørn A.J. ANGELSEN
Original assignee: Surf Technology AS
Current assignee: Surf Technology AS
Priority date: 2008-01-09
Filing date: 2009-01-09
Publication date: 2011-01-05
Also published as: CN101965232B; WO2009088307A1; CN101965232A

Abstract

Acoustic probes that transmits/receives acoustic pulses with frequencies both in a high frequency (HF), a and a selectable amount of lower frequency (LF1, LF2,..., LFn,...) bands, where the radiation surfaces of at least two of the multiple frequency bands have a common region. Several solutions for transmission (and reception) of HF, LF1, LF2,.... pulses and signals through the common radiation surface are given. The arrays and elements can be of a general type, for example annular arrays, phased or switched arrays, linear arrays with division in both azimuth and elevation direction, like a 1.5D, a 1.75D and a full 2D array, curved arrays, etc. The element division, array type, and array aperture sizes for the different bands can also be different. Electronic substrate layers with integrated electronic that connects to array elements can be stacked within the probe.

Description

MULTIPLE FREQUENCY BAND ACOUSTIC TRANSDUCER ARRAYS

1. Field of the invention The present invention is directed to technology and designs of efficient acoustic (sonic and ultrasonic) bulk wave transducers for operation in at least two frequency bands. Applications of the transducers are for example, but not limited to, medical ultrasound imaging, nondestructive testing, industrial and biological inspections, geological applications, and SONAR applications .

2. Background of the invention

The utilization of the nonlinear elasticity of tissue and ultrasound contrast agent micro-bubbles in medical ultrasound imaging provides improved images with less noise. The widest use is in the so-called harmonic imaging, where the 2^nd harmonic component of the transmitted frequency band is used for the imaging, extracted from the signal either through filtering or ^■ through the Pulse Inversion (PI) technique. A use of 3^rd and 4^th harmonic components of the transmitted pulse for imaging is also presented in US Pat 6,461,303.

US Pat applications 10/189,350 and 10/204,350 describe in depth different uses of dual band transmitted ultrasound and acoustic pulse complexes that provide images with reduced noise, images of nonlinear scattering, and quantitative object parameters that greatly enhance the use of ultrasound and acoustic imaging. The methods are applicable both with transmission and scatter imaging. For these applications one would transmit dual band pulse complexes as illustrated by the example in FIG. 1, where in FIG. Ia a high frequency (HF) pulse 101 rides on the peak pressure of a low frequency (LF) pulse 102. FIG. Ib shows another situation where the HF pulse 103 rides on the maximal gradient of the LF pulse 102. The ratios of the center frequencies of the LF and HF pulses can typically be in the range of 1:5 - 1:20, and at the same time the HF pulse must be found in defined intervals of the LF pulse throughout defined depth ranges of the images .

In other applications one wants from the same probe to transmit a low frequency (e.g. 0.5 - 2 MHz) wave for treatment of tissue (hyperthermia or cavitation destruction of tissue) or release of drug carried in nano or micro particles or bubbles, while being able to provide ultrasound imaging from the same probe surface at a higher frequency (e.g. 5 - 10 MHz). In yet another application one wants to have a probe for combined ultrasound treatment and imaging with 3 frequency bands, where for example a 2^nd lower frequency (LF2) band ~ 400 kHz is used to generate pulses for cavitation in the tissue, for example to break nano-sized liposome particles containing drugs for drug delivery to tumors, a 1^st low frequency (LFl) band ~ 3MHz is used for heating of the tissue for hyperthermia treatment of tumors, often referred to as HIFU - High Intensity Focused Ultrasound, or to increase blood flow in the tumor for improved oxygenation of the tumor or to improve the efficiency of the ~ 400 kHz breaking of drug carrying particle, and a high frequency (HF) band - 20 MHz is used for imaging, potentially also in combination with the ~ 3 MHz LFl band for nonlinear manipulation of object elasticity for imaging, for example according to US Pat applications 10/189,350 and 10/204,350.

In yet other applications one simply wants to have a larger selection of frequency bands available for imaging from the same probe for a large variation of depth ranges . For example in portable ultrasound imaging systems for emergency medicine, one wants to use center frequencies of 2.5 MHz for deep object imaging, and use the same probe to image at 7 - 10 MHz center frequencies for objects closer to the body surface. The arrays can for example be arranged as phased linear arrays, switched linear arrays, and curvilinear arrays . The need for multiband transducers is also found in many other applications of acoustic imaging, for example as in non-destructive testing (NDT) of materials, observations of geological formations with elastic waves, and SONAR measurements and imaging of fish, for example close to the sea bottom, the sea bottom, and objects like mines both on the sea bottom and buried under the sea bottom or in the soil on land. This both relates to nonlinear measurements and imaging with multiband pulse complexes, and the ability to select different frequency band pulses for different needs, such as different measurement ranges .

Dual band transmitted pulses were used in M-mode and Doppler measurements in BrHeart J.1984 Jan;51(l):61-9. Further examples are shown in US Pat 5,410,516 where sum and difference bands of the transmitted bands produced in the nonlinear scattering from contrast agent micro-bubbles where detected. A further development of this dual band transmission is done in US Pat 6,312,383 and US Pat application 10/864,992. The current invention presents several solutions to these challenges of transducer array designs. We do in the description most often consider the situation where the elastic waves are in the ultrasound frequency range, but it should be clear to any-one skilled in the art that the solutions according to the invention can be applied to any frequency range of acoustic waves, and also to shear waves in solids. 3. Summary ofthe invention

This summary gives a brief overview of components of the invention and does not present any limitations as to the extent of the invention, where the invention is solely defined by the claims appended hereto.

The invention presents solutions to the general need for an acoustic, often ultrasound, array probe that transmits /receives acoustic pulses with frequencies in separated multiple frequency bands through an at least partially common radiation surface. The common radiation surface has many advantages, for example to minimize the size of a dual or multi band probe to be used from the same instrument. In other situations one needs a common radiation surface for simultaneous transmission of a high frequency (HF) and a low frequency (LFl) pulse with low or controllable phase sliding between the HF and LFl pulses in an actual imaging range, so that the HF pulse is found in a defined region of the LFl pressure oscillation.

The invention also presents a general procedure to design an array with a freely selectable number of operating frequency bands. In special one presents solutions to transducer arrays for transmission and reception of 3 -band pulse complexes containing a high frequency (HF) , a 1^st lower frequency (LFl) and a 2^nd lower frequency (LF2) band, or transmission and potential reception of separate pulses in 3 different frequency bands (HF, LFl, and LF2 bands) . The invention provides solutions where the ratios of the center frequencies are in the range (HF: LFl) of ~ 3:1 - 20:1, with no defined upper or lower limit to the ratio. The ratio of the center frequencies of the LF1-.LF2 bands can have similar values. With the lowest separation of the center frequencies one for example obtains a probe for multi-selection of image frequency bands, for example center frequencies as 2 MHz, 5.5 MHz, and 15 MHz. With a larger separation of the center frequencies one obtains a probe for imaging with the methods described in US Pat applications 10/189,350 and 10/204,350 that also can include frequency bands for HIFU and cavitation treatment of the tissue.

To achieve transmission of multi-band pulses where at least a part of the radiation surfaces are common, the invention presents solutions with a group of arrays that are resonant for each frequency band, and that has at least partially a common radiation surface. The arrays can have a general arrangement of the elements, for example linear phased or switched arrays, or annular arrays. The arrays can be flat or curved, both concave and convex, in one or two dimensions. Element divisions of the linear arrays in the elevation direction to for example a 1.5D, a 1.75D and even a full 2D array are also embodiments according to the invention. One can further have different sizes, forms, and divisions of the array elements for the different bands, for example, but not limited to, an annular array for low frequency treatment pulses with a linear switched or phased array for imaging. The invention also provides solutions for efficient packaging of electronics related to the array beam forming, such as transmit and receive amplifiers for the individual array elements, sub-aperture beam former electronics that allow connection of a group of elements to the instrument via a single wire, electronic switches for connecting selected groups of array elements in electric parallel to beam former channels via a single wire, both for transmit and receive, etc.

In order to minimize relative position sliding between the different frequency band pulses with depth, and defeat diffraction to obtain adequately collimated low frequency (LFl₇ LF2, ..., LFn,...) beams at deep ranges, the invention presents a solution where the arrays for the different bands have a large common radiation surface, and where parts of the radiation surfaces of the lower frequency arrays can be outside the radiation surfaces of higher frequency arrays . To minimize the overlap between the different frequency band pulses in the near field, the invention also presents solutions where central parts of the lower frequency apertures are inactive. To efficiently select between different sizes and overlaps of the radiation surfaces of the different bands, the invention devices the use of different arrays for the different bands, with special solutions of the array constructions to provide common radiation surfaces of the different bands .

In one embodiment according to the invention to obtain a common HF and LFl array radiation surface, the HF and LFl pulses are generated with separate piezoelectric layers stacked in front of each other with the HF piezo-layer in the front, and an isolation section for HF vibrations to the front of the LFl piezo-layer. A load matching section of impedance matching layers is placed between the HF piezo-layer and the load material to the front. The isolation section is designed so that the reflection coefficient between the HF piezo-layer and the isolation section is high in the HF band so that the layers behind the HF piezo-layer has low influence on resonances in the electro-acoustic transduction of the HF piezo-layer in the HF band. The isolation section is also designed so that in the LFl band it cooperates with the probe layers in front of the isolation section to provide acoustic matching of the LFl piezo-layer to the load material. Close to unit reflection coefficient between the HF piezo- layer and the isolation section is obtained when the impedance seen into the isolation section from the front is low or high compared to the characteristic impedance of the HF piezo-layer. When the impedance into the isolation section from the front is low in the HF band, the HF piezo-layer will have a thickness resonance when it is approximately half a wavelength thick around the center of the HF band. When the impedance into the isolation section from the front is high in the HF band, the HF piezo-layer will have a thickness resonance when it is approximately a quarter wavelength thick around the center of the HF band. The quarter wave resonance generally allows wider bandwidth of the HF layer resonance, but with poorer phase angle of the electrical impedance compared to for half wave length resonance of the HF piezo-layer.

The invention provides special designs of the isolation section that provides either adequately high or adequately low impedance into the front of the isolation section in the HF band, with low sensitivity to the impedance seen from the back of the isolation section. This is especially important when the LFl piezo-layers are made as ceramic/polymer composites where one wants to minimize variations in the reflection coefficient from the HF layer towards the isolation section when the isolation section connects to polymer or ceramics in the composite. To achieve this reduced impedance sensitivity of the reflection coefficient, the invention provides solutions where the isolation section is composed of at least two acoustic layers.

In a 1^st embodiment of the isolation section according to the invention, the isolation section contains an impedance regularizing layer at the back of the isolation section that is adequately thin and heavy so that it in the HF band approximately represents a mass, adequately large, in series with the impedance to the back. This mass is then in series with the loading of the isolation section to the back, and makes the impedance transformation of the whole isolation section less dependent on whether the isolation section ends into polymer or ceramic in the LFl piezo- composite. The impedance regularizing layer is preferably a heavy material, for example Cu, Ag, Au, Pd, Pt, W, or alloys of such materials, or powders of such materials or their alloys sintered together or glued in a solvent such as a polymer. The thickness of the back layer can typically be of the order of λ_HF/30 or higher. Due to the large wave propagation velocity of Si (8.44 mm/μsec) , a Si layer can also be used for an impedance regularizing layer with adequate mass, although the mass density of Si is only 2330 kg/m³. The invention also presents a solution where the impedance regularizing layer of the isolation section is made of ceramics, where the ceramics layer can be part of the LFl piezo-layer. This ceramics back layer may conveniently be combined with a thin layer (the order of X_HF/30) of heavy material like Cu, Ag, Au, Pd, Pt, W, or alloys of such materials, or powders of such materials or their alloys sintered together or glued in a solvent such as a polymer. A low impedance into the isolation section can then for example be obtained with a matching layer in front of said impedance regularizing layer of large mass, where said matching layer has low characteristic impedance and is quarter wavelength thick around the center of the HF band. Said matching layer can preferably be made of polymer or similar material . An approximate analysis on how to match the LFl piezo-layer to the load in the LFl band can be done by realizing that both said isolation section matching layer and the HF piezo-layer with load matching layers will be thin compared to the wavelength in the LFl band. This allows a thin layer approximation where said low impedance matching layer behaves as an elastic spring in series with the mass of the HF piezo- and matching layers and the load impedance. The center frequency of the LFl band is then selected at the resonance between this spring and mass where the phase of the impedance into said isolation section matching layer seen from the back is zero. This resonance frequency can be tuned by varying the stiffness of said isolation section matching layer and the mass of the HF piezo and load matching layers . This mass can for example be tuned by varying the ceramic volume fill in the HF piezo-composite.

