WO2019038242A1

WO2019038242A1 - Ultrasound transducer array, device and system

Info

Publication number: WO2019038242A1
Application number: PCT/EP2018/072464
Authority: WO
Inventors: Nenad Mihajlovic; Martin Pekar; Peter Dirksen; David Hope Simpson
Original assignee: Koninklijke Philips N.V.
Priority date: 2017-08-21
Filing date: 2018-08-20
Publication date: 2019-02-28

Abstract

An ultrasound transducer (10) is disclosed comprising a substrate (101); and a plurality of CMUT (capacitive micromachined ultrasound transducer) cells (100, 100') formed on the substrate, each having substantially the same collapse voltage threshold and comprising a first electrode (110) coupled to the substrate; a flexible membrane (130, 130') separated from substrate by a cavity; and a second electrode (120) coupled to the membrane. The CMUT cells are arranged to form at least one array element (99), wherein within each array element the CMUT cells are arranged to be simultaneously activated by a single stimulus; and at least some of the CMUT cells have different peak resonance frequencies, and each CMUT cell having a lower peak resonance frequency has a larger distance to the centre of the array element in an extension direction of the array element than any CMUT cell having a higher peak resonance frequency. Also disclosed are an interventional medical device and an ultrasound imaging system including such an ultrasound transducer.

Description

ULTRASOUND TRANSDUCER ARRAY, DEVICE AND SYSTEM

FIELD OF THE INVENTION

The present invention relates to an ultrasound transducer array comprising a substrate and a plurality of CMUT (capacitive micromachined ultrasound transducer) cells formed on the substrate, each comprising a first electrode coupled to the substrate; a flexible membrane formed in spaced relationship to the first electrode; and a second electrode coupled to the membrane.

The present invention further relates to an invasive medical device comprising such an ultrasound transducer array.

The present invention further relates to an ultrasound imaging system comprising such an ultrasound transducer array.

BACKGROUND OF THE INVENTION

The ultrasonic transducers used for medical imaging have numerous characteristics which lead to the production of high quality diagnostic images. Among these are broad bandwidth and high sensitivity to low level acoustic signals at ultrasonic frequencies. Conventionally the piezoelectric materials which possess these characteristics and thus have been used for ultrasonic transducers have been made of PZT and PVDF materials, with PZT being the most preferred. However the ceramic PZT materials require manufacturing processes including dicing, matching layer bonding, fillers, electroplating and interconnections which are distinctly different and complex and require extensive handling, all of which can result in transducer stack unit yields which are less than desired.

Furthermore, this manufacturing complexity increases the cost of the final transducer probe.

As ultrasound system mainframes have become smaller and dominated by field programmable gate arrays (FPGAs) and software for much of the signal processing functionality, the cost of system mainframes has dropped with the size of the systems.

Ultrasound systems are now available in inexpensive portable, desktop and handheld form. As a result, the cost of the transducer probe is an ever- increasing percentage of the overall cost of the system, an increase which has been accelerated by the advent of higher element- count arrays used for 3D imaging. Accordingly, it is desirable to be able to manufacture transducer arrays with improved yields and at lower cost to facilitate the need for low-cost ultrasound systems.

Recent developments have led to the prospect that medical ultrasound transducers can be manufactured by semiconductor processes. Desirably these processes should be the same ones used to produce the circuitry needed by an ultrasound probe such as a CMOS process. These developments have produced micromachined ultrasonic transducers or MUTs. The individual MUT cells can have round, rectangular, hexagonal, or other peripheral shapes. MUTs have been fabricated in two design approaches, one using a semiconductor layer with piezoelectric properties (PMUTs) and another using a diaphragm and substrate with electrode plates that exhibit a capacitive effect (CMUTs). The CMUT transducers are tiny diaphragm-like devices with electrodes that convert the sound vibration of a received ultrasound signal into a modulated capacitance.

For transmission the capacitive charge applied to the electrodes is modulated to vibrate the diaphragm of the device and thereby transmit a sound wave. Since these devices are manufactured by semiconductor processes the devices generally have dimensions in the 10-200 micron range, but can range up to device diameters of 300-500 microns. Many such individual CMUTs can be connected together and operated in unison as a single transducer element. For example, four to sixteen CMUTs can be coupled together to function in unison as a single transducer element. A typical 2D transducer array currently will have 2000-10,000 piezoelectric transducer elements. When fabricated as a CMUT array, upwards of 50,000 CMUT cells will be used. Surprisingly, early results have indicated that the yields on semiconductor fab CMUT arrays of this size should be markedly improved over the yields for PZT arrays of several thousand transducer elements.

One characteristic of a transducer that impacts image quality is ultrasound beam-width. Ideally the ultrasound beam is narrow, allowing good resolution of small targets. Beam-width and resolution are a function of the aperture size, which affects the focal spot size b and hence the resolution. The beam and focal characteristics of a small aperture transducer 500 is illustrated in FIG. 1. The beam profile 520 is seen to converge on a focus spot size b at a distance D from the transducer or array of aperture size W. In general, the focal spot size b is related to the focal distance D, the aperture size W, the frequency of the ultrasonic wave f with a wavelength λ in accordance with: b ~ λϋ/W = λΡ Larger apertures have improved resolution, but at a cost of reduced depth-of- field (the range of depth D over which a narrow beam is achieved) as shown by the larger element or array 540 with its narrower focal spot size b as depicted in FIG. 2. Thus, conventional single element or one-dimensional transducer arrays trade off resolution to achieve good image quality over a wide range of depth.

To improve this trade off transducers and arrays can be constructed of multiple elements or element groups as illustrated in FIG. 3. The transducer depicted in this drawing has three elements or groups of elements, 600 in the center, 620 elevationally offset or surrounding the center element or group of elements 600, and an elevationally outer element or group of elements 640. In the near field, element or group 600 is actuated, producing a near field focus bl . At the mid-range the next ring or group of elements 620 is actuated, producing a mid-range focus b2. In the far field the outer ring or group 640 is actuated, producing a far field focus b3. This transducer can be operated in a zone focus mode in which the three elements or group of elements are actuated sequentially: transmit and receive in the near field with element or group 600, then transmit and receive in the mid-range with element or group 620, then transmit and receive in the far field with element or group 640. After signals have been acquired from all three ranges, a complete image is put together using the echoes acquired from all three ranges. This, of course, requires three transmit-receive cycles which reduces real time frame rate by one-third.

An alternative mode of operation is to transmit with one or several elements or group of elements, then dynamically vary the aperture and receive focus as echoes are received over the full depth of field. For example, elements or groups 600 and 620 can be connected together by closing switch 660 and transmitting a beam focused at the mid-range focal point b2. During receive, the active aperture and focal point are progressively changed by opening switch 660 to initially receive echoes from the near field with element or group 600, with a focal point at bl . After the near field echoes have been received with 600, switch 660 is closed to receive mid-range echoes with both the center and next element or group in elevation 620. The active aperture is now larger, being the combination of both 600 and 620, the latter having a mid-range focal point b2. After reception of the mid-range echoes switch 680 is also closed to receive with the largest aperture, the combination of 600, 620 and 640, receiving far field echoes with the elevationally outermost element or group with a far field focal point b3. A group of transducers (elements) such as 600, 620 or 640 within the array, which are activated to form a current aperture, are referred as an active aperture herein. A combination of all active apertures, which are activated to form a single ultrasound image, is referred as an aperture herein.

The focal points of array are illustrated to be located along a line orthogonal to a central region (surface area) of the array. It shall be understood to the skilled person that, if the array is arranged to perform an electronic beam steering, the focal points of the array can lie along directions which cross the surface area of the array at different to orthogonal angles. In addition a central region of the active apertures can be different from the central region the array.

Nevertheless, the transducer remains susceptible to the phenomenon of depth- dependent attenuation. Echoes from increasing depths contain progressively less signal and progressively more noise at the higher frequencies. This is because the transducer as described has no frequency selectivity, no ability to shape the passband for the lower frequency echo signals received by the elevationally outer elements at increasing depths.

One approach to addressing this problem is described in US Pat. 5,976,091 (Hanafy). This patent describes a transducer element in which the transducer crystal and its matching layers and lens have a variable thickness, tapering from thin at the center to thicker at the edges in the elevation dimension. The outer elevation portion of the crystal resonates at the lowest frequency, and is focused, in both transmission and reception, at the deepest part of the image where the low frequencies also provide better penetration. The central portion is thinner, resonating at higher frequencies and exhibits a shallower focus. Such a transducer arrangement is commonly referred to as a Hanafy lens. While this approach provides a desired frequency selectivity, it is extremely difficult and costly to fabricate properly tapered transducer layers and bond them together. Furthermore, in a multi-element transducer array, specialized switches are required to switch in each element at its proper phase of the transmit-receive cycle. Accordingly, an improved approach to providing frequency selectivity is desirable.

SUMMARY OF THE INVENTION

The present invention seeks to provide an ultrasound transducer array that provide such frequency selectivity in a more straightforward manner.

The present invention further seeks to provide an interventional medical device comprising such an ultrasound transducer array.

The present invention yet further seeks to provide an ultrasound imaging system comprising such an ultrasound transducer array. According to a first aspect, there is provided an ultrasound transducer comprising a substrate and a plurality of CMUT (capacitive micromachined ultrasound transducer) cells formed on the substrate, each having substantially the same collapse voltage threshold and comprising a first electrode coupled to the substrate; a flexible membrane separated from substrate by a cavity; and a second electrode coupled to the membrane;

wherein the CMUT cells are arranged to form at least one array element, wherein within each array element the CMUT cells are arranged to be simultaneously activated by a single stimulus; and at least some of the CMUT cells have different peak resonance frequencies, and each CMUT cell having a lower peak resonance frequency has a larger distance to the centre of the array element in an extension direction of the array element than any CMUT cell having a higher peak resonance frequency.

The present invention is based on the insight that a CMUT transducer including one or more array elements may be provided that can emulate the behaviour of a Hanafy lens in a straightforward manner. This is achieved by the provision of a plurality of CMUT cells in each array element that have substantially the same collapse voltage, i.e. the applied bias voltage at which the flexible membrane of the cell collapses onto the substrate, whilst the respective peak resonance frequencies of the CMUT cells in such an array element decreases with increasing distance from the central region of the array element in its extension direction, i.e. the direction in which the CMUT cells within the array are arranged, e.g. a linear arrangement in case of a linear array element. Hence, whilst the CMUT cells in each array element are simultaneously addressed, i.e. activated, by a single stimulus including a bias voltage component driving each of the cells within the array element into collapse mode and a RF excitation component such as a short enough pulse to generate the different peak resonance frequencies with the different CMUT cells within the array element, the center portion of the array element generates higher resonance frequencies than its peripheral portions, thereby providing an ultrasound transducer including one or more array elements that can be manufactured in a relatively straightforward manner and can emulate the spatial focal length variations of a Hanafy lens.

