Patent Application
20190008487
All publications and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication or patent application was specifically and individually indicated to be incorporated by reference.
This disclosure generally relates to ultrasound imaging systems and more particularly to systems and methods for calibrating a multiple aperture ultrasound probe.
In conventional ultrasonic imaging, a focused beam of ultrasound energy is transmitted into body tissues to be examined and the returned echoes are detected and plotted to form an image. While ultrasound has been used extensively for diagnostic purposes, conventional ultrasound has been greatly limited by depth of scanning, speckle noise, poor lateral resolution, obscured tissues and other such problems.
In order to insonify body tissues, an ultrasound beam is typically formed and focused either by a phased array or a shaped transducer. Phased array ultrasound is a commonly used method of steering and focusing a narrow ultrasound beam for forming images in medical ultrasonography. A phased array probe has many small ultrasonic transducer elements, each of which can be pulsed individually. By varying the timing of ultrasound pulses (e.g., by pulsing elements one by one in sequence along a row), a pattern of constructive interference is set up that results in a beam directed at a chosen angle. This is known as beam steering. Such a steered ultrasound beam may then be swept through the tissue or object being examined. Data from multiple beams are then combined to make a visual image showing a slice through the object.
Traditionally, the same transducer or array used for transmitting an ultrasound beam is used to detect the returning echoes. This design configuration lies at the heart of one of the most significant limitations in the use of ultrasonic imaging for medical purposes: poor lateral resolution. Theoretically, the lateral resolution could be improved by increasing the width of the aperture of an ultrasonic probe, but practical problems involved with aperture size increase have kept apertures small. Unquestionably, ultrasonic imaging has been very useful even with this limitation, but it could be more effective with better resolution.
A method of calibrating an ultrasound probe is provided, comprising the steps of placing a first array and a second array of the ultrasound probe in position to image a phantom, each of the first and second arrays having a plurality of transducer elements, imaging the phantom with the first array to obtain a reference image, wherein imaging is dependent on data describing a position of each transducer element of the first array, imaging the phantom with the second array to obtain a test image, wherein imaging is dependent on data describing a position of each transducer element of the second array, quantifying a first error between the reference image and the test image; iteratively optimizing the data describing the position of each transducer element of the second array until the first error is at a minimum.
In some embodiments, the method further comprises imaging the phantom with a third array of the ultrasound probe to obtain a second test image, the third array having a plurality of transducer elements, quantifying a second error between the reference image and the second test image and iteratively optimizing data describing a position of each element of the third array until the second error is minimized.
In some embodiments, the method further comprises storing raw echo data received while imaging the phantom with the second array.
In one embodiment, the iteratively optimizing step comprises adjusting the data describing the position of the transducer elements of the second array to create first adjusted position data, re-beamforming the stored echo data using the first adjusted position data to form a second test image of the reflectors, quantifying a second error between the second test image and the reference image, and determining whether the second error is less than the first error.
In one embodiment, adjusting the data describing the position of the transducer elements of the second array includes adjusting a position of a reference point of the array and an angle of a surface of the array, but does not include adjusting a spacing between the elements of the second array.
In some embodiments, the method further comprises, after a first iteratively optimizing step, performing a second iteratively optimizing step comprising adjusting the first adjusted position data, including adjusting a spacing between at least two transducer elements of the second array to create second adjusted position data, re-beamforming the stored echo data using the second adjusted position data to form a third test image of the reflectors, quantifying a third error between the third test image and the reference image, and determining whether the third error is less than the second error.
In one embodiment, iteratively optimizing the transducer element position data comprises optimizing using a least squares optimization process.
In other embodiments, quantifying the first error comprises quantifying a distance between positions of reflectors in the reference image relative to positions of the same reflectors in the test image. In some embodiments, quantifying the first error comprises quantifying a difference in brightness between reflectors in the reference image and reflectors in the test image. In additional embodiments, quantifying the first error comprises quantifying a difference between a pattern of reflectors and holes in the reference image compared with a pattern of holes and reflectors in the test image.
In one embodiment, the reference image and the test image are three-dimensional volumetric images of a three-dimensional pattern of reflectors, holes, or both reflectors and holes.
In other embodiments, wherein the phantom comprises living tissue.
In some embodiments, the method further comprises identifying positions of reflectors in the phantom and fitting a mathematically defined curve to a detected pattern of reflectors.
In one embodiment, the curve is a straight line.
In other embodiments, the step of quantifying a first error comprises calculating a coefficient of determination that quantifies a degree of fit of the curve to the pattern of reflectors.
A method of calibrating an ultrasound probe is provided, comprising the steps of insonifying a plurality of reflectors of a phantom with the ultrasound probe, receiving echo data with the ultrasound probe, storing the echo data, beamforming the stored echo data using first transducer element position data to form an image of the reflectors, obtaining reference data describing the reflectors, quantifying an error between the image and the reference data, and iteratively optimizing the transducer element position data based on the quantified error.
In some embodiments, the iteratively optimizing step comprises iteratively optimizing the transducer element position data with a least squares optimization process.
In one embodiment, the iteratively optimizing step comprises adjusting the transducer element position data, re-beamforming the stored echo data using the adjusted transducer element position data to form a second image of the reflectors, quantifying a second error based on the second image, and evaluating the second error to determine whether the adjusted transducer element position data improves the image.
In some embodiments, adjusting the transducer element position data comprises adjusting an array horizontal position variable, an array vertical position variable and an array angle variable. In other embodiments, adjusting the transducer element position data does not comprise adjusting a spacing between adjacent transducer elements on a common array.
In one embodiment, the reference data is based on physical measurements of the phantom.
In some embodiments, the method further comprises deriving the reference data from a reference image of the phantom.
In one embodiment, the reference image is obtained using a different group of transducer elements of the probe than a group of transducer elements used for the insonifying and receiving steps.
In additional embodiments, the step of iteratively optimizing the transducer element position data comprises using a least squares optimization process.