The embodiment can be modified to obtain a high impedance into the isolation section by adding a

2^nd λ_HF/4 matching layer with high characteristic impedance that connects to the HF piezo-layer in front of the 1^st λ_HF/4 matching layer with low characteristic impedance. With this solution, the impedance seen from the front into the isolation section is less dependent on the thin impedance regularizing layer described above, where this layer in many situations can be removed when two λ_HF/4 matching layers are used. The selection of characteristic impedances of the 1^st and 2^nd matching layers can be done through standard considerations of impedance matching known to anyone skilled in the art. In the LFl band the 1^st, low impedance matching layer will then behave approximately as a spring in series with the combined mass of said 2^nd, high impedance, matching layer and the HF piezo- and load matching layers , where the center frequency of the LFl band is selected at the resonance frequency of said spring and load system, where the material parameters of the spring and mass system can be tuned for resonance in the LFl band. In a less efficient embodiment to provide high impedance into the isolation section in the HF band, one can use a single X_HF/4 matching layer with high characteristic impedance to the front of said impedance regularizing layer of large mass. In the LFl band this single matching layer will approximately behave as a mass in series with the mass of HF piezo- and load matching layers and provide a load impedance seen from the LFl piezo-layer that has an inductive phase. This do not provide optimal impedance matching but a useful form of the LFl electro- acoustic transfer function is obtained.

We have above given some examples of interesting structures of the isolation section, but it is clear to any-one skilled in the art that different designs of the isolation section can be found according to principles of impedance matching known to any-one skilled in the art, where the essence of the invention is to use an isolation section of at least two layers. The impedance regularizing mass layer is very useful when the LFl piezo-layer is made as a ceramic polymer composite, but can be omitted when the LFl piezo-layer is made as a whole ceramic. This can for example be the situation when the LFl layer is used for high power therapy purposes without direction steering of the beam.

One can according to the invention add 2^nd , 3^rd , etc. lower frequency bands to the structure above by extending the piezo-layer structure backwards in front of the backing with sections containing an isolation section in front of a piezoelectric layer for each new low frequency band, where the resonance frequency for the piezo-layers has a monotone decrease with the position backwards in the structure. The isolation sections are designed according to the same principles as for the dual piezo-layer structure described above, where the reflection coefficient into the front of the isolation section is close to unity within the resonance band of the neighbor piezo-layer in front of said new section. Within the resonance frequency of said new piezo-layer, the new isolation section interacts with the layers in front of the isolation section to provide resonant impedance matching between the load and the new piezoelectric layer. The structure can hence be extended backwards with a new such combined isolation section and piezo-layer for each new lower frequency band, in principle ad infinitum, where most practical applications requires in total 2 or 3 lower frequency bands in addition to the HF band.

The structure typically ends with a backing material that has so high absorption that reflected waves in the backing material can be neglected. The last piezo-layer can attach directly to the backing material, or through back matching sections composed of impedance matching layers . The backing material can be used as acoustic power absorbant to reduce resonances in the electro-acoustic transfer functions. Resonances in any of the frequency bands can also be dampened with matching layers of absorbing materials, for example viscous damping polymer materials, and even adding particles to the polymer materials to increase absorption. Viscous damping polymer materials and particle filled polymer materials can also be used in the polymer fills of the ceramic/polymer composites of the piezoelectric layers. Solid/polymer composites can also be used for matching layers to tune the characteristic impedance, where viscous and/or particle filled polymers can be used for increased absorption in the matching layers .

Heavy layers or high-impedance layers of said isolation sections can conveniently be made of one or more electronic substrate layers (typically Si-layers) with electronic circuits, such as transmit and receive amplifiers for the array elements, channel number reducing circuits such as switches for electronically selectable connection of groups of array elements in electric parallel to beam former channels, sub-aperture beam forming for one or both of transmit and receive, so that groups of array elements can be connected to further processing, within the probe or in the instrument, via a reduced number of wires . The signals from groups of elements or groups of sub-apertures of elements may also be transmitted on a single wire by time-multiplexing samples of the signals from such groups, where the time- multiplexing circuits are integrated into said electronic substrate layers , to reduce the cable connections to the arrays .

The electronic substrate layers can conveniently be part of the heavy, impedance regularizing back layer of an isolation section, but also conveniently part of a high impedance front layer of an isolation section. In the latter situation, electronic circuits on the front of the substrate layer can connect directly to the array elements in front, for example HF array elements, through metal pads and known connection techniques such as anisotropic conducting polymer glue containing conducting particles, micro soldering, ultrasonic bonding, etc. Channel number reducing circuits are conveniently implemented in these front electronic substrate layers (e.g. switched element selection, sub-aperture electronics, etc.), to reduce the number of connections to further processing electronics that can comprise or be part of an impedance regularizing back layer of an isolation section. Such reduced number of connections through a matching layer with low electrical conduction can then be obtained via metal connectors through the layer that are so thin that they have minimal effect on the characteristic acoustic impedance of the matching layers . To extend the thickness of a back isolation section layer of Si substrates for increased processing and circuit complexity in these layers, the lower frequency array behind can conveniently be made as a ceramic/polymer composite with average characteristic acoustic impedance close to that of the electronic substrate (for Si substrate the characteristic impedance is approximately 19.7 MRayl) so that the electronic substrate layers participate in the definition of the resonance of said lower frequency array.

Substrate layers with electronics can also be placed in front of the HF array, behind the HF acoustic matching layers. With the front placement of the substrate layers, the HF array is conveniently made of piezo-ceramic/polymer composite with average characteristic impedance close to that of the electronic substrate layer, so that the substrate layers participate in the resonance definition of the HF array.

In another embodiment according to the invention to obtain common HF and lower frequency array radiation surfaces, the HF transduction is provided by vibrating membranes on a substrate activated by cmut/pmut technology, while the lower frequency pulses are generated with a piezo-layer to the back of said cmut/pmut structure. Behind the cmut/pmut substrate one can conveniently place several electronic substrate-layers with transmit and receive amplifiers, electronic switches, sub-aperture beam forming circuitry, etc. The high acoustic propagation velocity of Si (8.44 mm/μsec) , means that the total thickness of such layers can be a fraction of the LFl wave length in Si, and hence provide minimal modification of the lower frequency transmission through the Si layers. Such modification is further reduced by making the lower frequency piezo-layer closest to the substrates as a ceramic/polymer composite with characteristic impedance close to that of the electronic substrate layers . One can then according to the description above extend the structure backwards with more piezoelectric layers with resonances in lower frequency bands (LF2, LF3 , ...) including isolation sections for vibrations within the bands of the layers to the front, according to the principles of the invention described above . In yet another embodiment according to the invention, to obtain common HF and LFl array radiation surfaces , both the HF and the LFl pulses are generated with separate cmut/pmut membranes on a common substrate, either side by side of each other or the HF membranes on top of the LFl membranes. The HF LFl membranes are then optimized for operation in their respective frequency bads . In yet another embodiment according to the invention, to obtain common array radiation surfaces for a HF band and more than one lower frequency bands, both the HF and more than one lower frequency bands are generated with different cmut/pmut membranes for the different frequency bands on a common substrate. The membranes for the different frequency bands can be either placed side by side of each other or some or all of the membranes are stacked on top of others with increasing frequency band from the lowest to the top, while the rest of the membranes are placed directly on the substrate by the side of the stacked membranes. Also in these embodiments one can conveniently place several electronic substrate-layers with transmit and receive amplifiers and beam forming circuitry behind the cmut/pmut substrate, and the structure can be extended backwards with added lower frequency piezo-layers with isolation section in front, as described above. The arrays can be used for transmission and reception in each of the frequency bands. The methods cited in US Pat applications 10/189,350 and 10/204,350 would transmit dual band complexes and use only the received signal in the highest frequency band for processing to measurement or image signals. The frequency bands of the transmitted dual band complex can then be selected from any of the frequency bands in the probe.

The invention is also useful with sparse arrays, where the grating lobes from the HF aperture should be different from possible grating lobes of the lower frequency arrays, to suppress the effect of transmitted HF grating lobes for example with imaging methods and instruments according to US Pat applications 10/189,350 and 10/204,350. The invention also prescribes instruments that uses acoustic multiple band array probes according to the invention for different purposes, for example the use of the different frequency bands of the probe for imaging at different depths, or acoustic tissue treatment at different frequencies, or imaging according to the methods described in US Pat applications 10/189,350 and 10/204,350, or combined acoustic treatment and imaging with any method. The frequency bands are selected by the instrument, either automatically from the operational settings of the instrument, or manually by the instrument operator through instrument controls . For example can with the imaging methods described in US Pat applications 10/189,350 and 10/204,350 the radiation surfaces of the lower frequency apertures be selectably varied to be one of equal to the HF transmit aperture, and larger than the HF aperture where the HF radiation area is part of the lower frequency radiation areas, and the LFl and/or the HF apertures can be selected to have an inactive central region.

4. Summary ofthe drawings FIG. 1 shows examples of low frequency (LFl) and high frequency (HF) pulse complexes that one wants to transmit,

FIG. 2 shows example HF and LFl radiation surfaces according to the invention, and also for analysis of HF and LFl pulse phase relationships,

FIG. 3 shows a cross section of a dual and a tripple piezo-layer stack arrangement according to the invention that allows transmission and reception of a two and three frequency band pulses through a common front face,

FIG. 4 shows examples of other layer structures that participate in the isolation of the piezo- electric sections in FIG. 3, and also integrated circuit layers to be integrated in the acoustic stack, FIG. 5 shows a front view of a phased array probe according to the invention,

FIG. 6 shows an example of a dual piezolayer arrangement to reduce the electric impedance of array elements,

FIG. 7 shows a front view of a substrate with cmut/pmut micro-machined transduction cells, FIG. 8 shows a cross section of a transducer stack where the HF transduction is generated by cmut/pmut cells on a substrate in front of a piezolayer for LFl transduction, and also inclusion of substrate layers with integrated electronics,

FIG. 9 shows a cross section of a transducer stack where the LFl transduction is generated by cmut/pmut cells on a substrate in front of a piezolayer for HF transduction,

FIG. 10 shows a front and cross section view of a combined LFl and HF section implemented by cmut/pmut transduction cells micro-machined on a substrate, where the HF cells are placed on top of the LFl cells.

FIG. 11 shows a front view of a LFl and HF array arranged as a sparse array where the HF and LFl elements are placed between each other,

FIG. 12 shows a front view of a combined low and high frequency section implemented by cmut/pmut transduction cells micro-machined on a substrate, where the low and high frequency cells are placed side by side of each other.

FIG. 13 shows how a 3^rd electro-acoustic transduction band can be obtained with the cmut/pmut structures in FIG. 8 - 12. 5. Detailed description ofthe invention

Example embodiments of the invention will now be described in relation to the drawings. We start with describing solutions to dual frequency arrays, and describe how these designs can be extended with the same principle for operation in 3 or more frequency bands . Typical examples of dual frequency pulses that one wants to transmit are shown in FIG.1 as described above. The challenges in the design of the arrays lie both in the design of the radiation surfaces so that the HF pulse is kept within desired location of the LFl pulse for adequate image range while maintaining adequate amplitude of the LFl pulse, and in design of a vibration structure that allows transmission of LFl and HF pulses with such wide separation between the frequencies from the same surface.

In some of the applications it is important that the amplitude of the LFl pulse at the location of the HF pulse is as high and close to constant as possible throughout an adequate imaging range. This can require large apertures of the LFl radiation surface to avoid diffraction spread of the LFl beam due to the long wavelength of the LFl pulse compared to the HF pulse. The width of the HF transmission aperture can be limited by a requirement on the length of the HF transmit focal region. This gives situations where one would prefer a larger LFl aperture than the HF aperture, which introduces a sliding between the position of the HF pulse relative to the LFl pulse.

For further analysis of this sliding phenomenon we consider circular apertures because one have analytic expressions of the field on the axis of such apertures . FIG. 2a shows by way of example a circular HF transmit aperture 201 with diameter D_H0 = 2aκo and a concentric LFl transmit aperture 202 which for the example is shown as a ring with outer diameter D_L0 = 2a_LO and inner diameter D_LI = 2a_Lr- A cross section diagram shows the HF and LFl transmit apertures as 203, where they by way of example are curved to the same focus F, 204. The common focus for the HF and LFl transmit apertures is chosen by way of example, and one can in other situations also have different foci of the two apertures, where the LFl aperture also can be unfocused. The transmitted axial continuous wave field for the LFl and the HF apertures at a frequency ω is as a function of the axial distance z given as

(D

where Jc = ω/c and ft) is the angular frequency of the transmitted pulse and c is the acoustic propagation velocity. R_L0(z) shown as 205 is the distance from the outer edge of the LFl aperture to the point z (208) on the z-axis, R_LI(Z) shown as 206 is the distance from the inner edge of the LFl aperture to 208 on the z-axis, and R_H0{z) shown as 207 is the distance from the outer edge of the HF aperture to 208 on the z-axis, and R_Hi(z) is the distance from the inner edge of the HF aperture to 208 on the axis. As the HF aperture has no missing part in the center, we get R₁₁₁(Z) = z, but we shall also consider situations where a central part of the HF aperture with dimater D_HI = 2a_HI is missing. P_LO (Q)) is the LFl transmit pressure at the aperture while Pm(O)) is the HF transmit pressure at the aperture. An absorbing medium can be modeled by a complex wave vector where the imaginary part - Jed represents power absorption and the real part k_r represents wave propagation with an in general frequency dependent phase velocity c_p(ω) . The frequency variation of the phase velocity is produced by the absorption, and can for most situations in tissues and materials with similar absorption be neglected, i.e. c_p(ω) * c. The absorption coefficient is often, due to multiple relaxation phenomena, proportional to the frequency, i.e. k_d(ω) * αω.

