Systems and methods for concurrent ultrasound elastography and attenuation

Abstract

Example embodiments of the described technology provide systems and methods for ultrasound imaging. A method of ultrasound imaging may comprise inducing propagation of shear waves through a region of tissue. The method may also comprise acquiring a three-dimensional sweep of ultrasound data of the tissue region. The ultrasound data may be processed to identify voxels corresponding to a particular tissue type by tracking the propagation of the shear waves through the tissue region. Spatially registered elastography images and attenuation images of the tissue region may be generated from the acquired ultrasound data.

Description

SYSTEMS AND METHODS FOR CONCURRENT ULTRASOUND ELASTOGRAPHY AND ATTENUATIONCross-Reference to Related Applications

[0001] This application claims priority from, and for the purposes of the United States the benefit under 35 USC 119 in connection with, United States application No. 63 / 563299 filed 8 March 2024 which is hereby incorporated herein by reference.Field

[0002] The present disclosure relates to ultrasound systems and methods. Some embodiments provide systems and methods useful for concurrently measuring elastography and attenuation properties of tissue.Background

[0003] Ultrasound is commonly used to image tissue. To acquire image data, ultrasound pulses are transmitted into the tissue and reflected signals (i.e. “echos”) are received back from the tissue. Properties of the imaged tissue can be deduced from the received echo signals.

[0004] Tissue elasticity may be measured by acquiring ultrasound elastography data as described, for example, in: A. Baghani et aL, “Real-time quantitative elasticity imaging of deep tissue using free-hand conventional ultrasound” in Medical Image Computing and Computer-Assisted Intervention -MICCAI 2012, 2012, pp. 617-624 and in US patent No. 10667791 to BAGHANI et al. and entitled ELASTOGRAPHY USING ULTRASOUND IMAGING OF A THIN VOLUME.

[0005] The attenuation of ultrasound signals in tissue may be measured by acquiring ultrasound attenuation data as described, for example, in: Kuc, R., 1980, “Clinical application of an ultrasound attenuation coefficient estimation technique for liver pathology characterization”, IEEE Transactions on Biomedical Engineering, (6), pp.312-319; Ophir, J. et aL, 1984, “Attenuation estimation in reflection: progress and prospects” Ultrasonic Imaging, 6(4), pp.349-395; and US patent No. 4621645 to Flax and entitled METHOD OF ESTIMATING TISSUE ATTENUATION USING WIDEBAND ULTRASONIC PULSE AND APPARATUS FOR USE THEREIN.

[0006] There is a need for improved systems and methods for ultrasound imaging. There is also a need for improved systems and methods for measuring differenttissue properties of imaged tissue using ultrasound.Summary

[0007] This invention has a number of aspects. These include, without limitation:• systems and methods for ultrasound elastography imaging;• systems and methods for ultrasound attenuation imaging;• system and methods for ultrasound backscatter estimation;• systems and methods for conditioning acquired image data;• systems and methods for computing one or more metrics of health of a tissue;• system and methods for combining quantitative ultrasound methods to compute one or more characteristics of tissue, such as estimated fat content;• and• methods and apparatus for measuring quality of ultrasound imaging.

[0008] One aspect of the invention provides a method for ultrasound imaging. The method comprises: inducing propagation of shear waves through a region of tissue; acquiring a three-dimensional sample set of ultrasound data of the region responsive to the induced shear waves; processing the sample set to identify voxels of interest based on at least one of: propagation of the shear waves through the tissue region; and one or more quality metrics; and determining ultrasound elastography values, ultrasound attenuation values and ultrasound backscatter values corresponding to the identified voxels based on the sample set.

[0009] Determining ultrasound elastography values, ultrasound attenuation values and ultrasound backscatter values may occur based on the same sample set.

[0010] The method may comprise combining any plurality of the ultrasound elastography values, the ultrasound attenuation values and the ultrasound backscatter values at one or more voxels of interest to create a health metric for one or more voxels of interest or for the sample set.

[0011] Processing the sample set to identify voxels of interest may be performed prior to determining the ultrasound elastography values, ultrasound attenuation values and / or ultrasound backscatter values.

[0012] Processing the sample set to identify voxels of interest may be performed after determining the ultrasound elastography values, ultrasound attenuation values and / or ultrasound backscatter values.

[0013] Processing the sample set to identify voxels of interest may comprise determining the voxels of interest to correspond to a particular tissue type.

[0014] Processing the sample set to identify voxels of interest may comprise: determining, based on the sample set, that particular voxels correspond to blood vessels or tissue that is not of interest; and excluding, from the voxels of interest, such particular voxels.

[0015] Processing the sample set to identify voxels of interest may comprise: determining, based on the sample set, displacement of tissue induced by the shear waves; and applying a thresholding process to the displacement to identify the voxels of interest (or voxels that are not of interest).

[0016] Processing the sample set to identify voxels of interest may comprise: applying a thresholding process to the ultrasound elastography values to identify the voxels of interest (or voxels that are not of interest).

[0017] Processing the sample set to identify voxels of interest may comprise: using a trained machine learning (artificial intelligence) engine to identify the voxels of interest based on one or more of the sample set, the ultrasound elastography values, the ultrasound attenuation values and the ultrasound backscatter values.

[0018] The method may comprise using a trained machine learning (artificial intelligence ) engine to determine tissue type, one or more characteristics of shear wave propagation and / or another quality metric for one or more voxels of interest or for the sample set.

[0019] Processing the sample set to identify voxels of interest may comprise: determining, based on the sample set, a quality metric (e.g. the signal-to-noise ratio and / or the like) for particular voxels; and applying a thresholding process to the quality metric to identify the voxels of interest (or voxels that are not of interest).

[0020] Processing the sample set to identify voxels of interest may be performed at least partially concurrently with acquiring the sample set.

[0021] Processing the sample set to identify voxels of interest may be performed prior to determining the ultrasound elastography values, ultrasound attenuation values and ultrasound backscatter values. Determining the ultrasound elastography values, ultrasound attenuation values and ultrasound backscatter values may be performed only for the voxels of interest.

[0022] Acquiring the three-dimensional sample set may comprise: parsing a three- dimensional volume to be sampled into a plurality of frames, each frame comprising a plurality of lines; grouping the plurality of lines corresponding to each frame into a plurality of groups, each group comprising a plurality of lines; and repeating the following steps until all of the groups have been traversed: acquiring samples from each line in a particular group a plurality of times within one period of the shear waves; and changing the particular group.

[0023] Changing the particular group may comprise changing from a current particular group to a spatially sequential group.

[0024] Changing the particular group may comprise changing from a current particular group to a non-spatially sequential group.

[0025] Acquiring samples from each line in the particular group may comprise traversing the lines in the group in a spatially sequential order.

[0026] Acquiring samples from each line in the particular group may comprise traversing the lines in the group in a non-spatially sequential order.

[0027] Another aspect of the invention provides an ultrasound apparatus comprising: an exciter for inducing propagation of shear waves through a region of tissue; one or more ultrasound transducers for acquiring a three-dimensional sample set of ultrasound data of the region responsive to the induced shear waves; a controller configured to: process the sample set to identify voxels of interest based on the propagation of the shear waves through the tissue region; and determine ultrasound elastography values, ultrasound attenuation values and ultrasound backscatter values corresponding to the identified voxels based on the sample set.

[0028] The controller of the ultrasound apparatus may be configured to perform any of the method features, combinations of method features and / or sub-combinations of method features recited above.

[0029] Other aspects of the invention provide methods comprising any features, combinations of features, and / or sub-combinations of features disclosed herein and / or in the accompanying drawings.

[0030] Other aspects of the invention provide apparatus comprising any features, combinations of features, and / or sub-combinations of features disclosed herein and / or in the accompanying drawings.

[0031] Further aspects and example embodiments are illustrated in the accompanying drawings and / or described in the following description.

[0032] It is emphasized that the invention relates to all combinations of the above features, even if these are recited in different claims.Brief Description of the Drawings

[0033] The accompanying drawings illustrate non-limiting example embodiments of the invention.

[0034] Figure 1 is a schematic illustration of an ultrasound system according to an example embodiment of the invention, including a controller (e.g. one or more computer(s) and / or similar data processor(s)) for signal synchronization and signal processing.

[0035] Figure 2 is a schematic illustration of different transducer types.

[0036] Figure 3 is a schematic illustration of an example ultrasound transducer and example image frame acquired using the ultrasound transducer.

[0037] Figure 4A is a schematic illustration of how an exciter can be integrated into a bed.

[0038] Figure 4B is a schematic illustration of how an exciter can be integrated into or mounted on a transducer.

[0039] Figure 5 is a flow chart showing an example method for measuring one or more metrics of health of a desired tissue (e.g. liver tissue).

[0040] Figure 6A, 6B and 6C show an example ultrasound transducer and different methods for collecting data.

[0041] Figure 7 is a schematic illustration of a mask generated according to an example embodiment of the invention.

[0042] Figure 8 is a schematic block diagram of example steps in a method for determining tissue stiffness according to an example embodiment of the invention.

[0043] Figure 9 is a schematic block diagram of example steps in a method for determining ultrasound attenuation according to an example embodiment of the invention.

[0044] Figure 10 and 10A are schematic illustrations of a graphical user interfaceaccording to an example embodiment of the invention.

[0045] Figure 11 A and 1 1 B are plots of correlation between MRI-PDFF and VDFF for the training cohort and validation cohort respectively, showing particular experimental data.Detailed Description

[0046] Throughout the following description, specific details are set forth in order to provide a more thorough understanding of the invention. However, the invention may be practiced without these particulars. In other instances, well known elements have not been shown or described in detail to avoid unnecessarily obscuring the invention. Accordingly, the specification and drawings are to be regarded in an illustrative, rather than a restrictive sense.

[0047] One aspect of the invention described herein provides an ultrasound system that is configurable to acquire ultrasound image data of a tissue region of a patient and process the ultrasound image data to obtain co-registered elastography, attenuation and backscatter images. The co-registered images may be further processed to compute one or more metrics indicative of the health of tissues in the tissue region, such as the estimation of liver fat. The one or more measures may be based on, elastography, attenuation and / or backscatter properties of the imaged tissue. Using the same data set for generating elastography data, attenuation data and / or backscatter data advantageously causes the data to be automatically spatially registered. Using the same data set to generate these images may additionally reduce data collection and therefore examination time, improve accuracy of the measured properties, and reduce the likelihood of artifacts being introduced (e.g. during spatial registration). In some embodiments the ultrasound system is configured to generate and display a quality metric that indicates quality of the ultrasound data. The quality metric may be displayed to an operator of the ultrasound system in real time.

[0048] The present technology may be applied to any type of human or animal tissue. However, the inventors have determined that the tissue characteristics that can be measured by the present technology have particularly strong correlation to the health of liver tissue.

[0049] The liver performs several vital functions such as filtering blood from the digestive tract, detoxifying blood and metabolizing biochemical compounds (e.g. carbohydrates, proteins, fats, etc.). A patient suffering from liver disease typically has a liver which cannot properly perform these functions. As liver disease progresses, the liver’s ability to perform these functions is typically reduced.

