Optimal organ segmentation based on ultrasound

Abstract

A segmentation selection system includes a transducer (14) configured to transmit and receive imaging energy for imaging an object. A signal processor (26) is configured to process the received imaging data to generate processed image data. A segmentation module (50) is configured to generate multiple segments of the object based on features or combinations of features of the imaging data and / or the processed image data. A selection mechanism (52) is configured to select one of the multiple segments that best satisfies the criteria used to perform the task.

Description

[0001] This application is a divisional application of patent application 201680017996.4, filed on March 15, 2016, entitled "Optimal Organ Segmentation Based on Ultrasound". Technical Field

[0002] This disclosure relates to medical instruments and, more specifically, to systems, methods, and user interfaces for image component segmentation in imaging applications. Background Technology

[0003] Ultrasound (US) is a low-cost, easy-to-use imaging modality widely used for real-time guidance and monitoring of procedures. Ultrasound image segmentation is primarily driven by the clinical need to extract organ boundaries from B-mode images as a step toward dimensional measurements such as tumor size and extent. The visualization of the geometric boundaries of organs in US images relies heavily on acoustic impedance between tissue layers. Despite its low cost and ease of use, B-mode US imaging is not necessarily best suited for anatomical imaging. Compared to other imaging modalities such as magnetic resonance (MR) and computed tomography (CT), B-mode US images suffer from poor signal-to-noise ratio and background speckle noise. Existing ultrasound-based segmentation methods utilize pixel (or voxel) information from B-mode images as input to a metric calculator (where "metric" refers to the calculated quantity, such as signal-to-noise ratio (SNR), contrast, texture, etc.).

[0004] Due to the isoechoic pixel values ​​in B-mode images, US B-mode images are also affected by the poor contrast between organs and their adjacent surrounding tissues (e.g., prostate-bladder, prostate-rectal wall). This limits the robustness of many existing automated segmentation methods. Manual segmentation remains the only possible alternative, and the accuracy achievable using this method depends heavily on the clinician's skill level. The inter-operator variability in manual US segmentation is approximately 5 mm, with the Dice coefficient (similarity metric) being 20-30% lower than that of MRI segmentation. Summary of the Invention

[0005] According to this principle, a segmentation selection system includes a transducer configured to transmit and receive imaging energy for imaging an object. A signal processor is configured to process the received imaging data to generate processed image data. A segmentation module is configured to generate multiple segments of the object based on features or combinations of features of the imaging data and / or the processed image data. A selection mechanism is configured to select one of the multiple segments that best satisfies the criteria used to perform the task.

[0006] Another segmentation selection system includes an ultrasonic transducer configured to transmit and receive ultrasonic energy for imaging the object. A mode-B processor is configured to process the received imaging data to generate processed image data. A segmentation module is configured to generate multiple segments of the object based on one or more combinations of input data and segmentation metrics. A graphical user interface allows the user to select features or combinations of features of the imaging data and / or processed image data to generate multiple segments and select the segment that best meets the criteria used to perform the task.

[0007] A method for segmentation selection includes: receiving imaging energy for imaging an object; performing image processing on the received data to generate processed image data; generating multiple segments of the object based on features or combinations of features of the original imaging data and / or the processed imaging data; and selecting at least one segment from the multiple segments that best satisfies the segmentation criteria.

[0008] These and other objects, features, and advantages of this disclosure will become apparent from the following detailed description of its illustrative embodiments, which should be read in conjunction with the accompanying drawings. Attached Figure Description

[0009] This disclosure will be presented in detail with reference to the following drawings, in which: Figure 1 This is a block diagram / flowchart illustrating an imaging system with a segmentation selection module according to one embodiment; Figure 2 This is a block diagram / flowchart illustrating the data or signal inputs of a segmentation selection module for generating multiple segments according to one embodiment; and Figure 3 This is a flowchart illustrating a segmentation selection method according to an illustrative embodiment. Detailed Implementation

[0010] Based on this principle, a system and method are provided to estimate the optimal segmentation of multiple segments from a single imaging modality (e.g., ultrasound (US)). Multiple segments can be derived from, for example, raw beam-summed US radio frequency (RF) data and / or other data metrics formed along a B-mode (luminance mode) image, combined with various segmentation metrics. Both manual and automatic methods can be selected by the user through a user interface to choose the optimal segmentation.

