Systems, devices, and methods for evaluating livestock

Abstract

Livestock animal assessment models may be trained using ultrasound images of livestock animals taken over their lifespan and quality measurements of the same livestock animals taken after slaughter that provide quantitate indications of meat quality and / or tenderness. Once generated and / or trained, the livestock animal assessment models may be used to predict livestock quality and / or value during lifespan using ultrasound images of animal tissue and / or other pre-slaughter measurements.

Description

SYSTEMS, DEVICES, AND METHODS FOR EVALUATING LIVESTOCKRelated Application

[0001] This application is an INTERNATIONAL (PCT) application of, and claims priority to, United States Provisional Patent Application Number: 63 / 670,492, filed on 12 July 2024 and entitled “SYSTEMS, DEVICES, AND METHODS FOR EVALUATING LIVESTOCK,” which is incorporated herein by reference.Field of Invention

[0002] The invention pertains to the field of livestock animal quality assessment and / or management. In particular, the present invention is directed to systems, devices, and methods for imaging and / or analyzing livestock tissue to assess quality and grade according to one or more metrics such as regulatory and / or industry standards.Background

[0003] Currently, in the United States, quality assessment and pricing for livestock (e.g., beef, poultry, and / or pork) carcasses is primarily based on United States Department of Agriculture (USDA) Quality and Yield Grade assessments that are made following slaughtering of the animal. Quality assessment and pricing for livestock in countries outside the United States (e.g., Brazil, Australia, and / or Canada) is based on similar metrics. In particular, beef quality and pricing are assessed using grade quality, intermuscular fat (IMF), which is commonly considered “marbling,” to arrive at the following classifications, listed here in descending order of quality: prime, choice, select, standard, commercial, utility, cutter, and canner. Although tenderness of livestock may be quantitively assessed using, for example, the Warner-Bratzler shear force test, these tests are too costly and time consuming to be used at scale for quality grading of individual animals at the slaughterhouse. Thus, IMF is often used as a proxy assessment for tenderness with greater IMF being associated with greater beef tenderness.

[0004] Grade quality may be assessed using animal maturity (age) and a feature of the animal’s IMF feature (e.g., quantity, position, etc.). Yield grade assessments provide an estimate of cutability, or how much of the animal will be salable based on a measurement of a) a location-specific thickness of backfat (BF); b) ribeye area (REA); and internal fat (e.g., kidney, pelvic and heart (KPH) fat). Prior to harvest, producers (ranchers, feedlots, etc.) currently assess and predict livestock animal quality and yield grades while the animal is still “on the hoof’ by using ultrasound technology to assess BF, REA, and IMF at one or more points in the animal’s lifespan. However, this procedure requires skilled ultrasound technicians and remote analysis of ultrasound images that takes significant time, typically a week or more and does not yield results for tenderness or other metrics that may be useful in determining livestock animal quality.Summary

[0005] The systems and devices disclosed herein may be used measure or estimate current quality and / or predict a quality of a livestock animal either now / real-time and / or in the future (e.g., a week, a month, a year, and / or 5 years). The systems and devices may include a measurement device configured to generate one or more images (e.g., ultrasound) of livestock animal (e.g., cow, chicken, pig, etc.) tissue, a user interface configured to receive input from and / or display output to a user, a processor communicatively coupled to the measurement device, the user interface, and a memory, the processor being configured to execute one or more sets of instructions stored on the memory, a power source configured to provide power to the measurement device, the user interface, the processor, and the memory, and a housing configured to house the measurement device, the user interface, the processor, the memory, and the power source. In some embodiments, the device may include a communication interface (e.g., transceiver, antenna, etc.) for communicating with one or more external devices such as computers, mobile communication devices, machinery, and / or livestock sorting devices and / or systems via, for example, a near-field communication protocol, Wi-Fi, a cellular communication protocol, and / or a communication port.

[0006] The memory may be configured to store a livestock animal assessment model and a set of instructions thereon, which when executed by the processor cause the processor to receive the image (e.g., ultrasound) of livestock animal tissue from the measurement device, input the image into the livestock animal assessment model, receive an output from the livestock animal assessment model, the output including automated measurements of BF, REA and IMF, and predictions including one of a tenderness classification for the imaged livestock animal tissue, and provide an indication of the output to the user interface, a user, and / or an external computing device and / or livestock herd management system. In some cases, the livestock animal assessment model may be trained using feedback from the device itself and / or other devices and / or systems via, for example, software and / or firmware updates.