The ultrasound transducer may comprise a plurality of said array elements, wherein the respective array elements are individually addressable and wherein, the CMUT cells are arranged in a plurality groups extending across the plurality of array elements, and wherein the CMUT cells in a group have the same peak resonance frequency. This has the advantage that each array element is operable as a Hanafy lens, whilst beamforming may be achieved by addressing the different array elements with different time delays for instance.

The CMUT cells may be arranged in a plurality of staggered rows extending in said extension direction, wherein each group extends across each of said staggered rows. Such staggering may be utilized to increase the CMUT cell density of such an ultrasound transducer array.

In an embodiment, the CMUT cells are arranged in a plurality of columns and each group extends across each of said columns, the ultrasound transducer further comprising a plurality of silicon islands each carrying at least one of said columns; and a flexible foil retaining the respective silicon islands, the flexible foil comprising conductive interconnects such that the ultrasound transducer array may be wrapped around an object such as the body of an invasive medical device, e.g. a catheter or guide wire.

In another embodiment, each further group envelopes a group of said plurality of group, thereby providing a 2-D ultrasound transducer having improved beamforming characteristics.

In order to achieve the desired common collapse voltage threshold and different peak resonance frequencies of the CMUT cells within such an array element, several design parameters of the CMUT cells may be varied in a systematic manner. For example, the gap height in the absence of the single stimulus in the different cells may be different, in which case in order to keep a substantially constant collapse voltage for the CMUT cells in the array element the membrane thickness must be varied as well. Typically, an increase in gap height is compensated by a reduction in the thickness of the membrane to retain the same collapse voltage.

Alternatively or additionally, the thickness of a dielectric layer in between the first electrode and the second electrode, typically a dielectric layer forming part of the flexible membrane may be varied across the CMUT cells within each array element to vary the peak resonance frequencies of these elements whilst retaining the same collapse voltage threshold.

These variables may be combined with a change in membrane diameter across the CMUT cells of such an array element, wherein as will be understood by the skilled person these variables are chosen such that the collapse voltage threshold of the various CMUT cells within a single array element remains substantially the same, i.e. varies by less than 20%, preferably by less than 10%, more preferably by less than 5%. A particularly attractive way of altering the peak resonance frequency of the CMUT cells within a single array element whilst keeping their respective collapse voltage thresholds substantially constant is by controlling the mass of the membranes of the various groups of CMUT cells within the one or more array elements whilst keeping the spring constants of the membranes of the CMUT cells across the various groups substantially similar. This ensures that the CMUT cells of an array element can be addressed by a single stimulus, e.g. to simultaneously drive all cells into collapse mode, owing to the substantially similar spring constants of the CMUT cell membranes across the array element, whilst the systematically increased mass of the CMUT cell membranes at the cell group level in a direction away from the centre of the array element shifts the peak resonance frequency of the higher mass membranes to lower frequencies when activated by the RF component of the single stimulus.

In an embodiment, each of the flexible membranes comprises a layer stack, and the CMUT cells within each array element comprises a first CMUT cell proximal to the central region and a plurality of further CMUT cells at a greater distance from the central region than the first CMUT cell, wherein the layer stacks of the flexible membranes of at least some of the further CMUT cells include a layer of a material having a higher density than any of the layers in the layer stack of the flexible membranes of the first CMUT cell.

The provision of a (thin) layer of a high density material to selected CMUT cells in order to differentiate the CMUT cells of the transducer array may be used to significantly increase the mass of the membranes of the selected CMUT cells without significantly increasing the spring constant of these membranes. Consequently, CMUT cells of the transducer array may still collapse at approximately the same bias voltage yet exhibit different resonance frequencies due to the difference in mass of the membranes of the first group and second group of CMUT cells respectively, thereby yielding a transducer array that can be safely operated in collapse mode and that exhibits an increased bandwidth, i.e.

broadband characteristics, compared to CMUT transducer arrays in which all CMUT cells have the same geometry.

For this reason, the material preferably also has a lower Young's modulus compared to the material of any of the layers of the first layer stack, such that the high density material has a minimal contribution to the overall bending stiffness of the membranes of the further groups of CMUT cells. For example, the material having a higher density than any of the layers in the layer stack of the CMUT cells in the first group may have a density in excess of 7 g/cm³, preferably a density in excess of 10 g/cm³ and a Young's modulus of less than 200 GPa, preferably a Young's modulus of less than 100 GPa.

Preferably, the variation in membrane spring constants across each array element is no more than 20% and more preferably no more than 10% to ensure that all CMUT cells of the array element enter collapse mode at approximately the same bias voltage. In some embodiments, the spring constants of the membranes across each array element are identical. The tuning of the spring constant of the respective membranes may be achieved by selecting the materials and tuning the thickness of the layers of these materials in the membranes of the first CMUT cell and further CMUT cells respectively of an array element.

In an embodiment, the layer of a material having a higher density than any of the layers in the layer stack of the first CMUT cell is a metal layer, preferably a gold layer or a platinum layer. Metals may be used due to their high density, which makes these materials particular suitable for use as a high density material layer. In the context of the present application, the term metal also includes metal alloys.

The layer of said material may be a patterned layer. Patterning the high- density layer is a particularly advantageous way of tuning the mass of the membranes of the CMUT cells whilst avoiding height variations between membranes of different groups of CMUT cells including a layer of said material. Alternatively, the layer of said material may be present in the flexible membranes of each of the further CMUT cells within an array element, with the thickness of said layer increasing between said further CMUT cells with increasing distance from the central region.

In another embodiment, the layer stack of each flexible membrane of the first CMUT cell in an array element comprises a layer of a dielectric material to a first thickness; and the layer stack of each flexible membrane of the further CMUT cells in the array element comprises a layer of the dielectric material to a second thickness that is smaller than the first thickness. The second thickness may be chosen such that the combination of the dielectric material layer to the second thickness and the thickness of the layer of the high- density material in the membrane layer stacks of the further groups provides (approximately) the same contribution to the overall spring constant of the membranes of the CMUT cells in the second group as the layer of the dielectric material to the first thickness achieves for the membranes of the CMUT cells in the first group.

Alternatively, the layer stack of each flexible membrane of the first CMUT cell in the array element is the same as the layer stack of each flexible membrane of the further CMUT cells in the array element apart for the layer of said material having a higher density than any of the layers in the layer stack of each flexible membrane of the first CMUT cell being an additional layer. This provides an ultrasound transducer that may be

manufactured in a particularly straightforward manner, e.g. by the selective deposition of the layer of the high density material to the CMUT cells of the further groups.

In a further embodiment, the flexible membranes of the CMUT cells have a diameter, and the CMUT cells within an array element include a first set of CMUT cells proximal to said central region in the extension direction of the array element and a second set of CMUT cells distal to said central region in the extension direction, wherein the diameter of the flexible membranes of the CMUT cells in said first set is smaller than the diameter of the flexible membranes of the CMUT cells in said second set.. In collapse mode, the resonance frequency of the CMUT cells becomes largely independent of cell diameter, such that the addition of the high density material layer to the second group of CMUT cells introduces an alternative design freedom to achieve broadband characteristics in an array element operable in collapse mode having CMUT cells of different diameters.

According to another aspect, there is provided an interventional medical device comprising a flexible elongate body and the ultrasound transducer of any of the herein described embodiments arranged around said flexible elongate body such that the extension direction of the one ore more array elements is aligned with the elongation direction of the flexible elongate body, thereby providing such an interventional medical device with improved imaging capability in the field of view around the device.

According to yet another aspect, there is provided an ultrasound imaging system comprising a patient interface module and the ultrasound transducer or the interventional medical device according to any of the herein described embodiments. Such an ultrasound imaging system may further comprise a power supply adapted to simultaneously activate the CMUT cells of an array element in a collapse mode during at least one of an ultrasound transmission mode and an ultrasound reception mode by supplying each of the CMUT cells of an array element with the same stimulus including a bias voltage component and a RF excitation component. This therefore provides an ultrasound imaging system exhibiting improved depth of view characteristics by implementing the functionality of a

Hanafy lens as previously explained without required complex steering electronics to steer or drive the various CMUT cells in the ultrasound transducer into their respective collapse modes. BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention are described in more detail and by way of non- limiting examples with reference to the accompanying drawings, wherein:

FIG. 1 (prior art) illustrates a small aperture transducer element or array and its focal characteristic;

FIG. 2 (prior art) illustrates a large aperture transducer element or array and its focal characteristic;

FIG. 3 (prior art) illustrates the different focal characteristics of a transducer array with differently operating elements in the elevation dimension;

FIG. 4 schematically depicts a top view of an ultrasound transducer according to an embodiment;

FIG. 5 schematically depicts a top view of an ultrasound transducer according to another embodiment;

FIG. 6 schematically depicts a cross-section of part of an ultrasound transducer according to an embodiment;

FIG. 7 is a graph depicting an operating characteristic of an ultrasound transducer according to an embodiment;

FIG. 8 is an output pressure contour plot of a CMUT array with (right pane) and without (left pane) a high density layer of its CMUT membranes;

FIG. 9 schematically depicts a cross-section of part of an ultrasound transducer according to another embodiment;

FIG. 10 schematically depicts a top view of an ultrasound transducer according to another embodiment;

FIG. 1 1 schematically depicts a top view of an ultrasound transducer according to yet another embodiment;

FIG. 12 schematically depicts typical beam forming characteristics achieved with an ultrasound transducer according to an embodiment;

FIG. 13 schematically depicts an example manufacturing process for an ultrasound transducer according to embodiments of the present invention;

FIG. 14 is a plan view of a symmetrically arranged CMUT of rows and columns of CMUT cells according to an embodiment;

FIG. 15 is a plan view of a CMUT configured with staggered rows of cells with the cells of adjacent rows and columns interspersed within each other according to another embodiment; FIG. 16 illustrates steps in the fabrication of a flexible interconnect of adjacent cell silicon islands according to an embodiment;

FIG. 17 illustrates the operation of adjacent staggered rows of CMUTs as a single row of transducer elements in accordance with an embodiment;

FIG. 18 illustrates the CMUT array of FIG. 17 when wrapped around an invasive medical device; and

FIG. 19 is a block diagram of an ultrasound imaging system according to an embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

It should be understood that the Figures are merely schematic and are not drawn to scale. It should also be understood that the same reference numerals are used throughout the Figures to indicate the same or similar parts.

In the context of the present application, where reference is made to a membrane, this is a deformable structure that spans the gap or cavity over the substrate of a capacitive micromachined ultrasound transducer (CMUT), and that supports, e.g. embeds one of the electrodes of the CMUT, e.g. an electrode opposing a further electrode on the substrate and separated therefrom by a gap or cavity.

In the context of the present application, where reference is made to a membrane layer stack, this is intended to include membranes formed of a single layer as well as membranes formed of multiple layers, but excluding the electrode embedded in or otherwise supported by the membrane.