[0011] Ultrasound B-mode images are created by detecting the envelope of the raw RF data, followed by logarithmic compression and scan transformation. Logarithmic data compression allows for parallel visualization of a wide range of echoes. However, it also suppresses subtle variations in image contrast that are crucial for accurate segmentation. Based on this principle, segmentation metrics are computed in different US data formats (before generating conventional B-mode data). Segmentation is derived based on one or more metrics (texture, contrast, etc.). Multiple segmentation outputs are presented to the user based on the data matrix used and the segmentation metrics, where the optimal segment can be selected manually or automatically. The selected segment is overlaid on the B-mode image visualized on the screen without adding any complexity from the user's perspective. In general, unprocessed RF data can be rich in information about the boundaries between organs and tissues in a study. Including several different data streams that supplement the B-mode image during segmentation can lead to more accurate segmentation that is less sensitive to speckle noise. Therefore, multi-channel segmentation metrics are computed on pre-recorded compressed data. This principle increases the accuracy and robustness of the segmentation algorithm, thereby improving and streamlining clinical workflows.

[0012] It should be understood that the invention will be described in relation to a medical instrument used for US imaging; however, the teachings of the invention are broader and applicable to any imaging modality, for which multiple segmentation options are available. In some embodiments, the principles are used to track or analyze complex biological or mechanical systems. In particular, the principles are applicable to tracking processes within biological systems and can include processes in all areas of the body, such as the lungs, gastrointestinal tract, excretory organs, blood vessels, etc. The elements depicted in the figures can be implemented in various combinations of hardware and software and provide functionality that can be combined in a single element or multiple elements.

[0013] The functionality of the various elements shown in the figures can be provided by using dedicated hardware and hardware capable of executing software in conjunction with appropriate software. When provided by a processor, the functionality can be provided by a single dedicated processor, a single shared processor, or multiple separate processors, some of which can be shared. Furthermore, the explicit use of the terms "processor" or "controller" should not be construed as exclusively referring to hardware capable of executing software, and may implicitly include, but is not limited to, digital signal processor ("DSP") hardware, read-only memory ("ROM") for storing software, random access memory ("RAM"), non-volatile memory, etc.

[0014] Furthermore, all statements and specific examples detailing the principles, aspects, and embodiments of the invention herein are intended to cover their structural and functional equivalents. Moreover, it is intended that such equivalents include both currently known equivalents and those developed in the future (i.e., any element developed to perform the same function, regardless of its structure). Therefore, for example, those skilled in the art will recognize that the block diagrams presented herein represent conceptual views of illustrative system components and / or circuits embodying the principles of the invention. Similarly, it should be recognized that any flowchart, diagram, etc., representation can substantially represent various processes in a computer-readable storage medium and thereby executed by a computer or processor, whether or not such computer or processor is explicitly shown.

[0015] Furthermore, embodiments of the present invention may take the form of a computer program product accessible from a computer-usable or computer-readable storage medium that provides program code for use by or in conjunction with a computer or any instruction execution system. For the purposes of this specification, a computer-usable or computer-readable storage medium can be any means that may include storing, communicating, propagating, or transmitting a program for use by or in conjunction with an instruction execution system, apparatus, or device. The medium can be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system (or apparatus or device) or a propagation medium. Examples of computer-readable media include semiconductor or solid-state memory, magnetic tape, removable computer disks, random access memory (RAM), read-only memory (ROM), rigid disks, and optical discs. Current examples of optical discs include compact disc read-only memory (CD-ROM), compact disc read / write (CD-R / W), Blu-ray™, and DVD.