[0050] The cells of diseased liver tissue often contain more fat than the cells of healthy liver tissue and are often inflamed. Sustained inflammation of the liver cells typically results in scarring of the liver. Significant scarring of the liver may eventually lead to liver failure, which can be deadly if a liver transplant is not available. Significant scarring is also associated with increased risk of developing liver cancer.

[0051] The amount of fat present in liver tissue may be quantified by measuring attenuation and backscatter properties of ultrasound signals propagating in the liver tissue. Ultrasound signals interact differently with liver tissue having differing amounts of fat content. This difference in interactions can be quantified by measuring the attenuation and backscatter of the ultrasound signals. An increased fat content and chronic inflammation leads to tissue scarring. Scarring of the liver tissue results in increased amounts of fibrous tissue within the liver. Fibrous liver tissue is less elastic or stiffer than healthy liver tissue. Ultrasound elastography of liver tissue may therefore be correlated to the health of the liver tissue.

[0052] The progress of liver disease may be assessed based on measures of the fat content of liver tissue (which can in turn be determined by ultrasound attenuation and backscatter measurements) and the degree to which the liver tissue is affected by scarring (which can in turn be assessed by ultrasound elastography measurements). For example, if fat content is low and scarring is low it is likely that the liver tissue is healthy. As another example, if fat content is high but scarring is low, the liver may be in an early and more reversible stage of liver disease.

[0053] It would be beneficial to have simple metrics or “scores” that provide an indication of the overall health of liver tissue. Such metrics may be used to monitor changes in liver health over time and to assess the effectiveness of various treatments. The inventors have determined that such metrics may be generated by combining information from ultrasound elastography, ultrasound attenuation, ultrasound backscatter and / or other ultrasound property measurements.

[0054] Figure 1 schematically shows an pvamnip ultrasound system 10 operable toconcurrently provide a measure of elastography, attenuation, backscatter and other ultrasound properties of an imaged tissue region of patient P.

[0055] System 10 comprises an ultrasound unit 12 having a controller 13 for signal synchronization and processing. An ultrasound transducer 14 is coupled to controller and is coupled to, or forms part of, ultrasound unit 12. Transducer 14 is positioned adjacent to the skin of a patient P proximate to a region or volume of tissue that is to be imaged.

[0056] Figure 2 schematically shows non-limiting examples of transducers 14A, 14B, 14C any of which may be used to implement transducer 14 of ultrasound system 10. Transducer 14 may, for example, comprise a 1 D array of transducer elements 14A or 2D array of transducer elements (not shown). Transducer 14 may comprise a linear transducer face (as is the case with transducers 14B and 14C) or non-linear transducer face (as is the case with the convex transducer 14A).

[0057] Ultrasound unit 12 is operative to obtain ultrasound images of the region or volume of tissue of interest (e.g. all of, or a portion of, a person’s liver). The basic principles of ultrasound imaging are well known and are not described herein.Ultrasound unit 12 may, for example, comprise commercially available ultrasound imaging components.

[0058] Images of a 3D volume of tissue may be obtained, for example, by beam steering using a 2D transducer array or by moving a 1 D transducer array. A transducer 14 may be moved relative to patient P to acquire a larger volume of ultrasound data. Figure 3 schematically shows one method of how a transducer 14 may be pivoted angularly (e.g. over some arc or angular range T about an axis A) and / or translated (e.g. along a line 106), while in acoustic contact with the skin of patient P. The resulting 2D images 105 can be reconstructed into a volume 104.

[0059] In some embodiments, transducer 14 is positioned manually relative to patient P by an operator of system 10. In some embodiments transducer 14 is positioned and / or moved by a mechanical system. In some embodiments the positioning or transducer 14 is controlled by a robotic system. In some embodiments transducer 14 is positioned and / or moved by electronic beam steering using a 2D transducer array.

[0060] To measure elastography properties of the imaged tissue, ultrasound unit 12 tracks movements of the imaged tissue in response to shear waves which propagatethrough the tissue. The shear waves may be induced into the tissue by one or more exciters or activators 15 (see Figure 1 ). An exciter 15 is a device which can generate shear waves that may propagate through tissues of patient P. For example, exciter 15 may comprise: an electromagnetically driven vibration transducer, an unbalanced rotor, a vibrating table, and / or the like. For example, exciter(s) 15 may be constructed as described in International patent publication WO 2018 / 000103 A1 which is hereby incorporated herein by reference for all purposes.

[0061] Figure 4 shows two non-limiting examples of how exciter 15 may be positioned. An exciter 15 may be positioned proximate to the region of tissue to be imaged. For example, in Figure 4A, exciter 15 may comprise a vibrating pad which is positioned underneath patient P on a bed 17. In some embodiments, shown in Figure 4B, the exciter 15 is integrated into transducer 14.

[0062] Controller 13 controls transducer 14 and exciter 15 to acquire ultrasound imaging data according to a desired imaging plan. Controller 13 may also receive, and process received echo signals from transducer 14. If system 10 comprises sensing and / or positioning systems that are operative to sense and / or control the position of transducer 14, controller 13 may use sensors to track the position of the transducer 14 for image formation, and / or to control the positioning system to vary the position of transducer 14.

[0063] In some embodiments, system 10 comprises computing resources (not expressly shown in Fig. 1 ) and a display 16. Computing resources may comprise central processing units, graphics processing units, field programmable gate arrays, or application-specific integrated circuits, or any combination thereof, and may be used to process the echo data from the transducer 14, transducer location (position and orientation) data, and perform the control to vary the position of the transducer 14.Display 16 may display information which may include, without limitation, any one or more of:• acquired imaging data;• a subset of acquired imaging data (e.g. an enlarged region of interest);• a quality of acquired imaging data (e.g. based on a signal to noise ratio);• settings of system 10 (e.g. set parameters (frequency, amplitude, etc.) of pulse signals);• a desired imaging plane;• one or more measures of quality of the acquired data;• one or more measures of health of the imaged tissue;• patient particulars;• position of the transducer relative to the collected data;• position of the transducer relative to the patient;• resulting images of B-mode, elasticity, attenuation or backscatter;• etc.

[0064] Figure 5 is a flow chart showing an example method 20 for measuring one or more metrics of health of a desired tissue (e.g. liver tissue) of patient P according to a particular embodiment.

[0065] In block 21 A, exciter 15 is operated to transmit shear waves into patient P. In block 21 B transducer 14 is used to acquire image data 22 of the desired tissue region. The imaging in block 21 B may take place after exciter 15 has been operating long enough for the shear waves to reach a steady state.

[0066] For elastography measurements it is desirable to obtain plural images of the same region or volume of tissue so that tissue displacements resulting from the shear waves generated by exciter 15 may be characterized. Advantageously, as described elsewhere herein, elastography, attenuation, backscatter and other ultrasound properties may all be computed from image data 22 without having to spatially coregister different data sets together (i.e. if separate datasets were used to compute the different properties (e.g. elastography, attenuation and backscatter properties) and those properties were needed at the same location, the different datasets would have to be spatially registered to one another to find overlapping regions where all the measurements exist).

[0067] In block 23, image data 22 is processed to yield an elastography image 23A. In block 24, image data 22 is processed to yield an attenuation image 24A. In block 26, image data 22 is processed to yield a backscatter image 26A. Blocks 23 and 24 and 26 may be performed concurrently or in any order.

[0068] As discussed above, it can be desirable to provide a simple metric which is indicative of overall liver health or combine certain parameters to provide metrics such as an estimate of liver fat. Such a metric may be determined based onelastography image 23A and attenuation image 24A or determined based on attenuation image 24A and backscatter image 26A.

[0069] Although elastography image 23A and attenuation image 24A may image (e.g. capture) all or part of a tissue of interest (e.g. the liver), at least some pixels or voxels within acquired image data 22 typically correspond to tissues other than the tissue of interest (e.g. blood vessels within or outside of the liver or surrounding organs such as kidneys, gall bladder, etc., bodily fluids, cavities within the abdomen and / or the like) which are within the field of view of transducer 14. The properties of such voxels are typically not relevant to the properties of the tissue of interest. The inventors have determined that a metric that indicates overall liver health can be improved by omitting data corresponding to such non-relevant tissues.

[0070] Additionally or alternatively, some pixels or voxels within acquired image data 22 may have a low signal to noise ratio (e.g. as a result of imaging depth, a face of transducer 14 may not fully contact patient P’s skin surface, etc.). Measurements based on the values of such pixels or voxels may be unreliable. The inventors have determined that providing feedback to a user that indicates quality of the ultrasound data 22 being acquired can help the user to acquire higher quality ultrasound data 22 that, in turn, can provide higher quality results. The inventors have also determined that a metric that indicates overall liver health can be improved by omitting data derived from parts of ultrasound data that are of low quality (e.g. low signal to noise ratio).

[0071] In block 25 one or more masks 25A identifying the locations of pixels and / or voxels that do not correspond to tissues of interest and / or pixels and / or voxels for which the SNR in image data 22 (and / or some other metric of quality) is undesirably low are generated.

[0072] Mask(s) 25A may optionally be applied to elastography image 23A and / or attenuation image 24A or backscatter image 26A in block 27 to exclude the undesirable pixels or voxels.

[0073] One or more metrics of health of the desired tissue may optionally be computed in block 27 based on the values for pixels or voxels in one or more of elastography image 23A, attenuation image 24A and / or backscatter image 26A

[0074] Although the steps of method 20 have been described sequentially, it isunnecessary to perform method 20 sequentially in all cases. For example, elastography image 23A may be generated before, concurrently with or after images 24A and 26A.

[0075] In some embodiments, method 20 proceeds to process acquired image data 22 (e.g. compute parts of elastography image 23A, compute parts of attenuation image 24A, compute parts of mask(s) 25, etc.) before all of image data 22 has been acquired. This may advantageously provide interim results to an operator of system 10, allowing the operator to adjust the data acquisition if necessary. The interim results may include one or both of data corresponding to the imaged tissue and data corresponding to the quality of the data being acquired.

[0076] Examples of ways to perform individual steps of method 20 will now be described in more detail.Imaging Sequence

[0077] Figure 6A schematically shows an example ultrasound transducer 14 having a face 14A.

[0078] Example transducer 14 has a field of view 30. Field of view 30 may be imaged by sequentially transmitting ultrasound pulses along lines 30A-1 to 30A-N (generally or collectively lines 30A) and receiving echo signals corresponding to each of lines 30A-1 to 30A-N. The exact value of N can vary. N is usually in the range of 64 to 512. For ultrasound elastography, it may be desirable to analyze the shear wave pattern induced by exciter 15 and that this pattern be examined nearly simultaneously at different spatial locations. This is desirable, for example, to accurately find the wavelength of the shear wave, or to find the phase gradient over a relatively large spatial extent.

[0079] For example, where exciter 15 has a frequency of 50 Hz, the tissue will oscillate with a period of 20 ms. This can present a challenge because of the time involved in obtaining each line 30A-1 to 30A-N. For example, the time to obtain each line 30A-1 to 30A-N may be at least 130 ps (assuming that a length of each line is 20 cm and that the speed of sound in tissue is 1540 m / s). Therefore, even if N is only 128 it would take a minimum of about 17 ms to acquire one frame. There would not be enough time to acquire more than one frame in one period of the shear wave.