In some embodiments, the method further comprises identifying positions of reflectors in the phantom and fitting a mathematically defined curve to a detected pattern of reflectors. In one embodiment, the curve is a straight line.
In some embodiments, the step of quantifying a first error comprises calculating a coefficient of determination that quantifies a degree of fit of the curve to the pattern of reflectors.
A method of calibrating ultrasound imaging data is also provided, comprising the steps of retrieving raw echo data from a memory device, the raw echo data comprising a plurality of echo strings, each echo string comprising a collection of echo records corresponding to echoes of a single ultrasound ping transmitted from a single transmit aperture and received by a single receive element, retrieving first calibration data describing a position of each receive transducer element corresponding to each echo string, retrieving second calibration data describing a position of at least one transducer element corresponding to a transmitted ping associated with each echo string, forming a reference image by beamforming a first collection of echo strings corresponding to a first group of receive transducer elements, wherein beamforming comprises triangulating a position of reflectors based on the first and second calibration data, forming a test image by beamforming a second collection of echo strings corresponding to a second group of transducer elements that is not identical to the first group of transducer elements, quantifying first error between the reference image and the test image, adjusting first calibration data to describe adjusted positions for the elements of the second group, re-beamforming the test image with the adjusted positions for the elements of the second group to obtain a second test image, quantifying a second error between the second test image and the reference image, and evaluating the new error to determine whether the second error is less than the first error.
In some embodiments, the method is performed without any physical or electronic connection to a probe used to create the raw echo data.
In some embodiments, there is no ultrasound probe connected to the memory device.
An ultrasound probe calibration system is provided, comprising an ultrasound probe having a plurality of transmit transducer elements and a plurality of receive transducer elements, a phantom having a pattern of reflectors, a first memory device containing reference data describing the pattern of reflectors of the phantom, a second memory device containing transducer element position data describing a position of each transmit transducer element and each receive transducer element relative to a common coordinate system, and an imaging control system containing calibration program code configured to direct the system to insonify the phantom with the transmit transducer elements, receive echo data with the receive transducer elements, and store echo data in a third memory device, form a first image of the pattern of reflectors by beamforming the stored echo data using the transducer element position data, determine measurement data describing a position of the pattern of reflectors as indicated by the first image, quantify an error between the measurement data and the reference data, and iteratively optimize the transducer element position data based on the quantified error.
In some embodiments, the imaging control system is configured to iteratively optimize the phantom by adjusting the transducer element position data; forming a second image of the pattern of reflectors by re-beamforming the stored echo data using the adjusted transducer element position data quantifying a second error based on the second image and evaluating the second error to determine whether the adjusted transducer element position data improves the image.
In one embodiment, the reference data is based on physical measurements of the phantom.
In other embodiments, the reference data is based on a reference image.
In some embodiments, the imaging control system is configured to iteratively optimize the transducer element position data using a least squares optimization process.
In other embodiments, the phantom further comprises at least one region that absorbs ultrasound signals.
In some embodiments, the ultrasound probe comprises a plurality of transducer arrays. In another embodiment, the ultrasound probe comprises a single continuous transducer array. In one embodiment, the ultrasound probe comprises a transducer array with a concave curvature.
In some embodiments, the phantom comprises a pattern of pins.
In one embodiment, the phantom comprises living tissue.
In some embodiments, the calibration program code is configured to determine measurement data by fitting a curve to a detected pattern of reflectors.
In one embodiment, the calibration program code is configured to quantify an error by determining a coefficient of determination quantifying a degree of fit of the curve.
In another embodiment, at least two of the first memory device, the second memory device, and the third memory device are logical portions of a single physical memory device.
The novel features of the invention are set forth with particularity in the claims that follow. A better understanding of the features and advantages of the present invention will be obtained by reference to the following detailed description that sets forth illustrative embodiments, in which the principles of the invention are utilized, and the accompanying drawings of which:
The various embodiments will be described in detail with reference to the accompanying drawings. References made to particular examples and implementations are for illustrative purposes, and are not intended to limit the scope of the invention or the claims.
The various embodiments herein provide systems and methods for dynamically calibrating a multiple aperture ultrasound probe using a static phantom. Calibration of a multiple aperture ultrasound imaging probe may generally comprise determining an acoustic position of each transducer element in the probe. Some embodiments of a dynamic calibration process may generally include the steps of imaging a calibration phantom having a known pattern of reflectors, quantifying an error between known information about the phantom and information obtained from the imaging, and performing an iterative optimization routine to minimize an error function in order to obtain improved transducer element position variables. Such improved transducer element position variables may then be stored for use during subsequent imaging using the calibrated probe.
Although the various embodiments are described herein with reference to ultrasound imaging of various anatomic structures, it will be understood that many of the methods and devices shown and described herein may also be used in other applications, such as imaging and evaluating non-anatomic structures and objects. For example, the probes, systems and methods described herein may be used in non-destructive testing or evaluation of various mechanical objects, structural objects or materials, such as welds, pipes, beams, plates, pressure vessels, etc.
As used herein the terms “ultrasound transducer” and “transducer” may carry their ordinary meanings as understood by those skilled in the art of ultrasound imaging technologies, and may refer without limitation to any single component capable of converting an electrical signal into an ultrasonic signal and/or vice versa. For example, in some embodiments, an ultrasound transducer may comprise a piezoelectric device. In other embodiments, ultrasound transducers may comprise capacitive micromachined ultrasound transducers (CMUT).
Transducers are often configured in arrays of multiple individual transducer elements. As used herein, the terms “transducer array” or “array” generally refers to a collection of transducer elements mounted to a common backing plate. Such arrays may have one dimension (1D), two dimensions (2D), 1.X dimensions (1.XD) or three dimensions (3D). Other dimensioned arrays as understood by those skilled in the art may also be used. Annular arrays, such as concentric circular arrays and elliptical arrays may also be used. An element of a transducer array may be the smallest discretely functional component of an array. For example, in the case of an array of piezoelectric transducer elements, each element may be a single piezoelectric crystal or a single machined section of a piezoelectric crystal.