[0007] Exemplary outputs include, but are not limited to, current and / or future and predicted measurements of BF, REA, and IMF, a tenderness score, an estimated value of the livestock animal, a recommendation for the care and / or feeding of the livestock animal, and a sorting classification for the livestock animal. Exemplary sorting classifications include, but are not limited to, a numerical value (e.g., 1 -5), a predicted tenderness of meat generated by the livestock animal (e.g., very tender, tender, average, and tough), and / or a recommendation for treatment and / or care of the animal (e.g., ready for slaughter, feed for 3 months then slaughter, breed the animal, return animal to the pasture for 1 year, etc.).

[0008] In some embodiments, the device may further comprise an animal identification device configured to identify a livestock animal associated with the livestock animal tissue. Exemplary animal identification devices include, but are not limited to, a camera, an optical scanner, and / or a radio-frequency identification device.

[0009] In some embodiments, the device may be portable and / or handheld. Additionally, or alternatively, the device may be configured for use in, for example, the field (e.g., in a pasture as a standalone device) and / or in a livestock processing and / or feeding facility (e.g., chute-side) so that it may be used to evaluate and / or sorting livestock animals based on output from the livestock animal assessment model. In some embodiments, the device may be configured to communicate with and / or send instructions to machinery,computers, switches, and / or other devices that may be associated with, for example, livestock quality predictions and / or sorting categories and / or resident within in, for example, livestock housing, processing, and / or sorting facilities. For example, the device may be able to send a signal and / or instruction to a chute and / or sorting gate to open, close, remain open, or remain closed responsively to a determination of a category, or grade, for a livestock animal. In this way, the device may automatically sort livestock animals responsively to a quality prediction and / or category by, for example, directing the livestock animals to one or more pens or areas of a processing and / or feeding facility. The device of any of the above claims, wherein the livestock animal assessment model may be trained using ultrasound-shear force paired datasets.

[0010] In some embodiments, the systems, devices, and / or methods disclosed herein may receive an ultrasound image (e.g. ultrasound, hereafter generally referred to as “ultrasound” or “ultrasound” image without limiting solely to this technology) of livestock animal tissue, input the ultrasound image into a livestock animal assessment model, receive an output from the livestock animal assessment model that includes a prediction of a tenderness classification for the imaged livestock animal tissue, and provide an indication of the output to a user interface. In some cases, an identification of a livestock animal associated with the ultrasound image may also be received and associated with the output so that, for example, a quality of the meat generated by an identified livestock animal may be used to provide feedback to, and / or update the model.

[0011] The indication of the output may include, for example, a tenderness score, an estimated value of the livestock animal, a recommendation for the care and / or feeding of the livestock animal, and / or a sorting classification for the livestock animal. At times, the indication of the output to a livestock herd management system and / or livestock processing and / or sorting system.

[0012] In some embodiments, the method may be performed on living cows, chickens, pigs, deer, and / or buffalo as well as meat and / or or other post-slaughter or harvested products in livestock farming, sorting, feeding, harvesting, slaughtering, wholesale, foodservice, and / or retail settings. For example, in some embodiments, the method may be performed chute-side while evaluating and / or sorting live cattle based the output from the livestock animal assessment model.

[0013] In some embodiments, a livestock animal assessment model may be stored in a memory of a livestock evaluation device, which when executed and / or used by a processor of the livestock evaluation device cause the processor to receive a measurement and / or image of tissue from a livestock animal generated by a measurement device and process the measurement and / or image to extract one or more features indicative of livestock tissue characteristics. Exemplary features include, but are not limited to, dimensions, tissue type, and a volume, mass, and / or area of a particular tissue type shown in an image.

[0014] The one or more features may be input into the livestock animal assessment model and an associated output may be received. The output may include, for example, a set of predicted measurements for at least one of backfat thickness, ribeye area, intramuscular fat, and a tenderness classification, internal fat percentage, estimated animal age, yield grade classification, and carcass value estimation for the livestock animal.

[0015] The output may then be used to generate an overall livestock quality score and / or a recommendation for the care, feeding, and / or harvesting of the livestock animal. Optionally, the output may be provided to, for example, a user device and / or the livestock evaluation device.