[0080] This challenge may be addressed by sampling some or all of lines 30A beforecompleting a full frame that includes all of lines 30A. There are various ways to sequence the acquisition of data for lines 30A. For example:• Figure 6B shows a methods of dividing lines 30A into groups 60-1 , 60-2 ... 60- M (collectively, groups 60) of lines 30A, where each group 60 contains a number of lines that is small enough so that each line 30A in the group 60 may be acquired three or more times (t) within one period of the shear waves. After data for any particular group 60 of lines 30A has been acquired, data for another one of the groups 60 of lines 30A may be acquired. Each group 60 may, for example, be made up of 10 to 40 lines 30A. In some embodiments, lines 30A within a group 60 of data and / or groups 60 of data may be acquired sequentially (e.g. in some sort of spatial order). This is not necessary. Lines 30A within groups 60 may generally be acquired in any spatial sequence; lines 30A in different groups may be acquired in any spatial sequence; and different groups 60 may be acquired in any spatial sequence. Groups 60 may be selected such that groups 60 span the field of view of transducer 14 as shown in Figure 6C.

[0081] Image data 22 contains a plurality of sets of data for each line 30A, where each set of data is acquired from ultrasound pulses transmitted along the line 30A at a corresponding plurality of three or more different times (e.g. 30A-1 , Ti ; 30A-1 , T2; ...; 30A-1 , TK). Ideally the plurality of times are spaced apart sufficiently to sample different phases of a shear wave at locations along the line.

[0082] For each line 30A, the received echo data comprises spatial data corresponding to at least one value in the x (e.g. lateral) direction and a plurality of values in the / direction (along the line). The plural sets of data acquired for each line 30A (e.g. 30A-1 , Ti ; 30A-1 , T2; ... ; 30A-1 , TK) may be grouped together as echo data l(x,y,t) wherein (x, y) corresponds to positions of points along the individual line 30A and t corresponds to the time at which each spatial point was interrogated. The value of t corresponding to a specific point (x, y) depends both on the time when an ultrasound pulse was transmitted along the line containing point (x, y) and how far along that line (x, y) is located (since it takes transmitted ultrasound more time to reach points that are farther from transducer 14 due to the finite speed of sound in tissue).

[0083] For example, if plural sets of data are acquired for each individual line 30Abefore acquiring data from the next individual line 30A and the individual lines 30A are imaged sequentially, t may be as follows:• t = h, t2, ts for a point along the first line 30A-1 ;• t = ts+i, ts + 2, tes for a corresponding point along the second line 30A-2;• t = t2s+ 1, t2s+2, fes for a corresponding point along the third line 30A-3;• etc.

[0084] To acquire data corresponding to a three-dimensional volume of tissue of patient P (i.e. a plurality of two-dimensional frames corresponding to different fields of view 30), transducer 14 may be pivoted (e.g. about axis A by an angle T). In some embodiments, an amount by which transducer 14 has been pivoted may be measured by transducer 14 directly. For example, transducer 14 may comprise a gyroscope and accelerometer configured to measure changes in orientation of transducer 14.

[0085] Lines 30 may have different arrangements. For example, lines 30 may fan out from transducer 14 with a curvilinear array transducer, lines 30 may be closely spaced apart parallel lines, when using a linear array transducer and so on. It is desirable that the volume of tissue of interest (e.g. a volume that includes a significant portion of a patient’s liver) be imaged at a reasonably high resolution e.g. the resolution id desirably high enough to sample several times, 5 for example, for each wavelength. The wavelength may vary depending on the elasticity of the tissue imaged.

[0086] Transducer 14 may have any type of face (e.g. a convex face, a concave face, etc.). The field of view 30 may differ depending on factors such as the configuration of the face of transducer 14, beamforming applied to transducer 14, etc. For example, for a transducer 14 having a convex shape, field of view 30 may be fan shaped (e.g. lines 30A diverge outwardly to form the fan shape) rather than rectangular. The systems and methods described herein for imaging tissue may be applied using transducers having any of a wide variety of configurations.

[0087] In some embodiments, acquired sets of data for individual lines 30A are stored in one or more buffers. In some embodiments, system 10 comprises a buffer for each line 30A. Data stored in the buffers may be used to generate ultrasound images in real time. Once a sweep (e.g. acquiring a 3D sample set of the ultrasound (acoustic) energy) of the tissue region of interest is complete, data stored in the buffers may bestored as image data 22 (e.g. in a data store, in cloud-based memory, etc.).Generation of Mask

[0088] As described elsewhere herein, identifying the locations of pixels and / or voxels that do not correspond to tissues of interest and / or pixels and / or voxels for which the applied quality metric, e.g. signal-to-noise ration, in image data 22 is undesirably low may be desirable.

[0089] When imaging the liver, for example, vessels containing fluid such as blood vessels or bile ducts of the liver may be captured. Such fluid enclosures have different elastography and quantitative ultrasound properties than the liver tissue itself and may be beneficially excluded when determining a metric that indicates the overall health of the liver tissue. Different tissue types, other than liver tissue, may also have different elastography and quantitative ultrasound properties than the liver tissue itself and may be beneficially excluded when determining a metric that indicates the health of the liver tissue.

[0090] The methods of computing elastography and / or attenuation and / or backscatter described elsewhere herein may also exhibit reduced efficacy in regions where image data 22 is noisy or has no speckle. Typically, ultrasound data may become noisier with depth. Noisy data or data having no speckle may be identified and ignored.

[0091] Additionally, when an ultrasound transducer does not have good contact with the patient’s skin (e.g. poor acoustic contact) ultrasound waves cannot effectively propagate into the tissue and ultrasound image data corresponding to the tissue may not be properly acquired. Any data acquired in such cases is typically of low quality. Such data may be ignored.

[0092] Figure 7 schematically shows an example region 105 of a liver 108 to be imaged. Region 105 comprises liver tissue 106 and blood vessels 107. An example mask 25A may be generated to ignore pixels and / or voxels 113 corresponding to blood vessels 107. The illustrated example mask 25A is a binary mask that comprises pixels / voxels 112 corresponding to data to be included and pixels / voxels 113 corresponding to data (in this case blood vessels 107) to be ignored. While Figure 7 is shown in 2-dimensions, it will be appreciated that mask 25A may be a 3- dimensional mask with voxel elements.

[0093] In some embodiments, pixels / voxels to be indicated for exclusion by mask(s)25A are identified using a tissue displacement tracking algorithm. As described elsewhere herein, mechanical exciter 15 induces shear waves within the imaged tissue. However, shear waves do not propagate through fluids. Shear waves may also propagate differently within different tissue types. By processing image data 22 to track the movement of the induced shear waves and predict how the tissue of interest should respond to the movement, regions having no movement (e.g. corresponding to regions of the tissue having fluids and therefore corresponding to blood vessels, cysts, other fluid enclosures, etc.) and / or more or less movement than predicted (e.g. corresponding to a different tissue type such as a neighboring organ) can be identified. The identified pixels / voxels may be labeled as pixels / voxels for exclusion 113 in mask(s) 25A.

[0094] A mask 25A may, for example, be based on elastography data. It may be known that tissues of interest have elasticity that satisfy a criterion (such as having elasticity that is in a specific range or is above or below a threshold value). A mask 25A may be generated which includes all pixels / voxels that do not satisfy the criteria. For example, liver tissue can be expected to have elasticity in the range of about 2 kPa to about 30 kPa. Pixels / voxels for which the elasticity is outside of this range may be considered to correspond to other than liver tissue and may be labelled as pixels / voxels for exclusion 113 in mask(s) 25A.

[0095] In some embodiments, an image recognition system may be applied to image data 22, elastography image 23A and / or attenuation image 24A and / or backscatter image 26 to recognize boundaries of a region of interest. The image recognition system may comprise an Al based image recognition system, such as a trained neural network, image recognition software or any other system capable of processing images to identify structures shown within the images. For example, an image recognition system may be trained to recognize the liver or shear waves in ultrasound images as described in US patent No. 11672503 titled SYSTEMS AND METHODS FOR DETECTING TISSUE AND SHEAR WAVES WITHIN THE TISSUE, which is hereby incorporated herein by reference. Voxels / pixels that the image recognition system recognizes as not lying within the liver or as corresponding to blood vessels or other structures that are within the liver but are not of interest may be included as voxels / pixels for exclusion 113 in mask(s) 25A. For example, the image recognition system may be trained to recognize blood vessels in ultrasoundimages. Voxels / pixels that are inside of or part of blood vessels may be included as voxels / pixels for exclusion 113 in mask(s) 25A.

[0096] In some embodiments, pixels / voxels having a tracked movement that deviates from a predicted amount of movement by more than a threshold amount (e.g. 5%, 10%, 15%, etc.) are marked as pixels for exclusion 113 in mask 25A. In some embodiments, rather than being excluded in a binary mask, pixels / voxels having a tracked movement that deviates from a predicted amount of movement are assigned a weight that corresponds to the amount of deviation.

[0097] Additionally, or alternatively, mask(s) 25A may include pixels / voxels for exclusion 113 for which the quality metric (e.g. the signal-to-noise ratio) of data 22 is below a threshold. Signal-to-noise ratio may be calculated from the local power spectrum of the image data (e.g. image data 22) and is the log of the ratio of the power at center frequency to the other frequencies.

[0098] As described elsewhere herein, mask(s) 25A may be generated in real time as image data is being acquired. It is not necessary to acquire all of image data 22 before at least partially generating mask(s) 25A. Additional image data 22 that may be acquired after a mask 25A is generated may be used to validate mask 25A (e.g. confirm that in fact a blood vessel is found at that location, etc.).

[0099] In some embodiments, a measure of quality of the acquired image data 22 may be computed. For example, a measure of quality may be computed based on a comparison of how much of data 22 has been identified as undesirable and therefore ignored relative to how much of the data 22 has been identified as good data. The measure of quality may be provided for an operator of system 10 (e.g. displayed on display 16) in real time. Based on the measure of quality the operator may attempt to re-position transducer 14, vary settings of ultrasound system 12, etc. in order to achieve a higher measure of quality.Measure of Elastography

[0100] As described elsewhere herein, mechanical exciter 15 induces shear waves in the tissue that is being imaged by transducer 14. Propagation of the shear waves through the tissue causes the tissue to be displaced in a manner which generally corresponds to the waveforms of the shear waves. By tracking the propagation of the shear waves and / or the general tissue movement, it is possible to estimatecharacteristics of the shear waves such as wavelength, amplitude (how much the imaged tissue was displaced), phase and so on. Using the estimated characteristics of the shear waves, it is possible to compute one or more mechanical properties of the imaged tissue, such as elasticity of the imaged tissue and / or the like.