As used herein, the terms “transmit element” and “receive element” may carry their ordinary meanings as understood by those skilled in the art of ultrasound imaging technologies. The term “transmit element” may refer without limitation to an ultrasound transducer element which at least momentarily performs a transmit function in which an electrical signal is converted into an ultrasound signal. Similarly, the term “receive element” may refer without limitation to an ultrasound transducer element which at least momentarily performs a receive function in which an ultrasound signal impinging on the element is converted into an electrical signal. Transmission of ultrasound into a medium may also be referred to herein as “insonifying.” An object or structure which reflects ultrasound waves may be referred to as a “reflector” or a “scatterer.”
As used herein, the term “aperture” may refer to a conceptual “opening” through which ultrasound signals may be sent and/or received. In actual practice, an aperture is simply a single transducer element or a group of transducer elements that are collectively managed as a common group by imaging control electronics. For example, in some embodiments an aperture may be a physical grouping of elements which may be physically separated from elements of an adjacent aperture. However, adjacent apertures need not necessarily be physically separated.
It should be noted that the terms “receive aperture,” “insonifying aperture,” and/or “transmit aperture” are used herein to mean an individual element, a group of elements within an array, or even entire arrays with in a common housing, that perform the desired transmit or receive function from a desired physical viewpoint or aperture. In some embodiments, such transmit and receive apertures may be created as physically separate components with dedicated functionality. In other embodiments, any number of send and/or receive apertures may be dynamically defined electronically as needed. In other embodiments, a multiple aperture ultrasound imaging system may use a combination of dedicated-function and dynamic-function apertures.
As used herein, the term “total aperture” refers to the total cumulative size of all imaging apertures. In other words, the term “total aperture” may refer to one or more dimensions defined by a maximum distance between the furthest-most transducer elements of any combination of send and/or receive elements used for a particular imaging cycle. Thus, the total aperture is made up of any number of sub-apertures designated as send or receive apertures for a particular cycle. In the case of a single-aperture imaging arrangement, the total aperture, sub-aperture, transmit aperture, and receive aperture will all have the same dimensions. In the case of a multiple array probe, the dimensions of the total aperture may include the sum of the dimensions of all of the arrays.
In some embodiments, two apertures may be located adjacent one another on a continuous array. In still other embodiments, two apertures may overlap one another on a continuous array, such that at least one element functions as part of two separate apertures. The location, function, number of elements and physical size of an aperture may be defined dynamically in any manner needed for a particular application. Constraints on these parameters for a particular application will be discussed below and/or will be clear to the skilled artisan.
Elements and arrays described herein may also be multi-function. That is, the designation of transducer elements or arrays as transmitters in one instance does not preclude their immediate redesignation as receivers in the next instance. Moreover, embodiments of the control system herein include the capabilities for making such designations electronically based on user inputs, pre-set scan or resolution criteria, or other automatically determined criteria.
As used herein the term “point source transmission” may refer to an introduction of transmitted ultrasound energy into a medium from single spatial location. This may be accomplished using a single ultrasound transducer element or combination of adjacent transducer elements transmitting together as a single transmit aperture. A single transmission from a point source transmit aperture approximates a uniform spherical wave front, or in the case of imaging a 2D slice, a uniform circular wave front within the 2D slice. In some cases, a single transmission of a circular or spherical wave front from a point source transmit aperture may be referred to herein as a “ping” or a “point source pulse.”
Point source transmission differs in its spatial characteristics from a “phased array transmission” which focuses energy in a particular direction from the transducer element array. Phased array transmission manipulates the phase of a group of transducer elements in sequence so as to strengthen or steer an insonifying wave to a specific region of interest. A short duration phased array transmission may be referred to herein as a “phased array pulse.”
In some embodiments, multiple aperture imaging using a series of transmitted pings may operate by transmitting a point-source ping from a first transmit aperture and receiving echoes with elements of two or more receive apertures, one or more of which may include some or all elements of a transmit aperture. A complete image may be formed by triangulating the position of scatterers based on delay times between ping transmission and reception of echoes, the speed of sound, and the relative positions of transmit and receive transducer elements. As a result, each receive aperture may form a complete image from echoes of each transmitted ping. In some embodiments, a single time domain frame may be formed by combining images formed from echoes at two or more receive apertures from a single transmitted ping. In other embodiments, a single time domain frame may be formed by combining images formed from echoes received at one or more receive apertures from two or more transmitted pings. In some such embodiments, the multiple transmitted pings may originate from different transmit apertures.
In other embodiments, any other multiple aperture ultrasound imaging probe may be calibrated using the systems and methods described below. For example,
As used herein, the term “phantom” may refer to any substantially static object to be imaged by an ultrasound probe. For example, any number of phantoms designed for sonographer training are widely commercially available from various suppliers of medical equipment, such as Gammex, Inc. (gammex.com). Some commercially available phantoms are made to mimic the imaging characteristics of objects to be imaged such as specific or generic human tissues. Such properties may or may not be required by various embodiments of the invention as will be further described below. The term “phantom” may also include other objects with substantially static reflectors, such as a region of a human or animal body with substantially static strong reflectors. An object need not be purpose-built as a phantom to be used as a phantom for the calibration processes described herein.
With reference to
In some embodiments, the width of a receive aperture may be limited by the assumption that the speed of sound is the same for every path from a scatterer to each element of the receive aperture. In a narrow enough receive aperture this simplifying assumption is acceptable. However, as receive aperture width increases, an inflection point is reached (referred to herein as the “maximum coherent aperture width” or “coherence width”) at which the echo return paths will necessarily pass though different types of tissue having different speeds of sound. When this difference results in phase shifts in excess of 180 degrees, additional receive elements beyond the maximum coherent receive aperture width will actually degrade the image rather than improve it. The coherence width will vary depending on an intended imaging application and is difficult if not impossible to predict in advance.