[0016] In some embodiments, the livestock animal assessment model may be further configured to cause the processor configured to receive user-defined threshold values one or more type(s) of the predicted measurements and, in these embodiments, the generation of the overall livestock quality score may be further based upon the user-defined threshold values. At times, the set of user-defined threshold values may be configurable via a user interface of the livestock evaluation device to allow adjustment of how the predicted measurements are used to generate the overall livestock quality score.

[0017] In some embodiments, the livestock animal assessment model may be trained using a training dataset comprising, for each of a plurality of livestock animals one or more pre-slaughter ultrasound images of tissue, post-slaughter values for at least one of a shear-force measurement, a backfat thickness measurement, a ribeye area measurement, and an intramuscular fat content for the tissue included in the one or more pre-slaughter ultrasound images, and an overall livestock quality score derived from the post-slaughter values according to a predetermined scoring function or classification rule.

[0018] Exemplary methods for training a livestock animal assessment model may include receiving a plurality of training data samples that include one or more of a pre-slaughter ultrasound image of tissue from a livestock animal and corresponding post-slaughter measurement data comprising at least one of a backfat thickness measurement, a ribeye area measurement, an intramuscular fat (IMF) measurement, and a tenderness measurement derived from a shear-force test for one or more livestock animals, training the livestock animal assessment model using the pre-slaughter ultrasound images as input and the reference data and / or the overall livestock quality scores as output and storing the trained livestock animal assessment model in a memory for use in a livestock evaluation device. The livestock animal assessment model may be used to compute an overall livestock quality score for each training sample based on one or more user-defined or standard scoring thresholds applied to the post-slaughter measurement data.Brief Description of the Drawings

[0019] The present invention is illustrated by way of example, and not limitation, in the figures of the accompanying drawings in which:

[0020] FIG. 1 provides a drawing of an exemplary cow with a first, second, and third ultrasound scan path superimposed thereon.

[0021] FIG. 2A provides a photograph of a first ultrasound image that shows the ribeye area and rib fat of a cow that may be taken along a second ultrasound scan path of FIG. 1 , in accordance with some embodiments of the present invention.

[0022] FIG. 2B provides a photograph of a second ultrasound image that shows intramuscular fat of a cow that may be taken along a first ultrasound scan path of FIG. 1 , in accordance with some embodiments of the present invention.

[0023] FIG. 2C provides a photograph of a third ultrasound image, which shows a layer of animal hide, a layer of backfat, a series of striations, a sagittal image of the longissimus dorsi muscle, a rib, and area of intercostal connective tissue for an imaged cow, in accordance with some embodiments of the present invention.

[0024] FIG. 2D provides a photograph of a fourth ultrasound image, which shows layer of animal hide, layer of backfat, and a volume of animal tissue of FIG. 2C but with better resolution, in accordance with some embodiments of the present invention.

[0025] FIG. 2E provides a photograph of a view perpendicular to that of FIG. 2D in which a percentage of IMF may be evaluated, in accordance with some embodiments of the present invention.

[0026] FIG. 3A is a photograph of a ribeye steak with a fat thickness of 1 .1 inches, in accordance with some embodiments of the present invention.

[0027] FIG. 3B is a photograph of a ribeye steak with a fat thickness of 0.2 inches, in accordance with some embodiments of the present invention.

[0028] FIG. 3C is a photograph of a side of a slaughtered cow with a portion of the ribeye area exposed, in accordance with some embodiments of the present invention.

[0029] FIG. 3D is a close-up photograph of the ribeye area of FIG. 3C.Backfat may be measured at, or near, line 310, in accordance with some embodiments of the present invention.

[0030] FIG. 4 is a block diagram of an exemplary system that may be configured to execute one or more methods disclosed herein, in accordance with some embodiments of the present invention.

[0031] FIG. 5A is a block diagram of an exemplary computing system, in accordance with some embodiments of the present invention.

[0032] FIG. 5B is a block diagram of an exemplary livestock animal evaluation device, in accordance with some embodiments of the present invention.

[0033] FIG. 6 is a flowchart showing an exemplary method for training a livestock animal tissue recognition and measurement model, in accordance with some embodiments of the present invention.

[0034] FIG. 7 is a flowchart showing an exemplary process for using a measuring or predicting anatomical features of livestock animals using a livestock animal tissue recognition and measurement model, in accordance with some embodiments of the present invention.

