Segmentation tool for identifying flow regions in an imaging system

0001The present invention relates generally to ultrasound imaging systems, and more particularly relates to a system and method for optimizing an ultrasound imaging process.

0002Thanks to an ongoing series of technological advances, diagnostic ultrasound remains one of today's most important medical tools. Since the mid-1960's, continuous advances have improved the clinical value of ultrasound, expanding its capabilities, accuracy, and ease of use. Recent advances, such as real-time 3D imaging can be used to collect important details, such as blood flow and other motion data, during a relatively short exam. This type of data is particularly useful in areas such as cardiology, where an abnormality in the flow of blood throughout the heart may be an indicator of heart disease.

0003Unfortunately, because ultrasound data is collected using sound waves, it is subject to the physical limitations of the speed of sound in tissue. Specifically, ultrasound data is acquired with a transducer that transmits an acoustic pulse along a look direction or line, and then listens for echoes along the same line. Received echo information gathered from a set of adjacent lines can be processed and used, for instance, to form an image that can be displayed on a monitor. Depending on the particular implementation, the number and density of the lines will vary. In the case of a two-dimensional (2D) image the lines form a frame, and iii the case of a three-dimensional (3D) image the lines form a volume. Because ultrasound information is typically displayed in real-time as a series of frames or volumes, the time it takes to form an image (i.e., frame or volume) is critical for many applications. Specifically, if the time is too great, the frame rate or volume rate may be too slow for ultrasound imaging of moving tissue (for example, blood or fetal anatomy).

0004Color flow Doppler, which generates a color image that indicates velocity and direction of any flow within an image, is particularly susceptible to the above-mentioned problems. Motion is detected by analyzing differences in the received echo signal for multiple received echo lines formed along the same axis. This type of data can provide important analytical information, including blood flow velocity, regurgitation, etc. However, since the detection of flow along each look line requires the use of multiple transmit/receive cycles, the use of color flow Doppler significantly increases the time it takes to form an image, and thereby further reduces the frame or image rate. As such, the acquisition of color flow Doppler data throughout an entire image volume is not practical in many clinical contexts because of the substantial degradation in the acquisition rate due to the number of transmit/receive cycles required to obtain Doppler phase shift information.

0005If the frame rate is too slow the resulting ultrasound image may misrepresent the physiological condition by introducing imaging artifacts and distortions. Accordingly, for some applications involving color flow Doppler, such as analyzing regurgitent flow through the mitral valve of the heart, the side effects of an inadequate image rate may limit the diagnostic performance of the ultrasound imager.

0006Another proposed application for ultrasound that is also adversely affected by the physical limitation of the speed of sound in tissue is the acquisition of data for subsequent off-line or retrospective analysis. In this scenario, sufficient ultrasound data is acquired from a region within the patient so that a subsequent diagnosis may be made from that data. The benefits of such an approach are several-fold. In one case, the attending clinician need only locate the general region of interest, such as the patient's heart, from which to acquire ultrasound data. With the use of an ultrasound scanner capable of acquiring data from a 3D volume, the clinician performing the diagnosis may then navigate through the acquired data to obtain the specific components of the data required to form a diagnosis. In this fashion, the time required to acquire data from the patient is minimized by allowing the diagnosis to take place after the exam, rather than during the exam. This approach allows the diagnosing clinician to perform their diagnostic function both at a different time and a different location from the examination of the patient. Furthermore it is possible to employ an attending clinician with a lesser skill set than would be required were the diagnosis to be performed during the exam. The use of this approach requires that all the data required to form a diagnosis be acquired during the exam.

0007One important component of many exams is the acquisition of flow data from potentially diseased portions of the anatomy being imaged. Detection of such a region may be accomplished using an algorithm within the ultrasound scanner that detects blood flow conditions that correspond to the anatomy and physiology of interest. An example of this is the common finding of mitral valve regurgitation that has a common color flow Doppler signature. Once such a region has been identified using such an ultrasound scanner based algorithm, the ultrasound scanner can automatically be made to perform a specialized acquisition such as continuous (CW) or pulsed wave (PW) Doppler.

0008This application, which has been proposed in the field of cardiology, seeks to obtain spectral Doppler data and grayscale echo data from a patient's heart in a short period of time (e.g., several heart cycles), and then store the data for later analysis. Such a system would capture data from all relevant regions of interest, such as blood flow velocity at the mitral valve. Because a physician would not view the information until a later time, the examination time could be greatly reduced. Unfortunately, the application of spectral Doppler to collect data for an entire volume is impractical since, for instance, the placement of the Doppler sample volume (i.e. the point at which the Doppler data is acquired) is highly specific.