Therefore, in order to make use of a wide probe with a total aperture width greater than the maximum coherent width, the full probe width may be physically or logically divided into multiple apertures, each of which may be limited to a width less than the maximum coherent aperture width and small enough to avoid phase cancellation of received signals. The maximum coherent width can be different for different patients and for different probe positions on the same patient. In some embodiments, a compromise width may be determined for a given probe system. In other embodiments, a multiple aperture ultrasound imaging control system may be configured with a dynamic algorithm to subdivide the available elements in multiple apertures into groups that are small enough to avoid significant phase cancellation.
In some embodiments, it may be difficult or impossible to meet additional design constraints while grouping elements into apertures with a width less than the maximum coherent width. For example, if material is too heterogeneous over very small areas, it may be impractical to form apertures small enough to be less than the maximum coherent width. Similarly, if a system is designed to image a very small target at a substantial depth, an aperture with a width greater than the maximum coherent width may be needed. In such cases, a receive aperture with a width greater than the maximum coherent width can be accommodated by making additional adjustments or corrections may be made to account for differences in the speed-of-sound along different paths. Some examples of such speed-of-sound adjustments are provided herein.
With a multiple aperture probe using a point-source transmission imaging technique (also referred to as ping-based imaging), each image pixel may be assembled by beamforming received echo data to combine information from echoes received at each of the multiple receive apertures and from each of the multiple transmit apertures. In some embodiments of multiple aperture imaging with point-source transmission, receive beamforming comprises forming a pixel of a reconstructed image by summing time-delayed echo returns on receive transducer elements from a scatterer in the object being examined. The time delays may be determined by the geometry of the probe elements and an assumed value for the speed of sound through the medium being imaged.
The locus of a single reflector will lie along an ellipse with a first focus at the position of the transmit transducer element(s) and the second focus at the position of the receive transducer element. Although several other possible reflectors lie along the same ellipse, echoes of the same reflector will also be received by each of the other receive transducer elements of a receive aperture. The slightly different positions of each receive transducer element means that each receive element will define a slightly different ellipse for a given reflector. Accumulating the results by coherently summing the ellipses for all elements of a common receive aperture will indicate an intersection of the ellipses for a reflector, thereby converging towards a point at which to display a pixel representing the reflector. The echo amplitudes received by any number of receive elements may thereby be combined into each pixel value. In other embodiments the computation can be organized differently to arrive at substantially the same image.
Because the position of each transmit and receive element plays an important role in producing an image during ping-based ultrasound imaging, the quality of an image produced from ping-based imaging is substantially dependent on the accuracy of the information describing the relative positions of the transducer elements.
Various algorithms may be used for combining echo signals received by separate receive elements. For example, some embodiments may process echo-signals individually, plotting each echo signal at all possible locations along its ellipse, then proceeding to the next echo signal. Alternatively, each pixel location may be processed individually, identifying and processing all echoes potentially contributing to that pixel location before proceeding to the next pixel location.
Image quality may be further improved by combining images formed by the beamformer from one or more subsequent transmitted pings, transmitted from the same or a different point source (or multiple different point sources). Still further improvements to image quality may be obtained by combining images formed by more than one receive aperture. An important consideration is whether the summation of images from different pings, different transmit point-sources or different receive apertures should be coherent summation (phase sensitive) or incoherent summation (summing magnitude of the signals without phase information).
In some embodiments, multiple aperture imaging using a series of transmitted pings may operate by transmitting a point-source ping from a first transmit aperture and receiving echoes with elements of one or more receive apertures (which may overlap with the transmit aperture). A complete image may be formed by triangulating the position of scatterers based on delay times between transmission and receiving echoes and the known position of each receive element relative to each point-source transmit aperture. As a result, a complete image may be formed from data received at each receive aperture from echoes of each transmitted ping.
Images obtained from different unique combinations of a ping and a receive aperture may be referred to herein as image layers. Multiple image layers may be combined to improve the overall quality of a final combined image. Thus, in some embodiments, the number of image layers can be the product of the number of receive apertures and the number of transmit apertures (where a “transmit aperture” can be a single transmit element or a group of transmit elements). In other embodiments, the same ping imaging processes may also be performed using a single receive aperture.
Some embodiments of ultrasound probe calibration methods using a phantom may generally include the steps of characterizing the phantom using some known baseline reference data, then imaging the phantom with the probe to be calibrated. An error between the known reference data and data obtained from the generated image may then be quantified and an iterative optimization routine may be used to obtain improved transducer element position information. Such improved transducer element position variables may then be stored for use during subsequent imaging using the calibrated probe.
In some embodiments, reflectors 30 may be arranged in the phantom 20 in a pattern that may have characteristics selected to facilitate a calibration process. For example, a non-repeating reflector pattern will allow a calibration process to recognize an imaged position of the reflectors without confusion. For example, a complete grid pattern is highly repetitive because portions of the pattern are identically duplicated merely by shifting one full grid position. In some embodiments, the pattern of reflectors may also comprise a number of reflectors with displacement along the X axis 46 that is approximately equal to a number of reflectors with displacement along the Y axis 47. Thus, in some embodiments a pattern in the shape of a cross or a plus sign may be used. In other embodiments, reflectors may be positioned randomly or in other patterns, such as an X-shape, an asterisk, a sunburst, a spiral or any other pattern.
In some embodiments, reflectors may also have depth or distinguishable detail in the z-direction 48. For example, the reflectors 30 may be rods with longitudinal axes along the z-direction 48. Alternatively, the reflectors may be substantially spherical or uniform three-dimensional shapes. In other embodiments, an arrangement of intersecting wires or rods may be used to form a distinguishable pattern in three-dimensional space within a phantom.
The reflectors 30 in the calibration phantom 20 may be of any size or shape as desired. In some embodiments, the reflectors 30 may have a circular diameter that is on the same order of magnitude as the wavelength of the ultrasound signals being used. In general, smaller reflectors may provide better calibration, but in some embodiments the precise size of the reflectors need not be an important factor. In some embodiments, all reflectors 30 in the phantom may be the same size as one another, while in other embodiments, reflectors 30 may be provided in a variety of sizes.