[0035] FIG. 8 is a flowchart showing an exemplary method for training a livestock animal assessment model, in accordance with some embodiments of the present invention.

[0036] FIG. 9A provides a graph that plots a predicted tenderness score for a plurality of cuts of meat as a function of a Warner-Bratzler shear force test measurement in Newtons (N) for each respective cut of meat, in accordance with some embodiments of the present invention.

[0037] FIG. 9B provides a graph that plots a predicted tenderness score for a plurality of cuts of meat as a function of slice shear force test measurement in kilograms-force (KgF) for each respective cut of meat, in accordance with some embodiments of the present invention.

[0038] FIG. 10 is a block diagram depicting steps of the method of FIG. 8 and use of the second, or updated, version of the livestock animal assessment model when installed on a user device, in accordance with some embodiments of the present invention.

[0039] FIG. 11 is a flowchart showing an exemplary process for predicting livestock animal quality using a livestock animal assessment model such as the livestock animal assessment model generated the process shown in FIG. 8, in accordance with some embodiments of the present invention.

[0040] FIG. 12 provides a rendering of a graphic user interface, in accordance with some embodiments of the present invention.

[0041] FIG. 13 provides a rendering of a first graphic user interface, in accordance with some embodiments of the present invention.

[0042] FIG. 14 provides a rendering of a second graphic user interface, in accordance with some embodiments of the present invention.

[0043] Throughout the drawings, the same reference numerals, and characters, unless otherwise stated, are used to denote like features, elements, components, or portions of the illustrated embodiments. Moreover, while the user invention will now be described in detail with reference to the drawings, the description is done in connection with the illustrative embodiments. It is intended that changes and modifications can be made to the described embodiments without departing from the true scope and spirit of the user invention as defined by the appended claims.Written Description

[0044] In the livestock industry, the ability to accurately assess meat quality prior to slaughter is of high value for both producers and processors. Traditional meat tenderness assessments are performed post-slaughter using laboratory-based techniques such as Warner-Bratzler shear force tests. These tests, while accurate, are destructive and cannot be used on live animals. Thus, they are of limited use particularly for decision making regarding the care andfeeding of a particular animal during its lifespan and for the prediction of the value of an animal at slaughter during the animal's lifespan. The present invention overcomes these problems and directed to methods for training and evaluating a livestock animal assessment model that incorporates local model inference, RFID tracking, data logging, and / or continuous learning based on post-slaughter feedback. The present invention is also directed to methods for using a livestock animal assessment model to evaluate livestock animals over their lifespan according to various criteria to predict their quality and / or yield following slaughter, which may impact how the livestock animal is treated (e.g., type and / or frequency of feed) over its lifespan and / or when the a livestock animal may be optimally sold and / or slaughtered.

[0045] The systems, devices, and methods disclosed herein may be configured for use in-situ when an animal is under care or being processed at, for example, a ranch, feedlot, or other facility and may be configured to quickly provide an evaluation of one or more metrics for the animal so that a user may make informed decisions regarding treatment of the animal going forward to, for example, optimize production. The systems, devices, and methods disclosed herein may be configured to provide results that may be compliant and / or aligned with current regulatory and industry standards for livestock processing and / or grading. The systems, devices, and methods described herein may use sonography and other inputs to analyze and present the key metrics of a) a location-specific thickness of backfat (BF); b) ribeye area (REA); intermuscular fat (IMF), and / or internal fat (e.g., kidney, pelvic and heart (KPH) fat) measurements for livestock animals in a manner more relevant, informative, and timely to producers and operators along the life cycle prior to slaughter and / or USDA evaluation.

[0046] In some embodiments, the present invention provides, among other things, a chute-side, real-time (or nearly real-time), and / or portable solution to evaluate tenderness in live animals (e.g., beef tenderness in live cattle), thereby enabling data-driven decisions about sorting, marketing, care, feeding, and / or breeding of the animals proximate to where the livestock animals are. In some cases, the present invention may reduce the costs associated with raising and harvesting livestock and / or improve the quality of livestock by, for example, reducing a need for post-slaughter testing, enabling data-driven herd selection,and / or providing real-time animal sorting and feedback. In some embodiments, the systems, devices, and methods disclosed herein may be portable and capable of running off-line when, for example, access to the Internet is poor or non-existent. Additionally, or alternatively, the systems, devices, and methods disclosed herein may be configured to improve over time through, for example, iterative and / or reinforcement training and / or learning.