[0016] In this specification, references to "one embodiment" or "embodiment" and other variations thereof mean that a particular feature, structure, characteristic, etc., described in connection with the embodiment is included in at least one embodiment of the principle. Therefore, the appearance of the phrase "in one embodiment" or "in an embodiment" in various places throughout the specification, as well as any other variations, does not necessarily all refer to the same embodiment.

[0017] It should be recognized that, for example, in the cases of “A / B,” “A and / or B,” and “at least one of A and B,” any use of the following “ / ,” “and / or,” and “at least one” is intended to include a selection of only the first listed option (A), or a selection of only the second listed option (B), or a selection of both options (A and B). As another example, in the cases of “A, B, and / or C” and “at least one of A, B, and C,” such phrases are intended to include a selection of only the first listed option (A), or a selection of only the second listed option (B), or a selection of only the third listed option (C), or a selection of only the first and second listed options (A and B), or a selection of only the first and third listed options (A and C), or a selection of only the second and third listed options (B and C), or a selection of all three options (A, B, and C). As will be apparent to those skilled in the art and related fields, this can be extended for a large number of listed entries.

[0018] Now refer to the accompanying drawings, in which the same reference numerals denote the same or similar elements, and first refer to... Figure 1 An ultrasonic imaging system 10 constructed according to this principle is shown in the form of a block diagram. The ultrasonic system 10 includes a transducer device or probe 12 having a transducer array 14 for transmitting ultrasonic waves and receiving echo information. The transducer array can be configured as, for example, a linear array or a phased array, and can include piezoelectric elements or capacitive micromachined ultrasonic transducer (CMUT) elements. For example, the transducer array 14 can include a two-dimensional array of transducer elements capable of scanning in the pitch and azimuth dimensions for 2D and / or 3D imaging.

[0019] Transducer array 14 is coupled to microwave beamformer 16 in probe 12, which controls signal transmission and reception via transducer elements in the array. Microwave beamformer 16 can be integrated with flexible transducer device 12 and coupled to transmit / receive (T / R) switch 18, which switches between transmission and reception and protects main beamformer 22 from high-energy transmitted signals. In some embodiments, T / R switch 18 and other components in the system can be included in the transducer probe rather than in a separate ultrasound system substrate. Under the control of microwave beamformer 16, transmission of the ultrasonic beam from transducer array 14 is guided by transmit controller 20 and beamformer 22 coupled to T / R switch 18. Transmit controller 20 and beamformer 22 can receive input from user operation via user interface or control panel 24.

[0020] One function controlled by the transmit controller 20 is the direction in which the beam is manipulated. The beam can be manipulated to travel in a straight line (orthogonal to) the transducer array 14, or to be manipulated at different angles for a wider field of view. The partial beamforming signal generated by the microwave beamformer 16 is coupled to the main beamformer 22, where the partial beamforming signals from individual patches of the transducer elements are combined into a fully beamformed signal.

[0021] The beamforming signal is coupled to signal processor 26. Signal processor 26 is capable of processing the received echo signal in various ways, such as bandpass filtering, decimation, I and Q component separation, and harmonic signal separation. Signal processor 26 can also perform additional signal enhancement (e.g., speckle reduction, signal recombination, and noise cancellation). The processed signal is coupled to B-mode processor 28 or other mode processors (e.g., M-mode processor 29), which is capable of using amplitude detection to image structures within the body. The signals generated by mode processors 28 and 29 are coupled to scan converter 30 and multiplane reformer 32. Scan converter 30 arranges the echo signals in the spatial relationships in which they are received in the desired image format. For example, scan converter 30 can arrange the echo signals into a two-dimensional (2D) fan-shaped format or a pyramidal three-dimensional (3D) image. Multiplane reformer 32 is capable of converting echoes received from points in a common plane in a volumetric region of the body into an ultrasound image of that plane.