In the context of the present application, where reference is made to a CMUT element or transducer element, this may equate to a single CMUT cell or to a cluster of CMUT cells arranged to be operated in unison, e.g. arranged to be addressed by a single control signal. In particular, where reference is made to an array element, this is to be understood as a plurality of CMUT cells that are arranged to be operated in unison, without time delay between the respective CMUT elements. In order words, the CMUT cells of such an array element are typically addressed by a single stimulus comprising a bias voltage (DC) component, that drives each of the CMUT cells of such an array element into collapse mode and a RF (radio-frequency) (AC) component that causes the collapsed membranes of the different CMUT cells within an array element driven into collapse mode to vibrate at different peak resonance ferquencies as will be explained in more detail below. FIG. 4 schematically depicts a top view of an ultrasound transducer 10 according to an embodiment. The ultrasound transducer array 10 comprises a plurality of array elements 99, with each array element 99 comprising a plurality of CMUT cells. The CMUT cells of the ultrasound transdcuer 10 are divided in a plurality of groups of CMUT cells, including a first group Gl of CMUT cells 100 and a plurality of further groups G2-G4 of CMUT cells 100'. The CMUT cells 100 and 100' are shown to have a circular shape by way of non-limiting example only. Other shapes may be contemplated for the CMUT cells 100, 100'. The CMUT cells 100, 100' may be arranged in a staggered fashion, i.e. in staggered rows 150 as schematically shown in FIG. 5 to increase the cell density of the ultrasound transducer array 10, with each of the groups G1-G4 meandering across the staggered rows 150. Cells 100, 100' in a single row may be clustered, e.g. addressed by a single row address signal, as indicated by the horizontal arrows. In other words, each single row defines an array element, i.e. a cluster of CMUT cells 100, 100' arranged to be operated in unison, in which the CMUT cells 100, 100' are arranged in the extension direction of the array element 99. As can be seen from FIG. 4 and FIG. 5, the groups G1-G4 of CMUT cells 100, 100' typically extend in a direction perpendicular the the extension direction E of the array elements 99 such that each array element 99 comprises at least one member, i.e. one CMUT cell 100, 100' of each of the groups G1-G4. The array elements 99 preferably are symmetrical in the sense each array element 99 comprises a symmetry axis in its center running perpendicular to the extension direction E of the array element 99.

The plurality of array elements 99 may be individually addressable, e.g. to facilitate electronic beam steering with the ultrasound transducer 10. Alternatively, the array elements 99 may be operated in unison. It should furthermore be understood that the ultrasound transducer 10 alternatively may comprise only a single array element 99, e.g. implement a one-dimensional ultrasound transducer 10. Other arrangements will be apparent to the skilled person.

In accordance with the inventive principles of the present invention, each array element 99 comprises a plurality of CMUT cells 100, 100' belonging to different groups Gl- G4 in case of the presence of multiple array elements 99 in which the CMUT cells within each array element 99 has substantally the same collapse voltage threshold, i.e. a difference between the respective collapse voltage thresholds of the CMUT cells 100, 100' within a single array element 99 is less than 20%, preferably less than 10%, more preferably less than 5% such that a single bias voltage signal applied to such an array element 99 can

simultaneously drive all its CMUT cells 100, 100' into collapse mode. At the same time, the groups G1-G4 of CMUT cells 100, 100' distinguish themselves from each other in that with increasing distance from the central region of the one or more array elements 99 the CMUT cells 100, 100' in these groups exhibit a decrease in their peak resonance frequencies such that for a single array element 99 in which its CMUT cells 100, 100' are simultaneously driven into collapse mode, the central CMUT cells 100 have the highest peak resonance frequency, with the peak resonance frequencies of the CMUT cells 100' decreasing with increasing distance from the central region of the array element 99. In other words, all CMUT cells 100, 100' within a single group G1-G4 exhibit the same peak resonance frequency.

It is furthermore noted that this point that within a single array element 99 multiple CMUT cells having the same peak resonance frequency may be present. In such a case, the CMUT cells having the same peak resonance frequency are typically spatially clustered, such that the array element 99 comprises a plurality of such clusters, with the peak resonance frequency of the CMUT cells between clusters decreasing with increasing distance from the centre of the array element 99.

Hence, such an ultrasound transducer 10 comprises one or more array elements 99 implementing a Hanafy lens in which the CMUT cells 100 and 100' in each array element 99 are operated in unison using a single stimulus including a bias voltgae component driving the CMUT cells 100, 100' into collapse and an RF excitation component such that an outer portion of the array element 99, i.e. the group G4 of CMUT cells 100' resonates at the lowest frequency, and is focused, in both transmission and reception, at the deepest part of the image where the low frequencies also provide better penetration. The central portion of the array element 99, i.e. the group Gl of CMUT cells 100 resonates at higher frequencies and exhibits a shallower focus.

CMUT cells 100, 100' exhibiting the same or a substantially similar collapse voltage threshold but different peak resonance frequencies within a single ultrasound transducer 10, i.e. within a single array element 99, may be achieved by systematically varying a number of design parameters of the CMUT cells 100, 100' such as for example flexible membrane thickness, height of the cavity between the substrate and membrane, thickness and/or dielectric constant of a dielectric (electrically insulating) layer between the electrodes of the CMUT cell, membrane diameter and so on.

For example, it has been found that a combination of an increase of the gap height of the cavity, i.e. the gap height in the absence of the single stimulus, and a decrease in the thickness of the membrane of a CMUT cell does not alter the collapse voltage threshold of such a cell, but alters its peak resonance frequency, with a 10% increase in gap height typically resulting in a 10% reduction in peak resonance frequency.

Similarly, a reduction in thickness of a dielectric layer between the opposing electrodes of the CMUT cell does not (significantly) affect its collapse voltage threshold, but increases the peak resonance frequency of the cell due to the larger electric field between the electrodes, with a 10% reduction in the thickness of the dielectric layer typically leading to about a 10% increase in peak resonance frequency of the cell.

Another design freedom to tune the peak resonance frequency of the CMUT cell is the membrane diameter, with larger membranes typically exhibiting lower peak resonance frequencies. Of course, any combination of such design freedoms may be used to tune the peak resonance frequency of the CMUT cells of the ultrasound transducer 10, i.e. within an array element 99.

In the above design freedoms, the dimensions of existing structures within a typical CMUT cells are systematically varied in order to tune the peak resonance frequencies of such cells whilst retaining a common collapse voltage threshold for such cells. This can come at the expense of a more involved manufacturing process, where several additional (masking) steps may be required to obtain the different CMUT cells 100, 100' within a single ultrasound transducer 10.

FIG. 6 schematically depicts a cross-section of a CMUT cell 100 of the first group Gl and a CMUT cell 100' of one of the further groups G2-G4 respectively. Each CMUT cell comprises a first electrode 1 10 supported by a substrate 101 and a second electrode 120 opposing the first electrode 1 10 and separated therefrom by a cavity 105. It is reiterated for the avoidance of doubt that the groups G1-G4 may comprise one or more members, i.e. in the case of a single array element 99, each group may only contain a single member, although multiple members may be present in each group even in the case of a single array element 99 where such an array element has spatially clustered CMUT cells 100, 100' having the same peak resonance frequency in a cluster as previously explained.

In the first group of CMUT cells 100, the second electrode 120 is supported by a flexible membrane comprising a first layer stack 130, which for example includes a first layer 131 and a second layer 133 of an electrically insulating or dielectric material such as silicon oxide, silicon nitride and so on. In some applications, e.g. low-frequency applications, the first layer stack 130 may have a thickness well in excess of 1 micron, e.g. a thickness in the range of 5-20 microns. In some other applications, e.g. high-frequency applications, the first layer stack 130 may have a thickness of less than 2 microns. As explained above, the thickness of the layer stack may be varied between different groups G1-G4 of CMUT cells 100, 100' in combination with the gap height of the cavity 105 (see below) to tune the peak resonance frequency of the CMUT cells.

The second electrode 120 may be embedded in the flexible membrane comprising the first layer stack 130 such that the second electrode 120 is separated from the cavity 105 by a thin layer of dielectric material, for example to prevent a short circuit between the first electrode 1 10 and the second electrode 120 upon a central region of the flexible membrane contacting the substrate 101, e.g. during operation of the CMUT cells 100, 100' of the ultrasound transducer array 10 in collapse mode. As explained above, the thickness of this dielectric layer may be varied between different groups G1-G4 of CMUT cells 100, 100' to tune the peak resonance frequency of the CMUT cells.

In yet another embodiment, the CMUT cells 100', the second electrode 120 is supported by a flexible membrane comprising a second layer stack 130', which includes a layer 135 of a material having a higher density than any of the layers in the first layer stack 130. The layer 135 will be referred to as the mass layer 135 in the remainder of this description. The mass layer 135 is present in the membrane layer stack 130' of the further groups of CMUT cells 100' to increase the mass of the membrane layer stack 130' without significantly affecting its bending stiffness, which bending stiffness largely defines the spring constant of the CMUT cells 100, 100', in particular when the CMUT cells 100, 100' are operated in a collapse mode in which a central part of the membranes is permanently collapsed onto the substrate 101, with the peripheral portion of the membranes around the central portion oscillating to generate an ultrasound pulse having a desired frequency spectrum in a transmission mode of the ultrasound transducer 10 or resonating in response to a pulse echo having a desired frequency spectrum being received in a reception mode of the ultrasound transducer 10. As such collapse mode operation is well-known per se, this will not be explained in further detail for the sake of brevity only.

The center frequency F_c of a circular CMUT cell membrane may be approximated by the following equation (1):

In this equation, R is the membrane radius, p is the membrane density, D is the flexural rigidity (spring constant) of the membrane, t is the membrane thickness and x is the ratio Rc/R wherein R_c is the radius of the central portion of the membrane collapsed onto the substrate 101 (R_c = 0 when the membrane is not in collapse). As can be derived from this equation, the resonance frequency of the membrane depends on both the flexural rigidity (spring constant) D and the density p of the membrane. However, the energy or force required to collapse the membrane onto the substrate 101, e.g. to force the CMUT cell into collapse mode, is dominated by the flexural rigidity D of the membrane.

Therefore, embodiments of the present invention are directed to increasing the mass of the layer stack 130' of the further groups G2-G4 of CMUT cells 100' compared to the mass of the layer stack 130 of the first group of CMUT cells 100 whilst minimizing the difference in spring constants between the membranes comprising the first layer stack 130 and the second layer stack 130' respectively.

The flexural rigidity D of the membrane may be expressed by the effective Young's modulus E_eff of the membrane layer stack (equation (2)):

In this equation, μ is the Poisson ratio. The effect of the addition of a mass layer 135 to the effective Young's modulus may be -approximated by the following equation

E₁ * t₁ + E₂ * t₂

(t_{1 +} t₂)

In this equation, Ei and ti are the Young's modulus and thickness respectively of the membrane stack and E₂ and t₂ are the Young's modulus and thickness respectively of the mass layer 135. For example, for a (multi-layer) silicon nitride membrane stack having a thickness of 3 microns and a gold mass layer having a thickness of 1 micron, the effective Young's modulus compared to a membrane stack only containing the silicon nitride layer(s) to a thickness of 3 microns is increased by 20%.