We note from the 1^st lines of the expressions in Eqs . (Ia, b) that the pressure do in the near field break up into two pulses with delays R_LI{z)/c and R_LO(z)/c for the LFl pulse, and R_HI (z)/c (from the center) and R_H0(z)/c for the HF pulse. As z increases, the delay difference between these pulses reduces, so that the two pulses start to interfere, both for the LFl and HF waves . We then get a longer pulse than given by P_L0 (co) and P_H0 (ω) with complex central part due to interference between the edge pulses . The interference can introduce zeros in the middle of the LFl and HF pulses with destructive interference, and maxima with constructive interference. For z < F, the propagation distance to z on the axis from the outer edge is longer than the propagation distance from the inner edge, and for an absorbing medium one hence do not get complete destructive interference with zeros of the central part of the LFl and HF pulses. Apodization of the pressure drive amplitude across the array surface, so that the drive amplitude is reduced towards the edges, will also reduce in amplitude the pulses from the edges, i.e. with the delay R_w(z)/c for the LFl pulse, and R_H0(z)/c for the HF pulse.

In the focal region, Taylor expansion of the second lines of Eqs . (Ia, b) shows that interference between the two pulses produces a pulse which approaches the time derivative of the transmitted pulses P_L0 (co) and P_H0 (of) in the focus, and with a delay given by the phase terms. This situation is also found in the far-field of an unfocused aperture, and generally relates to the region where the beam width is limited by diffraction. The phase terms in Eq. (1) represent the average propagation lag from the LFl and HF apertures, respectively as

(2)

^_HF{Z) = ~{R_HO(.Z)+ R_HI{Z))

The differentiation of the transmitted LFl pulse P^ico) towards the focus produces an added time advancement of T_LF/4 of the LFl pulse oscillations, where T_LF is the temporal period of the LFl pulse center frequency, with minor effect on the pulse envelope. We hence see that in the focal region, the LFl and HF pulse lengths are given by the transmitted pulse lengths on the array surface, with a change in the oscillation phase of 90 deg due to the differentiation and propagation lags given by Eq. (2) . Due the differentiation of the LFl pulse, and in addition when D_w > O_HO, HF and LFl pulses will get z-dependent propagation delays that differ from each other, and the location of the HF pulse relative to the LFl pulse will slide with depth as exemplified in 209 - 211 for depths zl, z2 and z3. Albeit the above formulas are developed for circular apertures they illustrate a general principle for apertures of any shape, because the radiated beam originates as interference between spherical waves with origin at all points on the aperture (Huygen's principle) . Hence, the waves originating from points on the LFl aperture outside the HF aperture, will have longer propagation distance to the axis than points on the HF aperture. The difference between these propagation distances varies with depth z, which hence produces the position sliding between the HF and the LFl pulse.

We see that when the LFl and HF transmit apertures are equal, there is no sliding between the LFl and HF pulses in the focal region, but we get a T_LF/4 advancement of the LFl pulse oscillations from the near field to the focus, due to the temporal differentiation of the LFl pulse in the diffraction limited region. An LFl transmit aperture equal to the HF transmit aperture can in many situations be too small so that too high LFl beam divergence due to diffraction is found. Therefore it is often desirable to have a wider LFl transmit aperture than the HF transmit aperture. This produces some added sliding between the HF and LFl pulses with depth, which can be established between tolerable limits through the dimensioning of the transmit apertures. This sliding can also be utilized for different purposes, for example to compensate for variations in the LFl pulse amplitude so that the observed LFl pressure at the location of the HF pulse has less variation with depth than the LFl pulse amplitude.

To further analyze the situation when the LFl and HF apertures are different we continue with the circular apertures. For a common focal depth F, we get the distances from outer and inner edges of the LFl and HF apertures as

2

R_gO(z) = ylz² +2e_gO (F-z) e_g0 = F - ^F² - a_g ² ₀ - ^- g = L,H

2 ^{( 3 )}

R₈₁(Z) = Jz² +2e_gI (F-Z) e_gl = F-p² -a_g ² _l ~ ^L g = L,H

where D_L0 = 2a_LO, D_LI = 2a_LIf D_H0 = 2a_H0, and D_Hi = 2 a_Hχ. When the last term under the root sign is relatively small, we can approximate The z variation of the propagation lag difference between the LFl and HF pulses is then found by inserting Eq. (4) into Eq . ( 2 ) which gives

Aτ(z)= τ_LF(z)- τ_HF{z) = —-^^_L0+i_ι ~4o ~4_r) (5 )

Hence, by choosing we obtain, with accuracy within the approximation, zero sliding between the HF and LFl pulses in the focal range of the LFl pulse, even in the situation where the outer dimension of the LFl transmit aperture is larger than the outer dimension of the HF aperture. A disadvantage with the removed central part of the HF transmit aperture is that the side lobes in the HF transmit beam increase. However, these side lobes are further suppressed by a dynamically focused HF receive aperture. The approximation in Eq. (4) is best around the beam focus, and Eq. (6) do not fully remove phase sliding between the LFl and HF pulses at low depths. For other than circular apertures (for example rectangular apertures) one does not have as simple formulas for the axial field as in Eq. (1) but the analysis above provides a guide for a selection of a HF transmit aperture with a removed center, for minimal phase sliding between the LFl and the HF pulses with depth. With some two-dimensional arrays one can approximate the radiation apertures with circular apertures where Eq. (6) can be used as a guide to define radiation apertures with minimal phase sliding between the LFl and HF pulses .

Different measurement situations put different requirements on tolerable variations of the LFl amplitude and also position sliding between the HF and the LFl pulses, and one therefore often wants at least the LFl transmit aperture to be composed of elements so that the effective width of the LFl transmit aperture can be selected together with the relative transmit timing of the HF and LFl pulses so that in the desired range one gets best possible amplitudes and relative locations of the two pulses. The invention devices an instrument using such a probe, where the selection of the active LFl transmit aperture surface can be done automatically by the instrument depending on the application (e.g. suppression of multiple scattering noise or detection of contrast agent micro bubbles) and image depth, or manually by the instrument operator. One also wants to vary the HF transmit aperture, and during reception of the scattered HF signal one typically wants a receive aperture that increases dynamically with the focus to follow the scatterer depth. Hence, a preferred solution is a combined LFl and HF array with common radiation surfaces, but where the actual LFl and HF transmit apertures can be selected for the application, where the LFl transmit aperture is typically larger than the HF transmit aperture, while the HF receive aperture can be selected wide or possibly wider than the LFl transmit aperture at large depths, for example with dynamic receive aperture with depth.

In the above example, the LFl and HF transmit amplitudes have common foci, which is an advantage in some situations, but differences in LFl and HF transmit foci can also be utilized in the beam designs for different purposes . For example can one for practical purposes use a LFl array that is flat outside the HF aperture, and has the same curvature or lens focus as the HF array within the HF aperture. For some applications one can prefer to use an unfocused LFl aperture that is so wide that the actual imaging range is within the near-field region of the LFl aperture, to avoid phase changes of the LFl pulse due to the differentiation of the LFl pulse as one moves into the diffraction limited region (far-field, focal region) of the LFl beam. With a switched linear HF array where the HF beam directions are normal to the radiation surface (aperture) , the LFl aperture can for some applications be a single element array transducer with somewhat wider aperture than the linear HF array, so that the LFl near field region covers the whole HF image ranger, for example as illustrated in FIG. 2b. In this Figure 220 illustrates the front view of a single element LFl array, that produces a beam illustrated in side view as 221 up to the maximal image depth Z, which is within the near field of the LF aperture for this example. The front view of the radiating surface of a linear HF array is shown as 222, indicating the linear array elements 223, where a selected group of elements produces a selected HF transmit aperture 224 that produces the HF transmit beam 225. For imaging the HF transmit and receive beams are scanned within the rectangular image field 226 while the LFl beam covers the field 221 for all HF beams .

The example embodiment in FIG. 2b is useful to obtain low variation of the LFl pressure along the HF pulse propagation, which is useful for imaging of nonlinear scattering of micro-bubbles and hard scatterers, as described in US Pat applications 10/189,350 and 10/204,350. However, for improved suppression of multiple scattering noise, for example as described in the same applications, it is useful to have a LFl aperture that is inactive in a central region as indicated in FIG. 2c. This Figure shows a LFl aperture 220 that is composed of two elements, a central element 227 with an outer element 228 around. In this example embodiment the central element is larger than the HF aperture 222, but one can also see applications where the element 227 is narrower than the HF aperture in the elevation direction. For imaging of nonlinear scattering, the two LFl elements 227 and 228 would typically be coupled electrically in parallel to give an active LFl transmit aperture 220 as in FIG. 2b. For improved suppression of HF multiple scattering noise one could then use only the outer element 228 for transmission of the LFl pulses, which would reduce the nonlinear interaction between the HF and LFl pulses in the HF near field.

Hence, the invention provides solutions to different challenges for transmitting dual band pulse complexes, where one in general wants to select between a variety of radiation surfaces for the LFl and HF pulses, as conceptually illustrated in FIG. 2d. The form of the apertures are chosen circular for conceptual demonstration of the variations, where one can choose any form of the apertures, for example rectangular, elliptical, curved, etc. according to what suits the application best. In FIG. 2d 230 illustrates a concept where the HF aperture (235) is common to parts of the LFl aperture (236) in a common aperture 238, while the LFl aperture also extends outside the HF aperture. 231 shows a modified concept where the central part 237 of the LFl aperture is inactive as LFl radiation surface, for example to reduce the nonlinear interaction between the LFl and HF pulses in the HF near field. 232 illustrates a further modification to 231 where the inactive central part of the LFl aperture is extended to be larger than the HF aperture, while 233 shows a modification where the LFl and HF apertures are equal. In many situations one wants to have an array where one can select between two or more of these conceptual situations for different operations of the measurement or imaging. The selection of the apertures can for example be done automatically by the instrument depending on the application, or manually by the instrument operator to optimize the image quality in a given measurement situation.

Yet another example application of a dual or multiple frequency band array according to the invention, is to use the different frequency bands to image at different depth ranges with the same probe, for optimized selection of frequency for different image depths. One would then use the HF band to image at lower depths for improved resolution with focus in these depths, for example as a switched linear array operating at 10 MHz, and the LFl band to image at deeper depths with correspondingly deeper focus for improved penetration, for example as a linear phased array operating at 2.5 MHz. Such a probe is for example desirable with portable scanners, especially for emergency use, as one reduces the amount of probes to be carried around. By dividing the apertures into array elements, one can electronically steer the focal depths of both the LFl and HF apertures, and also the beam directions, according to known methods. Due to the larger wavelength of the LFl band, the array elements for the LFl band can have larger radiation surfaces with larger distance between neighboring element centers, than do the HF array elements within the common radiation surface, as for example discussed in relation to FIG. 5 below. In FIG. 2b we even use a LFl array composed of a single element, whereas the HF array has a large number of elements . In FIG. 2c we also see that the LFl and HF elements have different shapes. The invention hence presents a general solution for a combined LFl and HF array with a common radiation surface, also allowing the apertures, frequencies and foci to be electronically selectable for optimal measurements in different situations, either automatically by the instrument depending on the application, or manually by the instrument operator to optimize image quality.

The common radiation surfaces provide challenges in the structural design of electro/acoustic transduction due to the wide separation between the LFl and HF frequency bands, where the current invention provides several solutions to this problem. A first example of a stack of piezoelectric and acoustic layers that allows operations of a LFl and a HF pulse with widely separated frequencies from the same radiation surface, is shown in FIG. 3a. The Figure shows a cross section through a layered structure that radiates and receives both frequency bands through radiation surfaces that at least have a common region 302 in acoustic contact with the load material 301. For typical applications, both the LFl and the HF components might in addition be transmitted or received across separate surfaces outside the common surface. However, for equal LFl and HF transfer functions across the whole aperture, it is advantageous to use the same thickness stack across the whole aperture, and define the LFl and HF apertures by the areas of the active element electrodes as discussed below. The HF pulse is received and/or generated by the transducer array assembly 303 which in this example is composed of a piezoelectric layer 304 that is resonant in the HF band, with two acoustic matching layers 305 and 306 in front that acoustically connect to the load material 301. The acoustic contact can either be direct or through a fluid and a dome, all according to known methods. The piezoelectric layer 304 has a set of electrodes on the front and back faces that electrically define the array elements, where by example FIG. 3a shows the cross section of the electrodes 307 and 308 for one array element that generates the electric port 309 for that element. Driving the electric port 309 with a voltage signal V₀ in the HF band, will generate vibrations on the radiating surface 302 that generate a wave 310 propagating into the load material with frequencies in the high band. Similarly, an incoming wave 311 with frequencies in the high band will produce electrical voltage oscillations across the HF port 309.