[0022] Volume plotter 34 converts the echo signals of the 3D dataset into a projected 3D image as viewed from a given reference point. The 2D or 3D image from scan converter 30, multiplane reformer 32, and volume plotter 34 is coupled to image processor 36 for further enhancement, buffering, and temporary storage for display on image display 38. Graphics processor 40 is capable of generating graphic overlays for display alongside the ultrasound images. These graphic overlays or parameter blocks may contain, for example, standard identification information (e.g., patient name, date and time of image, imaging parameters, frame index, etc.). For these purposes, graphics processor 40 receives input (e.g., typed patient name) from user interface 24. User interface 24 is also capable of being coupled to multiplane reformer 32 for selecting and controlling the display of multiple multiplane reformatted (MPR) images.

[0023] According to this principle, ultrasound data is acquired and stored in memory 42. Memory 42 is depicted as being centrally located; however, memory 42 can store data and interact at any point in the signal path. Depending on the position of the elements in array 14, correction can be used as feedback for correcting the beam manipulation signal (beam manipulation).

[0024] Display 38 is included for viewing internal images of an object (patient) or volume. Display 38 may also allow a user to interact with system 10 and its components and functions or any other elements within system 10. This is further facilitated by interface 24, which may include a keyboard, mouse, joystick, haptic device, or any other peripheral device or control to allow user feedback from system 10 and user interaction with system 10.

[0025] Ultrasound B-mode images are output from B-mode processor 28 and created by detecting the envelope of the raw RF data, followed by logarithmic compression and scan conversion by scan converter 30. The logarithmic data compression of B-mode processor 28 allows for parallel visualization of a wide range of echoes.

[0026] Ultrasound M-mode images are output from the M-mode processor 29. The M-mode images can be processed (e.g., compressed) and sent for scanning conversion by the scan converter 30. In M-mode (motion mode) ultrasound, pulses are rapidly and continuously emitted while A-mode or B-mode images are acquired to record continuous images. This can be used to determine the velocity of specific organ structures because organ boundaries reflect relative to the probe 12.

[0027] (Before generating traditional B-mode or M-mode data) Segmentation metrics are calculated in different US data formats and received by segmentation module 50. Segmentation module 50 exports multiple segmented images generated based on one or more metrics (e.g., texture, contrast, etc.) measured from the US data. Based on the adopted data matrix and segmentation metrics, multiple segmentation outputs are generated using image processor 36 (or graphics processor 40) for display on display 38. The segmentation outputs are presented to the user on a graphical user interface (GUI) 52 generated by image processor 36 (or graphics processor 40).

[0028] The segmentation output is presented to the user, where the optimal segmentation can be selected manually or automatically. The selected segmentation can be overlaid on a B-mode image visualized on a display 38 by the image processor 36 (or graphics processor 40). The unprocessed RF data provided prior to B-mode processing 28 (or even prior to signal processing 26) may be rich in information about the boundaries between organs or regions and tissues in the study. Introducing several different data types into the segmentation process through the segmentation module 50 can complement the B-mode image from the B-mode processor 28 and lead to more accurate segmentation that is less sensitive to speckle noise or other image degradation. Therefore, the computation of multi-channel segmentation metrics on pre-recorded compressed data can increase the accuracy and robustness of the segmentation algorithm, thereby improving and streamlining clinical workflows.

[0029] Image processor 36 is configured to generate a graphical user interface (GUI) or other selection mechanism 52 for user selection of optimal segmentation. Segmentation module 50 provides different segmentation outputs obtained from multiple combinations of input data and segmentation metrics to determine the optimal segmentation for the application under consideration. Optimality can vary based on the organ being segmented and the task under consideration. Input data refers to the use of different forms of US data in the segmentation algorithm and includes, but is not limited to, the following potential possibilities: raw beam-summed RF data before envelope detection and logarithmic compression; envelope detection data formed from raw beam-summed RF data; B-mode image formed without logarithmic compression; B-mode image formed without logarithmic compression and without applying other filters; conventional B-mode image; M-mode image; etc. Segmentation metrics refer to quantities used to characterize tissue regions, such as signal-to-noise ratio (SNR), contrast, texture, model-driven algorithms, etc. In some cases, optimality can be a subjective measurement determined by the user.