[0101] Since different lines 30A (Figure 6A) are acquired starting at different times and the received ultrasound echo signals corresponding to different points along a line 30A also correspond to different times, data corresponding to different points on the same or different lines generally correspond to different phases of the shear waves. In some embodiments, the data is phase compensated. In some such embodiments, only a subset of the data for the frame is phase compensated. In some embodiments, all of the data for the frame is phase compensated. Using the phase compensated data, sinusoidal waves representative of the shear waves may be fit to the data and tissue displacement caused by the shear waves may be estimated. Based on the estimated tissue displacements, elasticity of the tissue may be estimated.

[0102] Example methods of elastography are described in US patent No. 10667791 to BAGHANI et al. and titled ELASTOGRAPHY USING ULTRASOUND IMAGING OF A THIN VOLUME and in A. Baghani, H. Eskandari, W.Wang, D. D. Costa, M. N. Lathiff, R. Sahebjavaher, S. E. Salcudean, and R. Rohling, “Real-time quantitative elasticity imaging of deep tissue using free-hand conventional ultrasound,” in Medical Image Computing and Computer-Assisted Intervention -MICCAI 2012, 2012, pp. 617-624 which are hereby incorporated by reference for all purposes.

[0103] An example embodiment of a method for computing a measure of elastography is shown in Figure 8.

[0104] Image data 22 (Figure 5) is processed to estimate characteristics of induced shear waves. In some embodiments, a position of a point 31 on an individual line 30A is tracked as a function of time. A sine wave 32 may be fit to the different positions of the point 31 . In block 72, from the sine wave a phasor representation of the amplitude and relative phase of a shear wave may be computed. By processing multiple phasors corresponding to different points on the same or different lines 30A, a wavelength of the shear waves and / or a direction of propagation of the shear waves can be determined. In block 73, the estimated characteristics of the induced shear waves are phase compensated.

[0105] Mask 25A may optionally be applied either before block 72 or in block 74 to filter out undesirable pixels / voxels from the computed elastography data as described elsewhere herein.

[0106] In block 75 an elastography image (e.g. elastography image 23A) or images are generated from the computed elastography data (e.g. from the masked phasor image 74).

[0107] Additionally, or alternatively, the elastography data 23A may be further processed in block 75 to compute a representative measure of elastography for the tissue. For example, the representative measure of elastography determined in block 75 may correspond to an average, weighted average, median, etc. of the computed elastography data. In some embodiments, statistical characteristics (e.g. standard deviation, etc.) may be computed for the measured elastography data.

[0108] In some embodiments, method 70 is continuous (e.g. occurs continuously while image data 22 is being acquired).

[0109] In some embodiments, individual measures of elastography are computed for smaller regions of interest of the imaged tissue.

[0110] In some embodiments, less elastic tissue is identified from the tracked movement of the shear waves and such areas are highlighted (e.g. using a special color or other indicator) for an operator of system 10.Measure of Attenuation

[0111] Acquired image data 22 can also be processed to generate an attenuation image (e.g. attenuation image 24A described elsewhere herein).

[0112] To compute a useful overall measure of attenuation, values for acoustic attenuation at pixels / voxels of the attenuation image may be combined to provide a representative attenuation value, for example by averaging, weighted averaging, taking a median value, removing outliers and then averaging, weighted averaging or taking a median value, etc. In some embodiments, the attenuation data is averaged or summed together into a single overall measure of attenuation. In some embodiments, the attenuation data is averaged or summed together to provide a representative attenuation value for one or more particular regions of the imaged tissue (e.g. attenuation data may be averaged or summed together into a single measure of attenuation for a particular reaion of imaaed liver tissue).

[0113] Figure 9 is a flow chart of an example method for computing a measure of attenuation of a desired tissue (e.g. the liver) for which image data 22 has been acquired according to an example embodiment.

[0114] In block 81 , sets of attenuation data for individual lines (e.g. individual lines 30A or groups of lines) of an acquired frame are computed. In some embodiments, several individual lines may be combined to form a group and then used to compute the attenuation data 24A. The sets of attenuation data 24A for each line, or group of lines, comprise data corresponding to an amount by how much individual points along the line, or group of lines, attenuate transmitted ultrasound pulses. Attenuation data 24A for an individual line, or group of lines, may be computed in a number of ways. For example:• a single set of acquired data corresponding to the individual line may be processed to compute attenuation data 24A for that line;• plural sets of data corresponding to the individual line may be processed individually to compute plural sets of attenuation data for that line and the plural sets of attenuation data may be averaged or summed togetherto yield attenuation data 24A;• plural sets of data corresponding to the individual line may be summed averaged together before processing the summed / averaged data to compute attenuation data 24A for that line;• etc.

[0115] In block 81 A, mask 25A is optionally applied to attenuation data 24A before or after attenuation computation to filter out undesirable points along the individual lines as described elsewhere herein.

[0116] In block 82 an average attenuation value for each line (e.g. each individual line 30A, or group of lines) of ultrasound data in an acquired frame is computed. An average attenuation value for a line, or group of lines, of ultrasound data in the frame may be computed in a number of ways. For example:• a single set of attenuation data corresponding to the individual line may be processed to compute an average attenuation value for that line;• plural sets of attenuation data corresponding to the individual line may be processed individually to compute plural average or summed attenuation values for that line and the plural average or summed attenuation values maybe averaged or summed together;• plural sets of attenuation data corresponding to the individual line may be summed or averaged together before processing the summed or averaged attenuation data to compute a summed or average attenuation value for that line;• plural lines (spatially) used to sum or average the power at a given point.• lines for other frames, spatially close to each other within a collected volume, may be summed or averaged together to compute a summed or average attenuation value for that group.• etc.Block 82 shows an average attenuation value for each line in the illustrated frame.

[0117] Additionally or alternatively, summed or averaged values may be determined for pixels or regions comprising groups of pixels within a frame as shown in block 83. In general, other groupings within a frame of image data may be summed or averaged in an analogous manner.

[0118] In block 84, the average values for each frame (e.g. from blocks 82 and / or 83) are summed or averaged together to compute a total average attenuation value for the frame. Block 85 schematically shows a stack of frames of image data 22, each having an average attenuation for the frame. Block 86 schematically shows the result of summing or averaging the values corresponding to the frames in block 85 into a single attenuation value for a volume of tissue.

[0119] The computed total average for the volume of tissue (e.g. generated in block 85) may be output as measure of attenuation.

[0120] In some embodiments, method 80 is continuous (e.g. occurs continuously while image data 22 is being acquired).

[0121] In some embodiments, attenuation values corresponding to different regions of the tissue may be averaged differently. For example, attenuation values for smaller sections of a particular region of interest may not be averaged together (e.g. individual attenuation values corresponding to smaller sections of the region of interest may be computed rather than one total attenuation value) while attenuation values for the remaining tissue surrounding the region of interest may be averaged together.

[0122] In some embodiments, the computed attenuation data is processed to ascertain how homogenous the imaged tissue is. The homogeneity of the tissue may be linked to the type of disease or state of disease progression, i.e. some diseases or disease states may cause patchy changes in tissue properties, while others are more homogeneous. If the tissue is highly homogenous (e.g. different regions in the tissue attenuate ultrasound the same or within a threshold degree of variance), computing an overall average attenuation value for the imaged tissue may be simplified by omitting data from the averaging. For example, if an average attenuation value for an adjacent line, frame, etc. varies from the preceding line, frame, etc. by less than a threshold amount (e.g. less than 1%, less than 5%, etc.), the adjacent line, frame, etc. may be excluded from further processing.

[0123] In some embodiments, different attenuation values may be displayed to an operator differently. For example, different attenuation values may be shown using different colors or different shades of the same color. As another example, higher attenuation values may be shown with a higher intensity than lower attenuation values.

[0124] In some embodiments, only attenuation measurements above a threshold (e.g. a threshold minimum attenuation value for having a fatty liver) may be specially displayed for an operator. For example, this may allow the operator to visualize what sections of the imaged tissue may be unhealthy.Measure of Backscatter

[0125] Acquired image data 22 can also be processed to generate a backscatter image (e.g. backscatter image 24B described elsewhere herein).

[0126] The methods of data collection, image generation and summing or averaging described for the creation of an attenuation image 24A may be used for the backscatter images, as they are also derived from individual lines or groups of lines of ultrasound data.Overall Metric(s) of Health

[0127] In some embodiments, one or more overall metrics of health of the imaged tissue (e.g. the liver) is / are computed based on a measure of tissue elasticity computed from elastography image 23A and a measure of attenuation computed fromattenuation image 24A and / or a measure of backscatter from the backscatter image 24B. Example non-limiting metrics of health which may be computed include:• a percentage value of how healthy the patient’s tissue is compared to completely healthy tissue;• an estimate of an amount ofliver fat;• tracking changes in the tissue over time, or in response to intervention;• a likelihood of a patient responding to an intervention;• a likelihood of progression of a disease of the tissue (e.g. liver disease);• a likelihood of irreversibly harming the tissue;• a likelihood of the tissue failing;• a comparison of the health of the patient’s tissue to representative population samples;• etc.

[0128] Elasticity, attenuation and backscatter data may be combined using linear or non-linear combinations. These metrics may also be combined with patient data, such as age, body mass index or diabetes status to create better metrics of disease, or disease change / progression.

[0129] Determining metrics such as estimated liver fat may comprise calibration using external measures, such as magnetic resonance imaging, histology measures and / or the like.Example Graphical User Interface (GUI)

[0130] In some embodiments, display 16 displays a graphical user interface (GUI) 90 for an operator of system 10. GUI 90 may allow the operator to visualize results of the imaging. Additionally, or alternatively, GUI 90 may display information assisting the operator to adjust one or more parameters to improve the quality of acquired image data. For example, GUI 90 may display results corresponding to the health of the imaged tissue, one or more measures of quality of the acquired image data, one or more settings of system 10 and / or the like.

[0131] Figure 9 shows a portion 90A of GUI 90 according to a particular example embodiment.

[0132] GUI 90 comprises an image disolav 91 in which images corresponding toacquired data 22, elastography image 23A and / or attenuation image 24A and / or backscatter image 24B may be displayed. The displayed images may be static (e.g. a single frame selected by the operator is shown) or may be a live feed of acquired data 22.

[0133] An operator may optionally vary one or more properties of the displayed images using image settings tool bar 92. Image settings such as image contrast, what frame is to be displayed, what data is to be displayed (e.g. just elastography data, just attenuation data, just backscatter data, both elastography data and attenuation data), how the data is to be displayed (e.g. elastography data in one color and attenuation data in another color), etc. In some embodiments image settings tool bar 92 is partially or fully integrated into image display 91 . For example, image display 91 may support “pinch to zoom” functionality, “panning” allowing the operator to view different regions of the imaged tissue and / or the like.

[0134] Quality measure display 93 may display a measure of the quality of the acquired ultrasound data. Quality measure 93 may display a quantitative indication of the quality (e.g. a number) and / or a qualitative graphical element (e.g. a smiley face for good data and a frowny face for bad data). Quality measure 93 may also display a message indicating that data acquisition will fail based on the current settings of system 10 and patient P’s body characteristics. In some cases, ultrasound signals may have difficulty propagating through a patient’s tissue - e.g. if the patient has a high volume of fat.