The LFl pulse is in this example embodiment generated by the transducer array assembly 312, which is composed of a piezoelectric layer 313 that is resonant in the LFl band, covered on the front with a layered section 317 for acoustic isolation of HF vibrations in the HF structure from the LFl structure. The isolation section is designed so that the reflection coefficient between the HF assembly 303 towards the isolation section is close to unity in the HF band to avoid interference from the LFl structure on vibrations of the HF structure in the HF band. The isolation section is also designed so that in the LFl band it cooperates with the probe layers in front of the isolation section to provide acoustic matching of the LFl piezo-layer 313 to the load material. When the LFl piezo-layer is made as a ceramic/polymer composite it is advantageous that the isolation section 317 is made of at least two layers, where the back layer, or group of layers, 318 of this section preferably is a heavy, impedance regularizing structure for the reasons described below. The whole transducer assembly is mounted on a backing material 320 with so high absorption that reflected waves in the backing material can be neglected. In some embodiments one can have impedance matching layers between the LFl layer 313 and the backing 320 to increase the acoustic coupling, according to known methods. The Figure also shows a cross section of the electrodes 314 and 315 for a particular LFl array element, or parts of the LFl array element as the LFl array element often is wider than the HF array element. The electrodes constitute a LFl electric port 316, where driving this port with an electric voltage signal Vi in the LFl band produces LFl vibrations on the array front face 302 that radiates a wave 310 into the load material 301.

Close to unit reflection coefficient between the HF piezo- layer and the isolation section is obtained when the impedance seen into the isolation section from the front is low or high compared to the characteristic impedance of the HF piezo-layer. When the impedance into the isolation section from the front is low in the HF band, the HF piezo-layer will have a thickness resonance when it is half a wavelength (or whole number of half wavelengths, where the half wave length is the most efficient) thick around the center of the HF band. When the impedance into the isolation section from the front is high in the HF band, the HF piezo-layer will have a thickness resonance when it is a quarter of a wavelength thick (or an odd number of quarter wavelengths) around the center of the HF band. The quarter wave resonance generally allows wider bandwidth of the HF layer resonance with poorer phase angle of the electrical impedance compared to for half wavelength resonance of the HF piezo-layer.

The thickness of the HF piezo-layer 304 is lower than the thickness of the LFl piezo-layer 313 due to the separation of the HF and the LFl frequencies. For this reason the cuts between elements or in the composite of the LFl layer require a thicker saw blade than for the cuts in the HF layer. It can hence in the practical manufacturing situation be difficult to control whether the ceramic posts of the HF layer connect to ceramics or polymer fill in the LFl piezo-layer. To make the HF isolation properties of the matching section 317 have enough low sensitivity to a connection into LFl ceramic or polymer fill, the invention devices that the back layer or group of layers 318 of the section 317 close to the LFl piezo- layer 313 to be made of heavy materials with high acoustic impedance, for example metals like Ag, Cu, Au, Pd, Pt, and W, or even a ceramic material or integrated electronic substrates as discussed below. Large shear stiffness of the layer (s) 318 will also help in reducing the sensitivity to connection of 317 into ceramic or polymer fill, but large shear stiffness of 318 would also introduce lateral vibration coupling between LFl array elements, and hence the thickness of this layer should be limited, while still making the impedance seen from the front into the section 317 adequately insensitive to connection into ceramic or polymer fill on the back side. Thicknesses of layer (s) 318 less than λ_HF/20 are found useful, as discussed below. Of the listed metals, Ag, Au, Pd, and Pt have the lowest shear stiffness and still a high mass density which makes the materials most efficient for reducing the sensitivity to connection into ceramic or polymer fill with lowest lateral coupling between LFl array elements .

The other layers of the isolation section 317 are typically chosen with λ_HF/4 thickness at the high frequency. A low impedance into the isolation section 317 can for example be obtained with a matching layer in front of said impedance regularizing layer 318, where said matching layer has low characteristic impedance and is quarter wavelength (λ_HF/4) thick at the center of the HF band. Said matching layer can preferably be made of polymer or similar material. A high impedance into the isolation section can for example be obtained with a 1^st

X_HF/4 matching layer with low characteristic impedance to the front of said impedance regularizing layer 318 of large mass. This 1^st matching layer connects into a 2^nd λ_HF/4 matching layer with high characteristic impedance that connects to the HF piezo-layer. The selection of characteristic impedances of the 1^st and 2^nd matching layers can be done through standard considerations of impedance matching known to anyone skilled in the art. When the characteristic impedance of the 2^nd λ_HF/4 matching layer is adequately high, it is also possible to omit the impedance regularizing structure 318 without large modification of the HF electro-acoustic transfer function.

An example of the effect of layers (s) 318 on the impedance seen into the section 317 from the front, is shown in FIG. 3b-d. In FIG 3b the isolation section 317 is composed of a single polymer layer that is λ/4 thick at 10 MHz. The curve 321 shows the acoustic impedance from the front into 317 as a function of frequency when the layer connects to the ceramic on the back. The impedance into the ceramics of layer 313 oscillates between a low value of the backing impedance Z_B when the LFl ceramic is a whole number of λ/2 thick and a high value (Z_cer)²/Z_B > Z_B when the LFl ceramic is an odd number of λ/4 thick. Z_cer is the characteristic impedance of the ceramic. The λ_HF/4 polymer layer 317 then transforms this impedance into the curve 321 that oscillates with the frequency where close to 10 MHz we get a minimum value close to (Z_Poi/Z_cer) ²*Z_B and peak values close to Z_Poi²/Z_B, where Z_poi is the characteristic impedance of the X_HF/4 polymer layer. The curve 322 shows the impedance from the front into 317 as a function of frequency when the section connects to the polymer fill between the LFl ceramic posts. The impedance into the polymer fill in layer 313 oscillates between a high value of the backing impedance Z_B when the fill is a whole number of λ/2 thick, and a low value (Z_fm) ²/Z_B < Z_B when the fill is an odd number of λ/4 thick. Z_fill is the characteristic impedance of the polymer fill between the ceramic posts in the ceramic/polymer composite of layer 313. The λ/4 polymer layer 317 then transforms this impedance into an oscillating variation 322 where close to 10 MHz the peak values are close to (Z_poi/Z_fin) ²*Z_B and minimum values are close to Z_Poi²/Z_B.

FIG. 3c shows the impedance seen from the front into section 317 when a Cu layer 318 of 20 μm thickness (about λ/25 of Cu at 10 MHz) is introduced on the backside of the λκF/4 polymer layer described in FIG. 3b. The curve 323 shows the impedance seen from the front into the section 317 when the Cu layer is connected to the polymer fill between the LFl ceramic posts. The Cu layer of this thickness gives an added inductive impedance of the mass load of the Cu seen into the fill, which increases the impedance seen from the λκF/4 layer towards the back, and the λ_HF/4 layer inverts this impedance into an impedance < 2 MRayl in the band 7 - 13 MHz which gives a very good isolation from the HF to the LFl section in this band. The curve 324 shows the impedance seen into section 317 when the section is connected to the LFl ceramic posts. We note that the effect of the Cu layer makes less modification from the curve 324 from 321 than of the curve 323 from 322 when connecting to the polymer fill . The reason is that because the ceramic has a high characteristic impedance, the Cu layer mainly changes the frequencies of the low and the high impedance seen from the back of the λ_HF/4 layer, and not so much the value of the low and the high impedance. However, by using a sufficiently high backing impedance, for example Z_B = 5 MRayl in this example, the maximal impedances seen into the isolation section 317 when connected to ceramic is still below 2 MRayl in the 7 - 13 MHz band, which gives a high isolation seen from the HF section in this band.

The effect of the Cu layer on the HF electro-acoustic transfer function is shown in FIG. 3d. The curve 325 shows the HF transfer function when isolation section 317 is composed of a single λ_HF/4 polymer layer as in FIG. 3b and connected to the polymer fill on the back. We note that this curve shows resonances due to internal HF reflections in the LFl section 312 because the impedance curve 322 do not provide adequate reflection at the back of the HF piezo-layer 304. Introducing a layer 318 of 20 μm Cu changes this transfer function to curve 326 where the resonances due to reflections in the LFl section have disappeared. The curve 328 shows the transfer function without the layer (s) 318 and when the section 317 is directly connected to ceramics, where this curve moves to 327 when the Cu layer is introduced. We note that the Cu layer removes the resonances in curve 325 and makes the transfer function 326 for connection into polymer fill and 328 for connection into ceramic of the LFl section close to equal . This Figure hence demonstrates that introducing the Cu layer makes the HF electro-acoustic transfer function insensitive to whether the isolation section connects to polymer fill or ceramics in the layer 313. The dual band electro-acoustic transfer function can then typically take the form as in FIG. 3d where 331 shows the transfer function for the LFl port and 332 shows the transfer function for the HF port. We should note that the important effect of this thin Cu layer is its mass, i.e. pL where p is the layer mass density and L is the layer thickness, that introduces an inductive impedance. The layer is therefore conveniently made of any heavy material, such as Cu, Ag, Au, Pd, Pt, W, and ceramics, or alloys of these materials powders of these materials or alloys cintered together or glued in a solvent. The heaviest materials allows the thinnest layers, and as stated above the materials Ag, Au, Pd, and Pt have the lowest shear stiffness for their mass density and therefore produces the least lateral coupling between the LFl elements . The wave propagation velocity for Si is 8.44 mm/μsec and for Al it is 6.4 mm/μsec. This allows quite thick (L) layers while still L « λ_HF so that the layer has the effect of a mass load. One hence can get adequate masses pL for both Si and Al layers also, as described below.

The layer (s) 318 can also include part of the ceramics in layer 313 as illustrated in FIG. 4a where the labeling for the same layers follows that in FIG 3a. The polymer filled cuts 401 in the LFl piezo-layer 313 are diced from the back of the layer but not diced completely through the LFl ceramic layer 313 so that a complete ceramic layer 402 is left and included in the layer (s) 318 of the HF isolation section 317. The LFl front electrode 315 can also be made so thick that it has an acoustic effect in the HF band and also can be included as part of the layers 318.

An approximate analysis on how to match the LFl piezo- layer to the load in the LFl band can be done by realizing that both said isolation section matching layers and the HF piezo-layer with load matching layers will be thin compared to the wavelength in the LFl band. A thin low impedance layer between high impedance layers will then approximately behave as an elastic spring in series with the rest of the structure, while the thin high impedance layers will behave as a series mass . When the isolation section 317 is composed of a single X_HF/4 low impedance matching layer in front of the impedance regularizing layer 318, for low impedance into the isolation section in the HF band, the LFl piezo-layer 313 will to the front observe the elastic spring of the low impedance X_HF/4 layer in series with the mass of the HF section 303 that is dominated by the mass of the HF piezo-layer 304. When the isolation section has a 2^nd λ_HF/4 high impedance matching layer to obtain a high impedance into the isolation section as described above, this high impedance X^/A. matching layer will give an added mass in series with the spring of the low impedance X_HF/4 matching layer. The center frequency of the LFl band may preferably then be selected around the resonance between this spring and mass system where the phase of the impedance into said isolation section matching layer seen from the back is zero. This resonance frequency can be tuned by varying the stiffness of said low impedance λ_HF/4 matching layer and the mass density of the HF piezo and load matching layers (and high impedance X_HF/4 matching layer of 317) . This mass density can for example be tuned by varying the ceramic volume fill in the HF piezo-composite. In a less efficient design to provide high impedance into the isolation section in the HF band, one can use a single X_HF/4 matching layer with high characteristic impedance to the front of said impedance regularizing layer of large mass . In the LFl band this single matching layer will approximately behave as a mass in series with the mass of HF piezo- and load matching layers and provide a load impedance seen from the LFl piezo-layer that has an inductive phase. This matching system does not provide optimal LFl impedance resonant matching, but a useful form of the LFl electro-acoustic transfer function is obtained.

Using the method of an isolation section between piezoelectric layers, one can add piezoelectric layers at lower resonance frequencies backwards, in principle ad infinitum, for most applications with one or two layers, where FIG. 3e illustrates the general principle by adding one more lower frequency layer to the structure in FIG. 3a. In FIG. 3e a 2^nd lower frequency section 340, referred to as LF2 is added to the back of the 1^st lower frequency section 312, referred to as LFl. The layers of the LFl and the HF section 303 are given the same labeling as in FIG. 3a. The LF2 section is composed of a piezo-layer 341 with an isolation section 342 to the front. The purpose of the isolation section is to isolate vibrations in the LFl band in the section 312 in front to propagate backwards into the LF2 section 340, to suppress the interference of section 340 with vibrations in the LFl band in section 312, in the same manner as discussed for the HF isolation section 317 above. The front and back of the piezo-layer

341 are then covered with electrodes 344 and 345 to form the electric port 346 of an element of the LFl array, where the Figure illustrates single array elements or parts of a LF2 and a LFl array elements when these are wider than the HF array elements.