[0030] The different segments obtained will be presented to the user on display 38 via, for example, a GUI 52, for manual selection of the optimal segment. Automatic selection of the optimal segment is also anticipated, as will be described. Image processor 36 (or graphics processor 40) provides visualization of the segmentation. Different segmentation outputs are overlaid on the original B-mode image to allow the user to select the best segment. In one embodiment, image processor 36 automatically cycles through different segmentation results periodically. In another embodiment, all outputs may be visualized simultaneously in color-coded or other visual formats. When performing automatic selection of the optimal segment, the selected optimal segment will be overlaid on the B-mode image displayed on display 38.

[0031] refer to Figure 2 The block diagram / flowchart illustratively illustrates two sets of inputs to segmentation module 50, depicting input data 102 and segmentation metric 104. The inputs to segmentation module 50 represent the pipeline for US B pattern image formation, where uncompressed data is used at various stages of the pipeline. Specific combinations of input data 102 and segmentation metric 104 will provide “optimal” segmentation for specific organ sites, US system capabilities, etc. Input data 102 may include raw RF data 106 (from...) Figure 1The input data includes anywhere in the signal path before the B-mode processor 28, envelope detection data 108 (from the signal envelope (carrier)), and data from the B-mode display 114 (B-mode image or image data). M-mode images or data 115 (or other image modes) can also be used as input data. The B-mode or M-mode image can be derived from a scan transformation 112 with or without logarithmic compression 110. Segmentation metrics 104 can include statistical models 116 or the volume of the US imaging, SNR 118, contrast 120, texture 122, edge detection 124, or any other image characteristics.

[0032] For input data 102, logarithmic compression of the data is avoided to capture the full range of information for calculating the segmentation metric 104. Multiple segmentations can be provided, and the segmentation module 50 can select the "optimal" or "best" segment from among them. This selection can be performed automatically or manually.

[0033] To automatically select the optimal split, appropriate criteria can be defined. Examples of possible criteria include, for instance, whether the split volume is at least "x" cm. 3 The following metrics are used to evaluate the quality of segmentation: Does the segmented volume contain any pre-labeled anatomical landmarks? Does the segmented volume differ from the population-based average segmentation model by more than "x"%? Does the segmented shape differ from the population-based average shape model by more than "x"%? These metrics, not used to generate the segmentation, can be used to rate the quality of the segmentation.

[0034] To manually select the optimal segment, different segments can be presented to the user by overlaying different segments onto (one or more) B-mode images. For example, all segments can be overlaid, each with a different color / line style. Individual segments can be overlaid at any given time, and the user has the ability to cycle through all available segments by mouse clicks and / or keyboard shortcuts.

[0035] An illustrative GUI 52 according to an exemplary embodiment includes an image panel, wherein the image may include overlaid candidate segments. Each candidate segment may be displayed as an overlay on a B-mode image. Candidate segments may be displayed all simultaneously or sequentially (or any combination thereof). A user can select input factors such as the organ to be segmented by selecting checkboxes. A user can also select data to perform segmentation by selecting checkboxes. A user can also select data to perform segmentation using a segmentation metric by selecting checkboxes. The final segmentation may be selected by a button.

[0036] If segmentation is performed automatically, the segmentation obtained from each combination of input factors is shown to the user for selection of the "optimal" segment. If segmentation is to be performed manually by the user, multiple images (e.g., beam-added RF images, envelope-detected images, B-mode images with and without filtering, etc.) are shown to the user sequentially. The user performs manual segmentation on the displayed images and has the option to accept or reject it. The user can choose between automatic and manual operation modes. After the user selects the optimal segmentation, it is finally completed by clicking the "Complete Segmentation" button.

[0037] It should be understood that the GUI 52 described above is for illustrative purposes. The interface can be developed or extended as needed to include more functionality. The types and locations of features on GUI 52 can be changed or reorganized as needed or desired. Additional buttons or controls can be used, or some buttons or controls can be removed.