0009One solution is to allow the technician to selectively identify the placement of the Doppler sample volume at various points of interest, and then collect data from only those points. However, such a process is limited to the technician's ability to accurately place the Doppler sample volume. If the technician fails to precisely capture the point of interest, the results would not be known until a later time when an off-line analysis took place, and the test would have to be rescheduled. <patcit id="pcit0001" dnum="US6245017B1"><text>US-6245017B1</text></patcit> discloses a 3-D ultrasonic diagnostic apparatus, wherein an image of a volume is surveyed to collect motion data.

0010In a first aspect, the invention provides a method according to claim 1.

0011In a second aspect, the invention provides an ultrasound system, according to claim 12.

0012In a third aspect, the invention provides an ultrasound system according to claim 20.

0013The features of this invention will be more readily understood from the following detailed description of the various aspects of the invention taken in conjunction with the accompanying drawings in which: <ul id="ul0001" list-style="none" compact="compact"><li><figref idref="f0001">Figure 1</figref> depicts an ultrasound system in accordance with the present invention.</li><li><figref idref="f0002">Figure 2</figref> depicts a volume containing a vessel having a flow and non-flow region.</li><li><figref idref="f0002">Figure 3</figref> depicts a volume containing a heart.</li><li><figref idref="f0003">Figure 4</figref> depicts a volume containing a heart wherein the mitral valve has been automatically detected and imaged with a scan line in accordance with the present invention.</li><li><figref idref="f0003">Figure 5</figref> depicts a volume containing a heart wherein the mitral valve has been automatically detected and imaged with a multi-line scan beam in accordance with the present invention.</li></ul>

System Overview

0014Referring now to the drawings, <figref idref="f0001">Figure 1</figref> depicts an ultrasound system 10 for imaging a target volume 32, and more specifically allows for the application of an ultrasound modality to a region of interest (ROI) 33 in the target volume 32. In the embodiments described herein, the applications generally provide some type of "flow" imaging for collecting color flow data, and include color flow Doppler and spectral Doppler. However, any other type of ultrasound modality could likewise be utilized, e.g., B-FLOW™, time domain correlation, speckle tracking, strain imaging, other Doppler techniques etc., and can be implemented in manners not described herein. It should be understood that the present invention uses the term "flow" to describe any type of movement, including blood flow, tissue motion, target motion, etc., and the specific use of such terms herein are not intended to limit the scope of the invention.

0015The present invention facilitates the use of ultrasound modalities, including spectral Doppler and color flow Doppler, by first segmenting the target volume 32 into flow and non-flow regions. To achieve this, ultrasound system 10 is provided with a survey system 12, a segmentation system 18, and one or more applications 20. Ultrasound system 10 acquires ultrasound data from image 32 using an imaging acquisition system 11. Imaging acquisition system 11 may include any mechanisms known in the art for collecting and processing ultrasound data, such as one or more transducers, related hardware, software, input devices, monitor, etc. In addition, ultrasound system may create an output 34. Output 34 may include, for instance, a stream of images that can be viewed in real-time, an electronic/digital file for storing image data that, e.g., allows a physician to retrospectively study the scanned image, etc. Images may be collected and/or processed as a 2D slice (i.e., frame) or a 3D volume, and the concepts described herein are applicable to both.

0016As noted above, it is often the case that a target volume 32 being scanned will have one or more specific regions of interest 33, e.g., the mitral valve in the heart, the wall of the aorta, some other important vascularity, a point, etc. However, given the limitations mentioned above, it may be impractical to scan the entire volume 32 using an imaging technique such as color flow Doppler or spectral Doppler. Because the region of interest 33 typically involves some motion or flow, the present invention automatically segments the flow region(s) from the non-flow region(s). The regions may be segmented in any shape or volume, including, but not limited to, a 3D pie slice, a cube, an arbitrary shape, a collection of shapes, etc. Once the flow region is identified, a flow imaging technique can be restrictively applied to the regions of interest.

0017To implement the above, ultrasound system 10 of the present invention includes a survey system 12 that "surveys" the volume to collect motion data. Survey system 12 can be implemented in any manner to collect any type of "survey" data that can help indicate a region of interest 33, namely, motion or flow. Once identified, a segmentation system 18 may be implemented to store the information in a motion map 20 that delineates the flow regions 22 from the non-flow regions 24. This information can then be used by one or more applications 25, as described in further detail below. While the present invention is generally described in terms of identifying motion or flow, the invention may be implemented in terms of detecting the absence of motion or flow, and such an embodiment is believed to fall within the scope of the invention.