In an embodiment, the respective layer compositions of the first layer stack 130 and the second layer stack 130' are chosen such that the membranes of the first group Gl of CMUT cells 100 have a first spring constant (effective Young's modulus) and the membranes of the further groups G2-G4 of CMUT cells 100' have a second spring constant (effective Young's modulus) that is no more than 20% different to the first spring constant. More preferably, the membranes of the further groups G2-G4 of CMUT cells 100' have a second spring constant that is no more than 10% different to the first spring constant. In some embodiments, the first spring constant and the second spring constant are approximately the same, i.e. are less than 1 % different to each other. This is particularly advantageous where the CMUT cells 100 and 100' are typically addressed by a single control signal, e.g. a single bias voltage to simultaneously drive the respective CMUT cells 100, 100' of the ultrasound transducer array 10 into collapse mode, as the comparable spring constants of the CMUT cells 100 and 100'ensure that all cells of the transducer element can remain collapsed onto the substrate 101 during operation of the transducer element, e.g. in a transmission mode or a reception mode of the ultrasound transducer array 10, i.e. exhibit comparable dynamic 'collapse' behaviour.

More specifically, whilst the respective spring constants of the flexible membranes of the CMUT cells 100, 100' of the different groups G1-G4 vary by less than 20% as explained above, the mass of the flexible membranes of the CMUT cells of a group within said plurality of groups increases with increasing distance of said group from the central region, i.e. in the elevation direction E of the CMUT array 10. In other words, the central region of the CMUT array 10 may be defined as the region in between the

neighboring groups Gl of CMUT cells 100, with the mass of the flexible membranes of the CMUT cells 100, 100' increasing from group to group with increasing distance of the group in the elevation direction E from this central region. This may for example be achieved by increasing the thickness of the mass layer 135 between the further groups G2-G4 in the elevation direction E and/or by using higher density materials for the mass layer 135 in order to increase the mass of the mass layer 135 for a further group G2-G4 at a greater distance from the central region of the ultrasound transducer array 10.

Any suitable 'heavy' material, i.e. a material having a relatively high density, such as a density of at least 7 g/cm³, preferably a density of at least 10 g/cm³, more preferably a density of at least 15 g/cm³ may be used as the mass layer 135. The mass layer preferably further has a Young's modulus of less than 200 GPa Dense materials having a Young's modulus of less than 100 GPa are particularly preferred. The Young's modulus may be determined (e.g. by room temperature) using a nano- indentation measurement as per ISO 14577 or ASTM E2546-07, for example. The material may be an elemental material or a composite or alloy material. The material for example may be a metal (or metal alloy) having a high density and a low Young's modulus, e.g. gold or platinum, with gold being considered particularly suitable. It is nevertheless reiterated for the avoidance of doubt that the mass layer 135 is not necessarily a metal or metal alloy; the use of any suitable material having a suitably high density and low Young's modulus may be contemplated.

At this point it is noted that four groups G1 -G4 or CMUT cells 100, 100' are shown by way of non-limiting example only and that the ultrasound transducer array 10 may have any suitable number of groups of CMUT cells 100, 100'.

As shown in FIG. 6, the second layer stack 130' may be identical to the first layer stack 130 but further include the mass layer 135. In such an embodiment, the spring constant of the second layer stack 130' will differ by a small amount from the spring constant of the first layer stack 130 due to the presence of the additional mass layer 135. For this reason, the thickness of the mass layer 135 preferably is kept as small as possible to minimize the impact on the bending stiffness, i.e. spring constant, of the second layer stack 130' .

However, it should be understood that the layer composition of the first layer stack 130 may be more substantially different to the layer composition of the second layer stack 130'; for example, the first layer stack 130 may comprise one or more layers of a dielectric material having a different thickness than the corresponding layers in the second layer stack 130' and/or the first layer stack 130 may comprise one or more layers of a dielectric material not present in the second layer stack 130' or vice versa. Although this may require a more involved manufacturing process, in this manner the respective bending stiffnesses, i.e. spring constants, of the first layer stack 130 and the second layer stack 130' including the mass layer 135 may be tuned to have a minimal difference, e.g. be

(approximately) identical. As a simple example, the first layer stack 130 may comprise a layer of a dielectric material to a first thickness and the second layer stack may comprise a layer of the same dielectric material to a second thickness that is smaller than the first thickness, wherein the second thickness may be chosen such that the combination of the layer of the dielectric material to the second thickness and the mass layer 135 exhibit a bending stiffness that is (approximately) the same as the bending stiffness of the layer of the dielectric material to the first thickness in the first layer stack 130.

FIG. 7 depicts the FEM-calculated frequency response of an ultrasound transducer array 10 of CMUT cells 100 all having the same membrane without a mass layer (dashed line), the frequency response of an ultrasound transducer array of CMUT cells 100' all having this membrane onto which a mass layer 135 consisting of a 1.5 micron thick gold layer is formed (thin solid line) and the frequency response of an ultrasound transducer array comprising an alternating pattern of such CMUT cells 100 and CMUT cells 100' (thick solid line). The frequency response was obtained by driving the respective transducer arrays in collapse mode with a stimulus comprising a DC bias voltage component and an AC RF component to the collapsed CMUT cells of the respective transducer arrays.

As can be seen from FIG. 7, the application of the mass layer 135 shifts the peak resonance frequency of the CMUT cells 100 from about 13 MHz to about 7MHz for the CMUT cells 100' including the mass layer 135. The combination of the CMUT cells 100 and the CMUT cells 100' in the single ultrasound transducer array identified by the thick solid line exhibits a frequency response having a bandwidth that is nearly doubled compared to the ultrasound transducer arrays comprising CMUT cells 100 or CMUT cells 100' only. This clearly demonstrates that the presence of the first group of CMUT cells 100 as well as the second group of CMUT cells 100' within a single array element 99 improves the broadband characteristics of this array element. More specifically, due to the spatial orientation of the CMUT cells 100, 100' of the different groups G1-G4 within such an array element 99, the array element 99 acts as a Hanfy lens due to the spatial distribution of the respective peak resonance frequencies of the CMUT cells 100, 100' across the array element 99 as previously explained.

FIG. 8 is a contour plot of the output pressure of a transducer array 10 solely comprising the first group of CMUT cells 100 (left pane) and of a transducer array 10 solely comprising the second group of CMUT cells 100' including the mass layer 135 in their respective membranes (right pane). The horizontal axis depicts the applied bias voltage [V] component and the vertical axis depicts the generated pulse length [ns] by the applied RF frequency component. The lower dashed line horizontally extending through both contour plots denotes the central frequency of the first group of CMUT cells 100 and the upper dashed line horizontally extending through both contour plots denotes the central frequency of the second group of CMUT cells 100'. As indicated by the block arrow, the central frequency of such CMUT cells may be effectively shifted to a lower value by the inclusion of a mass layer 135 in such cells, thereby demonstrating that the broadband characteristics of an array element 99 may be increased by the inclusion of multiple groups of CMUT cells within the array element, wherein only some of the groups have membranes including such a mass layer 135.

More generally speaking, where the mass of the flexible membranes of the CMUT cells 100, 100' increases from group to group, i.e. from the flexible membranes of the CMUT cells 100 in group Gl having the lowest mass to the flexible membranes of the CMUT cells 100' in group G4 having the highest mass, a spatial distribution of peak resonance frequencies of the CMUT cells 100, 100' in the respective groups G1-G4 is achieved in which the highest peak resonance frequency is exhibited by the CMUT cells 100 of group Gl, with the peak resonance frequencies of the CMUT cells 100' in the groups G2- G4 decreasing with increasing distance from the central region of the ultrasound transducer array. Formulaicly, M(G1)<M(G2)<M(G3)<M(G4), whereas f(Gl)>f(G2)>f(G3)>f(G4), in which M is the mass and f is the peak resonance frequency of the flexible membranes of the CMUT cells 100, 100' of the identified group.

Consequently, because the spring constants of the flexible membranes of the CMUT cells 100, 100' are substantially the same, i.e. differ by less than 20%, preferably by less than 10%, more preferably by less than 1%, between the various groups G1-G4 of

CMUT cells 100, 100', all the CMUT cells can be simultaneously driven into collapse mode by a single bias voltage, i.e. the same bias voltage may be simultaneously applied to all CMUT cells 100, 100' in order to improve the resolution (image quality) of the ultrasound transducer array 10 as will be explained in further detail below.

As will be understood, the number of groups G of CMUT cells 100' that differ from each other in terms of the mass of the flexible membranes of these CMUT cells as previously explained to approximate the continuous spatial focal depth variation across the elevation of a traditional Hanafy lens, i.e. across the extension direction E of each array element 99, with a larger number of groups G providing a closer approximation of such a lens, at the cost of a more involved manufacturing process. Therefore, a trade-off may be chosen between the closeness of the approximation of a Hanafy lens and the manufacturing complexity of the ultrasound transducer 10 according to embodiments of the present invention.

FIG. 9 schematically depicts a cross-section of another embodiment of a CMUT cell 100', in which the mass layer 135 is a patterned layer. This for example may be advantageous to tune the mass of the flexible membranes of the CMUT cells 100' in the various groups G2-G4. Rather than having to deploy mass layers 135 of different thicknesses to achieve such different masses, the respective mass layers 135 may be formed to a single thickness, with the mass layers 135 of the CMUT cells 100' of the second sub group being patterned in order to reduce the mass of these layers. As will be immediately apparent to the skilled person, the pattern density and/or shape may be controlled in order to tune the reduction in mass of the mass layers 135. The patterned layer in some embodiments may be a pattern of concentric rings or a radial pattern of spokes or strips radially extending from the center of the membrane upper surface. A particular pattern may be chosen to impart a desirable stress profile onto the membrane, as will be readily understood by the skilled person.

In an alternative embodiment, which is schematically depicted in FIG. 10, at least some of the groups of CMUT cells 100' have a larger (membrane) diameter. The use of groups G1-G4 of CMUT cells 100, 100' having different (membrane) diameters provides additional design freedom when adding the mass layer 135 to selected group(s) of the CMUT cells. This may be used to compensate for the fact that during collapse mode, the center frequency of CMUT cells becomes largely independent of membrane radius. This can be understood from Equation (1), where the collapse radius Rc for a small radius membrane is much smaller than for a large radius membrane, such that the term (1-x²) compensates the term R² in equation (1), thus leading to the center frequency becoming largely insensitive to the membrane radius when the CMUT cells are operated in collapse mode. It is preferred that, in particular in ultrasound imaging applications, the maximum membrane diameter does not extend beyond the minumum pitch for the CMUT 100, 100' required for imaging quality purposes, in order to safguard the quality of the ultrasound image produced with the ultrasound transducer 10. It should furthermore be understood that in FIG. 10 all the CMUT cells 100, 100' are stacked in a linear manner, e.g. in a cubic packing by way of non-limiting example. It is for instance equally feasible to stack at least the innermost groups of CMUT cells 100, 100', i.e. the cells having the smallest diameters, in a staggered manner, e.g. a hexagonal closest packing to increase signal density produced with the central regions of the one or more array elements 99.