[0135] Figure 10A shows another portion 90B of a GUI 90 according to an example embodiment. In some embodiments, a desired imaging plan 95 may be displayed for the operator in imaging plane display 94. Imaging plane display 94 may display a sequence of images that are to be acquired of patient P, a graphical representation 96 of different orientations of transducer 14 which are to be followed by the operator, etc. In some embodiments, the operator’s current location within the imaging plan is displayed. In some embodiments, the user can choose which plane to display by adjusting slider (or other input) 98. One or more corrective measures may also be displayed if an operator deviates from a desired imaging plan. By way of non-limiting example, corrective measures could include:• displaying the imaging plan in red if the operator deviates from the plan too much;• displaying the imaging plan in green once the operator is back on track to follow the imaging plan;• displaying arrows indicating in which direction the operator should pivot transducer 14 to correct a position of transducer 14;• etc.

[0136] Computed measures of health corresponding to the tissue may be displayed in display 97. For example, the displayed measures of health may include one or more of:• elastography properties of the imaged tissue;• attenuation properties of the imaged tissue;• backscatter properties of the imaged tissue;• estimated fat fraction of the imaged tissue;• metrics corresponding to an overall health of the imaged tissue;• etc.Example Clinical Application

[0137] In one example case, system 10 is used as described herein to monitor the health of the liver of a patient suffering from liver disease and the effectiveness of their prescribed treatment.

[0138] The patient may periodically visit, for example, a clinic for a scan of their liver using system 10 to be completed. The patient may have a scan of their liver done daily, weekly, monthly or at any other set time interval. Computed measures of elastography, attenuation and backscatter may be recorded and displayed to the patient.

[0139] In some cases, system 10 autonomously analyzes the computed measures of health, such as estimated liver fat content. In some cases a medical professional analyzes data produced by system 10. For example, the test results for the patient may be graphically visualized (e.g. the computed measures of health may be plotted on a graph for a medical professional to be able to visualize any trends that may be present in the data).

[0140] In some cases, the patient (or the patient’s medical professional) may receive a message from system 10 notifying the patient of how their treatment is progressing. For example, system 10 may display a message indicating that the health of thepatient’s liver is improving. As another example, system 10 may sound an alarm if the patient’s liver has deteriorated to a point that they should seek immediate medical attention.

[0141] In some cases, the patient’s prescribed treatment is synced with the measures of health computed by system 10. The computed measures of health may, for example, be communicated to the patient’s physician (e.g. via a cloud-based system) allowing the physician to monitor the progress of the patient’s treatment in real time.Non-limiting Example Embodiment(s)

[0142] Non-limiting example embodiments of various non-limiting aspects of the invention are described below.

[0143] Training and Validation of Velacur Determined Fat Fraction in patients with MASLD

[0144] Introduction:

[0145] As prevalence of patients with steatoic liver diseases increases throughout the world, it is necessary to have accurate and accessible methods to estimate liver fat content. Using quantitative ultrasound parameters, such as attenuation and backscatter, it is possible to estimate liver fat, with MRI proton density fat fraction as the reference standard. Velacur determined fat fraction (VDFF) is a new output measurement available on Velacur (Sonic Incytes Medical Corp, Vancouver, BC).

[0146] Methods:

[0147] This study described the results of training and validation of VDFF. Patients were recruited from sites within the US and Canada. All patients had contemporaneous Velacur and MRI proton density fat fraction scans. Training was completed using k-fold approach, and a separate cohort was used for validation. The AUG for detection of 5% liver fat based on MRI-PDFF and the correlation between MRI-PDFF and VDFF was measured in both cohorts.

[0148] Results:

[0149] VDFF had an AUROC of 0.97 for the detection of MRI-PDFF > 5% in the training cohort, and 0.99 in the validation cohort. The correlation [95% Cl] between MRI-PDFF and VDFF was r = 0.84 [0.78 - 0.89] for the training cohort and r = 0.90 [0.82 - 0.95] for the validation cohort.

[0150] Conclusion:

[0151] The Velacur Determined Fat Fraction (VDFF) is an accurate and accessible way to estimate steatosis as confirmed by MRI-PDFF. Velacur can fill the unmet need of an accurate means to diagnosis hepatic steatosis as an alternative to biopsy or MRI-PDFF.

[0152] Abbreviations

[0153] 1. Introduction

[0154] Although liver biopsy remains the current gold standard for the diagnosis ofmany chronic liver diseases including Steatotic Liver Disease (SLD), Metabolic Dysfunction Associated Liver Disease (MASLD) and metabolic associated steatohepatitis (MASH), liver biopsy has risks of complication [1], [2] and misclassification due to sampling error and inter-observer and intra-observer variation [3], [4], With the growing prevalence of SLD in the general population, and especially within patients with diabetes, there is a need for more assessable and accurate non- invasive methods for steatosis evaluation [5], [6], [7], New pharmaceutical options that will soon be available for SLD, MASLD and / or MASD are also accentuating the unmet need for assessment of hepatic steatosis[8].

[0155] MRI Proton Density Fat Fraction has emerged as the leading non-invasive method for steatosis assessment [9],

[0010] ,

[0011] ,

[0012] ,

[0013] . This method is able to measure the amount of water vs triglyceride protons within the tissue to estimate the overall fat fraction. Although highly accurate, this method is not assessable at the point of care for most patients and, given the high burden of disease in the general population, is not practical for patient screening.

[0156] Multiple society guidelines advocate for the use of non-invasive tests (NITs) for the assessment of patients with SLD or for those who are at risk of MASLD and MASH based on the presence of diabetes and / or metabolic syndrome [1],

[0014] ,

[0015] ,

[0016] . The 2023 American association for the study of liver diseases (AASLD) Practice Guidance on the clinical assessment and management of non-alcoholic fatty liver disease specifically outlines a tiered approach of assessment using both blood-based markers and imaging of liver stiffness and fat [1]. NIT’s, specifically ultrasound attenuation, are recommended to be used to quantify liver fat. The liver society guidance documents commonly cite > 5% liver fat on MRI-PDFF as the definition of steatosis [1],

[0014] ,

[0015] ,

[0016] . This means that the accurate detection of greater than 5% liver fat is an essential part of the diagnosis of liver steatosis.

[0157] Ultrasound is an accessible modality with the promise to quantify hepatic steatosis

[0017] ,

[0018] ,

[0019] ,

[0020] ,

[0021] ,

[0022] , Quantitative ultrasound (QUS), or measuring parameters that can be derived directly from the ultrasound signal, separate from the created image, has been used for this purpose

[0022] , Ultrasound attenuation has been shown to correlate with liver fat from biopsy

[0021] ,

[0023] ,

[0024] and with MRI-PDFF

[0025] ,

[0026] ,

[0027] , Parameters of qualitative ultrasound, such as backscatter coefficient and attenuation, can be further combined to better correlate with MRI-PDFF. This hasbeen shown in the literature

[0022] ,

[0028] and similar method of combining quantitative ultrasound parameters to estimate liver fat has been commercially available on the Siemens Acuson Sequoia system (Ultrasound Derived Fat Fraction)

[0028] .

[0158] Quantitative ultrasound methods have also been compared with MRI-PDFF for both adults and children, including the Siemens ultrasound derived fat fraction (UDFF)

[0020] ,

[0028] , FibroScan® Controlled Attenuation Parameter (CAP) (EchoSens)

[0026] , backscatter alone

[0017] and other quantitative ultrasound parameter combinations

[0029] -

[0159] Velacur® is a point of care ultrasound-based device that is specifically designed to be used for liver assessment. This system was designed to incorporate both liver stiffness and attenuation measurements. The aim of this study was to define a third output for Velacur, the estimate of hepatic steatosis known as Velacur Determined Fat Fraction (VDFF). VDFF uses quantitative ultrasound parameters to estimate hepatic steatosis.

[0160] 2. Methods

[0161] 2.1 Study Design

[0162] Two cohorts of patients were included in this analysis. The first cohort included patients consecutively or conveniently recruited at 4 sites in the US and Canada (University of California San Diego, Southern California Research Center, Beth Israel Deaconess Medical Center, University of British Columbia). This cohort is used to train the VDFF parameters and will be referred to as the training cohort. A second cohort of patients recruited from two separate sites in the US (Washington University School of Medicine) and Canada (Vancouver Coastal Health Research Institute) was used as validation, and will be referred to as the validation cohort.

[0163] All participants had suspected or diagnosed MASLD or MASH at the time of enrollment and underwent both the Velacur and MRI scans. Data obtained during Velacur scans performed on participants in the Training cohort was used to calibrate and measure VDFF. The study was conducted in accordance with Good Clinical Practices. Informed consent, in writing, was obtained from each patient before their participation in the study and the study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the appropriate institutional review committees.

[0164] 2.2 Inclusion and Exclusion Criteria

[0165] The training cohort consisted of participants between the ages of 19 and 75 years. Only patients with MASLD and / or suspected MASH were enrolled in the training cohort. Excluded patients included those with viral hepatitis, other known causes of chronic liver disease, decompensated cirrhosis, serum ALT or AST > 5 x ULN on historical blood work within the past 3 months, individuals with history of persistent ethanol abuse, or patients who were pregnant or planning to become pregnant during the study. Persistent ethanol use was defined as individuals with history of persistent ethanol abuse (consumption > 20 g ethanol / day for women, > 40 g ethanol / day for men) for more than 3 months in the past year. Patients with BMI greater than 40 kg / m2(or as determined by the MRI bore size) were also excluded.

[0166] The validation cohort consisted of patients with the above inclusion criteria (6 / 38) and participants recruited at the Washington University School of Medicine, St. Louis, MO (32 / 38). Those from St. Loius were aged between 18-70 years of age, with a BMI no greater than 50 kg / m2and included people with and without MRI-diagnosed hepatic steatosis.

[0167] 2.3 MRI-PDFF

[0168] MRI-PDFF was used as the reference standard for liver fat fraction for this study. MRI-PDFF is available on most MRI systems. The standard MRI-PDFF protocol for each manufacturer was used.

[0169] MRI scans were segmented manually by a radiologist or by AMRA®. The radiologist segmented areas of at least 2cm in diameter within the right lobe of the liver were used to measure the MRI-PDFF results. Segmenters avoided vessels and areas near the liver boundary. The mean of all segmented areas or reported by AMRA® was used as the final result.

[0170] 2.4 Velacur Determined Fat Fraction (VDFF)

[0171] VDFF is a measurement derived from quantitative measurements from the radio frequency data collected during the Velacur scans. As part of the measurement of tissue displacements for stiffness measurements, Velacur utilizes a method of high-speed imaging to measure each block of tissue multiple times. The multiple measurements (up to 25 measurements per line) allow for many more repeated measures to be used in the quantitative assessment of tissue. These repeatedmeasurements allow for more data to be averaged during the calculation of each parameters in each location creating better overall estimates.