The isolation is obtained when the impedance into the isolation section from the front is either much higher than or much lower than the characteristic impedance of the neighboring piezo-layer 313 in front as discussed for the HF isolation section 317. For a high impedance into

342 from the front, the piezo-layer 313 would operate at λ_LF/4 resonance, while with a low impedance into 342 from the front, the piezo-layer 313 would operate at λ_LF/2 resonance. The λ_LF/2 can be preferred at high medical ultrasound frequencies ( ~ 10 MHz and upwards) as this gives thicker piezo-layers that simplifies machining, while for lower medical and SONAR frequencies the λ_LF/4 resonance can be preferred as this gives wider bandwidth and requires less piezoceramic material that is expensive. If the piezo-layer 341 is made as a composite, it can be advantageous that the isolation section 342 is composed of at least two layers, where the back layer 343 is a heavy, impedance regularizing layer thinner than the LFl wave length, similar to 318, to reduce the difference in impedance when the ceramic posts of the LFl piezo-layer 313 connects to ceramic posts or polymer fill in the layer 341. In the LF2 band the layers in front of the LF2 section are so thin that they function approximately as a spring or mass in series. The low impedance layers of the isolation section 342 then generally functions as the spring in series with the mass of the layers in front, and the center of the LF2 band is selected at the resonance of this system as discussed for the LFl band above. The backing material can be used as acoustic power absorbant to reduce peaking resonances in the electro-acoustic transfer functions. For improved acoustic coupling to the backing, one can also introduce acoustic matching layers between 340 and the backing 320 according to known methods . Resonances in any of the frequency bands can also be dampened with matching layers of absorbing materials, for example viscous polymer materials, and even adding particles to the polymer materials to increase absorption. Viscous polymer materials and particle filled polymer materials can also be used in the polymer fills of the ceramic/polymer composites of the piezoelectric layers. Solid/polymer composites can also be used for matching layers to tune the characteristic impedance, where viscous and/or particle filled polymers can be used for increased absorption in the matching layers .

It should now be clear that the procedure could be repeated by adding further lower frequency sections to the back, each section includes a piezo-layer for acousto- electric coupling and an isolation section for vibrations in the band of the neighbor section to the front. The procedure can hence be repeated in principle ad infinitum, where most applications would require only a single or a dual lower frequency band.

FIG. 3a, e, and FIG. 4a show thickness structures for example elements or parts of elements of the arrays according to the invention, where it is clear to anyone skilled in the art that the invention can be used to build acoustic arrays of any organization, for example annular arrays, linear phased, linear switched arrays, or linear arrays with divisions in the elevation direction of many scales from 1.5D via 1.75D up to 2D arrays for full 3D steering of the beams. The lateral width (radiation surface) of an array element is typically limited by a ratio to the wavelength in the object. As the LFl wavelength is larger than the HF wavelength, one would often use wider LFl array elements (larger element radiation surface) than HF array elements. The isolation section in FIG. 3a and FIG. 4a then allows independent selection of the LFl and HF array elements, as the HF isolation is practically independent of whether the isolation section terminates into ceramics or polymer. This for example also allows that the arrays for the different bands are of different nature, for example, but not limited to, a 1.5D linear switched array for the HF band and a linear phased array for the LFl band. When the LF2 array is used for therapy, one do not in some applications have to steer the beam direction, and the LF2 array can be made as a single element with fixed focus, or annular elements to steer the depth of the focus. With a single LF2 element composed of whole ceramic, the heavy back layer of the isolation section in front of the LF2 layer can then be left out, as the ceramic posts to the front would end in ceramics regardless of their lateral position.

When a multiple frequency probe according to the invention is used for imaging at multiple depth ranges at different frequencies, the front HF array can often be used as a switched linear (or curvilinear) array, while the LFl array is used for phased array imaging. The required element pitch of the HF and the LFl arrays can then be the same, for example 0.3 mm for a 7 MHz switched HF array, where the same pitch is Xι^/2 for a phased array at 2.5

MHz. The structures of the isolation section given above is however still useful as one would like to have more dense cuts in the HF ceramic/polymer composite than the LFl composite, and the isolation section 317 as described above also allows less accurate lateral positioning between the HF and lower frequency arrays . The larger LFl wavelength also favors the use of larger LFl than HF transmit apertures as discussed above. For large depths, the HF receive aperture can however be larger than the LFl transmit aperture, where in general one would favor a design with the same thickness structure throughout the whole array, and the size of the transmit and receive apertures can be varied by electrically selecting the elements that participate in the apertures (radiation surfaces) . Which of the arrays (HF, LFl, LF2 , ....) that is connected to the instrument beam former can be selected through electronic switches, but also through electric filters that would guide the different frequency transmit pulses to the array for the frequency, and similarly to guide the receive signals from the actual frequency band array to the beam former, all according to known methods. The arrays of any frequency band would show some sensitivity in the lower frequency bands, which can be suppressed by electrical filtering at the electric port. Sensitivity to the higher frequency bands is suppressed by the acoustic isolation sections, so that one can omit the filter to the lowest frequency band.

For the large number of elements that are found with some linear arrays but specially with 1.5D, 1.75D and full 2D arrays, one can reduce the number of wires that connect the probe to the instrument by including in the probe electronic circuits, for example electronic switches that electronically selects and connects sub-groups of elements to the instrument beam former, or sub-aperture electronics that delay and combine the signal from several array elements into a single sub-aperture signal that connects to single channels in the instrument beam former, etc., all according to known principles. The signals from groups of elements or groups of sub-apertures of elements may also be transmitted on a single cable by time-multiplexing samples of the signals from such groups, where the time- multiplexing circuits are integrated into said electronic substrate layers, to reduce the cable connections to the arrays. The HF elements are generally more numerous than the LFl elements, and more difficult to connect to electrically in the structure of FIG. 3a and e. Electrical connection with electronic switches and/or sub-aperture electronics and/or time multiplexing for large element number HF arrays can conveniently be done with electronics on substrate layers as shown in FIG. 4b - d below. For special high frequencies with less number of HF elements the structures are also useful for amplifiers only, preferably receive amplifiers but in special situations also transmit amplifiers.

To further illustrate this situation by way of example we analyze a 2D array concept probe according to the invention, illustrated in FIG. 4c, operating at a HF frequency of 3.5MHz and a LFl frequency of 0.5MHz. With a λ_HF/2 pitch of 0.22mm, one gets a 20mm HF aperture with 90 HF elements in a diameter. With the hexagonal form of the aperture, this gives a total number of HF elements in the 2D array of approximately 90²*3root (3) /8 = 5,261 elements. Using sub-apertures of 5*5 = 25 elements, the total HF aperture is supported with 210 sub-apertures, which is a convenient number for cable connection to an instrument for final beam forming. The λ_LF/2 pitch for a LFl frequency of 0.5MHz is 1.54 mm, and one fills a 20mm LFl aperture diameter with 13 elements . For the hexagonal aperture the total number of LFl elements in the 2D array is then approximately 13²*3root (3) /8 = 110 elements which is conveniently operated via a cable from an instrument with a LFl transmit beam-former in the instrument. For abdominal applications one could increase the diameter to 40 mm and the frequency to 5 MHz with \_w/2 pitch of

0.154mm with a diameter of 256 elements and a total of 256²*3root (3) /8 = 42,566 elements. With 7*7 = 49 elements per sub-aperture we get in total 868 sub-apertures, and using a time multiplex factor of 7 per electric cable, we can connect to the HF array with 128 coax cables with 7x multiplex per cable. The LFl array will then get a similar increase in number of elements .

In FIG. 4b substrate layers with integrated electronics are included in the HF isolation section 317, Si substrate layers are commonly used for integrated electronics and have a convenient characteristic impedance of 19.7 MRayl, which is a convenient value for a high impedance λ_HF/4 matching layer. Other substrate materials with high characteristic impedance, like GaAs, can also be used. In more detail, FIG. 4b shows the HF isolation section 317 composed of Si-substrate layers 405 and 406 included in the impedance regularizing section 318, a 1^st low impedance X_HF/4 layer 407, typically made of polymer, a 2^nd high impedance λ_Hp/4 layer 408 composed of two Si-substrate layers 409 and 410. The LFl front electrode 315 can also be made so thick that it gives acoustic contribution to the function of the layers 318. The isolation function of the section 317 with this structure is described above. Taking the example of the 2D 3.5/0.5 MHz array above, we note that λ_Si/4 at 3.5 MHz is 0.6mm, which gives 0.3 mm thickness of the two Si-substrates 409 and 410, which is a convenient thickness for integrated circuit electronics. One could even use lower thicknesses for more Si-layers or for higher frequencies. The thickness of each substrate could for example be reduced to 0.2 mm which would allow for 3 Si-substrate layers within the X_HF/4 high impedance layer 408. At 10 MHz the λ_HF/4 length in Si is 0.211mm that allows for a single Si substrate layer of this thickness in 408 at 10 MHz.

With 0.2 mm thickness of the Si-substrates 405 and 406 the layers 318 will approach λ_HF/4 in thickness, which reduces the impedance regularizing effect of 318, but with the structure of a 1^st low impedance λ_HF/& layer 407 and a 2^nd high impedance λ_HF/4 layer 408 one would still have a high impedance into the section 317 from the front. The thickness of the section 318 could be reduced by using only one or even zero Si-substrate layers, depending on how much processing electronics one want to put into the probe. The section 318 could even be made thicker with more Si-substrate layers to allow for more processing electronics in the probe, where one conveniently would match the characteristic impedance of the LFl ceramic/polymer layer 313 to the impedance of the layer

318 so that they together define the resonance of the LFl layer .

The front substrate layer 410 can by example contain receiver preamplifiers for the HF elements. The outputs of said pre-amplifiers can by example connect to the electronics in the 2^nd substrate layer 409 that can for example contain sub-aperture beam forming electronics that delays and combines the signals from several HF elements into a single sub-aperture channel that considerably reduces the number of connections required to the instrument or further substrate layers of electronics. A reduced number of sub-aperture channels could then be transported to the instrument for final beam forming according to known methods . The final beam forming can typically also include corrections for wave front aberrations due to spatial variations in propagation velocity, according to known methods. The sub-aperture dimensions are then limited by the correlation length of the aberrations along the array surface. The electronics in layers 410 or 409 could also contain switches that select subgroups of HF elements to the instrument beam former, for example as a switched array, or combining selected groups of 2D elements into linear elements of selectable direction as described in FIG. 4d. With the structure in PIG. 4b one obtains direct electrical connection between the HF array element electrodes and the front layer 410, where element electrodes could connect to metal pads on the layer 410. Micro-soldering, ultrasonic bonding, anisotropic conducting polymer glue with conducting particles are all known and useful methods for the connection. With conducting polymer glue the max thickness of the glue must be limited to minimize wave reflection between the substrate layers. A polymer glue between the Si layers can also be used to reduce the composite acoustic impedance of the substrate layers with glue. Electric connections between the stacked substrate layers can be obtained through via-holes in the substrates or with bonding at the edges of the layers, all according to known methods.

Electrical connections through an isolating layer, like the low impedance λ_EF/4 layer 407 can be obtained via metallic connectors 411 through the layer, where said metallic connectors are so thin that they have minor effect on the characteristic acoustic impedance of said low impedance layer 407. In the example embodiment, the number of required connections through the isolating layer 407 can be greatly reduced by the circuits in layers 409 and 410, which by the example array above is a reduction from 5261 to 210 connections through sub-aperture circuits. This shows the great advantage of channel reducing electronics in the high impedance section 408.

The HF acousto-electric transfer function is shown as 412 in FIG. 4b. The relative - 3dB bandwidth is ~ 70 %, a high value that is partly achieved by the λ_HF/4 resonance of the HF piezo-layer 304 that is produced by the high impedance into the isolation section 317 in the HF band, produced by the high impedance λ_HF/4 matching layer 408. At higher frequencies one might want a λ_KF/2 resonance of the HF piezo-layer to obtain a thicker layer that is easier to manufacture and handle. The resonance of the HF piezolayer can be considered to be a λ_HF/2 resonance of the composite HF piezo-layer 304 and the matching layer 408. One can hence increase the thickness of HF piezo-layer 304 at the expense of reducing the thickness of layer 408, for , example by reducing the number of generating substrate layers, while maintaining the same center frequency of the HF band. With higher frequencies one might not be using a full 2D HF array, rather a ID, 1.25D, 1.5D or 1.75D switched array, which all have less total number of elements. One could then even find it practical to connect the HF elements directly through the isolating layer 407 to electronic layers in the structure 318 via thin connectors as 411. This allows us to remove the high impedance layer 408 which produces full λ_HF/2 resonance of the HF piezo-layer.