In the foregoing embodiments, the ultrasound transducer 10 is embodied as a (pseudo) 1-D array for 2-D ultrasound imaging, in which the groups G1-G4 are arranged along the extension direction E of the one or more array elements 99. However, it should be understood that embodiments of the present invention are not limited to such ultrasound transducers. FIG. 1 1 schematically depicts a top view of an ultrasound transducer 10 in which the further groups G2-G4 of CMUT cells 100' envelope or surround the first group Gl of CMUT cells 100 proximal to the central region of the ultrasound transducer array 10. Such a 2-D ultrasound array 10 may have any suitable outline, e.g. a polygonal outline such as a rectangular outline as shown in FIG. 1 1, a circular outline and so on.. Specifically, although this enveloping arrangement of groups G1-G4 is shown for a rectangular array element 99, it should be understood that it is equally feasible for the groups G1-G4 to define circular patterns, e.g. rings of CMUT elements 100, 100' enveloping each other (apart from the innermost circular pattern of CMUT cells 100 of course). FIG. 13 schematically depicts a non-limiting example embodiment of a CMUT transducer array manufacturing method in which a mass layer 135 is included in some of the flexible membranes of the CMUT cells 100, 100' as previously explained. The method proceeds in step (a) with the provision of a substrate 101, which may be any suitable substrate such as a silicon substrate, a silicon-on-insulator substrate, a silicon germanium substrate, a gallium nitride substrate and so on. A silicon-based substrate may for instance be used in a CMOS manufacturing process. The substrate 101 may comprise several structures, such as semiconductor devices, a metallization stack interconnecting the semiconductor devices and/or the CMUT cells, a passivation stack over the metallization stack and so on. The substrate 101 may for instance be the substrate of an application specific integrated circuit (ASIC) including the CMUT cells 100, 100' on its layer stack, e.g. passivation and/or planarization stack, wherein the CMUT cells may be connected to signal processing circuitry on the substrate 101 by the metallization stack. The provision of such substrates 101 is well- known per se and belongs to the routine skills of the skilled artisan such that the provision of suitable substrates 101 will not be discussed in further detail for the sake of brevity only.

A first electrode 1 10 is formed on the substrate 101, which electrode may be formed from any suitable electrically conductive material, e.g. metals or metal alloys, doped semiconductor materials such as doped poly-silicon, (semi)conducting oxides and so on. It is for instance particularly advantageous to use metals that are readily available in the manufacturing technology of choice, as this requires minimal redesign of the manufacturing flow, which is attractive from a cost perspective. For example, in a CMOS process, conductive materials such as Al, W, Cu, Ti, TiN, Mo and so on, as well as combinations of such materials, e.g. AlCu, AINd, AISi, TiW may be used to form the first electrode 1 10. This may include layer stacks of such conductive materials, such as for example an AINd/TiW layer stack for such an electrode.

The formation of the first electrode 1 10 may form part of the formation of a first electrode arrangement over the substrate 101, which first electrode arrangement includes the respective first electrodes 1 10 of the CMUT cells 100, 100' .

The first electrode 1 10, and the substrate 300 may subsequently optionally be covered by an electrically insulating (dielectric) material layer 1 1 1. This is shown in step (b). Such a dielectric layer 1 1 1 for instance may be used to electrically insulate the first electrode 1 10 from its counter electrode 120 (see below) to reduce the risk of short circuits between the electrodes during the operation of the CMUT cell. In addition, the dielectric layer 1 1 1 may be used to protect the first electrode 1 10 and the substrate 101 from damage during the removal of the sacrificial material to form the cavity over the first electrode 1 10.

Although the dielectric layer 1 1 1 is shown to cover the entire substrate surface 101, it is equally feasible to provide a patterned dielectric layer 1 1 1 in which only certain parts of the substrate 101 together with the first electrode 1 10 are covered by the dielectric layer 1 1 1. Any suitable dielectric material may be used for the protection of the first electrode 1 10 and the substrate 101, e.g. one or more materials selected from silicon nitride (S13N4), silicon oxide (S1O2), aluminium oxide (AI2O3), hafnium oxide (Hf0₂) or the like, although it is emphasized that the suitable dielectric materials are not limited to these example materials, although ALD-deposited S1O2 is specifically mentioned. In addition, mixtures or laminates of the aforementioned dielectric materials may be used for the protection of the first electrode 1 10. As such a dielectric layer 1 1 1 may be formed in any suitable manner, e.g. using suitable deposition techniques such as ALD, (PE)ALD, PLD, PVD, LPCVD and PECVD, its formation will not be explained in further detail for the sake of brevity.

In step (c), a sacrificial material is formed, e.g. through a suitable deposition technique, on the substrate 101 including the first electrode 1 10 and the optional dielectric layer 1 1 1. The sacrificial material is patterned to include a first region 1 12 over the first electrode 1 10 from which the cavity is formed and may further comprise a second region 1 12' outside the intended cavity area acting as a channel through which the sacrificial material may be removed. The height of the sacrificial material layer corresponding to the gap height of the cavity to be formed is typically in the range of 100-1,000 nm although it should be understood that values outside this range may also be contemplated.

In an embodiment, the first region 1 12 is a circular region with the second region 1 12' extending from the first region 1 12 in the form of one or more teeth-like protrusions, e.g. 2-8 of such protrusions. A top-view of such a sacrificial material portion is shown in step (c'), in which four of such protrusions are shown by way of non-limiting example only. The teeth-like second regions 1 12' are typically used as cavity access platforms outside the membrane to be formed through which access to the first portion 1 12 can be provided for opening or releasing the cavity.

In principle, any suitable sacrificial material may be used, although for device performance reasons it is preferable to use sacrificial materials that can be effectively removed in a subsequent etching step. For instance, the use of metals such as Al, Cr and Mo, Ti and (Ti)W or non-metals such as amorphous silicon or silicon oxide may be contemplated. Materials such as Al, amorphous silicon and silicon oxide are for instance readily available in CMOS processes, and of these materials Al can be particularly effectively removed by etching. The patterned sacrificial material may be formed in any suitable manner, e.g. using suitable deposition and patterning techniques and its formation will not be explained in further detail for the sake of brevity.

It will be understood that the diameter of the first region 1 12 defines the diameter of the cavity of a CMUT cell 100, 100' to be formed. In an embodiment, the diameter is selected in a range of 20-500 micron, more preferably in a range of 50-300 micron, although it should be understood that larger diameters may also be contemplated, e.g. diameters up to 1,000 micron.

In step (d), a first dielectric layer 131 of the membrane to be formed is deposited over the first region 1 12 and the second region 1 12' of the sacrificial material and the exposed portions of the dielectric layer 1 1 1 if present. As the first dielectric layer 131 and the dielectric layer 1 1 1 are both exposed to the etch recipe for removing the sacrificial layer, the first dielectric layer 131 and the dielectric layer 1 1 1 may be of the same material, although it is of course also plausible to use different materials for the first dielectric layer 131 and the dielectric layer 1 1 1 respectively.

In an embodiment, the first dielectric layer 131 and the dielectric layer 1 1 1 each comprise at least one layer formed any suitable dielectric material, such as a silicon oxide layer, e.g., S1O2, a silicon nitride layer, e.g., S13N4 or the like, an aluminium oxide (A1₂0₃) layer, a hafnium oxide (Hf0₂) layer and so on. Many other suitable dielectric layer materials will be apparent to the skilled person. Preferably, deposition techniques such as PECVD and ALD are used to form the dielectric layers as these techniques can be performed at temperatures below 400°C, which makes them compatible with CMOS manufacturing processes. The first dielectric layer 131 may be formed as a layer stack, e.g. an oxide-nitride stack or an oxide-nitride-oxide stack. Similarly, the optional dielectric layer 1 1 1 may be formed as such a stack. It is reiterated that any suitable dielectric material may be used for the optional dielectric layer 1 1 1 and the first dielectric layer 131. In addition, mixtures or laminates, e.g. ALD laminates, of the aforementioned dielectric materials may be used for these dielectric layers.

After the formation of the first dielectric layer 131, the second electrode arrangement including the second electrodes 120 is formed on the first dielectric layer 131 as shown in step (e) such that each second electrode 120 is oriented opposite a first electrode 1 10. The second electrode arrangement preferably is formed of the same electrically conductive material as the first electrode arrangement, although it should be understood that the second electrode arrangement and the first electrode arrangement alternatively may be formed of different materials. The second electrode arrangement may for instance be formed from any suitable electrically conductive material such as Al, W, Cu, Ti, TiN and so on, as well as combinations of such materials. The second electrode arrangement may be formed using well-known techniques that are not further explained for the sake of brevity only. The first electrode arrangement including the first electrodes 1 10 and the second electrode arrangement including the second electrodes 120 may be formed to any suitable thickness, e.g. 50-250 nm thickness. Other suitable thicknesses may be contemplated, e.g. depending on the application domain.

After the formation of the second electrode 120, the method proceeds as shown in step (f), in which the second dielectric layer 133 is formed. In an optional embodiment, the second dielectric layer 133 is formed to a first thickness tl, which exceeds the thickness of the first portion 1 12 of the sacrificial material in between the first electrode 1 10 and the second electrode 120 such that upon formation of the cavity 130 the height g of the cavity gap is substantially smaller than the thickness tl, i.e. g/tl « 1. Preferably tl > 3g, more preferably tl > 5g. This ensures that during the release of the cavity 105 in step (g), i.e. by formation of the access or via 1 16 and the subsequent removal of the first portion 1 12 and the second portion 1 12' of the sacrificial material, the membrane exhibits excellent membrane robustness during the cavity release step as g«tl at the stage of removal of the sacrificial material to form the cavity 105. Moreover, because the second dielectric layer 133 is formed, e.g. deposited, prior to the release of the cavity 105, a membrane with excellent flatness characteristics is obtained as the presence of the sacrificial material prevents deformation of the first dielectric layer 131 during the formation of the second dielectric layer 133.

The first portion 1 12 and the second portions 1 12' of the sacrificial material are subsequently removed as shown in step (g) by the formation of the access or via 116 using a suitable etch recipe to form the cavity 105 in between the first electrode 1 10 and the second electrode 120 embedded in between the first dielectric layer 313 and the second dielectric layer 315 of the membrane layer stack 130 of the CMUT cells 100, 100'. Suitable etch recipes for such conventional sacrificial materials are well-known per se and the skilled person will have no difficulty selecting an appropriate etch recipe using his common general knowledge. The thickness of the layer stacks 130 may be further increased during the sealing of the access or via 1 16 in step (h) by the formation of the further dielectric layer 134 including the plug 1 18 in the access or via 1 16. The further dielectric layer 134 may be substantially thinner than the second dielectric layer 133. The further dielectric layer 134 may be formed to a thickness of at least twice the height, e.g. 3-4 times the height, of the cavity 105 to effectively seal the access or via 1 16.