[0172] In addition to the ability to average multiple lines, Velacur incorporates the use of a machine learning based model to define the location of liver and shear waves within the images

[0030] ,

[0031] . Only data from the liver and with high quality shear waves, as defined by the Velacur system quality algorithms, was used in the analysis. VDFF is a linear combination of backscatter and attenuation parameters calculated from the collected Velacur ultrasound data

[0032] , Attenuation measures the energy loss as the ultrasound wave travels through tissue and is higher in fatty tissue. Backscatter is a measure of the amount of scattering which occurs as the ultrasound waves interacts with tissue microstructure. Tissue with high levels of backscatter will look brighter in the B-Mode image and this increase in brightness is a marker of liver steatosis as seen by sonographers

[0032] , Each of these parameters was calibrated and validated against phantoms of known values before being applied to this patient data set.

[0173] The contribution of each parameter to the final VDFF result was defined using a leave-some-out approach. In this case, multiple rounds of fitting were used. During each round, approximately 80% of the data was used to set the parameters, using PDFF as the reference standard. 80% of the data was randomly selected for each round and this was repeated 1000 times. The mean value of each parameter was used as the final parameter weight. The final VDFF was then applied to all the participants.

[0174] 2.5 Study Objectives

[0175] The primary objective of this study was to determine the accuracy of VDFF in patients with MASLD, using the measure 5% MRI-PDFF as the reference standard. The secondary objective was to determine the accuracy at higher levels of MRI- PDFF, at 11% and 17%. These cut-off points were chosen to approximate the histological steatosis grades. Although the MRI-PDFF cut-off points that correspond to histology steatosis grades differ across studies, these cut-offs were defined by Imajo et al.

[0023] .

[0176] The VDFF parameters are first set using the MRI-PDFF measurements form patients within the training cohort. The final VDFF output is then tested against the MRI-PDFF measurements from the validation nnhnrt

[0177] An exploratory objective of the study was to compare the performance of CAP to VDFF in the training cohort of patients, who also received contemporaneous FibroScan exams.

[0178] 2.6 Statistical Plan

[0179] Receiver operator curve (ROC) analysis is used to measure the performance of VDFF in the detection of MASLD (MRI-PDFF > 5%) and more advanced disease (MRI-PDFF >11% and > 17%). For each MRI-PDFF cut-off, the mean area under the curve (AUG) and 95% confidence intervals (Cl), sensitivity, specificity are calculated. This cut-offs are based on those reported by Imajo et al.

[0023] .

[0180] Pearsons correlation coefficient are used to measure the overall correlation between VDFF and MRI-PDFF. Lin’s Concordance Correlation Coefficient was used to measure the strength of the 1 -to-1 relationship between MRI-PDFF and VDFF

[0033] .

[0181] Results from both the training and validation cohorts are reported.

[0182] 3. Results

[0183] 3.1 Patient Characteristics

[0184] A total of 113 participants with paired MRI-PDFF and Velacur scans were included in the training cohort with 38 participants with paired scans included in the Validation Cohort. The overall participant characteristics are outlined in Error! Reference source not found.. Participants in the training cohort were 51% female and 37.1% White, with a mean age of 57.4 year and mean body mass index (BMI) of 30 kg / m2. Fifteen percent (17 / 113) had an MRI-PDFF of less than 5%, 26% (30 / 113) had a MRI-PDFF of 5-11%, 24% (28 / 113) from 11 -17% and 34% (38 / 113) with a MRI- PDFF of 17% or greater.

[0185] The validation cohort was more homogeneous, with 73.7% white and 84% female. In addition, there were a higher percentage of participants (50%) with MRI- PDFF less than 5%. The average MRI-PDFF and VDFF were significantly lower in the validation cohort than the training cohort, at 8.3% vs 14% for PDFF, and 10.7% and 13.4% for VDFF respectively (p < 0.05 using unpaired t-test).

[0186] Table 1 : Summary of Patient Demographic InformationAbbreviation: ALT, alanine transaminase; AST, aspartate aminotransferase; BMI, body mass index; T2DM, type 2 diabetes mellitus.

[0187] 3.2 Detection of MASLD

[0188] 3.2.1 Training Cohort

[0189] The distribution of training patients into each MRI-PDFF grouping is shown in Table 2. VDFF had an AUROC of 0.97 for the detection of MRI-PDFF > 5%. The resulting AUG (mean [95% Cl]) for the other MRI cut-off of 11% and 17% was 0.91 [0.85 - 0.95] and 0.91 [0.83 - 0.95] respectively. Overall, the AUG for VDFF is excellent at each of the MRI cut-offs. The sensitivity and specificity of 5% VDFF was 98% for VDFF and the specificity was 41%.

[0190] Table 2: Training Cohort. Resulting AUG for each of the MRI cutoffs. 5% defines the diagnosis of any steatosis. Using the VDFF values (5%,11 % and 17%), the sensitivity and specificity of the VDFF cut points is shown for the training cohort.

[0191] 3.2.2 Validation Cohort

[0192] Error! Reference source not found, outlines the results of VDFF when applied to this validation cohort. The overall AUC’s are similar to or greater than those of the training cohort.

[0193] Table 3: Validation Cohort. Resulting AUG for each of the MRI cut-offs. 5% defines the diagnosis of any steatosis. Using the VDFF values (5%, 11 % and 17%), the sensitivity and specificity of the VDFF cut points is shown for the validation cohort.

[0194] 3.3 Correlation Between VDFF and MRI-PDFFThe overall correlation between MRI-PDFF and VDFF was 0.84 [0.78 - 0.89] for the training cohort (Error! Reference source not found.1 A left) and 0.90 [0.82 - 0.95] for the validation cohort (Error! Reference source not found.1 B right).

[0195] 4.1 Comparison to Fibroscan CAP

[0196] As an exploratory objective of the study, the results for FibroScan CAP were also analysed in the training cohort of participants (the validation cohort did not have FibroScan data available). In line with other literature discussed below, the CAP AUG was 0.85 [0.76, 0.92], 0.80 [0.71 , 0.87] and 0.81 [0.71 , 0.89] for the MRI cut-offs of 5,11 , and 17% respectively. Using DeLong’s test to test for significance, shows that all the AUC’s are significantly different (p < 0.05) and higher than CAP

[0034] ,

[0197] The overall correlation between CAP and MRI-PDFF in this cohort was 0.60 (0.47-0.71 ), which is also lower than the correlation between VDFF and MRI-PDFF (0.84 [0.78-0.89]).

[0198] 4. Discussion

[0199] In this cohort of patients with a broad range of MRI-PDFF values, the Velacur Determined Fat Fraction (VDFF) was shown to be an accurate estimation of MRI- PDFF, with AUC’s greater than 0.9 in both the training and validation cohorts, at all levels of MRI-PDFF. VDFF has the potential to be used in point of care settings, providing an alternative to MRI-PDFF in the diagnosis of patients with MASLD and MASH. As the number of patients in the US and around the world increases, it is important to have options for screening and diagnose patients with hepatic steatosis in order to provide correct and timely interventions. Neither MRI-PDFF nor biopsy are able to scale to the degree needed.

[0200] Although both attenuation and backscatter, as well as speed of sound, have been shown to correlate with MRI-PDFF and hepatic fat as measured from biopsy, VDFF adds to the current literature. Velacur offers a novel method of both collecting multiple measurements, and the collection of many more samples at the same location through the use of the high-speed imaging used.

[0201] This results are as good as or better than those previously reported in the literature. Paige et al. reviewed a number of studies using ultrasound attenuation from several manufacturers, and CAP specifically, using MRI-PDFF or biopsy as the reference

[0018] . The average AUG in this study for all steatosis cut-offs across all listed studies using ultrasound attenuation was about 0.89, and 0.83 for CAP. The UDFF results showed an AUG for detection of 5% steatosis of 0.75, vs conventional ultrasound at 0.53

[0028] . In pediatrics, the UDFF AUG for 6% fat on MRI was 0.95

[0020] . In a direct comparison between one specific QUS combination and CAP using MRI- PDFF as the reference standard, QUS had a significantly higher AUG for diagnosis of MASLD (5% on MRI PDFF) than CAP (0.92 vs 0.79)

[0029] , Using these past results as basis for comparison, the VDFF compares favourability, with an average AUG of 0.93 in this cohort of patients.

[0202] It is interesting to note that the results in the validation cohort where actually better than those from the training cohort. This may be due to the fact that the validation was very homogeneous, or that there were more patients with lower MIR- PDFF values. This proves that VDFF can be used in a variety of contexts, especially for screening patients in a population with less disease burden.

[0203] Although the parameters for VDFF were trained using a k-fold approach, and validated on a small external cohort, further validation of the measure is needed using a more diverse cohort of patients or on patients from other backgrounds.

[0204] As more and more quantitative ultrasound techniques are developed, it is important to be able to compare and standardize the measurements for multiple techniques. There are several initiatives, including those spearheaded by the Quantitative Imaging Biomarker Alliance (QIBA), to standardize the measurements of attenuation, backscatter and speed of sound across multiple ultrasound manufacturers

[0035] ,

[0036] ,

[0037] ,

[0205] 5. Conclusions

[0206] The Velacur Determined Fat Fraction (VDFF) is an accurate and accessible way to estimate MRI-PDFF in patients with a wide range of MRI-PDFF. The high AUC’s for detection of MRI-PDFF cut offs (>0.9) show that the VDFF is an accurate estimate of liver fat. Velacur, and other quantitative ultrasound methods, can fill the unmet need of an accurate means to diagnosis hepatic steatosis. As a low cost, non- invasive measurement, it is an excellent alternative to biopsy or MRI-PDFF.

[0207] Further studies should further validate VDFF in additional patient cohorts and other disease populations. When measured in patients undergoing lifestyle modification or therapeutic interventions VDFF should show changes over time that correlation with MRI-PDFF responses.

[0208] 6. References

[0209] [1] M. E. Rinella et aL, “AASLD Practice Guidance on the clinical assessment and management of nonalcoholic fatty liver disease,” Hepatology, vol. 77, no. 5, pp. 1797-1835, May 2023, doi: 10.1097 / HEP.0000000000000323.[2] D. C. Rockey, S. H. Caldwell, Z. D. Goodman, R. C. Nelson, and A. D. Smith, “Liver biopsy,” Hepatology, vol. 49, no. 3, pp. 1017-1044, 2009, doi:10.1002 / HEP.22742.[3] A. A. Bravo, S. G. Sheth, and S. Chopra, “Liver Biopsy,” https: / / doi.org / 10.1056 / NEJM200102153440706, vol. 344, no. 7, pp. 495-500, Feb. 2001 , doi: 10.1056 / NEJM200102153440706.[4] V. Ratziu et aL, “Sampling Variability of Liver Biopsy in Nonalcoholic Fatty Liver Disease,” Gastroenterology, vol. 128, no. 7, pp. 1898-1906, Jun. 2005, doi: 10.1053 / J.GASTRG.2005.03.084.[5] Z. M. Younossi, G. Wong, Q. M. Anstee, and L. Henry, “The Global Burden of Liver Disease,” Clinical Gastroenterology and Hepatology, Apr. 2023, doi: 10.1016 / J.CGH.2023.04.015.[6] Z. M. Younossi, A. B. Koenig, D. Abdelatif, Y. Fazel, L. Henry, and M. Wymer, “Global epidemiology of nonalcoholic fatty liver disease — Meta-analytic assessment of prevalence, incidence, and outcomes,” Hepatology, vol. 64, no. 1 , pp. 73-84, Jul. 2016, doi: 10.1002 / HEP.28431 .[7] Z. M. Younossi et aL, “The global epidemiology of NAFLD and NASH in patients with type 2 diabetes: A systematic review and meta-analysis,” J Hepatol, vol.71 , no. 4, pp. 793-801 , Oct. 2019, doi: 10.1016 / J.JHEP.2019.06.021 .[8] R. Vuppalanchi, M. Noureddin, N. Alkhouri, and A. J. Sanyal, “Therapeutic pipeline in nonalcoholic steatohepatitis,” Nature Reviews Gastroenterology & Hepatology 2021 18:6, vol. 18, no. 6, pp. 373-392, Feb. 2021 , doi: 10.1038 / s41575- 020-00408-y.[9] J. Gu et aL, “Diagnostic value of MRI-PDFF for hepatic steatosis in patients with non-alcoholic fatty liver disease: a meta-analysis,” Eur Radiol, vol. 29, no. 7, pp. 3564-3573, Mar. 2019, doi: 10.1007 / S00330-019-06072-4 / METRICS.