Real-time Flow Imaging

0018As noted above, the ability to effectively collect and display real-time flow imaging data, such as color flow Doppler data, is limited by a minimum acquisition frame or volume rate (e.g., 15 - 100 Hz) needed to adequately sample the physiology. However, such a required rate of acquisition may be unachievable if an entire image is scanned using flow imaging. For instance, consider the volume 32 depicted in <figref idref="f0002">Figure 2</figref>. Volume 32 generally comprises a flow region 42, and a non-flow region 44. The flow region 42 comprises a vessel, e.g., a carotid artery, while the non-flow region comprises, e.g., muscle, fat, connective tissue, etc. If the entire image were scanned using color flow Doppler, which requires an ensemble of transmit/receive cycles (e.g., 5-12) for each line, an effective frame rate could not be maintained and aliasing errors and the like could be introduced into the display. Thus, accurate information regarding the velocity and direction of the blood flow through the vessel could not be obtained.

0019To overcome this, the present invention first segments the image into flow and non-flow regions, and then limits the use of flow imaging to the flow regions, i.e., a region of interest. Using the ultrasound system 10 depicted in <figref idref="f0001">Figure 1</figref>, survey system 12 is first applied to collect "motion" data to help indicate motion or flow. It should be recognized that any type of data indicative of motion could be collected. In one exemplary embodiment, a color flow Doppler sampling system 14 is provided that collects color flow data from the entire volume 32 at some predetermined time interval, e.g., every nth frame. In another exemplary embodiment, a contour analysis system 16 may be implemented to identify a feature (e.g., a mitral valve) around or through which flow or motion is typical. <patcit id="pcit0002" dnum="US6447453B"><text>U.S. Patent 6,447,453</text></patcit>, discloses such a system. In this case, the motion data may comprise one or more identified contours or patterns within the image.

0020Once collected, the motion data is analyzed by segmentation system 18 to specifically identify which regions within volume 32 contain flow. The presence of flow can be identified in any known manner. For instance, conventional color flow techniques can be used to determine velocities within an image, and a velocity threshold can be established that separates flow from non-flow regions. Alternatively, the power of an image signal can be analyzed to identify a flow, and a power threshold can be established that separates flow from non-flow regions. Further, in the case of a contour analysis system 16, segmentation system 18 may include a pattern recognition system. Thus, certain identified contours can be recognized as being associated with flow, while others can be recognized as being associated with non-flow.

0021Segmentation system 18 may generate a motion map 20 in the form of a 2D frame or 3D volume that indicates flow regions 22 and non-flow regions 24. This motion map 20 can then be utilized by various applications 25. In this exemplary application, the motion map can be utilized by the flow imaging acquisition system 26 to restrict flow imaging to the identified region(s) of interest 33 within the volume 32. The non-flow portion of the image can be scanned with standard grayscale imaging.

0022In one exemplary embodiment, flow acquisition system 26 may include a control system 27 that tells the image acquisition system 11 to use color flow Doppler scanning only for the region(s) of interest 33 within image 32 and grayscale for the non-flow regions. Control system 27 may, for example, comprise a system for automatically setting a focal zone position based on the color flow data and a system for automatically setting an image depth based upon the location of a peak motion signal within the color data to limit the collection of high-density data to the region of interest 33. While this embodiment is described with reference to color flow Doppler scanning, any imaging technique may be utilized, e.g., color, B-FLOW, power motion imaging, tissue Doppler imaging, pulse wave, continuous wave, etc. Because flow data collection is limited to a relatively small region of interest 33, real-time 2D or 3D color imaging for that specific region can be effectively achieved, (i.e., an adequate acquisition rate can be maintained).

0023Flow acquisition system 26 can adjust a set of acquisition parameters to effectively scan the region of interest 33. Such parameters may include, e.g., b-mode line densities, color flow line densities, pulse repetition frequency, and ensemble length.

0024As is evident, survey system 12 is only concerned with detecting the presence of motion, as opposed to, e.g., making an accurate estimate of velocity. In one embodiment, survey system could be implemented utilizing: (1) a relatively low sampling of the spatial frequencies present in the image 32; (2) a relatively low density of scan lines (i.e., lines per millimeter or degree) relative to what would be typical to form an image; and/or (3) a lower than normal ensemble or number of transmit/receive cycles per line (e.g., 2 or 3). Accordingly, survey system 12 may generally utilize a relatively non-quantitative analysis, whose properties would potentially be inadequate for clinical flow imaging.