In step (i), a mass layer 135 is applied to selected membrane layer stacks 130, e.g. of the CMUT cells 100' in groups G2-G4, in order to form the membrane layer stacks 130' of the further groups of CMUT cells 100' . Such a mass layer 135 may be selectively applied in any suitable manner, e.g. by applying a mask layer across the membrane layer stacks 130, and patterning this mask layer to expose the membrane layer stacks 130 to which the mass layer 135 is applied, after which application the mask layer may be removed. Other suitable techniques for selectively applying the mass layer 135 will be immediately apparent to the skilled person. Although not explicitly shown, in order to achieve the increasing mass of the membrane stacks of the CMUT cells 100' going from group G2 to group G4 as previously explained, the selected mass layers 135 may be patterned in patterns having different pattern densities as previously explained. Alternatively or additionally, the thickness and/or the density of the mass layer 135 may be increased going from group G2 to group G4 in order to increase the mass of the membrane stack of the CMUT cells 100' between these different groups, e.g. by using different density materials for the mass layer 135.

At this point it is emphasized that steps (a)-(i) schematically depict an advantageous but non-limiting example of forming one or more CMUT cells 100, 100' on a substrate 101. Many alternative routes will be apparent to the skilled person. It is furthermore noted that although in this example embodiment the CMUT cells 100 in the first group Gl of CMUT cells do not comprise a mass layer 135 in their membrane stacks, it is of course equally feasible to also include such a mass layer 135 in the CMUT cells 100, as long as the mass m of the flexible membranes in the various groups remains ordered as m(Gl) < m(G2) <m (G3) < m(G4) such that the peak resonance frequency f remains ordered as f(Gl) > f(G2) > f(G3) > f(G4) as explained above.

Another notable process variation is that the access or via 1 16 may be sealed in any suitable manner using any suitable material, e.g. by depositing and patterning a dedicated sealing layer such as a metal or dielectric layer to form the plug 1 18. It is furthermore noted that the cavity 105 may be released at any suitable point in the CMUT manufacturing process, e.g. prior to the formation of the second dielectric layer 133. Yet another notable process variation is that the first electrode 1 10 and/or the second electrode 120 may or may not be separated from the cavity 130 by a dielectric layer, as this is a typical design choice. As previously mentioned, a dielectric layer, i.e. an electrically insulating layer, may be provided over the first electrode 1 10 and/or the second electrode 120 to prevent direct contact between the first electrode 1 10 and the second electrode 120 during operation of the CMUT cells 100, 100' and 100" if present.

The membrane stacks 130 may be formed in any suitable manner, e.g. by a single dielectric layer rather than a stack of dielectric layers, and so on. Such process choices fall within the routine skills of the skilled person and will therefore not explicitly mentioned in detail for the sake of brevity only. As previously mentioned, not all CMUT cells 100, 100' may have the same membrane stack 130. It is equally feasible to, in addition to the presence of the mass layer 135 in the groups of CMUT cells 100', to have different layers and/or different layer thicknesses between the layer stack 130 and 130' respectively.

Also, it should be understood that alternative designs of the individual CMUT cells 100 are of course equally feasible. The design of the CMUT cells 100 is not particularly relevant to the present invention, and any suitable design of the cells may be contemplated; for example, 3-electrode CMUT cells 100 in which an intermediate electrode is located between the bottom electrode 1 10 and the cavity 105 are equally feasible. Such 3-electrode CMUT cells for instance may be contemplated to provide a stimulus and a bias voltage through separate electrodes, e.g. to reduce the risk of membrane sticking to the bottom of the CMUT cell.

At this point it is further noted that although not shown in the various embodiments, it should be understood that the CMUT cells 100, 100' and 100" if present manufactured in accordance with embodiments may comprise additional circuit elements, which may be integrated on the substrate 101 or may be provided on a separate substrate and integrated into a single package with one or more of the CMUT devices from a wafer manufactured in accordance with embodiments of the present invention. Such additional circuitry may be instance be an IC, e.g. an ASIC, for controlling the one or more CMUT cells 100, 100' and/or processing the signals generated by the one or more CMUT cells 100, 100' if present, e.g. to control transmission and/or reception modes of the one or more CMUT cells 100, 100'. Other suitable embodiments of the CMUT cells 100, 100' and/or an ultrasound transducer array 10 comprising such cells will be immediately apparent to the skilled person. It is furthermore noted that in the aforementioned manufacturing process, a wafer processed during the manufacturing process may contain a single die, i.e. a single device, in which case the substrate 101 corresponds to the wafer, or a plurality of dies that may be singulated in any suitable manner, e.g. diced, after the completion of the device manufacturing process, in which case the substrate 101 corresponds to a part of the wafer, an example embodiment of which will be described in more detail below.

It is reiterated at this stage that embodiments of the present invention are not limited to ultrasound transducers 10 having one or more array elements 99 in which the peak resonance frequencies of the CMUT cells 100, 100' within the array element 99 are tuned using mass layers 135 as explained in more detail above. However, as the process flows for the manufacture of CMUT cells 100, 100' is well-known per se, these flows are not explained in further detail for the sake of brevity only for those embodiments in which the peak resonance frequencies of the CMUT cells 100, 100' within the array element 99 are tuned by systematically varying at least of the gap height of the cavity 105, membrane thickness, the thickness of a dielectric layer between the electrodes of the CMUT cells, cell diameter and so on, whilst ensuring that the collapse voltage of such different CMUT cells 100, 100' within each array element 99 remains substantially the same, as the implementation of the process variations to facilitate such different CMUT cells 100, 100' falls within the routine skill person of the skilled person.

FIG. 14 is a plan view of a two dimensional transducer array 10 of the previously described groups of circular CMUT cells 100, 100' according to an embodiment. The array is configured in the conventional manner of symmetrically aligned rows 56 and columns 58 of CMUT elements 100, 100', with the rows 56 defining the various groups Gl - G4 of CMUT elements 100, 100' as previously explained, and the columns defining one or more array elements 99. In this example each column 58 is covered with an integral flexible foil containing embedded metal tracks, which allow the columns to be bent in a cylindrical shape. The flexible foil will be described in greater detail below. In this example the array is dimensioned to have the same pitch in both the row and column directions, as indicated by arrow 52 which denotes the pitch in the column direction and arrow 54 which denotes the pitch in the row direction. The columns 58 may be formed as separate substrate islands, e.g. separate silicon islands or strips, such that the transducer array 10 may be wrapped around a three dimensional body, e.g. a cylindrical body of an invasive medical device such as an external sheath of a catheter such that the array elements 99 extend along the elongation direction of such a three dimensional body, i.e. have their extension directions E aligned with the elongation direction.

FIG. 15 is a plan view of a two-dimensional ultrasound transducer array 10 that is configured in accordance with a preferred embodiment. As shown in FIG. 10, the rows 56 and columns 58 of the groups G1-G4 of CMUT elements 100, 100' are staggered in alignment, as is well-known per se. The staggered alignment in this example is

accommodated by increasing the spacing 55 between CMUT cells 100, 100' in the columnar direction, which enables adjacent columns and rows to be further interspersed with each other. In an embodiment, the spacing 55 is at least the diameter D of a CMUT cell 100, 100'. In such a transducer array 10, the pitch between CMUT cells 100, 100' within a single group Gl, G2, G3 or G4 preferably is constant such that the array exhibits uniform broadband characteristics over the total transducer area of the array.

In the illustrated example the CMUT cells 100, 100' are so tightly interspersed that a tangential line from cell to cell in the column or row direction would actually intersect a cell of the adjacent staggered row or column. The interspersion of CMUT cells 100, 100' allows for an increase in the density of CMUT elements within the transducer array 10 without requiring an increase in the vertical spacing (i.e. in the column direction) between CMUT cells 100, 100', at least up to the point where the closest packing of such cells is achieved. Beyond this point, the pitch between successive CMUT cells 100, 100' in the columnar direction may be increased as arrow 55 shows to facilitate a further decrease in horizontal spacing, as indicated by arrows 57 and 59, but this will reduce the overall CMUT density of the transducer array 10. At least at its closest packing, the CMUT transducer array 10 of FIG. 15 has a greater cell density than the CMUT transducer array 10 of FIG. 14.

In FIG. 15, each column 58 of CMUT cells 100, 100' is located on a separate substrate island, i.e. a separate piece of silicon die. As before, each column 58 may carry at least one array element 99. The respective substrate islands, e.g. silicon islands, are characterized by having a meandering edge structure in the length direction of, i.e. along, the columns 58, with edge portions 58A meandering outwardly around CMUT cells 100, 100' and edge portions 58B meandering inwardly into a space between neighboring CMUT cells 100, 100' in a column 58. In other words, the columns 58 have wave-shaped opposing edges in the column direction where the wave peaks coincide with the CMUT cells 100, 100' and the wave valleys coincide with the spacings 55 in between the CMUT cells 100, 100' .

A neighboring column 58 is arranged such that an outwardly meandering edge portion of its silicon island aligns with, i.e. slots into, an inwardly meandering edge portion of a neighboring silicon island, thereby forming the staggered rows of CMUT cells 100, 100' by the staggered alignment of the CMUT cells 100, 100' between neighboring columns 58. Neighboring silicon islands are typically separated by a gap 57, to facilitate out-of-plane bending of the silicon islands respective to each other, e.g. when wrapping the CMUT transducer array 10 around a three dimensional body of an invasive medical device such as a cylindrical body, e.g. a catheter sheath.

In order to retain the relative positions of the substrate islands respective to each other, the CMUT transducer array 10 further comprises a flexible foil 60 onto which the substrate islands are mounted. The flexible foil 60 for instance may comprise a so-called flex -to-rigid foil in which a metal layer or metal layer stack, e.g. metal tracks, is embedded in or covered by a polymer layer or polymer layer stack, which polymer typically is electrically insulating in order to protect the metal layer from accidental short circuits. A non-limiting example of a suitable polymer for such a flexible foil 60 is polyimide, as it is well-known per se that polyimide is compatible with many semiconductor manufacturing processes such as a CMOS manufacturing process. Other suitable polymers, e.g. parylene, will be immediately apparent to the skilled person. A non-limiting example of a suitable metal is aluminium or any other metal commonly used in semiconductor manufacturing processes. The

compatibility of such materials with existing semiconductor manufacturing processes facilitates the manufacture of the CMUT transducer array using existing semiconductor manufacturing processes rather than having to redesign or redevelop such manufacturing processes, which would increase the cost of the CMUT transducer array 10.

The provision of the CMUT cells 100, 100' on a plurality of adjacent meandering substrate islands interconnected via a flexible foil 60 allows for the out-of-plane bending of the CMUT transducer array 10 in the row direction of the array whilst providing structural integrity in the column direction of the array, which for example is particularly advantageous when wrapping the array around an invasive medical device such as a catheter, e.g. an intra-vascular catheter or an intra-cardiac catheter. For example, the CMUT transducer array 10 may be wrapped around the external sheath of such a catheter, with the silicon island columns 58 aligning in a length direction of the catheter, i.e. CMUT transducer array 10 being bent out-of-plane and is wrapped around the catheter sheath in its row direction. Due to the provision of a large number of relatively narrow silicon islands, a near- cylindrical configuration of the CMUT transducer array 10 may be achieved when wrapping the CMUT transducer array 10 around a cylindrical body such as a catheter sheath, with the further advantage that such a CMUT transducer array is continuous over the entire surface of such a body, e.g. does not contain discontinuities between adjacent rectangular silicon islands forming part of the CMUT transducer array, as for instance is the case in EP 2 455 133 Al .