[0047] Turning now to the figures, FIG. 1 provides a drawing of an exemplary cow 100 with a first, second, and third ultrasound scan path 110, 120, and 130, respectively, superimposed thereon. First ultrasound scan path 110 corresponds to an IMF measurement position, second ultrasound scan path 120 corresponds to a rib fat thickness measurement position, and third ultrasound scan path 130 corresponds to a rump measurement position. In some embodiments, cow may be scanned with ultrasound equipment along and / or one or more point positions along a first, second, and third ultrasound scan path 110, 120, and 130. Additionally, or alternatively, a cow may be scanned proximate to one or more landmarks (e.g., tattoos, anatomical markers (e.g., joints, bones, and / or between bones (e.g., ribs)).

[0048] FIGs. 2A-2E provide photographs of ultrasonically-generated images of livestock, in this case, cows taken along first, second, or third ultrasound scan path(s) 110, 120, and 130. The images of FIGs. 2A-2E may be displayed on, for example, a user device like user device 415, a display device like display device 420 as shown in FIG. 4 and discussed below. Additionally, or alternatively, the images of FIGs. 2A-2E may be displayed on a screen provided by a livestock animal evaluation device such as livestock animal evaluation device 440, which is discussed below with regard to FIGs. 4, 5A and 5B. In particular, FIG. 2A provides a photograph of a first ultrasound image 201 that shows the ribeye area and rib fat of a cow that may be taken along second ultrasound scan path 120. FIG. 2B provides a photograph of a second ultrasound image 202 that shows intramuscular fat of a cow that may be taken along first ultrasound scan path 110. FIG. 2C provides a photograph of a third ultrasound image 203, which shows a layer of animal hide 220, a layer of backfat 222, a series of striations 224, a sagittal image of the longissimus dorsi muscle 226, a rib 226, and area of intercostal connective tissue 230 for an imaged cow. FIG. 2D provides a photograph of a fourth ultrasound image 204, which shows layer of animal hide 220, layer of backfat 222, and a volume of animal tissue 232 of FIG. 2C but with better resolution and FIG. 2E provides a photograph of a view perpendicularto that of FIG. 2D in which a percentage of IMF may be evaluated. Dimensions and other characteristics of first-fifth images may be used to determine one or more characteristics of a cow.

[0049] FIGs. 3A-3D are photographs showing different cuts of meat from different slaughtered animals, wherein FIG. 3A is a photograph of a ribeye steak with a fat thickness of 1 .1 inches; FIG. 3B is a photograph of a ribeye steak with a fat thickness of 0.2 inches; FIG. 3C is a photograph of a side of a slaughtered cow with a portion of the ribeye area exposed; and FIG. 3D is a close-up photograph of the ribeye area of FIG. 30. Backfat may be measured at, or near, line 310.

[0050] FIG. 4 is a block diagram of an exemplary system 400 that may be configured to execute one or more methods disclosed herein. System 400 includes an optional external ultrasound device 450, an optional external measurement device 410, an optional user device 415, a display device 420, a communication network 425, a computing platform 430, an optional database 435, a livestock animal evaluation device 440, and a livestock animal management system 445. External ultrasound device 450 may be, for example, a handheld or portable device and / or may have characteristics similar to, for example, the Butterfly iQ+, Butterfly iQ3, and / or the Butterfly iQ+ Vet manufactured by Butterfly Network, Inc. and / or a device installed in a livestock animal processing and / or sorting facility. Communication network 425 may be any network enabled to facilitate communication between two or more components of system 400, such as the Internet. Display device 420 may be any screen or other display device configured to display information to a user and, on some occasions, a display device 420 may be resident within ultrasound device 450, livestock animal evaluation device 440 and / or user device 415.

[0051] User device 415 may be any device configured to accept input (e.g., data and / or instructions) from a user, communicate that input to another component of system 400, and / or provide output from one or more components of system 400 to the user. Exemplary user devices 415 include, but are not limited to, computers, smart phones, tablet computers, and the like.