0025Alternatively, survey system 12 could utilize a very high spatial density scan and/or high sensitivity scan as a means for collecting motion data. While such a process would take additional time, it would only need to be done once (or relatively infrequently) to accurately capture the flow fields within an image.

0026Control system 27 may also include a tracking system that allows the survey system 18 to automatically re-survey the image every so often in a continuous mode to account for movement of the thing being imaged, movement of the transducer, etc. Thus, real-time adjustments could be made, e.g., every nth frame, to ensure proper tracking of the flow and non-flow regions. Alternatively, a one-button push system could be utilized to allow the technician to manually decide when motion data was to be collected.

0027Furthermore, once the target volume 32 has been segmented, further imaging may be applied more specifically to, for example, image the general non-flow region with standard b-mode scanning and the vascular flow region with targeted color flow. The net result of this approach is a substantial improvement in frame rate. Additionally, the viewable image benefits from reduced transmission of unnecessary flow pulses and the user benefits from an automatic isolation of the flow region.

0028A further application for using the segmented data may comprise a plaque/clutter analysis system 28 that automatically adjusts the gain of the imaging acquisition system 11. In imaging vessels with some low level echoes, it would be advantageous to determine whether such echoes stem from soft plaque, or from clutter (i.e., reverb). If the echoes stem from plaque, it would be useful to automatically increase the 2D gain to make the plaque more visible. On the other hand, if the low level echoes stem from clutter, it would be useful to automatically reduce the overall gain. Using only gray-scale data, it is difficult to determine the nature of these low level echoes.

0029To address this, the present invention provides a plaque/clutter analysis system 28. For distinguishing plaque from clutter when low level echoes are present in vessel interiors, the motion signals present in the same locations as the low level echoes can be analyzed. If there is no flow signal, the low level echoes are likely to be plaque, and an increase in gain is implemented to highlight those echoes. Alternately, if flow is present where the low level echoes exist, the echoes are likely to be clutter or reverb, and a reduction in gain is then automatically propagated to achieve automatic clutter suppression.

0030Furthermore, the segmented data may be utilized by any other imaging optimizations 30. For instance, in areas like vascular imaging, where the object of interest is a vessel whose interior is anechoic, ultrasound systems tend to key off the surrounding musculature. This means that, while the muscle layers become appropriately gained, the walls of the vessels become either over-gained or under-gained. Furthermore, in over-gained situations, clutter is introduced into the lumen. The sonographer typically does not care about the presentation of the musculature but focuses only on the vessel walls and interior. In this instance, the map 20 can be used to define a tight region of interest, centered around the vessel, on the regular 2D frame that may be input to an optimization algorithm. When the optimization works on this tight region, echoes from outside muscle layers beyond the vessels can be excluded, hence reducing the instances where the bright or dark echoes from the musculature inappropriately influence the optimization of the 2D echoes in the vessel walls and interior.

Retrospective Analysis

0031As noted above, an important potential application of retrospective analysis involves fields such as cardiology, where the challenge is to accurately acquire data, such as spectral Doppler data around the valves of the heart. As noted above, in order to retrospectively analyze collected spectral Doppler data, it is critical that the sample volume be placed at a point of interest so that the data of interest can be accurately and precisely captured, e.g., a typical Doppler sample volume is 1mm in diameter and 3 mm long. This process is typically done during a lengthy exam, because the placement of the sample volume is specific to the valve being studied and is also specific to the patient. An exemplary embodiment of an application utilizing retrospective analysis is further described with reference to <figref idref="f0002 f0003">Figures 3-5</figref>, which depict a scan of a heart 40 within a 3D volume 62.

0032The present invention utilizes the concepts described above to automatically implement the process of collecting a sample volume of spectral Doppler data for retrospective analysis. To achieve this, survey system 12 of ultrasound system 10 is utilized to collect motion data. Once collected, segmentation system 18 can specifically identify and map flow locations, namely, the point of interest. Finally, the flow acquisition system 26 can be utilized to obtain a sample volume containing spectral Doppler data at the point of interest, which can be stored for later analysis.