A schematic 3D rendering of such a probe with a 2D array according to the invention, is illustrated in FIG. 4c, where the HF 2D elements are indicated as 415 on the front faces shining through the HF acoustic matching layers 305 and 306. The layers are given the same labeling as in FIG. 4b. Connection between the electronic substrate layers (405, 406, 409, 410) and the instrument can for example be obtained through connecting pads 416 at the edges of one or more of the substrate layers. Flex print circuits 413 are then conveniently connected to these pads and brought along the side faces of the array structure and behind the backing where it can be connected to a flexible cable that connects to the instrument according to known methods. The connections can for example be obtained through micro- soldering, ultrasonic bonding, anisotropic conducting glue with conducting particles , etc . , according to known methods. The flex prints conveniently follow the flat side surfaces of the probe that gives a minimal added thickness to the probe.

Amplifiers, both transmit and receive, and sub-aperture circuits for the LFl array can be placed in the substrate layers in front of the LFl piezo-layer, typically behind potential substrate layers with electronics for the HF array (e.g. layers 405, 406), and as part of the isolation group of layers 318. Depending on the space available, electronics for the LFl and HF arrays could be placed on same substrate layers . One would then typically locate the substrate with LFl electronics closest to the LFl array. Substrate layers with electronics for the LFl array can also be placed at the back-side of the LFl piezo-layer in front of the backing material 320. With this last placements of the electronics, the connection to the cable can be done with wires through the backing material, where said wires are so thin that they do not propagate acoustic waves through the backing. The connection from the circuits to the cable is also conveniently done with for example flex print circuits on the side of the structure, as for the HF electronics described above. When placing the electronic substrates to the back of the array, the characteristic impedance of the piezo-layer is preferably close to that of the substrate to minimize reflections between the substrate and the piezo-layer so that the substrate layers participate in the definition of the LFl resonance together with the LFl piezo-layer, as discussed for the front placement above. The net acoustic impedance of the substrate layers can also be reduced by thin intermediate layers of lower characteristic impedance, for example an anisotropic polymer glue as described above. By connecting to the lowest frequency elements with wires through the backing 320, one can place potential amplifiers, switching circuits, and sub-aperture circuits for the lowest frequency elements behind of the backing, potentially located in stacked substrate layers with electronics, or with other arrangements according to known methods. One do generally have adequate space available in the probe handle, so that this solution can be simpler than circuit layers stacked together with the LFl array. The LFl elements are however larger and fewer with lower frequency than the HF array, and the pay-off is therefore less for using amplifiers and sub-aperture electronics in the probe itself, where one for many embodiments according to the invention will not use such circuits for the LFl array in the probe.

Instead of using the sub-aperture method to form beams within a volume sector in front of the probe, it is also interesting with a 2D array structure to use electronic switches in the electronic layers (409, 410) that connect groups of HF 2D elements into linear elements. An example embodiment is shown in FIG. 4d, where the HF 2D array 420 is composed of triangular elements 421 that can be connected to sets of linear elements 422, 423, 424 that with phased array steering can be used to produce 2D scan planes in different directions illustrated as 425, 426, 427. With the 3.5/0.5 MHz example array above, it would then be sufficient with 96 - 128 channels in the HF phased array beam former. Selectable linear arrays with different directions of the linear elements could also be implemented with a dual piezoelectric layer structure as described in US Pat appl 10/387,775.

Combination of 2D LFl elements into linear LFl elements could also for example be done in the electronic layers 405 and 406, or in electronics of other arrangements, or via a dual layer structure as described in US Pat appl 10/387,775. For the LFl linear array beam former in the example array above it would then be sufficient with 13 channels. However, the number of total LFl array elements in the example array described above is only 110, so that one could also connect all LFl elements to the instrument and do the LFl element combination in the instrument. This would provide full flexibility in the use of the LFl array as a 2D volume scanning array or as a linear array with selectable 2D scan directions . It would then be convenient that the electronics in layers 405 - 410 would also include both sub-aperture connection to the 2D HF array for full volume sector scanning of the HF beam together with such scanning of the LFl beam, and connection of 2D elements into linear elements for 2D sector scanning of the HF beam together with the LFl beam.

LFl transmit beam former electronics in the probe is especially interesting when the LFl array is used for transmit only, as described with the methods in US Pat applications 10/189,350 and 10/204,350, where only a transmit sub-aperture beam former is needed. For transmit amplifiers that switches the element signals to positive and negative power voltages, the power losses can be made so low that the whole transmit beam former with amplifiers can be integrated into the probe. Such a probe would have a simplified connection to existing scanners, for direct field upgrade to existing scanners with the methods described in the cited US Pat applications. Making the piezo-composite 304 with close to the same characteristic impedance as the Si-substrates, one could also place Si-substrate layers to the front of the HF piezo-layer 304 as illustrated in FIG. 4e where the Si- substrate layers are labeled as 430, 431, 432. The HF resonance is then defined by the combined thickness of the piezo-layer 304 and the Si-substrate layers, i.e. the structure 433. A simulated electro-acoustic HF transfer function for this structure is shown as 434. The example placements of electronic substrates within the acoustic structure in FIG. 4b-e can also be combined and modified in various forms for simplified connection between the HF and LFl array elements, the substrate electronics, and the instrument beam former. The probe could typically also contain electronic circuits in the handle, behind the backing material.

By example we show in FIG. 5 another linear phased array according to the invention, seen from the front face, where 501 indicates the elements of the phased array HF aperture, where λ_x is the HF wave length with a pitch of the HF elements of λχ/2. With proper steering of the signal on each element according to known methods, such an array allows steering of the beam direction within a sector in the azimuth direction. Steering in the elevation direction requires division of the elements also in the elevation direction into a two-dimensional (2D) array, and we will at this point emphasize that the basic methods of the invention is also applicable to 2D arrays.

The center of the HF band of this linear array is by example fi = 3.5 MHz which suggests a high frequency element pitch of λχ/2 ~ 0.22 mm. 84 high frequency elements then produces a total aperture of 18.48mm. With a center of the low frequency band of fo = 0.5 MHz we get λo/2 ~ 1.54 mm, which suggests 12 of the low frequency elements 502 that also produces a total aperture of 18.48mm. For better collimation of the LFl beam one could add extra LFl elements to each side of the HF elements, where the Figure shows by way of example two elements 503 increasing the LFl azimuth aperture to 14 elements ~ 21.56 mm. To increase the LFl elevation aperture one could similarly expand the LFl aperture by the elements in the elevation direction, where the Figure shows by way of example the elements 504. As follows from the analysis in relation to FIG. 2a-d, one would in some situations like to use the same transmit aperture of the LFl and HF radiation surfaces when it is critical that the phase between the HF and LFl pulses has minimal sliding with depth, while for higher LFl amplitude at large depths it can be desirable that the LFl transmit aperture is larger than the HF transmit aperture to reduce diffraction broadening of the LFl beam with depth. To reduce the nonlinear manipulation by the LFl pulse in the propagation and scattering of the HF pulse close to the array, one could like to remove the central radiation surface of the array. This can be obtained by further dividing the LFl elements into the sub elements 505. The LFl array in FIG. 5 then allows selection of the size of the LFl aperture, for example as one. of 1) to be equal to the HF aperture, 2) to be larger than the HF aperture either in the azimuth and elevation directions separately or in both the azimuth and elevation direction, and 3) a LFl aperture with an inactive area in the center of the HF aperture. We also point out that such variation of the LFl aperture relative to the HF aperture is obtained with other array configurations, for example 2D arrays, annular arrays, etc. where anyone skilled in the art can apply the essentials of this invention to all array configurations . For many applications one would like to use 2) only, which is achieved by combining the elements 502/504/505 into a single LFl element with elevation dimension equal to or larger than the HF aperture, and add extra LFl elements in the azimuth direction (503/504) to obtain a LFl aperture that is larger than the HF aperture in the azimuth direction. To get the same vibration conditions for the LFl elements over their whole area, one could typically use a stack like in FIG. 3 and 4 for the whole array area, and define the LFl and HF elements by the element electrodes and cuts in the piezoelectric layers as described above. It would also be advantageous to use ceramic/polymer composites for both HF and LFl piezoelectric layers, where the element dimensions could be defined by the electrodes. The HF radiation area could then for example be defined by a common ground electrode on the front side which would define the elevation width of the elements both through electrical coupling but also by defining the areas of the ferroelectric ceramic that is polarized to show piezoelectric properties. The azimuth width of the HF elements are then defined by the back side hot electrodes which can conveniently be extended to the edge of the assembly for electrical connection to the cable as the electro-acoustic coupling outside the ground electrode is low, both due to reduced electric field and reduced electric polarization of the ferroelectric ceramic material .

The isolation section 317 in FIG. 3 and 4 then makes accurate position matching between cuts in the HF and LFl piezo-layers less critical, as the impedance seen into the section 317 from the front has little variation with termination into polymer or ceramic, as for example discussed in relation to FIG.3b-d. This reduced sensitivity allows dicing of the LFl layer with thicker saw than the HF layer, and also reduces requirements for accurate lateral positioning between the HF and LFl layers.

The HF array in FIG. 5 could also be used as a switched linear array where the HF beams would be normal to the HF aperture. It is then in some applications useful to make the LFl array as a single element, that provides an unfocused LFl aperture. The LFl aperture would then be chosen so large, that the whole HF imaging depth would be within the near field of the LFl aperture, as described above. For suppression of multiple scattering noise, for example as described in US Pat Appl 11/189,350, one could then also remove from the LFl transmit aperture composed of the elements 502 and 505. For flexibility, the LFl aperture could then be composed of two elements: i) A central element composed of the elements 502 and 505 in parallel, and ii) an outer element composed of the elements 504 and 503 in parallel . For nonlinear imaging one would use both the central and outer LFl elements in parallel for the LFl transmit aperture, while for suppression of multiple scattering noise one could take the central element from the LFl transmit aperture. It is also known that the piezo-layers 304, 313 and 341 can be made of multiple layers, both piezoelectric and non-piezoelectric to alter and increase the bandwidth of the electro/acoustic transfer functions and reduce the electric impedance of the electric ports . Adding the stacks of substrate layer exemplified in FIG. 4b and 4e can be viewed as a way to add a non-piezoelectric layer that interferes with the resonance definition, for example as described in US Pat 6,645,150. To obtain lower electric impedance of the array elements for example, especially the lower frequency elements to be able to transmit high pressures with manageable drive voltage amplitudes, one could conveniently make one or more of the piezo-sections 304, 313 and 341 as stacked piezoelectric layers covered with electrodes.

FIG. 6 shows an example embodiment of two layers 601 and 602. The layers are covered with the electrodes 603, 604, and 605, where typically one would galvanically connect electrodes 603 and 605 to ground where the electrode 604 would be used as the hot electrode. The two piezoelectric layers would then have opposite polarization directions 606 and 607, so that the electrode coupling would provide an electrical parallel coupling of the layers 601 and 602 to provide a lower electric impedance port 608, which allows driving the low frequency array with lower voltages for the high pressures. For improved bandwidth of the layers, one can introduce a high impedance layer in front of the active piezolayers, as presented in US Pat 6,645,150. Parallel coupling of more layers can be done for even lower electric port impedance, also for the high frequency layer 304, according to known methods. In US Pat appl 10/387,775 it is also described how one with dual layers can obtain linear arrays with selectable direction of the electrodes, for electronic rotation of the 2D scan plane. This solution is interesting both for the higher and lower frequency arrays within the structure of this invention. It is also possible to base the electro-acoustic transduction on micro-machined transduction cells on the surface of a substrate, for example a Si (silicon) substrate, or other substrate of other materials such as Cu and Al. With these techniques, increased vibration of the surface is obtained by vibrating membranes on the substrate surface, with gas or vacuum behind the membrane, where the membrane connects to the acoustic load material either directly or through acoustic layers . The electromechanical coupling can either be obtained by capacitive coupling from the membrane to a reference electrode, referred to as cmuts (capacitive micro-machined ultrasound transducers) , or through piezoelectric films on the membranes, referred to as pmuts (piezolayer micromachined ultrasound transducers) . Examples of such membranes are illustrated as 701 shown from the front radiation face in FIG. 7, mounted on the front surface of the substrate 700. The dimensions and thicknesses of the membranes determine the resonant band where the transduction is most efficient, and several of the cmut/pmut cells are usually coupled together electrically to form one array element. In the current invention we are concerned with inventive implementations of the cmut/pmut techniques to transmit dual or triple band pulses from essentially the same radiating surfaces, where the Figures show inventive steps to achieve the dual or triple band function, and where details of the membranes, electrodes, and electrical connections are left out as many solutions for this are presented in the literature. We shall in the following refer to this technology as cmut/pmut transducers, cmut/pmut cells, and cmut/pmut membranes.