In accordance with a further aspect of the present invention, advantage is taken of this decreased spacing 57 and 59 by operating the array of FIG. 10 so that a row (group) of CMUT cells 100, 100' is not a horizontal row 56 of cells but an interspersed combination of two (or more) adjacent staggered rows. In the example of FIG. 15, a group G1-G4 of elements 100, 100' is formed by staggered rows of elements. For instance, one group in FIG. 15 comprises transducer elements 62i, 62₂, 62₃, 62₄, ... 62N of two adjacent staggered rows, i.e. the M¹ group comprises the M^th CMUT element 50 of each column 58 of CMUT cells 100, 100', with M being a positive integer, with each group typically forming a meandering annular row when wrapped around an invasive medical device such as a catheter. The tighter spacing of the staggered rows enables a group of 96 cells to be provided where the standard symmetrical alignment would only accommodate 64 cells, for instance.

In the embodiment of FIG. 15, the respective substrate, e.g. silicon, islands are retained by a continuous flexible foil 60. In an alternative embodiment, the flexible foil 60 may be patterned such that the flexible foil 60 comprises a plurality of recesses aligning with the gaps 57, with respective bridge portions or bridges extending across the gaps 57 to interconnect different regions of the flexible foil 60, e.g. different regions retaining different substrate (silicon) islands. This further increases the flexibility of the CMUT transducer array 10 but may be less robust.

FIG. 16 illustrates several process steps in the formation of a flexible foil bridge joining two silicon islands on which CMUT cells 100, 100' are located. FIG. 16(a) shows a silicon wafer 70 with thermal silicon dioxide layers 72 grown on the top and bottom sides. Patterned aluminum areas 81 are sputtered on the top side using standard lithography. A patterned area of polyimide 74 is laid over one of the aluminum areas on the top side, which pattern defines the bridges in the flexible foil. In case of a continuous flexible foil, the polyimide 74 may be a continuous sheet. An aluminum layer 80 is deposited over the polyimide 74 and a second polyimide layer 76 is laid over the aluminum. Another layer of aluminum 82 is patterned over the aluminum layer 80 for use as a mask during etching, all as shown in FIG. 16(b).

Finally, as shown in FIG. 16(c), the silicon wafer 70 is etched away from the back in areas outside the masked by thick resist areas 84, both under CMUT location 88 and under the flexible bridge 74, 80, 76. The polyimide layer 76 on either side of the flexible bridge 90 on the top side is patterned away on either side of etch mask layer 82, which is then itself etched away. The result is two separate silicon islands 92 and 94, joined by a flexible bridge 90. The flexible bridge 90 and others like it enable an array of such CMUT-populated islands to be wrapped in a cylindrical shape, fitting the needs of an intra-cardiac catheter transducer.

As previously mentioned, a group 200 of acoustic transducer elements 50 may be formed, not by a straight line of transducer elements as in the conventional manner, but by two or more adjacent staggered rows 202 and 204 of CMUT cells 100, 100'. FIG. 17 schematically depicts an alternative embodiment of an ultrasound transducer array 10 in which each silicon island column 58 comprises a pair of array elements 99 comprising CMUT cells 100, 100' arranged in a staggered arrangement, i.e. the area of a CMUT cell

100, 100' in a first column extends into the space between neighboring CMUT cell 100, 100' in the neighboring column, preferably such that a tangent between these neighboring CMUT cells intersects the area of the CMUT cell 100, 100' extending into the space between these neighboring CMUT cells.

As before, the silicon island columns 58 have a meandering edge structure in the length direction of, i.e. along, the columns 58, with edge portions meandering outwardly around CMUT cells 100, 100' and edge portions meandering inwardly into a space between neighboring CMUT cells 100, 100' in a column 58. A neighboring column 58 is arranged such that an outwardly meandering edge portion of its silicon island aligns with, i.e. slots into, an inwardly meandering edge portion of a neighboring silicon island, thereby forming the staggered rows of CMUT cells 100, 100' by the staggered alignment of the CMUT cells 100, 100' between neighboring columns 58. Neighboring silicon islands are typically separated by a gap 57, to facilitate out-of-plane bending of the silicon islands respective to each other, e.g. when wrapping the CMUT transducer array 10 around an invasive medical device such as a catheter sheath as previously explained.

This embodiment has the advantage of providing larger, i.e. wider, silicon islands, which improves the structural rigidity of such islands, whilst still providing an ultrasound transducer array with excellent flexibility in the row direction. This embodiment is particularly advantageous where the circumference of a body of an invasive medical device, e.g. a catheter sheath, around which the transducer array is to be wrapped, is many times the width of a single silicon island, such that many silicon islands are to be wrapped around the body, and such that substantially continuous transducer rows are provided around the body, thereby providing a plurality of individually adressable array elements 99 each operating as a Hanafy lens, extending in the elongation direction of such a body. The separate silicon islands are overlaid with a flexible foil 60, e.g. a continuous foil as shown in the drawing or a patterned foil containing bridge portions 90 across the gaps between neighboring silicon islands, to retain the respective orientation of the silicon islands and to enable the two dimensional transducer array 10 to be bent into a cylindrical shape around a distal tip 210 of an invasive medical device 250 as shown in FIG. 18. It should be understood that the array is wrapped around the distal tip 210 by way of non- limiting example only; it is for instance equally feasible to wrap the transducer array 10 around any other part of the invasive medical device 250, even though it is preferred that the array 10 is located proximal to the distal tip of the catheter 250. The ultrasound transducer array 10 typically is wrapped around a flexible elongate body of the invasive medical device 250 that can be flexibly inserted into a vein, artery, digestive system and so on of a patient, such that the surroundings of the flexible elongate body can be imaged with the ultrasound transducer array 10, e.g. for guidance or diagnostic purposes.

In some embodiments, the invasive medical device 250 may comprise a further ultrasound transducer array (not shown) on the distal tip 210, e.g. a planar ultrasound transducer array having a circular circumference in addition to the wraparound ultrasound transducer array 10, such that the invasive medical device 250 can generate images of a body portion ahead of the invasive medical device as well as around the invasive medical device, which for instance is particularly advantageous in intra-cardiac imaging. In some

embodiments, the invasive medical device 250 therefore may be an intra-cardiac or intravascular catheter.

FIG. 19 schematically depicts an example embodiment of an ultrasonic diagnostic imaging system 1 with an ultrasound transducer 10, e.g. an array of ultrasound transducer element tiles (transducer elements) comprising multiple CMUT cells 100 and 100' arranged in groups G1-G4 across one or more array elements 99 as previously explained. The array 10 may form part of an ultrasound probe or an invasive medical device as previously explained. In FIG. 19, the transducer array 10 is provided for transmitting ultrasonic waves and receiving echo information. As previously explained, the transducer array 10 may be a one- or a two-dimensional array of ultrasound transducer element tiles capable of scanning in a 2D plane or in three dimensions for 3D imaging.

The transducer array 10 may be coupled to a microbeam former 12, which may be integrated in a probe or invasive medical device 250, which controls transmission and reception of signals by the ultrasound transducer cells 100, 100' (or clusters thereof).

Microbeam formers are capable of at least partial beam forming of the signals received by groups or "patches" of transducer element tiles for instance as described in US patents US 5,997,479 (Savord et al.), US 6,013,032 (Savord), and US 6,623,432 (Powers et al.)

The microbeam former 12 may be coupled by a probe cable, e.g. coaxial wire, to a terminal, e.g. a patient interface module or the like, comprising a transmit/receive (T/R) switch 16 which switches between transmission and reception modes and protects the main beam former 20 from high energy transmit signals when a microbeam former is not present or used and the transducer array 10 is operated directly by the main system beam former 20. The transmission of ultrasonic beams from the transducer array 10 under control of the microbeam former 12 may be directed by a transducer controller 18 coupled to the microbeam former by the T/R switch 16 and the main system beam former 20, which receives input from the user's operation of the user interface or control panel 38. One of the functions controlled by the transducer controller 18 is the direction in which beams are steered and focused. Beams may be steered straight ahead from (orthogonal to) the transducer array 10, or at different angles for a wider field of view. The transducer controller 18 may be coupled to control the voltage source 45 for the ultrasound transducer array 10. For instance, the power supply 45 sets the DC and AC bias voltage(s) that are applied to CMUT cells of a CMUT array 10, e.g. to operate the one or more CMUT cells 100, 100' of the CMUT elements in collapse mode, as is well-known per se. In a particularly advantageous embodiment, the power supply 45 is configured to supply all the CMUT cells 100, 100' of the one or more array elements 99 of the ultrasound transducer array 10 with the same stimulus including a bias voltage component and a RF frequency component, which bias voltage component drives the CMUT cells 100, 100' of such an array element 99 into collapse mode whilst the RF frequency component causes the CMUT cells 100, 100' to vibrate with different peak resonance frequencies. Where the ultrasound transducer array 10 comprises a plurality of such array elements 99, these array elements 99 preferably are individually addressable, e.g. to facilitate electronic beam steering using the different array elements 99 as is well-known per se.

The power supply 45 may optionally comprise separate stages for providing the DC and AC components respectively of the stimulus of the CMUT cells 100, e.g. in transmission mode. A first stage may be adapted to generate the static (DC) voltage component and a second stage may be adapted to generate an alternating variable voltage component having a set alternating frequency, which signal typically is the difference between the overall drive voltage, i.e. stimulus, and the aforementioned static component thereof. The static or bias component of the applied drive voltage preferably meets or exceeds the threshold voltage when forcing the CMUT elements into their collapsed states, i.e. when operating the CMUT elements in collapsed mode. This has the advantage that the first stage may include relatively large capacitors, e.g. smoothing capacitors, in order to generate a particularly low-noise static component of the overall voltage, which static component typically dominates the overall voltage such that the noise characteristics of the overall voltage signal will be dominated by the noise characteristics of this static component.

Other suitable embodiments of the power supply 45 should be apparent, such as for instance an embodiment in which the power supply 45 contains three discrete stages including a first stage for generating the static DC component of the CMUT drive voltage, a second stage for generating the variable DC component of the drive voltage and a third stage for generating the frequency modulation component of the signal, e.g. a pulse circuit or the like. It is summarized that the power supply 45 may be implemented in any suitable manner.

The partially beam-formed signals produced by the microbeam former 12 may be forwarded to the main beam former 20 where partially beam- formed signals from individual patches of transducer elements are combined into a fully beam-formed signal. For example, the main beam former 20 may have 128 channels, each of which receives a partially beam-formed signal from a patch of dozens or hundreds of ultrasound transducer cells 100, 100'. In this way the signals received by thousands of transducer cells of a transducer array 10 can contribute efficiently to a single beam- formed signal.