[0052] Measurement device 410 may be an analog or digital measurement device configured to, for example, measure dimensions within an ultrasound image, measure dimensions (e.g., thickness and / or volume) of an animal or portions of tissue within an animal before slaughter and / or after slaughter, weigh an animal before slaughter and / or after slaughter, and / or determine characteristics (e.g.,weight, volume, types of tissue, etc.) for a piece of a slaughtered animal. Exemplary measurement devices 410 include, but are not limited to, rulers, calipers, scales, measuring tapes, digital measuring tools, and laser-based measurement tools. In some embodiments, measurement device 410 may include one or more devices or components used to perform a Warner-Bratzler shear force test. Additionally, or alternatively, measurement device 410 may be an oximeter and / or a device configured to measure tissue oximetry and / or oxygen saturation of tissue and / or tissue layers of an animal while alive or dead. Additionally, or alternatively, measurement device 410 may be a device configured to measure density and / or water content of animal tissue. Additionally, or alternatively, measurement device 410 may be a CT scanner, X-ray machine, and / or high spectral scanner configured to X-ray or scan an animal and / or a portion thereof before and / or after slaughter. In some embodiments, measurements from measurement device 410 may be used to train one or more machine learning models such as a livestock animal assessment model and / or as feedback for the livestock animal assessment model.

[0053] Computing platform 430 may be any computer or set of computers configured to execute one or more methods disclosed herein. In some embodiments, computing platform 430 and / or a portion thereof may be resident within a cloud, or remote, computer environment. At times, computing platform 430 may include a machine learning architecture such as a deep neural network configured to develop, train, and / or test one or more algorithms or processes as, for example, described herein.

[0054] Optional database 435 may be configured to, for example, store information about individual livestock animals (e.g., identifying information, age, breed, sex, type of feed used, medications administered to the animal, disease state of the animal, overall health of the animal, ranch where it was raised, and / or lineage). Additionally, or alternatively, database 435 may be configured to store information about a group of animals (e.g., all animals of a particular shipment, all animals of a particular age or breed, ranch where they were raised, feedlot used, etc.). Additionally, or alternatively, database 435 may be configured to store information about individuals or organizations involved in raising, feeding, or otherwise caring for a particular animal or group of animals. Additionally, or alternatively, database 435 may be configured to store information generated by execution of one or more processes and / or process steps disclosed herein.

[0055] Livestock animal management system 445 may any physical, virtual, and / or software-based system or device configured to manage livestock embodied as, for example, herd management software, sorting gates, pens for holding livestock, and / or livestock feeding and watering systems.

[0056] FIG. 5A provides an example of a system 500 that may be representative of any of the computing systems (e.g., user device 415, measurement device 410, computing platform 430, ultrasound device 450, and / or livestock animal evaluation device 440) discussed herein. Note, not all of the various computer systems disclosed herein have all of the features of system 500. For example, certain ones of the computer systems discussed above may not include a display inasmuch as the display function may be provided by a client computer or device communicatively coupled to the computer system or a display function may be unnecessary. Such details are not critical to the present invention.

[0057] System 500 includes a bus 502 or other communication mechanism for communicating information, and a processor 504 coupled with the bus 502 for processing information and / or executing instructions according to, for example, one or more methods disclosed herein. Computer system 500 also includes a main memory 506, which may be a random access memory (RAM) or other dynamic storage device, coupled to the bus 502 for storing information and instructions to be executed by processor 504. Main memory 506 also may be used for storing temporary variables or other intermediate information during execution of instructions by processor 504. Computer system 500 further includes a read only memory (ROM) 508 or other static storage device coupled to the bus 502 for storing static information and instructions for the processor 504. A storage device 510, for example a hard disk, flash memory-based storage medium, or other storage medium from which processor 504 can read and / or write to, is provided, and coupled to the bus 502 for storing information and instructions (e.g., operating systems, applications programs, and the like). In some embodiments, storage medium 510, memory 506, and / or ROM 508 may store one or more livestock evaluation models disclosed herein and / or sets of instructions for using, generating, training, and / or updating the one or more livestock evaluation models disclosed herein.

[0058] Computer system 500 may be coupled via the bus 502 to a user interface 512, which may be embodied as, for example, a display device like a flat panel display and / or touch screen display, for displaying information to a computeruser, an array of indicator lights, a speaker, and / or a microphone. An input device 514, such as a keyboard including alphanumeric and other keys, mouse, track pad, and / or a touch screen, may be coupled to the bus 502 for communicating information, command selections, directional information, gestures, and controlling cursor movement of / input by the user to the processor 504.