0033In one embodiment of the invention, based on the identification of a point of interest, the look direction in the volume acquisition is determined automatically and spectral Doppler data is acquired axially along that look direction at multiple ranges. In addition, the use of multi-line beamforming to provide multiple look directions, either around the same point or potentially at arbitrary look directions may also be incorporated. To demonstrate this teclmique, <figref idref="f0002">Figure 3</figref> shows a heart 40 imaged as a volume 62 using real time ultrasound imaging.

0034Typically, a technician acquires a series of 2D images from various perspectives to help reconstruct in their mind a 3D view. The goal for 3D is the full acquisition of a sample volume in a relatively short time. However, there is no equivalent of 3D for spectral Doppler, which continues to be an important part of an ultrasound exam in cases where high flow sensitivity and temporal resolution are required. Because spectral Doppler consists of interrogating (essentially) a point in space, skill and time are required by the technician to obtain specific data. The present invention addresses this by automatically acquiring an "acquisition volume" of spectral Doppler data (including phase information) that covers a larger region than a single point of interest. This allows the technician to be less skillful, and still capture the point of interest within the acquisition volume. Then, retrospectively, either the technician or control system 27 can identify the sample volume within the acquisition volume and generate spectral Doppler for the sample volume of interest. One method for achieving this involves capturing the received data from several "range gates" and analyzing each one independently to derive the associated spectral data. Such a methodology is described in <patcit id="pcit0003" dnum="US5365929A"><text>U.S. Patent 5,365,929</text></patcit>. Because the exact range/depth of the sample volume for which the spectral data is desired is uncertain, data is captured over a wider range than would otherwise be normal. The actual sample volume can later be defined retrospectively.

0035In order to determine the proper look direction, the region of interest, e.g., a valve, must be automatically identified. One method of identifying a valve is to use survey system 12 to identify motion data as an area of high velocity flow. This can be achieved with a 3D color Doppler image that would generate a motion map 20 to indicate where flow was occurring. Another method, described above with reference to <patcit id="pcit0004" dnum="US6447453B"><text>U.S. patent 6,447,453</text></patcit>, utilizes a contour analysis system to identify motion data, e.g., using pattern recognition to detect a mitral valve. Once identified, the look direction and range of the scan line can be automatically determined by control system 27.

0036<figref idref="f0003">Figure 4</figref> shows a mitral valve 46 with a line of acquisition 48 placed automatically. Note that in this case, data is acquired for an acquisition volume 50 that includes the left ventricle and the left atrium to ensure that sufficient data is available for retrospective analysis. Note also that as the heart moves as a function of the heart cycle, the location of the Doppler imaging line of sight can be repositioned automatically by the tracking feature of control system 27.

0037In a further embodiment, the acquisition volume for Doppler data can be expanded into a conical zone 52 by the use of multi-line beamforming, as illustrated in <figref idref="f0003">Figure 5</figref>. Multi-line beamforming is a technique for receiving (focusing and steering) more than one receive beam from one transmit event (e.g., beam). Such a process is described in <patcit id="pcit0005" dnum="US6471650B2"><text>U.S. Patent 6,471,650 B2</text></patcit>. In this case, a multi-line bundle (for example, 2x2 or 4x4) covers a finite value of interest around the orifice 46. The acquisition would store the received data in such a manner that the phase information is preserved, for instance the radio frequency data or basebanded IQ data etc., enabling a flexible retrospective review of the entire acquisition volume.

0038It is understood that the systems, functions, mechanisms, methods, and modules described herein can be implemented in hardware, software, or a combination of hardware and software. They may be implemented by any type of computer system or other apparatus adapted for carrying out the methods described herein. A typical combination of hardware and software could be a general-purpose computer system with a computer program that, when loaded and executed, controls the computer system such that it carries out the methods described herein. Alternatively, a specific use computer, containing specialized hardware for carrying out one or more of the functional tasks of the invention could be utilized. The present invention can also be embedded in a computer program product, which comprises all the features enabling the implementation of the methods and functions described herein, and which - when loaded in a computer system - is able to carry out these methods and functions. Computer program, software program, program, program product, or software, in the present context mean any expression, in any language, code or notation, of a set of instructions intended to cause a system having an information processing capability to perform a particular function either directly or after either or both of the following: (a) conversion to another language, code or notation; and/or (b) reproduction in a different material form.

0039The foregoing description of the preferred embodiments of the invention has been presented for purposes of illustration and description. They are not intended to be exhaustive or to limit the invention to the precise form disclosed, and obviously many modifications and variations are possible in light of the above teachings. Such modifications and variations that are apparent to a person skilled in the art are intended to be included within the scope of this invention as defined by the accompanying claims.