In some embodiments, the physical size and location of the reflectors in the phantom 20 may be determined by mechanical measurement of the phantom (or by other methods, such as optical measurement or ultrasonic measurement using a known-calibrated system) prior to, during or after construction of the phantom. Reflector position reference data may then by obtained by storing the reflector location information within a memory device accessible by software or firmware performing a calibration process. Such reference data may include information such as the position, size, orientation, arrangement or other information about the reflectors and/or holes in the phantom. Reference data may be represented or stored as a reference image or as a series of data points. Alternatively, reference data may be extracted from a reference ultrasound image.
In some embodiments, a reference image of the phantom may be obtained using a probe or an array within a probe that is known to be well-calibrated. In other embodiments, a reference image of the phantom may be obtained using a selected group of elements of the probe. Reflector size and/or location information may then be determined from the reference image for use in calibrating remaining elements of the probe or a different probe.
Therefore, in some embodiments a reference image may be obtained by retrieving previously-determined reflector position data from a memory device. In other embodiments, a reference image may be obtained by imaging the phantom using a sub-set of all elements in a probe. In some embodiments, it may be desirable to obtain a reference image using an aperture that is no wider than an assumed maximum coherence width (as described above). This allows for a reference image to be formed without the need to correct for speed-of-sound variations along different ultrasound wave paths. If the phantom is known to have a uniform speed-of-sound (except for reflectors and/or holes), then the coherence width may be as large as an entire total aperture of a multiple aperture probe. In such embodiments, obtaining a reference image with a receive aperture smaller than the coherence width for an intended imaging application may be useful as a starting point.
For example, when calibrating a three-array probe such as that shown in
As shown for example in
In some embodiments, the process 400 may be entirely automated in software or firmware. In other embodiments, at least some steps may involve human participation, such as to identify or to quantify an error between an obtained image and a reference image. In other embodiments, a human user may also be called upon to determine whether a resulting image is “good enough” or whether the calibration process should be repeated or continued.
In various embodiments, the process 400 may be used to calibrate the position of one or more test transmit apertures, one or more test receive apertures, or both. The choice of which type of aperture to calibrate may depend on factors such as the construction of the probe, the number of transmit or receive apertures, or other factors. The definitions of test transmit apertures and test receive apertures used for the calibration process may be, but need not necessarily be the same as the definition of apertures used for normal imaging with the probe. Therefore, the phrase “test aperture” as used herein may refer to either a transmit test aperture or a receive test aperture unless otherwise specified.
In some embodiments, the test transmit aperture and the test receive aperture used during the process 400 of
In some embodiments, a single transmit test aperture may be used to obtain both a reference image and data from which a test image may be formed. In such embodiments, a first receive aperture may be used to form a reference image, and a second (or third, etc.) receive aperture may be used to form or obtain test image data. Similarly, a single receive aperture may be used for obtaining both a reference image and data for a test image if different transmit apertures are used for the reference image and the test image data. Thus, the test transmit aperture and the test receive aperture need not necessarily be near one another. In other embodiments, reference images may be obtained using transmit and receive elements of a first array, while data for test images may be obtained using transmit and receive elements of a second array, where the second array is a test array to be calibrated.
As described above, in some embodiments, the step of obtaining reference data 402 may comprise retrieving reference data from a data storage device. Such a data storage device may be physically located within a calibration controller, within an ultrasound imaging system, within a probe, or on a separate storage device that may be accessible via a wired or wireless network connection. Alternatively, the step of obtaining reference data 402 may comprise imaging the phantom with a reference group of transducer elements.
In some embodiments, the step of insonifying the phantom with a test transmit aperture 404 may comprise transmitting one or more pings from one or more transmit elements of a transmit aperture. A single transmit aperture may typically comprise one, two, three or a small number of adjacent elements.
After each transmitted ping, returning echoes may be received by all receive elements of the test receive aperture, and the echo data may be digitized and stored 406 in a digital memory device. The memory device may be any volatile or non-volatile digital memory device in any physical location that is electronically accessible by a computing device performing the imaging and calibration processes.
The received echo data may then be beamformed and processed to form a test image 408. In some embodiments, the steps of insonifying the phantom from a test transmit aperture 404 and receiving echoes with a test receive aperture 405 may be repeated using multiple combinations of different transmit apertures and/or receive apertures, and images obtained 408 from such transmitting and receiving may be combined in a process referred to as image layer combining prior to proceeding to subsequent steps of the process 400.
In various embodiments, the error function may be determined from some difference between the phantom reference data (e.g., information known about the position of reflectors in the phantom) and an image of the phantom obtained with the test receive aperture. In some embodiments, the choice of error function may be based on characteristics of the phantom used, available processing capabilities, a chosen optimization method or many other factors.
In some embodiments, a modified least squares optimization method may be used to minimize an error function based on the square of an aggregated straight-line error distance between the expected reflector center and an imaged reflector center. For example, after forming an image of the phantom with the echoes received at a test receive aperture, the system may identify the location of each reflector in the image by identifying the brightest point in the image of approximately the expected size in approximately the expected location of each known reflector. Once each reflector is identified, an error between the imaged position and the expected position of each reflector may be determined. In some embodiments, these individual reflector-position errors may then be aggregated into a collective reflector pattern error, such as by summing all individual reflector errors. Alternatively, the individual errors may be aggregated using any other function, such as taking a maximum error, an average, or a weighted sum of individual errors. For example, if a phantom has some reflectors that are more difficult to detect than others, difficult-to-detect reflectors may be given less weight in the aggregate error function so as to obtain a more balanced result. In various embodiments, such individual and/or aggregate errors may be either scalar or vector quantities.