[0038] This principle provides segmentation techniques applicable to various fields. Examples of accurate segmentation leading to accurate image registration are applicable to fields such as adaptive treatment planning for radiation therapy, in-process treatment monitoring (e.g., brachytherapy, RF ablation), real-time biopsy guidance, and so on. It should be understood that although described in relation to US imaging, the segmentation selection aspects based on this principle can be used in place of US for other imaging modalities. For example, segmentation selection can be applied to MRI, CT, or other imaging modalities.

[0039] This embodiment can be used to provide enhanced organ segmentation capabilities, complementing tracking technologies used in interventional devices (e.g., EM tracking, ultrasound tracking). Furthermore, the segmentation method and user interface can be integrated into existing commercial systems without providing user access to the raw data. The automated segmentation capability based on this principle can be adopted with improved accuracy in any clinically available imaging system, and particularly in US imaging systems.

[0040] refer to Figure 3 A method for segmentation selection is illustrated according to an illustrative embodiment. In block 302, imaging energy for imaging an object is received. The imaging energy may include ultrasound, but other imaging modalities and energy types may be employed. In block 304, the received data is subjected to image processing to generate processed image data. The processed image data may include logarithmic compression or other filtering or compression. The processing may include scan conversion and / or B-mode processing of the data. Other forms of processing are also anticipated.

[0041] In box 306, multiple segments for the object are generated based on features or combinations of features from the raw imaging data and / or processed imaging data. The raw imaging data may include raw radio frequency data, envelope detection data, signal-to-noise ratio data, contrast data, texture data, edge detection data, etc. The processed imaging data may include statistical model comparisons, compressed data, transformed data, B-mode processed display data, etc.

[0042] Segments can be generated based on different combinations of raw and processed data to produce segments that differ from each other. Generating multiple segments can include generating multiple segments based on one or more combinations of input data and segmentation metrics. For example, a segment can be generated using a specific segmentation metric and a specific type of input data. Input data can include raw RF data, envelope detection data, B-mode display data, etc. Segmentation metric information can include statistical model comparison data, signal-to-noise ratio data, contrast data, texture data, edge detection data, etc. Other segmentations can combine input data with segmentation metrics or combine aspects of the input data with aspects of the segmentation metric. As an example, raw RF data can be combined with contrast and texture data to generate a segment. Other combinations are anticipated.

[0043] In box 308, at least one segment from a plurality of segments is selected that best satisfies the segmentation criteria. The selection criteria may include using desired aspects or automatic criteria. Multiple segments may be generated using features or combinations of features from the raw imaging data and / or processed imaging data. The segment that best satisfies the user-defined criteria may be selected manually or automatically based on programmed criteria such as contrast or pixel thresholds, best-fit images with statistical model shapes, etc.

[0044] In box 310, a graphical user interface is generated and multiple segments are displayed. The segments are preferably displayed on a B-mode image (background), wherein the segmented images are displayed simultaneously or sequentially according to user preference.

[0045] In box 312, the selected segmentation is used to perform operational procedures or other tasks.

[0046] When interpreting the appended claims, it should be understood that: a) The word “comprising” does not exclude the presence of any other elements or actions besides those listed in the given claim; b) The word "one" or "a" preceding an element does not preclude the existence of multiple such elements; c) Any reference numerals in the claims do not limit the scope of the claims; d) Several “units” can be represented by the same items or by a structure or function implemented in hardware or software; and e) Unless otherwise specifically instructed, this is not intended to require a particular order of actions.

[0047] Preferred embodiments for optimal organ segmentation based on ultrasound have been described (these are intended to be illustrative and not restrictive). It should be noted that modifications and variations can be made by those skilled in the art based on the above teachings. Therefore, it should be understood that changes can be made to specific embodiments of the disclosed content, which are within the scope of the embodiments disclosed herein as described in the appended claims. Thus, the details and specificities required by patent law have been described, and the contents claimed and desired for protection under a patent license are set forth in the appended claims.