The characteristic impedance of Si is 19.7 MRayl and Al is 17.4 MRayl, which gives an interesting possibility of transmitting the LFl wave through a HF substrate. By example FIG. 8a shows a cross section of a structure with a cmut/pmut HF section 806 mounted in front of a LFl section made by a piezo-layer 801 with electrodes 803 and 804 generating the LFl element electric port 805. Details of the cmut/pmut drums with electrodes and electrical coupling are not shown as several examples exist in the literature per the discussion above. The total structure is in this example mounted on a backing material 802 (which can be low impedance or air) and a protection structure 808 is placed in front of the cmut/pmut drums 807. The protection structure can contain one or more layers designed for acoustic impedance matching between the load 301 and the cmut/pmut array, and absorption layers to reduce lateral coupling between array elements along the substrate, and it can also contain an acoustic lens that focuses the acoustic beams, etc..

The Figure also shows an optional absorbing layer 812 to reduce lateral acoustic coupling in the Si or Al substrate between the HF array elements and also between the substrate and the LFl section in the HF frequency band. The drums 807 reduces the effective acoustic impedance of the layer 806 below that of Si/Al, and by making the piezolayer 801 as a ceramic/polymer composite, the acoustic impedances of layer 801 and 806/808 can be matched so that the reflection coefficient between the piezoelectric layer 801 and the cmut layer 806 is low for improved bandwidth of the LFl port. The acoustic velocity of Si is 8.4 mnα/μsec and for Al it is 6.4 mm/μsec. One can therefore add further substrate layers of electronics (typically Si-substrate layers) between the cmut/pmut substrate and the piezo-layer 801 and the thickness of the total Section 806 can still be a fraction of the LiFl wavelength in the layers . This is illustrated in FIG. 8b where the Section 806 is by example composed of the cmut/pmut layer 820 mounted on 3 Si layers with integrated electronics, where by example 821 can be a receiver amplifier layer that is mounted on a sub-aperture beam forming layer 822 and a transmitter amplifier layer 823, similar to the substrate layers in FIG. 4b-e. Electrical connection between the different layers can be obtained with via-holes and connecting pads according to known methods in integrated circuit technology, or one can use known bonding techniques for example between connections at the edges of the substrates, as discussed in relation to FIG. 4b - e. With a layer thickness of 0.2 mm the total thickness of the Section 806 is 0.8mm, less than λ_si/8 for LFl frequencies less than 1.319 MHz.

The structure is for example useful for a 2D array probe similar to the one shown in FIG. 4c. One would typically use the similar apertures, number of elements, and number of sub-apertures, where one also would like to connect 2D array elements into linear array elements as described in relation to FIG. 4d. A perspective view of a 2D array concept probe with integrated electronics as part of the acoustic design, is illustrated in FIG. 8c. The cmut and integrated circuit layers are shown as 806 mounted on the LFl piezo-array 801 and the backing 802 with the front radiating surface 810. Connection between the instrument cable and the electronic circuits can for example be done with flex print circuitry from the back of the assembly to the edges of the circuit substrates as indicated in FIG. 4c.

The electronic layers closest to the LFl piezo layer, starting with 823, can be electrically connected to the LFl array elements 801, where one can integrate LFl switches, amplifier and sub-aperture electronics. The LFl array can also be connected to amplifier and sub-aperture electronics for example at the front of the LFl array, at the back of the LFl array, or inside the probe behind the backing, as discussed in relation to FIG. 4c. This can be especially interesting when the LFl array is used for transmit only, as described with the methods in US Pat applications 10/189,350 and 10/204,350, where only a transmit sub-aperture beam former is needed. For transmit amplifiers that switches the element signals to positive and negative power voltages, the power losses are so low that the whole transmit beam former with amplifiers can be integrated into the probe. For cardiac applications, the aperture dimension is limited by the distance between the ribs, but for abdominal applications one could double the HF aperture diameter, which would increase the number of HF and LFl elements by a factor of 4. This increases the number of HF sub-apertures to 840, and the LFl number of elements to 440, which also can be handled with cable connections to an instrument for beam forming. One could also use time multiplex of samples of the signals from several sub- apertures along a single wire, as discussed above, to reduce the number of wires required to connect to the instrument, where 8x time multiplex would require 105 wires to connect the HF array to the instrument.

Per the discussion above, one would often use a LFl radiation aperture that is wider than the HF radiation aperture. For parallel receive beam-forming one would typically use less than the whole HF aperture for transmit of the HF pulses to obtain a wide enough HF transmit beam. To further increase both the HF and LFl apertures without increasing the number of instrument channels too much, one could use sparse arrays as discussed in relation to FIG. 11, where not all element sites are connected electrically. This introduces grating lobes, but designing the sparse arrays so that potential grating lobes from the LFl and HF apertures do not overlap, one can with the imaging methods described in US Pat applications 10/189,350 and 10/204,350 highly suppress the effect of grating lobes in the images . Another example in FIG. 9 shows a cmut/pmut LFl section 906 in front of the HF piezolayer 901 made of ceramic/polymer composite mounted on a backing material 902. The element electrodes 903 and 904 constitute the HF element electric port 905. The LFl transduction is provided by the cmut/pmut drums 907 on the substrate layer 906. Details of the cmut/pmut drums with electrodes and electrical coupling are not shown as several examples exist in the literature per the discussion above. By making the piezolayer 901 as a ceramic/polymer composite one can match the acoustic impedance of this layer to the effective acoustic impedance of the Si layer 906 with drums 907 to define the HF acoustic resonance. In front of this structure there are placed acoustic impedance matching layers (typically one or two) 908 that connect the HF and LFl sections acoustically to the load material 301 for transmission (310) and reception (311) of dual band pulse waves in the load material. These layers can also be used to reduce lateral coupling between the LFl array elements through absorption. The acoustic matching section is together with the cmut/pmut layer 906 used to increase the bandwidth of the HF electro/acoustic transfer function, and will at the low frequency function as an acoustically thin protection cover layer for the LFl array 906, where the stiffness of the cmut/pmut membranes is tuned to the acoustic layer/load transfer. Due to the high longitudinal wave velocity of Si (8.44 mm/μsec) , the thickness of the Si substrate can be made adequately thin for acceptable effect on the HF electro/acoustic transfer function. To further limit lateral coupling inside the Si substrate one can also use an optional absorbing isolation layer 912 at the back of the substrate, the isolation being made adequately thin at the high frequencies to have limited effect on the HF transfer function.

The layered structure in PIG. 9 has interesting advantages for 2D arrays for three-dimensional (3D) beam steering and imaging, where one have electrical access to the large number (~ 3000) of HF elements from the back of the array structure for simplest connection to cable or sub-aperture beam forming electronics . The LFl 2D array will have much fewer elements (1/50 - 1/100 of the HF number) simplifying the connection to the LFl elements, for example with thin wires through the backing material 902, where also simplified connection techniques are available with the cmut/pmut manufacturing technology.

The invention further presents a solution to the LF/HF transduction with common radiation surface where high frequency cmut/pmut cells are mounted on top of low frequency cells, for example as illustrated in FIG. 10. FIG. 10a shows the substrate front 1000 with one low frequency cell 1001, and several high frequency cells 1002 on top of the low frequency cell . As the low frequency allows large dimensions of the low frequency cell, this cell might be micro-machined from the back side of the substrate as indicated in FIG. 10b which shows a cross section through the substrate 1003 where etching from the substrate back side provides the thin low frequency membrane 1004 in capacitive interaction with an electrode 1005 that is mounted on or part of a 2^nd substrate 1006 that is attached to the substrate 1003 through gluing or other bonding techniques . On the front of the low frequency membrane 1004 is micro-machined several high frequency cells 1002 from the front side of the substrate. With more complex manufacturing techniques, both the low and the high frequency cells can be manufactured from the front side. As for the other cmut/pmut solutions we have not shown details of the electrode arrangements and possible placement of piezoceramic elements, as several examples of such are given in the literature, and we stress in this description essential features of the design to be able to transmit both the LFl and HF pulses from the same radiation surface. However, when Si is used as a substrate, the Figure indicates LFl electrode solutions where a front layer 1007 of the Si-substrate is highly n-doped (n++) to provide a common ground electrode for the LFl and HF cells. The hot LFl electrode could similarly be obtained by high n-doping of a region 1005 of the 2^nd Si-substrate 1006.

Dual frequency band operation with widely separated frequency bands can also conveniently be implemented as sparse arrays, where the low and the high frequency elements are placed at different locations on the array surface, but sufficiently close so that at outside a certain distance from the array, the two frequency beams appear to originate from at least partially the same radiating surface. 2D sparse arrays are especially useful for 3D acoustic imaging where the sparse arrays allow two- dimensional direction steering of the acoustic beam with a reduced number of elements ( ~ 1000) . 2D sparse arrays are also useful for corrections for wave front aberrations and pulse reverberations, both with 2D and 3D beam scanning. An example illustration is shown in FIG. 11 which shows a part of the array surface 1100 with four LFl array elements 1101 with open space 1102 in between for placement of HF array elements in a sparse array pattern. Sparse arrays produce grating lobes in off-set directions from the beam main lobe, where the transmit and receive apertures must be designed for non-overlapping directions of the grating lobes . For imaging methods that are based on the nonlinear interaction between the dual frequency beams, for example as described in US Pat applications 10/189,350 and 10/204,350, one gets improved suppression of the grating lobes in the image when the grating lobes for the LFl and HF beams are non-overlapping. In fact, because of the large wavelength of the low band (λ~ 3mm @ 500 kHz) , it is possible to design an array with small low frequency array elements that do not have low frequency grating lobes but still with so large distance between the elements ( ~ 2 mm) that one can place many high frequency elements between the low frequency elements .

With resonant bulk piezo-ceramic elements for the electro- acoustic transduction similar to FIG. 3, one can with the sparse arrays for example manufacture a high frequency array with division of all its elements, and then select a subgroup of these element locations for the LFl elements which are produced by attaching a piezo-ceramic slab at the back of said selected HF elements and do electrical connection between the front electrode of the high frequency element, which is commonly the ground electrode, and a back electrode of said attached piezo-ceramic slab. With less electro-acoustic transduction efficiency, one can reduce the resonance frequency for the LFl elements by attaching a mass of a heavy and stiff material, for example metals like Cu, Ag, Au, Pd, Pt, or W at the back of the selected HF elements, and use the surface electrodes of the high frequency piezo-ceramic element for transduction .

Micro machined transduction elements on the front of a Si- substrate are also well suited for sparse array implementation of the dual frequency array, as the large low frequency cells and the smaller high frequency cells are machined at different locations on the array surface, as for example shown in FIG. 12, where 1201 shows cmut/pmut cells for the low frequency band on the substrate 1200, encircled by cmut/pmut cells 1202 for the high frequency band. High frequency band cells are electrically connected to form high frequency elements, while the low frequency band cells are connected to form larger low frequency elements, for example as illustrated for the phased array in FIG. 5. Typically several cells are electrically connected for each array element.

Acousto-electric transduction in a 2^nd LF2 band can also be obtained with the cmut/pmut solutions in FIG. 8 - 12 for a HF and a 1^st LFl band, and adding structures 1301 for the LF2 band as illustrated in FIG. 13a. The structure to represent the HF and LFl transduction according to one of FIG. 8 - 12 is indicated by 1321, with acoustic coupling to the load material 301, and with the HF electric port

1309 and LFl electric port 1316. The LF2 electro acoustic transduction is in this embodiment according to the invention obtained with the piezo-layer 1302 mounted on the backing material 1320 with an isolation section 1303 to the front with the same functionality as in relation to FIG. 3e, producing the LF2 electric port 1307 which couples acoustically to the load through the HF/LFl structure 1321. An acoustic transducer array probe with 3 band operation can also be obtained with the structure in FIG. 13a with a triple membrane cmut/pmut solution similar to FIG. 10 and 12, where the LFl band is operated with the piezo-layer at the location of 1302 and the LF2 band is operated by the large membranes 1001 in FIG. 10 or 1201 in FIG. 12. Yet another embodiment is a structure as in PiG. 10 and 12 where one of the membranes 1001/1002 in FIG. 10 or 1201/1292 in FIG. 12 has dual resonance frequencies, so that three resonance frequencies are obtained with two membranes. Typically could the large membranes 1001 or 1201 operate both the LFl and LF2 bands, while the smaller membranes 1002 or 1202 operate at the HF band, or the smaller membranes 1002 or 1202 operate the HF and LFl bands, while the large membranes 1001 or 1201 operates the LF2 band.

Yet another embodiment is to use three separate types of membranes for the LF2 , the LFl, and the HF bands, for example as shown in FIG. 13b, where 1343 shows the HF membranes mounted on top of the LFl membranes 1342 which again are mounted on top of the LF2 membranes 1341 on the common substrate 1340. Alternatively one could mount all the membranes by the side of each other similar to that in FIG. 12, or one membrane type mounted on the top of one of the other types, while the third membrane type is mounted by the side of the others.