In some embodiments, reflector images may be sought within a predetermined search area surrounding the expected location of each reflector. The shape and size of a search area may be defined based on the known pattern of reflectors and the distance between reflectors. In some embodiments, images of reflectors may be identified by artificial intelligence or probability analysis using information about nearby reflectors and the known pattern of reflectors. In other embodiments, the search area surrounding each reflector may comprise a circular, rectangular or other geometric area centered on the point of a center of an expected reflector position. The size of a search area may be selected to be larger than the imaged reflectors, but typically small enough that adjacent search areas do not overlap.
In some embodiments, when the actual positions of reflectors in the phantom are known, this knowledge may be used to greatly simplify the process of forming an image of the phantom. For example, forming an image 408 may be limited to beamforming only echoes representing search areas surrounding the expected positions of reflectors in the phantom (rather than beamforming an entire image field). In other embodiments, beamforming may be limited to a search area defining the overall pattern of reflectors. For example, this may be accomplished in some embodiments by beamforming vertical and horizontal pixel bands slightly wider than the expected position of the pins in
In some embodiments, the error function may be defined based on one or more simplifying assumptions. For example, instead of detecting and optimizing based on the two-dimensional or three-dimensional position of each individual reflector, a line or curve may be fit to the series of reflectors. For example, using the phantom layout shown in
In other embodiments, the error function may be defined as some quantity other than reflector position. For example, in some embodiments, an error function may be defined as a sum of absolute value differences in brightness of the individual imaged reflectors relative to a reference image. In another embodiment, an error function may be defined based on a complete collective reflector pattern. For example, a phantom may be designed to contain an array of reflectors representing a reference number in binary form (i.e., a reflector may represent a ‘1’ and the absence of a reflector at a grid position may represent a ‘0’). In such embodiments, a calibration process may be configured to ‘read’ the binary values, and the error function may be defined as the number of bits different from the expected reference number. In further embodiments, an error function may be at least partially based on a pattern of “holes”—regions of the phantom that absorb the ultrasound energy. Many other error functions may also be used.
In some embodiments, adjustments to the element position variables may be essentially random in each iteration (i.e., with no connection to adjustments made in prior iterations). Such random adjustments may be made within a predetermined range of values relative to current element position data based on expectations of the possible degree of mis-calibration of existing element position data. In the case of random adjustments, an error function obtained from each iteration may be stored, and a minimum error function may be identified by comparing the results of all iterations.
In other embodiments, adjustments may be directly based on information from previous iterations, such as an evaluation of the magnitude and/or direction of a change in the error value. For example, in some embodiments, if the new error function E1 is less than the initial error function E0, then the adjustment made in step 452 may be determined to be a good adjustment and the process may repeat for more iterations making further incremental adjustments to the position variable(s). If the new error function E1 obtained in the first iteration is not less than the initial error function E0 (i.e. E1≥E0), then it may be assumed that the adjustment of step 452 was made in the wrong direction. Thus, in a second iteration, during step 452, the original element position variable(s) P0 may be adjusted in a direction opposite to that tried during the first iteration. If the resulting new error function E2 is still not smaller than the initial error function E0, then the error function is at a minimum (at least with respect to the adjusted element position variable(s)). In such a case, the error minimization process may be stopped, and the last good position variables may be stored as the new transducer element positions.
In some embodiments, the process 414 may be repeated through as many iterations as needed until the error function is minimized. In other embodiments, the process 414 may be stopped after a fixed number of iterations. As will be clear to the skilled artisan, multiple ‘optimum’ solutions may exist. As a result, in some embodiments, the iterative calibration process may be repeated multiple times, and the results of the several calibrations may be compared (automatically using image processing techniques or manually by a person) to identify a suitable solution. In any event, it is not necessary to identify the absolute optimal result.
In various embodiments, the position of transducer elements may be described by multiple variable quantities. Ultimately, it is desirable to know the acoustic position (which may be different than the element's apparent mechanical position) of each transducer element relative to some known coordinate system. Thus, in some embodiments, the acoustic position of each transducer element may be defined by an x, y, and z position (e.g., with reference to a Cartesian coordinate system 45 such as that shown in
Performing the optimization process by adjusting the x, y and z position of each transducer element may be somewhat computationally intensive, since a single aperture may contain hundreds of individual elements. This may result in the iterative adjustment of several hundred if not thousands of variables. This is particularly true for probes with 2D arrays (i.e., those with transducer elements spaced from one another in X and Z directions), curved 1D or 2D arrays (i.e., arrays with curvature about either the X or the Z axis), and 3D arrays (i.e., probes with curvature about two axes). While potentially computationally intensive, the various embodiments herein may be used to calibrate any ultrasound probe with large continuous planar or curved 1D or 2D arrays as well as large continuous 3D arrays with curvature about two axes.
As an alternative, some embodiments may employ one or more simplifying assumptions. For example, in some embodiments it may be assumed that element position relationships within a single array remain fixed relative to one another such that an array with a common backing block will only move, expand or contract uniformly. In some embodiments, it may also be assumed that the elements are uniformly distributed across the array. Using such assumptions, locating a center point of an array, a width of the array and an angle of the array surface relative to a known datum may provide sufficient information about the acoustic position of each element. For example (with reference to
In some embodiments two or more optimization processes may be combined in parallel or sequential processes in order to improve processing efficiency, calibration precision, or both. For example, in one embodiment, a two-stage optimization process may be used in which a first stage provides a coarse improvement to element position variables while relying on one or more simplifying assumptions. A second stage may then provide a more detailed improvement to the element position variables while relying on fewer simplifying assumptions, but starting from the improved information obtained during the first stage. During a first stage of one example embodiment, multiple reflectors may be represented with a single geometric shape such as a line, and the spacing between transducer elements may be treated as fixed (i.e., such values are not varied during the optimization). A second stage process may then be performed, in which the position of each pin is optimized by varying element position variables including the spacing between transducer elements.