Claims

1. A segmentation selection system, comprising: A transducer (14) is configured to transmit and receive ultrasound imaging energy for generating imaging data to image an object; At least one signal processor (26) is configured to process the received imaging data to generate processed image data; A segmentation module (50) is configured to generate multiple segments of the object based on one or more combinations of input data and at least one segmentation metric, wherein the input data includes the imaging data and the processed image data, wherein the calculation of the at least one segmentation metric is provided on the imaging data before generating the processed image data, wherein the segmentation metric is a calculated image quantity characterizing a tissue region of the object, and wherein the imaging data includes at least one of raw radio frequency data and envelope detection data; and The selection mechanism (52) is configured to select at least one segment from the plurality of segments that best satisfies the criteria for performing the task.

2. The system according to claim 1, wherein, The transducer (14) includes an ultrasonic transducer and the at least one signal processor includes a mode-B processor (28).

3. The system according to claim 2, wherein, The imaging data also includes at least one of signal-to-noise ratio data, contrast data, texture data, and / or edge detection data.

4. The system according to claim 2, wherein, The processed imaging data includes at least one of statistical model comparison and / or B-mode display data.

5. The system according to claim 1, wherein, The selection mechanism (52) includes an image processor (36) configured to automatically select segments that best meet the criteria for performing the task, based on programmed criteria.

6. The system according to claim 1, wherein, The selection mechanism (52) includes a graphical user interface (52) that allows a user to select features or combinations of features of the imaging data and / or the processed image data to generate the plurality of segments and allows the user to manually select the segment that best satisfies the criteria.

7. The system of claim 1, further comprising a display (38) for displaying the plurality of segmented images, wherein, The image is displayed on the B-mode image in one of the following ways: simultaneously or sequentially.

8. A segmentation selection system, comprising: An ultrasonic transducer (14) is configured to transmit and receive ultrasonic energy for generating imaging data to image an object. A mode B processor (28) is configured to process the received imaging data to generate processed image data; A segmentation module (50) is configured to generate multiple segments of the object based on one or more combinations of input data and at least one segmentation metric, wherein the input data includes the imaging data and the processed image data, wherein, prior to generating the processed image data, a calculation of the at least one segmentation metric is provided on the imaging data, wherein the segmentation metric is a calculated image quantity characterizing a tissue region of the object; and A graphical user interface (52) allows a user to select features or combinations of features of imaging data and processed image data to generate the plurality of segments and allows the user to select the segment that best meets the criteria for performing the task, wherein the imaging data includes at least one of raw radio frequency data and envelope detection data.

9. The system according to claim 8, wherein, The input data includes at least one of raw radio frequency data, envelope detection data, and / or B-mode display data.

10. The system according to claim 8, wherein, The segmentation metric includes at least one of statistical models, signal-to-noise ratio data, contrast data, texture data, and / or edge detection data.

11. The system of claim 8 further includes an image processor (36) configured to automatically select segments that best satisfy the criteria for performing the task based on programmed criteria.

12. The system of claim 8, further comprising a display (38) for displaying the plurality of segmented images, wherein, The image is displayed on the B-mode image in one of the following ways: simultaneously or sequentially.

13. A method for segmentation selection, comprising: Receive (302) ultrasonic imaging energy for imaging the object; The received imaging data is processed (304) to generate processed image data; Multiple segments of the object are generated (306) based on one or more combinations of input data and at least one segmentation metric, wherein the input data includes the imaging data and the processed image data, wherein the calculation of the at least one segmentation metric is provided on the imaging data before generating the processed image data, wherein the segmentation metric is a calculated image quantity used to characterize the tissue region of the object, wherein the imaging data includes at least one of raw radio frequency data and envelope detection data; and Select at least one segment from the plurality of segments (308) that best satisfies the segmentation criterion.

14. The method according to claim 13, wherein, The selection (308) includes automatically selecting segments based on programming criteria.

Images