FIG. 13a and b show example structures that allow electro- acoustic transduction in 3 frequency bands. One typically wants to divide the radiation surfaces into arrays of elements, for steering of the focus and/or the direction of the beams at all three frequency bands . Typical arrays could be annular, linear, 1.5D, 1.57D, and 2D arrays. The lateral width of the elements (element radiation surface) is then related to the acoustic wavelength in the object 301 for the different frequency bands. The HF array would then require the lowest element width, with intermediate width elements for the LFl array, and largest width for the LF2 array, and so on. One would then typically use a layer structure as in FIG. 13a, b across the whole array width, and define array elements for each frequency band with electrodes and cuts in ceramic layers. With the two layer isolation structures 317 (HF) and 1303 one is less sensitive to location of cuts between the HF, LFl, and LF2 ceramic layers (See discussion in relation to FIG. 3b-e) . Typically one would make each of the piezo-layers as piezo-ceramic/polymer composites, and the elements of these layers would then be defined by the division between the electrodes on the composite surfaces at the cuts through the ceramic layers. This would allow different dimensions and even different shapes of the elements for the different frequency bands, as described above.

Thus, while there have shown and described and pointed out fundamental novel features of the invention as applied to preferred embodiments thereof, it will be understood that various omissions and substitutions and changes in the form and details of the devices illustrated, and in their operation, may be made by those skilled in the art without departing from the spirit of the invention. It is also expressly intended that all combinations of those elements and/or method steps which perform substantially the same function in substantially the same way to achieve the same results are within the scope of the invention. Moreover, it should be recognized that structures and/or elements and/or method steps shown and/or described in connection with any disclosed form or embodiment of the invention may be incorporated in any other disclosed or described or suggested form or embodiment as a general matter of design choice. It is the intention, therefore, to be limited only as indicated by the scope of the claims appended hereto.

Claims

6. Claims

1. An acoustic transducer array probe for transmission from a front face of said probe of acoustic waves in separate high frequency (HF) and at least a 1^st lower frequency

(LFl) bands, and reception of acoustic waves at least in the HF band, characterized by

- different arrays of transducer elements configured for respective HF and the LFl electro-acoustic transduction, the array for the HF electro-acoustic transduction having HF array elements and the array for the LFl electro-acoustic transduction having LFl array elements, and

- radiation surfaces for the HF band and the LFl band, wherein at least a portion of said radiation surfaces for the HF and LFl bands is common in a common radiation surface, wherein the LFl array elements within the common radiation surface have larger radiation surfaces with larger distance between neighboring element centers, than do the HF array elements within the common radiation surface, and - at least one electronic substrate layer with integrated electronics connecting to array elements where said at least one electronic substrate layer is one or both of i) stacked within said acoustic vibration structure of the probe, and ii) mounted in the probe behind said acoustic vibration structure.

2. An acoustic transducer array probe according to claim 1, where the LFl transmit surface has a central region without active LF transmit.

3. An acoustic transducer array probe according to claim 1 or 2, where the thickness structure for the LFl and HF transduction is the same throughout the whole array surface, and the size of HF and LFl transmit and receive apertures can selectably be varied by selectable electric connection to array elements

4. An acoustic transducer array probe according to claim 3, where one can select the HF receive surface to be wider than the LFl transmit surface.

5. An acoustic transducer array probe according to claim 1, where in addition acoustic pulses in one or more lower frequency bands (LF2, LF3 , ...) can be transmitted and received through at least said common radiation surface, where the electro-acoustic transduction for said one or more lower frequency bands are obtained with electro- acoustic transduction structures for each of said one or more lower frequency bands, where at least parts of the radiation surface of said one or more lower frequency bands are common with the radiation surface of the higher frequency bands .

6. An acoustic transducer array probe according to claim 5, where the electro-acoustic transduction arrays for each of said further lower frequency bands are obtained with a piezo-layer placed behind the transduction structures for higher frequency bands .

7. An acoustic transducer array probe according to claim 6, where an acoustic isolation section is placed to the front of each said piezo-layer, where said isolation section provides backwards acoustic attenuation for vibrations in the resonant band of the transduction structure to the front of said each piezo-layer.

8. An acoustic transducer array probe according to claim 1, where the LFl transmit surface is selectable between at least two of a) at least in a region there is a LFl transmit surface outside the HF transmit surface, and b) the LFl transmit surface is equal to the HF transmit surface, and c) the LFl transmit surface has a central region with no LFl transmit, and d) a combination of a) and c) .

9. An acoustic transducer array probe according to claim 1, where the HF and LFl arrays are independently arranged as one of

- a single element array, and

- an annular array of transducer elements, and

- a linear array of transducer elements, and

- a two dimensional array of transducer elements, and - a composition of transducer elements of any other form.

10. An acoustic transducer array probe according to claim 1, where - at least one of said LFl and HF transmit and/or receive surfaces is made as a sparse array of elements, and where

- said sparse arrays are designed so that potential grating lobes of said sparse HF transmit array are directed relative to the LF beam so that the LF pulse pressure at the HF pulse in said HF grating lobes is so low that the nonlinear manipulation of the object elasticity by the LF pulse at the HF pulse is negligible along said HF grating lobes compared to along the HF main lobe.

11. An acoustic transducer array probe according to claim 1, where the LF and HF arrays are stacked behind at least said common transmit surface.

12. An acoustic transducer array probe according to claim 7, where the thickness structure of the array is the same throughout the whole array surface, and the size of LFl and HF transmit and receive apertures are defined by the electrical connection to array element electrodes .

13. An acoustic transducer array probe according to claim 1, where

- the LFl and HF electro acoustic transduction is obtained with separate piezoelectric layers where said

HF piezolayer is stacked in front of said LFl piezolayer in a multilayered structure, and where

- an acoustic isolation section composed of at least two acoustic layers is placed between said HF and LFl piezolayers, and where

- an acoustic matching section for the HF band is placed in front of said HF piezolayer.

14. An acoustic transducer array probe according to claim 13, where said isolation section is composed of a back layer with characteristic acoustic impedance greater than 17 MRayl and at least one layer with characteristic impedance less than 5 MRayl.

15. An acoustic transducer array probe according to claim 13, where said isolation section is made of one of the materials Cu, Ag, Au, Pd, Pt, W, and alloys of these materials, and powders of one of these materials or their alloys sintered or glued together.

16. An acoustic transducer array probe according to claim 13, where said isolation section is composed of a ceramic layer to the back.

17. An acoustic transducer array probe according to claim 16, where said LFl piezoelectric layer is made as a ceramic/polymer composite where the ceramic is diced from the back not fully through said LF piezoelectric layer, so that a front portion of said LF ceramic piezoelectric layer forms a ceramic layer that forms said back layer of said isolation section.

18. An acoustic transducer array probe according to claim 17 , where a 2^nd layer from the back of said isolation section is made of one of the materials Cu, Ag, Au, Pd, Pt, W, and alloys of these materials, and powders of one of these materials or their alloys sintered or glued together.

19. An acoustic transducer array probe according to claim 13, where one or both of the LFl and HF piezoelectric layers are composed of at least two piezoelectric sub- layers on top of each other with electrodes on the surfaces of said sub-layers, and where the electrodes from different sub-layers are coupled together in relation to the electric polarization of said sub-layers so that said sub-layers are electrically coupled in parallel, while the thickness vibration of said sublayers are in series, to reduce the electric impedance of the array elements .

20. An acoustic transducer array probe according to claim 13, where at least one electronic substrate layer with integrated electronics connecting to array elements is stacked in the layer structure as at least one of

- part of said acoustic isolation section, and

- to the back of said HF piezo-layer, and - to the front of said HF piezo-layer, and

- to the front of said LFl piezo-layer, and

- to the back of said LFl piezo-layer, and

- to the back of a backing material .

21. An acoustic transducer array probe according to claim 13, where acoustic waves in further lower frequency bands (LF2, LF3 , ...) can be transmitted and received through at least said common radiation surface, where

- the electro-acoustic transduction for each of said further lower frequency bands are obtained with an added electro-acoustic transduction structure composed of a piezo-layer with an acoustic isolation section to the front, and where

- said isolation section is placed in acoustic contact with the back-side of the transduction section for the next higher frequency band, and where

- said isolation section provides backwards isolation of vibrations in the next higher frequency band of the neighbor transduction section to the front.

22. An acoustic transducer array probe according to claim 1, where

- one of the LFl and HF electro acoustic transduction is obtained by a piezoelectric layer, and

- the other of said HF and LFl electro acoustic transduction is obtained by a substrate layer with cmut/pmut based transduction membranes on the front face,

- said layer based on cmut/pmut transduction membranes is placed to the front of said piezolayer.

23. An acoustic transducer array probe according to claim 22, where said LFl electro acoustic transduction is obtained by said piezolayer.

24. An acoustic transducer array probe according to claim 22, where said HF electro acoustic transduction is obtained by said piezolayer.

25. An acoustic transducer array probe according to claim 22, where at least one electronic substrate layer with integrated electronics connecting to array elements are stacked in the layer structure as at least one of

- electronic substrate layers are mounted to the back of said substrate layer with cmut/pmut membranes, and

- electronic substrate layers are mounted to the back of said LF piezo-layer, and

- electronic substrate layers are mounted to the back of a backing material .

26. An acoustic transducer array probe according to claim 22, where acoustic waves in further lower frequency bands (LF2, LF3 , ...) can be transmitted and received through at least said common radiation surface, where

27. An acoustic transducer array probe according to claim 22, where the thickness structure for the LF and HF transduction is the same across the whole probe maximal transmit/receive surface, and the active LFl and HF apertures for radiation and reception can selectably be varied by selectable electric connection to array element electrodes.

28. An acoustic transducer array probe according to claim 1, where both the LFl and HF electro/acoustic transduction are obtained by cmut/pmut based membrane transducer technology on the same substrate where separate membranes are used for the LFl and HF transduction that are separately optimized for vibrations within the LFl and HF bands .

29. An acoustic transducer array probe according to claim 28, wherein at least one electronic substrate layer with integrated electronics connecting to array elements is placed to the back of said substrate layer with cmut/pmut membranes.

30. An acoustic transducer array probe according to claim 28, where the HF transduction membranes are placed on top of the LFl transduction membranes.

31. An acoustic transducer array probe according to claim 28, where the HF transduction membranes are placed side by side of the LFl transduction membranes, so close to each other that after a certain depth from the probe surface the LFl and HF beams appear to originate from radiation surfaces that at least have a common region.

32. An acoustic instrument utilizing an acoustic probe according to claim 28, where in addition acoustic pulses in a 2^nd lower frequency (LF2) band can be transmitted and received through the common radiation surface, where the LFl cmut/pmut membranes also have resonances in the LF2 band.

33. An acoustic instrument utilizing an acoustic probe according to claim 28, where in addition acoustic pulses in a 2^nd lower frequency (LF2) band can be transmitted and received through the common radiation surface, where the electro-acoustic transduction for said LF2 band is obtained with cmut/pmut membranes on the same substrate, where one of

- the HF and LFl membranes are placed on the top of the LF2 membranes, and

- the LF2 membranes are placed by the side of the HF and LFl membranes .

34. An acoustic instrument utilizing an acoustic probe according to claim 28, where in addition acoustic pulses in a 2^nd lower frequency (LF2) band can be transmitted and received through the common radiation surface, where the electro-acoustic transduction for said LF2 band is obtained with a LF2 piezo-layer placed behind said HF/LFl common substrate, so that the LF2 radiation surface at least has a part common to the common radiation surface of said HF and LF2 array.

35. An acoustic instrument utilizing an acoustic probe according to claim 34, where acoustic waves in further lower frequency bands (LF3, LF4, ...) can be transmitted and received through at least said common radiation surface, where

36. An acoustic transducer array probe according to claim 13 or 22 or 28, where said LFl and HF radiation surfaces are equal .

37. An acoustic transducer array probe according to claim 22 or 28, where the LFl and HF transmit and receive apertures can selectably be varied by selectable electric connection to array element electrodes.

38. An instrument utilizing an acoustic probe according to claim 1, where the active LFl transmit aperture is selected through one of a) automatic selection by the instrument depending on the image range and the ultrasound imaging modality and application, and b) directly by the instrument operator through instrument controls.

39. An acoustic transducer array probe according to claim 1, where said at least one electronic substrate layer with electronic circuits include one or more of a) receiver pre-amplifiers connected to array elements, and b) transmitter amplifiers connected to array elements, and c) electronic switches that connect selectable groups of array elements to a single wire that further connects to an instrument, and d) sub-aperture circuits connecting to a group of elements and adding delays to the individual sub- aperture signals before summing to sub-aperture signals that each connect to a single wire that further connects to an instrument, and e) time multiplex circuits that time multiplex samples of signals from groups of elements or groups of sub- apertures of elements on single wires that further connect to an instrument.