In some embodiments, a similar calibration process may be used to calibrate a probe 55 with a large continuous array 18, such as that illustrated in
Regardless of the number of variables to be optimized in the iterative error function minimizing process 414, element position variables may be adjusted 452 either in series or in parallel. For example, in embodiments in which position variables are to be adjusted in series, only one variable may be adjusted during each iteration. In some embodiments of serial optimization, a single variable may be optimized (i.e., the error function may be minimized by adjusting only that single variable) before proceeding to the next variable. In embodiments in which two or more position variables are to be adjusted in parallel, the two or more variables may each be adjusted during each iteration. In some embodiments, those two variables may be optimized before proceeding to optimization of other variables. Alternatively, all variables may be optimized in parallel. In other embodiments, position variables may be optimized using a combination of series and parallel approaches. It should be noted this distinction between series and parallel optimization approaches should not be confused with parallel computer processing. Depending on computing hardware used, even optimizations performed in series as described above may be computed simultaneously using separate threads in parallel processors.
After completing calibration of a first array or aperture, the process of
In some embodiments, transducer element position adjustments may be obtained and stored in the form of new corrected element position coordinates. In other embodiments, position adjustments may be obtained and stored as coefficients to be added to or multiplied with previous element position coordinates. For example, in some embodiments “factory” element position data may be stored in a read-only memory device in a location readable by an ultrasound system, such as a ROM chip within a probe housing. Such factory position data may be established at the time of manufacturing the probe, and subsequent calibration data may be stored as coefficients that may be applied as adjustments to the factory position data.
In some embodiments, adjusted element position data for each transducer element in a probe may be stored in a non-volatile memory device located within a probe housing. In other embodiments, adjusted element position data may be stored in a non-volatile memory device located within an imaging system, on a remote server, or in any other location from which the information may be retrieved by an imaging system during image beamforming.
In some embodiments, a calibration process using the methods described above, may be particularly useful in rapidly re-calibrating an adjustable probe such as that illustrated in
In some embodiments, one or more of the arrays in an adjustable probe may be permanently secured to the housing in a fixed orientation and position (e.g., the center array or the left or right end array), while the remaining arrays may be movable to conform to a shape of an object to be imaged. The fixed array would then be in a permanently known position and orientation. Alternatively, the position and orientation of one or more arrays may be known based on one or more position sensors within an adjustable probe. The known-position array(s) may then be used to obtain a reference image of a phantom (or even a region of an object or patient to be imaged), and an optimization process may be used to determine an adjusted position of the movable arrays. For example, a sonographer may adjust the adjustable arrays of an adjustable probe to conform to a patient's anatomy. Then, during normal imaging, a reference image may be obtained using the known array, and positions of the remaining arrays may be determined by an optimization routine configured to minimize an error function (e.g., using an optimization routine as described above) defining an error between the reference image obtained from the center array and images obtained from each adjustable array.
In other embodiments, a sonographer may adjust the arrays of an adjustable probe to conform to a patient's anatomy. The sonographer may then place the probe onto a phantom that includes a conformable section configured to receive the probe in its adjusted position. For example, a conformable section may include a flexible bag containing a liquid or gel selected to transmit ultrasound signals at substantially the same speed of sound as the material of the phantom. A calibration process may then be initiated, and the position of each adjustable array may be determined by an iterative optimization routine in which reference data describing the phantom is compared with images of the phantom obtained with each array.
In some embodiments, the element-position information may change between performing a calibration operation and capturing raw ultrasound data. For example, a probe may be dropped, damaged or may be otherwise altered (such as by thermal expansion or contraction due to a substantial temperature change) before or during a raw sample data capture session. In some embodiments, the probe may be re-calibrated using captured, stored raw echo data as described below.
In other embodiments, a calibration system may be incorporated into an ultrasound imaging system. In some embodiments, as shown for example in
In some embodiments, a calibration system may be provided independently of an imaging system. In such embodiments, components such as the video subsystem 512 may be omitted. Other components shown in
In practice, the transmit control subsystem 504 may direct the probe to transmit ultrasound signals into a phantom. Echoes returned to the probe may produce electrical signals which are fed into the receive sub-system 508, processed by an analog front end, and converted into digital data by an analog-to-digital converter. The digitized echo data may then be stored in a raw data memory device 502. The digital echo data may then be processed by the beamformer 520 in order to determine the location of each reflector so as to form an image. In performing beamforming calculations, the beamformer may retrieve calibration data from a calibration memory 530. The calibration data may describe the position of each transducer element in the probe. In order to perform a new calibration, the calibration processor may receive image data from the image formation block 520 or from an image buffer memory device 526 which may store single image frames and/or individual image layers.
The calibration processor may then perform an optimization-based calibration routine. Once a calibration process is complete, new calibration information may be stored in the calibration memory device 530 for use in subsequent imaging processes or in additional calibration processes.
Using such a system, raw echo data of a phantom may be captured and stored along with raw echo data from a target object imaging session (e.g., with a patient). Capturing and storing raw echo data of a phantom before and/or after an imaging session may allow for later optimization of the imaging-session data. Such optimization may be applied at any point after the imaging session using the stored raw data and the methods described above.
As shown in
The transmission of ultrasound signals from elements of the probe 506 may be controlled by a transmit controller 504. Upon receiving echoes of transmit signals, the probe elements may generate time-varying electric signals corresponding to the received ultrasound vibrations. Signals representing the received echoes may be output from the probe 506 and sent to a receive subsystem 508. In some embodiments, the receive subsystem 508 may include multiple channels. Each channel may include an analog front-end device (“AFE”) 509 and an analog-to-digital conversion device (ADC) 511. In some embodiments, each channel of the receive subsystem 508 may also include digital filters and data conditioners (not shown) after the ADC 511. In some embodiments, analog filters prior to the ADC 511 may also be provided. The output of each ADC 511 may be directed into a raw data memory device 502. In some embodiments, one independent channel of the receive subsystem 508 may be provided for each receive transducer element of the probe 506. In other embodiments, two or more transducer elements may share a common receive channel.
In some embodiments, the ultrasound imaging system may store digital data representing the timing, phase, magnitude and/or the frequency of ultrasound echo signals received by each individual receive element in a raw data memory device 502 before performing any further beamforming, filtering, image layer combining or other image processing.