[0010] S. B. Reeder, H. H. Hu, and C. B. Sirlin, “Proton Density Fat-Fraction: A Standardized MR-Based Biomarker of Tissue Fat Concentration,” J Magn Reson Imaging, vol. 36, no. 5, p. 1011 , Nov. 2012, doi: 10.1002 / JMRL23741 .

[0011] H. H. Hu et aL, “Linearity and bias of proton density fat fraction as a quantitative imaging biomarker: A multicenter, multiplatform, multivendor phantom study,” Radiology, vol. 298, no. 3, pp. 640-651 , Mar. 2021 , doi:10.1148 / RADIOL.2021202912 / ASSET / IMAGES / LARGE / RADIOL.2021202912.TBL6. JPEG.

[0012] E. Yoshizawa and A. Yamada, “MRI-derived proton density fat fraction,” Journal of Medical Ultrasonics, vol. 48, no. 4, pp. 497-506, Oct. 2021 , doi:10.1007 / S10396-021 -01135-W / METRICS.

[0013] T. Yokoo et aL, “Nonalcoholic Fatty Liver Disease: Diagnostic and Fat-Grading Accuracy of Low-Flip-Angle Multiecho Gradient-Recalled-Echo MR Imaging at 1.5 T1 ,” https: / / doi.org / 10.1148 / radiol.251 1080666, vol. 251 , no. 1 , pp. 67-76, Apr. 2009, doi: 10.1148 / RADIOL.2511080666.

[0014] F. Bril and K. Cusi, “Management of Nonalcoholic Fatty Liver Disease in Patients With Type 2 Diabetes: A Call to Action,” Diabetes Care, vol. 40, no. 3, pp. 419-430, Mar. 2017, doi: 10.2337 / DC16-1787.

[0015] K. Cusi et aL, “American Association of Clinical Endocrinology Clinical Practice Guideline for the Diagnosis and Management of Nonalcoholic Fatty Liver Disease in Primary Care and Endocrinology Clinical Settings: Co-Sponsored by the American Association for the Study of Liver Diseases (AASLD),” Endocrine Practice, vol. 28, no.5, pp. 528-562, May 2022, doi: 10.1016 / J.EPRAC.2022.03.010.

[0016] F. Kanwal et aL, “Clinical Care Pathway for the Risk Stratification and Management of Patients With Nonalcoholic Fatty Liver Disease,” Gastroenterology, vol. 161 , no. 5, pp. 1657-1669, Nov. 2021 , doi: 10.1053 / J.GASTRG.2021 .07.049.

[0017] S. C. Lin et aL, “Noninvasive Diagnosis of Nonalcoholic Fatty Liver Disease and Quantification of Liver Fat Using a New Quantitative Ultrasound Technique,” Clinical Gastroenterology and Hepatology, vol. 13, no. 7, pp. 1337-1345. e6, JuL 2015, doi: 10.1016 / J.CGH.2014.11.027.

[0018] J. S. Paige et aL, “A pilot comparative study of quantitative ultrasound, conventional ultrasound, and MRI for predicting histology-determined steatosis grade in adult nonalcoholic fatty liver disease,” American Journal of Roentgenology, vol.208, no. 5, pp. W168-W177, May 2017, doi:10.2214 / AJR.16.16726 / ASSET / IMAGES / LARGE / 05_16_16726_04B.JPEG.

[0019] G. Polti et aL, “Quantitative ultrasound fatty liver evaluation in a pediatric population: comparison with magnetic resonance imaging of liver proton density fat fraction,” Pediatr Radiol, vol. 53, no. 12, pp. 2458-2465, Nov. 2023, doi:10.1007 / S00247-023-05749-9 / TABLES / 2.

[0020] M. Zalcman, R. A. Barth, and E. Rubesova. “Real-time ultrasound-derived fatfraction in pediatric population: feasibility validation with MR-PDFF,” Pediatr Radiol, vol. 53, no. 12, pp. 2466-2475, Nov. 2023, doi: 10.1007 / S00247-023-05752- 0 / METRICS.

[0021] M. Sasso et al., “Controlled Attenuation Parameter (CAP): A Novel VCTETM Guided Ultrasonic Attenuation Measurement for the Evaluation of Hepatic Steatosis: Preliminary Study and Validation in a Cohort of Patients with Chronic Liver Disease from Various Causes,” Ultrasound Med Biol, vol. 36, no. 11 , pp. 1825-1835, Nov. 2010, doi: 10.1016 / J.ULTRASMEDBIO.2010.07.005.

[0022] M. L. Oelze and J. Mamou, “Review of quantitative ultrasound: envelope statistics and backscatter coefficient imaging and contributions to diagnostic ultrasound,” IEEE Trans Ultrason Ferroelectr Freq Control, vol. 63, no. 2, p. 336, Feb. 2016, doi: 10.1109 / TUFFC.2015.2513958.

[0023] K. Imajo et aL, “Magnetic Resonance Imaging More Accurately Classifies Steatosis and Fibrosis in Patients With Nonalcoholic Fatty Liver Disease Than Transient Elastography,” Gastroenterology, vol. 150, no. 3, pp. 626-637. e7, Mar. 2016, doi: 10.1053 / J.GASTRG.2015.11 .048.

[0024] V. de Ledinghen, J. B. Hiriart, J. Vergniol, W. Merrouche, P. Bedossa, and V. Paradis, “Controlled Attenuation Parameter (CAP) with the XL Probe of the Fibroscan®: A Comparative Study with the M Probe and Liver Biopsy,” Dig Dis Sci, vol. 62, no. 9, pp. 2569-2577, Sep. 2017, doi: 10.1007 / S10620-017-4638-3.

[0025] F. Deeba et aL, “SWTV-ACE: Spatially Weighted Regularization Based Attenuation Coefficient Estimation Method for Hepatic Steatosis Detection,” Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11768 LNCS, pp. 610-618, 2019, doi:10.1007 / 978-3-030-32254-0 68 / CGVER.

[0026] C. Caussy et aL, “Optimal threshold of controlled attenuation parameter with MRI-PDFF as the gold standard for the detection of hepatic steatosis,” Hepatology, vol. 67, no. 4, pp. 1348-1359, Apr. 2018, doi: 10.1002 / HEP.29639 / SUPPINFO.

[0027] V. de Ledinghen, J. Vergniol, J. Foucher, W. Merrouche, and B. le Bail, “Non- invasive diagnosis of liver steatosis using controlled attenuation parameter (CAP) and transient elastography,” Liver International, vol. 32, no. 6, pp. 911-918, JuL 2012, doi: 10.1111 / J.1478-3231 .2012.02820.X.

[0028] R. De Robertis et aL, “Ultrasound-derived fat fraction for detection of hepatic steatosis and quantification of liver fat content,” Radiologia Medica, vol. 1 , pp. 1-7, Aug. 2023, doi: 10.1007 / S11547-023-01693-8 / FIGURES / 5.

[0029] J. Jung et aL, “Direct Comparison of Quantitative US versus Controlled Attenuation Parameter for Liver Fat Assessment Using MRI Proton Density Fat Fraction as the Reference Standard in Patients Suspected of Having NAFLD,” Radiology, vol. 304, no. 1 , pp. 75-82, Jul. 2022, doi:10.1148 / RADIOL.211 131 / ASSET / IMAGES / LARGE / RADIOL.211131 .FIG4.JPEG.

[0030] M. Honarvar, J. Lobo, and C. Schneider, “Machine learning algorithm to detect shear waves during VelacurTM exams,” NASH-TAG Preceding. Park City, UT, 2023.

[0031] C. Schneider et aL, “Mo1360: ACCURACY OF VELACURTM Al POWERED LIVER GUIDE IN IDENTIFICATION OF LIVER IN ULTRASOUND IMAGES,” Gastroenterology, vol. 162, no. 7, p. S754, 2022.

[0032] J. Park, J. M. Lee, G. Lee, S. K. Jeon, and I. Joo, “Quantitative Evaluation of Hepatic Steatosis Using Advanced Imaging Techniques: Focusing on New Quantitative Ultrasound Techniques,” Korean J Radiol, vol. 23, no. 1 , p. 13, Jan.2022, doi: 10.3348 / KJR.2021 .0112.

[0033] L. L-K. Lin, “A Concordance Correlation Coefficient to Evaluate Reproducibility,” Biometrics, vol. 45, no. 1 , p. 255, Mar. 1989, doi: 10.2307 / 2532051.

[0034] E. R. DeLong, D. M. DeLong, and D. L. Clarke-Pearson, “Comparing the Areas under Two or More Correlated Receiver Operating Characteristic Curves: A Nonparametric Approach,” Biometrics, vol. 44, no. 3, p. 837, Sep. 1988, doi: 10.2307 / 2531595.

[0035] G. Ferraioli, V. Kumar, A. Ozturk, K. Nam, C. L. de Korte, and R. G. Barr, “US Attenuation for Liver Fat Quantification: An AIUM-RSNA QIBA Pulse-Echo Quantitative Ultrasound Initiative,” Radiology, vol. 302, no. 3, pp. 495-506, Mar. 2022, doi:10.1148 / RADIOL.210736 / ASSET / IMAGES / LARGE / RADIOL.210736TBL2.JPEG.

[0036] D. T. Fetzer et aL, “Pulse-Echo Quantitative US Biomarkers for Liver Steatosis: Toward Technical Standardization,” Radiology, vol. 305, no. 2, pp. 265- 276, Nov. 2022, doi:10.1148 / RADIOL.212808 / ASSET / IMAGES / LARGE / RADIOL.212808.FIG7.JPEG.