The beam-formed signals are coupled to a signal processor 22. The signal processor 22 can process the received echo signals in various ways, such as bandpass filtering, decimation, I and Q component separation, and harmonic signal separation which acts to separate linear and nonlinear signals so as to enable the identification of nonlinear (higher harmonics of the fundamental frequency) echo signals returned from tissue and microbubbles.

The signal processor 22 optionally may perform additional signal enhancement such as speckle reduction, signal compounding, and noise elimination. The bandpass filter in the signal processor 22 may be a tracking filter, with its passband sliding from a higher frequency band to a lower frequency band as echo signals are received from increasing depths, thereby rejecting the noise at higher frequencies from greater depths where these frequencies are devoid of anatomical information.

The processed signals may be forwarded to a B-mode processor 26 and optionally to a Doppler processor 28. The B-mode processor 26 employs detection of an amplitude of the received ultrasound signal for the imaging of structures in the body such as the tissue of organs and vessels in the body. B-mode images of structure of the body may be formed in either the harmonic image mode or the fundamental image mode or a combination of both for instance as described in US Patents US 6,283,919 (Roundhill et al.) and US 6,458,083 (Jago et al.)

The Doppler processor 28, if present, processes temporally distinct signals from tissue movement and blood flow for the detection of the motion of substances, such as the flow of blood cells in the image field. The Doppler processor typically includes a wall filter with parameters which may be set to pass and/or reject echoes returned from selected types of materials in the body. For instance, the wall filter can be set to have a passband characteristic which passes signal of relatively low amplitude from higher velocity materials while rejecting relatively strong signals from lower or zero velocity material.

This passband characteristic will pass signals from flowing blood while rejecting signals from nearby stationary or slowing moving objects such as the wall of the heart. An inverse characteristic would pass signals from moving tissue of the heart while rejecting blood flow signals for what is referred to as tissue Doppler imaging, detecting and depicting the motion of tissue. The Doppler processor may receive and process a sequence of temporally discrete echo signals from different points in an image field, the sequence of echoes from a particular point referred to as an ensemble. An ensemble of echoes received in rapid succession over a relatively short interval can be used to estimate the Doppler shift frequency of flowing blood, with the correspondence of the Doppler frequency to velocity indicating the blood flow velocity. An ensemble of echoes received over a longer period of time is used to estimate the velocity of slower flowing blood or slowly moving tissue.

The structural and motion signals produced by the B-mode (and Doppler) processor(s) are coupled to a scan converter 32 and a multiplanar reformatter 44. The scan converter 32 arranges the echo signals in the spatial relationship from which they were received in a desired image format. For instance, the scan converter may arrange the echo signal into a two dimensional (2D) sector-shaped format, or a pyramidal three dimensional (3D) image.

The scan converter can overlay a B-mode structural image with colors corresponding to motion at points in the image field with their Doppler-estimated velocities to produce a color Doppler image which depicts the motion of tissue and blood flow in the image field. The multiplanar reformatter 44 will convert echoes which are received from points in a common plane in a volumetric region of the body into an ultrasonic image of that plane, for instance as described in US Patent US 6,443,896 (Detmer). A volume renderer 42 converts the echo signals of a 3D data set into a projected 3D image as viewed from a given reference point as described in US Pat. 6,530,885 (Entrekin et al.)

The 2D or 3D images are coupled from the scan converter 32, multiplanar reformatter 44, and volume renderer 42 to an image processor 30 for further enhancement, buffering and temporary storage for display on an image display 40. In addition to being used for imaging, the blood flow values produced by the Doppler processor 28 and tissue structure information produced by the B-mode processor 26 are coupled to a quantification processor 34. The quantification processor produces measures of different flow conditions such as the volume rate of blood flow as well as structural measurements such as the sizes of organs and gestational age. The quantification processor may receive input from the user control panel 38, such as the point in the anatomy of an image where a measurement is to be made.

Output data from the quantification processor is coupled to a graphics processor 36 for the reproduction of measurement graphics and values with the image on the display 40. The graphics processor 36 can also generate graphic overlays for display with the ultrasound images. These graphic overlays can contain standard identifying information such as patient name, date and time of the image, imaging parameters, and the like. For these purposes the graphics processor receives input from the user interface 38, such as patient name.

The user interface is also coupled to the transmit controller 18 to control the generation of ultrasound signals from the transducer array 10 and hence the images produced by the transducer array and the ultrasound system. The user interface is also coupled to the multiplanar reformatter 44 for selection and control of the planes of multiple multiplanar reformatted (MPR) images which may be used to perform quantified measures in the image field of the MPR images.

As will be understood by the skilled person, the above embodiment of an ultrasonic diagnostic imaging system 1 is intended to give a non-limiting example of such an ultrasonic diagnostic imaging system. The skilled person will immediately realize that several variations in the architecture of the ultrasonic diagnostic imaging system are feasible without departing from the teachings of the present invention. For instance, as also indicated in the above embodiment, the microbeam former 12 and/or the Doppler processor 28 may be omitted, the ultrasound transducer array 10 may not have 3D imaging capabilities and so on. Other variations will be apparent to the skilled person.

The ultrasound transducer array 10 according to embodiments of the present invention may form part of an invasive medical device such as a catheter as previously explained. As will be readily understood by the skilled person, such a transducer array may form part of any type of investigative device, e.g. an ultrasound probe, an ultrasound matrix probe, an ultrasound catheter, an ultrasound needle, and so on. Such an ultrasound transducer array 10 may be used in any suitable ultrasound imaging technique.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design many alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps other than those listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several distinct elements. In the device claim enumerating several means, several of these means can be embodied by one and the same item of hardware. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

Claims

CLAIMS:

1. An ultrasound transducer (10) comprising:

a substrate (101); and a plurality of CMUT (capacitive micromachined ultrasound transducer) cells (100, 100') formed on the substrate, each having substantially the same collapse voltage threshold and comprising:

a first electrode (1 10) coupled to the substrate;

a flexible membrane (130, 130') at least partailly separated from the substrate by a cavity; and

a second electrode (120) coupled to the membrane;

wherein the CMUT cells are arranged to form at least one array element (99), wherein within each array element:

the CMUT cells are arranged to be simultaneously activated by a single stimulus; and

at least some of the CMUT cells have different peak resonance frequencies, and each CMUT cell having a lower peak resonance frequency has a larger distance to the centre of the array element in an extension direction of the array element than any CMUT cell having a higher peak resonance frequency.

2. The ultrasound transducer (10) of claim 1, comprising a plurality of said array elements (99), wherein the respective array elements are arranged to be individually activated by a single stimulus and wherein the CMUT cells are arranged in a plurality groups (G1-G4) extending across the plurality of array elements, and wherein the CMUT cells in a group have the same peak resonance frequency.

3. The ultrasound transducer (10) of claim 2, wherein the CMUT cells (100, 100') are arranged in a plurality of staggered rows extending in said extension direction, wherein each group (G1-G4) extends across each of said staggered rows.

4. The ultrasound transducer (10) of claim 2 or 3, wherein the CMUT cells (100,

100') are arranged in a plurality of columns (58) and each group (G1-G4) extends across each of said columns, the ultrasound transducer further comprising:

a plurality of silicon islands each carrying at least one of said columns; and a flexible foil (60) retaining the respective silicon islands, the flexible foil comprising conductive interconnects.

5. The ultrasound transducer (10) of claim 2, wherein each further group (G2- G4) envelopes a group of said plurality of groups.

6. The ultrasound transducer (10) of any of claims 1 -5, wherein different CMUT cells (100, 100') within an array element exhibit at least one of:

- a different flexible membrane diameter;

- a different gap height in the absence of the single stimulus and a different membrane thickness; and

- a different thickness of a dielectric layer between the first electrode and the second electrode.

7. The ultrasound transducer of any of claims 1-6, wherein each of the flexible membranes (130, 130') comprises a layer stack, and wherein:

the CMUT cells (100, 100') within each array element (99) comprises a first CMUT cell (100) proximal to the central region and a plurality of further CMUT cells (100') at a greater distance from the central region than the first CMUT cell, wherein the layer stacks of the flexible membranes of at least some of the further CMUT cells include a layer (135) of a material having a higher density than any of the layers in the layer stack of the flexible membranes of the first CMUT cell; and

respective spring constants of the flexible membranes of the different CMUT cells within each array elements vary by less than 20%, preferably by less than 10% whilst the mass of the flexible membranes of the different CMUT cells increases with increasing distance of said CMUT cell from the central region of the array element.

8. The ultrasound transducer (10) of claim 7, wherein the layer (135) of said material has a density in excess of 7 g/cm3, preferably a density in excess of 10 g/cm3 and a Young's modulus of less than 200 GPa, preferably a Young's modulus of less than 100 GPa; and/or wherein the layer (135) of said material is a metal layer or metal alloy layer, preferably wherein the metal layer is a gold layer or a platinum layer.

9. The ultrasound transducer (10) of claim 7 or 8, wherein the layer (135) of said material is a patterned layer.

10. The ultrasound transducer (10) of any of claims 7-9, wherein the layer (135) of said material is present in the flexible membranes (130') of each of the further CMUT cells

(100') within an array element (99), the thickness of said layer increasing between said further CMUT cells with increasing distance from the central region.

1 1. The ultrasound transducer (10) of any of claims 7-10, wherein the layer stack of each flexible membrane (130) of the first CMUT cell (100) in an array element (99) comprises a layer (131, 133) of a dielectric material to a first thickness; and

the layer stack of each flexible membrane (130') of the further CMUT cells (100') in the array element (99) comprises a layer of the dielectric material to a second thickness that is smaller than the first thickness; or

the layer stack of each flexible membrane (130') of the first CMUT cell (100') in the array element (99) is the same as the layer stack (130) of each flexible membrane (130) of the further CMUT cells (100) in the array element apart for the layer (135) of said material having a higher density than any of the layers in the layer stack of each flexible membrane of the first CMUT cell being an additional layer.

12. The ultrasound transducer (10) of any of claims 1-1 1, wherein the flexible membranes (130, 130') of the CMUT cells (100, 100') have a diameter, and the CMUT cells within an array element (99) includes a first set of CMUT cells (100) proximal to said central region in the extension direction (E) of the array element (99) and a second set of CMUT cells (100') distal to said central region in the extension direction, wherein the diameter of the flexible membranes of the CMUT cells in said first set is smaller than the diameter of the flexible membranes of the CMUT cells in said second set.

13. An interventional medical device (250) comprising a flexible elongate body and the ultrasound transducer (10) of any of claims 1-12 arranged around said flexible elongate body such that the extension direction (E) of each array element (99) is aligned with the elongation direction of the flexible elongate body.

14. An ultrasound imaging system (1) comprising a patient interface module and the ultrasound transducer (10) of any of claims 1-12 or the interventional medical device of claim 13.

15. The ultrasound imaging system (1) of claim 14, further comprising a power supply (45) adapted to operate the at least one array element (99) in a collapse mode during at least one of an ultrasound transmission mode and an ultrasound reception mode by supplying each of the CMUT cells (100, 100') within each array element with the same stimulus.