In addition to received echo data, information about one or more ultrasound transmit signals that generated a particular set of echo data may also be stored in a memory device, such as the raw data memory device 502 or another memory device. For example, when imaging with a multiple aperture ping ultrasound method as described above, it is desirable to know information about a transmitted ping that produced a particular set of echoes. Such information may include the identity and/or position of one or more a transmit elements as well as a frequency, magnitude, duration or other information describing a transmitted ultrasound signal. Transmit data is collectively referred herein to as “TX data”. In some embodiments, such TX data may be stored explicitly in the same raw data memory device in which raw echo data is stored. For example, TX data describing a transmit signal may be stored as a header before or as a footer after a set of raw echo data generated by the transmit signal. In other embodiments, TX data may be stored explicitly in a separate memory device that is also accessible to a system performing a beamforming process. In embodiments in which transmit data is stored explicitly, the phrases “raw echo data” or “raw data” may also include such explicitly stored TX data.
TX data may also be stored implicitly. For example, if an imaging system is configured to transmit consistently defined ultrasound signals (e.g., consistent magnitude, shape, frequency, duration, etc.) in a consistent or known sequence, then such information may be assumed during a beamforming process. In such cases, the only information that needs to be associated with the echo data is the position (or identity) of the transmit transducer(s). In some embodiments, such information may be implicitly obtained based on the organization of raw echo data in a raw data memory.
For example, a system may be configured to store a fixed number of echo records following each ping. In such embodiments, echoes from a first ping may be stored at memory positions 0 through ‘n’ (where ‘n’ is the number of records stored for each ping), and echoes from a second ping may be stored at memory positions n+1 through 2n+1. In other embodiments, one or more empty records may be left in between echo sets. In some embodiments received echo data may be stored using various memory interleaving techniques to imply a relationship between a transmitted ping and a received echo data point (or a group of echoes). In general, a collection of echo records corresponding to echoes of a single transmitted ping received by a single receive element may be referred to herein as a single “echo string.” A complete echo string may refer to all echoes of the single ping received by the receive element, whereas a partial string may refer to a sub-set of all echoes of the single ping received by the receive element.
Similarly, assuming data is sampled at a consistent, known sampling rate, the time at which each echo data point was received may be inferred from the position of that data point in memory. In some embodiments, the same techniques may also be used to implicitly store data from multiple receive channels in a single raw data memory device.
In other embodiments, the raw echo data stored in the raw data memory device 520 may be in any other structure as desired, provided that a system retrieving the echo data is able to determine which echo signals correspond to which receive transducer element and to which transmitted ping. In some embodiments, position data describing the position of each receive transducer element may be stored in the calibration memory device along with information that may be linked to the echo data received by that same element. Similarly, position data describing the position of each transmit transducer element may be stored in the calibration memory device along with information that may be linked to the TX data describing each transmitted ping.
In some embodiments, each echo string in the raw data memory device may be associated with position data describing the position of the receive transducer element that received the echoes and with data describing the position of one or more transmit elements of a transmit aperture that transmitted the ping that produced the echoes. Each echo string may also be associated with TX data describing characteristics of the transmitted ping.
In some embodiments, a probe may be calibrated using raw echo data stored in a memory device without raw data of a phantom image. Assuming at least one array (or one portion of an array) is known or assumed to be well-calibrated, nearly any image data with a pattern of strong reflectors may be used to calibrate second, third or further arrays or array segments. For example, echo data from the known-calibrated aperture, array or array segment may be beamformed to obtain a reference image. Stored echo data from the remaining apertures/arrays may then be calibrated using any of the methods described above to calibrate the position of the remaining arrays, apertures or array segments relative to the first. By performing a calibration process using stored echo data, a probe may be calibrated even when neither the probe itself nor the patient (or other imaged object) is physically present proximate to the device performing the re-beamforming and image processing. In such embodiments, the steps of insonifying a phantom 404, and receiving echoes 405 may be omitted from the process 400 of
Although this invention has been disclosed in the context of certain preferred embodiments and examples, it will be understood by those skilled in the art that the present invention extends beyond the specifically disclosed embodiments to other alternative embodiments and/or uses of the invention and obvious modifications and equivalents thereof. Various modifications to the above embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, it is intended that the scope of the present invention herein disclosed should not be limited by the particular disclosed embodiments described above, but should be determined only by a fair reading of the claims that follow.
In particular, materials and manufacturing techniques may be employed as within the level of those with skill in the relevant art. Furthermore, reference to a singular item, includes the possibility that there are plural of the same items present. More specifically, as used herein and in the appended claims, the singular forms “a,” “and,” “said,” and “the” include plural referents unless the context clearly dictates otherwise. As used herein, unless explicitly stated otherwise, the term “or” is inclusive of all presented alternatives, and means essentially the same as the commonly used phrase “and/or.” Thus, for example the phrase “A or B may be blue” may mean any of the following: A alone is blue, B alone is blue, both A and B are blue, and A, B and C are blue. It is further noted that the claims may be drafted to exclude any optional element. As such, this statement is intended to serve as antecedent basis for use of such exclusive terminology as “solely,” “only” and the like in connection with the recitation of claim elements, or use of a “negative” limitation. Unless defined otherwise herein, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
This application is a continuation of U.S. application Ser. No. 15/400,826, filed Jan. 6, 2017, now U.S. Pat. No. 10,064,605; which application is a continuation of U.S. application Ser. No. 13/964,701, filed Aug. 12, 2013, now U.S. Pat. No. 9,572,549, which application claims the benefit of U.S. Provisional Application No. 61/681,986, filed Aug. 10, 2012, titled “Calibration of Multiple Aperture Ultrasound Probes”, the contents of which are incorporated by reference herein.
| Number | Date | Country | |
|---|---|---|---|
| 61681986 | Aug 2012 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 15400826 | Jan 2017 | US |
| Child | 16121303 | US | |
| Parent | 13964701 | Aug 2013 | US |
| Child | 15400826 | US |