[0037] K. A. Wear et al., “US Backscatter for Liver Fat Quantification: An AIUM-RSNA QIBA Pulse-Echo Quantitative Ultrasound Initiative,” Radiology, vol. 305, no. 3, pp. 526-537, Dec. 2022, doi:10.1148 / RADIOL.220606 / ASSET / IMAGES / LARGE / RADIQL.220606. FIG12. JPEG.Interpretation of Terms

[0210] Unless the context clearly requires otherwise, throughout the description and the claims:• “comprise”, “comprising”, and the like are to be construed in an inclusive sense, as opposed to an exclusive or exhaustive sense; that is to say, in the sense of “including, but not limited to”;• “connected”, “coupled”, or any variant thereof, means any connection or coupling, either direct or indirect, between two or more elements; the coupling or connection between the elements can be physical, logical, or a combination thereof;• “herein”, “above”, “below”, and words of similar import, when used to describe this specification, shall refer to this specification as a whole, and not to any particular portions of this specification;• “or”, in reference to a list of two or more items, covers all of the following interpretations of the word: any of the items in the list, all of the items in the list, and any combination of the items in the list;• the singular forms “a”, “an”, and “the” also include the meaning of any appropriate plural forms.

[0211] Words that indicate directions such as “vertical”, “transverse”, “horizontal”, “upward”, “downward”, “forward”, “backward”, “inward”, “outward”, “left”, “right”, “front”, “back”, “top”, “bottom”, “below”, “above”, “under”, and the like, used in this description and any accompanying claims (where present), depend on the specific orientation of the apparatus described and illustrated. The subject matter described herein may assume various alternative orientations. Accordingly, these directional terms are not strictly defined and should not be interpreted narrowly.

[0212] Embodiments of the invention may be implemented using specifically designedhardware, configurable hardware, programmable data processors configured by the provision of software (which may optionally comprise “firmware”) capable of executing on the data processors, special purpose computers or data processors that are specifically programmed, configured, or constructed to perform one or more steps in a method as explained in detail herein and / or combinations of two or more of these. Examples of specifically designed hardware are: logic circuits, application-specific integrated circuits (“ASICs”), large scale integrated circuits (“LSIs”), very large scale integrated circuits (“VLSIs”), and the like. Examples of configurable hardware are: one or more programmable logic devices such as programmable array logic (“PALs”), programmable logic arrays (“PLAs”), and field programmable gate arrays (“FPGAs”). Examples of programmable data processors are: microprocessors, digital signal processors (“DSPs”), embedded processors, graphics processors, math coprocessors, general purpose computers, server computers, cloud computers, mainframe computers, computer workstations, and the like. For example, one or more data processors in a control circuit for a device may implement methods as described herein by executing software instructions in a program memory accessible to the processors.

[0213] Processing may be centralized or distributed. Where processing is distributed, information including software and / or data may be kept centrally or distributed. Such information may be exchanged between different functional units by way of a communications network, such as a Local Area Network (LAN), Wide Area Network (WAN), or the Internet, wired or wireless data links, electromagnetic signals, or other data communication channel.

[0214] For example, while processes or blocks are presented in a given order, alternative examples may perform routines having steps, or employ systems having blocks, in a different order, and some processes or blocks may be deleted, moved, added, subdivided, combined, and / or modified to provide alternative or subcombinations. Each of these processes or blocks may be implemented in a variety of different ways. Also, while processes or blocks are at times shown as being performed in series, these processes or blocks may instead be performed in parallel or may be performed at different times.

[0215] In addition, while elements are at times shown as being performed sequentially, they may instead be performed simultaneously or in differentsequences. It is therefore intended that the following claims are interpreted to include all such variations as are within their intended scope.

[0216] In some embodiments, aspects of the invention may be implemented in software. For greater clarity, “software” includes any instructions executed on a processor and may include (but is not limited to) firmware, resident software, microcode, and the like. Both processing hardware and software may be centralized or distributed (or a combination thereof), in whole or in part, as known to those skilled in the art. For example, software and other modules may be accessible via local memory, via a network, via a browser or other application in a distributed computing context, or via other means suitable for the purposes described above.

[0217] Where a component (e.g. a software module, processor, assembly, device, circuit, etc.) is referred to above, unless otherwise indicated, reference to that component (including a reference to a “means”) should be interpreted as including as equivalents of that component any component which performs the function of the described component (i.e. , that is functionally equivalent), including components which are not structurally equivalent to the disclosed structure which performs the function in the illustrated exemplary embodiments of the invention.

[0218] Specific examples of systems, methods and apparatus have been described herein for purposes of illustration. These are only examples. The technology provided herein can be applied to systems other than the example systems described above. Many alterations, modifications, additions, omissions, and permutations are possible within the practice of this invention. This invention includes variations on described embodiments that would be apparent to the skilled addressee, including variations obtained by: replacing features, elements and / or acts with equivalent features, elements and / or acts; mixing and matching of features, elements and / or acts from different embodiments; combining features, elements and / or acts from embodiments as described herein with features, elements and / or acts of other technology; and / or omitting combining features, elements and / or acts from described embodiments.

[0219] Various features are described herein as being present in “some embodiments”. Such features are not mandatory and may not be present in all embodiments. Embodiments of the invention may include zero, any one or any combination of two or more of such features. This is limited only to the extent thatcertain ones of such features are incompatible with other ones of such features in the sense that it would be impossible for a person of ordinary skill in the art to construct a practical embodiment that combines such incompatible features. Consequently, the description that “some embodiments” possess feature A and “some embodiments” possess feature B should be interpreted as an express indication that the inventors also contemplate embodiments which combine features A and B (unless the description states otherwise or features A and B are fundamentally incompatible).

[0220] It is therefore intended that the following appended claims and claims hereafter introduced are interpreted to include all such modifications, permutations, additions, omissions, and sub-combinations as may reasonably be inferred. The scope of the claims should not be limited by the preferred embodiments set forth in the examples but should be given the broadest interpretation consistent with the description as a whole.

Claims

WHAT IS CLAIMED IS:

1. A method of ultrasound imaging, the method comprising: inducing propagation of shear waves through a region of tissue; acquiring a three-dimensional sample set of ultrasound data of the region responsive to the induced shear waves; processing the sample set to identify voxels of interest based on at least one of: propagation of the shear waves through the tissue region; and one or more quality metrics; and determining ultrasound elastography values, ultrasound attenuation values and ultrasound backscatter values corresponding to the identified voxels based on the sample set.

2. The method of claim 1 or any other claim herein wherein determining ultrasound elastography values, ultrasound attenuation values and ultrasound backscatter values occurs based on the same sample set.

3. The method of any one of claims 1 to 2 or any other claim herein comprising combining any plurality of the ultrasound elastography values, the ultrasound attenuation values and the ultrasound backscatter values at one or more voxels of interest to create a health metric for one or more voxels of interest or for the sample set.

4. The method of any one of claims 1 to 3 or any other claim herein wherein processing the sample set to identify voxels of interest is performed prior to determining the ultrasound elastography values, ultrasound attenuation values and ultrasound backscatter values.

5. The method of any one of claims 1 to 3 or any other claim herein wherein processing the sample set to identify voxels of interest is performed after determining the ultrasound elastography values, ultrasound attenuation values and ultrasound backscatter values.

6. The method of any one of claims 1 to 5 or any other claim herein wherein processing the sample set to identify voxels of interest comprises determining the voxels of interest to correspond to a particular tissue type.

7. The method of any one of claims 1 to 6 or any other claim herein wherein processing the sample set to identify voxels of interest comprises: determining, based on the sample set, that particular voxels correspond to blood vessels or tissue that is not of interest; and excluding, from the voxels of interest, such particular voxels.

8. The method of any one of claims 1 to 7 or any other claim herein wherein processing the sample set to identify voxels of interest comprises: determining, based on the sample set, displacement of tissue induced by the shear waves; and applying a thresholding process to the displacement to identify the voxels of interest (or voxels that are not of interest).

9. The method of any one of claims 1 to 7 or any other claim herein wherein processing the sample set to identify voxels of interest comprises: applying a thresholding process to the ultrasound elastography values to identify the voxels of interest (or voxels that are not of interest).

10. The method of any one of claims 1 to 9 or any other claim herein wherein processing the sample set to identify voxels of interest comprises: using a trained machine learning (artificial intelligence) engine to identify the voxels of interest based on one or more of the sample set, the ultrasound elastography values, the ultrasound attenuation values and the ultrasound backscatter values.11 . The method of any one of claims 1 to 10 or any other claim herein comprising using a trained machine learning (artificial intelligence ) engine to determine tissue type, one or more characteristics of shear wave propagation and / or an other quality metric for one or more voxels of interest or for the sample set.

12. The method of any one of claims 1 to 11 or any other claim herein wherein processing the sample set to identify voxels of interest comprises: determining, based on the sample set, a quality metric (e.g. the signal-to- noise ratio and / or the like) for particular voxels; and applying a thresholding process to the quality metric to identify the voxels of interest (or voxels that are not of interest).

13. The method of any one of claims 1 to 12 or any other claim herein wherein processing the sample set to identify voxels of interest is performed at least partially concurrently with acquiring the sample set.

14. The method of any of any one of claims 1 to 13 or any other claim herein wherein: processing the sample set to identify voxels of interest is performed prior to determining the ultrasound elastography values, ultrasound attenuation values and ultrasound backscatter values; and determining the ultrasound elastography values, ultrasound attenuation values and ultrasound backscatter values is performed only for the voxels of interest.

15. The method of any one of claims 1 to 14 or any other claim herein wherein acquiring the three-dimensional sample set comprises: parsing a three-dimensional volume to be sampled into a plurality of frames, each frame comprising a plurality of lines; grouping the plurality of lines corresponding to each frame into a plurality of groups, each group comprising a plurality of lines; repeating the following steps until all of the groups have been traversed: acquiring samples from each line in a particular group a plurality of times within one period of the shear waves; and changing the particular group.

16. The method of claim 15 or any other claim herein wherein changing the particular group comprises changing from a current particular group to aspatially sequential group.

17. The method of claim 15 or any other claim herein wherein changing the particular group comprises changing from a current particular group to a non- spatially sequential group.

18. The method of any one of claims 15 to 17 or any other claim herein wherein acquiring samples from each line in the particular group comprises traversing the lines in the group in a spatially sequential order.

19. The method of any one of claims 15 to 17 or any other claim herein wherein acquiring samples from each line in the particular group comprises traversing the lines in the group in a non-spatially sequential order.

20. An ultrasound apparatus comprising: an exciter for inducing propagation of shear waves through a region of tissue; one or more ultrasound transducers for acquiring a three-dimensional sample set of ultrasound data of the region responsive to the induced shear waves; a controller configured to: process the sample set to identify voxels of interest based on the propagation of the shear waves through the tissue region; and determine ultrasound elastography values, ultrasound attenuation values and ultrasound backscatter values corresponding to the identified voxels based on the sample set.21 . The ultrasound apparatus according to claim 20 wherein the controller is configured to perform any of the features, combinations of features and / or subcombinations of features recited in any of claims 1 to 19.

22. Methods comprising any features, combinations of features, and / or subcombinations of features disclosed herein and / or in the accompanyingdrawings.

23. Apparatus comprising any features, combinations of features, and / or subcombinations of features disclosed herein and / or in the accompanying drawings.

Images