[0059] The processes referred to herein may be implemented by processor 504 executing appropriate sequences of computer-readable instructions contained in main memory 506. Such instructions may be read into main memory 506 from another computer-readable medium, such as storage device 510, and execution of the sequences of instructions contained in the main memory 506 causes the processor 504 to perform the associated actions. In alternative embodiments, hardwired circuitry or firmware-controlled processing units may be used in place of, or in combination with, processor 504 and its associated computer software instructions to implement the invention. The computer-readable instructions may be rendered in any computer language.

[0060] In general, all of the process descriptions provided herein are meant to encompass any series of logical steps performed in a sequence to accomplish a given purpose, which is the hallmark of any computer-executable application.Unless specifically stated otherwise, it should be appreciated that throughout the description of the present invention, use of terms such as “processing”, “computing”, “calculating”, “determining”, “displaying”, “receiving”, “transmitting” or the like, refer to the action and processes of an appropriately programmed computer system, such as computer system 500 or similar electronic computing device, that manipulates and transforms data represented as physical (electronic) quantities within its registers and memories into other data similarly represented as physical quantities within its memories or registers or other such information storage, transmission or display devices.

[0061] Computer system 500 also includes a communication interface 518 coupled to the bus 502. Communication interface 518 may provide a two-way data communication channel with a computer network, which provides connectivity to and among the various computer systems discussed above. For example, communication interface 518 may be a local area network (LAN) card to provide a data communication connection to a compatible LAN, which itself is communicatively coupled to the Internet through one or more Internet service provider networks, anantenna, and / or a transceiver. The precise details of such communication paths are not critical to the present invention. What is important is that computer system 500 can send and receive messages and data through the communication interface 518 and in that way communicate with hosts accessible via the Internet. Power source 520 may be any device configured to store and / or provide electrical power to system 500. Examples include batteries, rechargeable batteries, and / or a coupling to an electrical main. It is noted that the components of system 500 may be located in a single device or located in a plurality of physically and / or geographically distributed devices.

[0062] FIG. 5B provides a block diagram of exemplary livestock animal evaluation device 440 that includes some components of system 500 as well as additional components specific to livestock animal evaluation system 440. For example, livestock animal evaluation device 440 may include a housing 505, a measurement device 525, user interface 512, processor 504, memory 506, ROM 508, storage device 510, power source 520, a communication interface 518, an identification device 540, and / or one or more communication and / or power ports 545. Housing 505 may be a ruggedized, water-resistant, and handheld housing configured to house the components of livestock animal evaluation device 440. Measurement device 525 may be any device configured and enabled to take one or measurements of livestock animal tissue, typically beneath the hide or skin of the livestock animal. On some occasions, measurement device 525 may be an ultrasound imaging device configured to generate ultrasound images of animal tissue. Additionally, or alternatively, measurement device 525 may be a hyperspectral imaging device, a multispectral imaging device, a device to extract blood or take a tissue sample from a livestock animal, a device to perform a measurement on the blood and / or tissue sample (e.g., scan for bacteria and / or viruses, determine lipid and / or cholesterol content, etc.), and / or an oximeter. User interface 512 may be any device, or combination of devices, configured to receive user input and / or provide output to a user and / or another device (e.g., a component of system 400 like computer platform 430 for updating the livestock animal assessment model ) in communication with livestock animal evaluation device 440. In some embodiments, user interface 512 may include keyboard 514 and / or mouse 516.

[0063] Optionally, livestock animal evaluation device 440 may include identification device 540, which may be configured to obtain identification information from / about a livestock animal and / or a facility caring for, feeding, and / or housing a livestock animal. Identification device 540 may be embodied as, for example, a radio frequency identifier (RFID) reader, an optical scanner, and / or a camera. Communication and / or power ports 545 may be embodied as, for example, power couplings, USB ports, USB-C ports and the like. In some embodiments, communication interface 518 may be configured to, for example, communicate with external hardware and / or software (e.g., herd management software) and / or provide instructions thereto. For example, when processor 504 has determined a category for a particular livestock animal, it may provide instructions (via, for example, communication interface 518) to livestock animal management system 445, which may be embodied as a sorting gate proximate to the livestock animal so that the livestock animal may be automatically sorted and / or directed to a position within a livestock animal facility according to the animal’s categorization.

[0064] FIG. 6 is a flowchart showing an exemplary process 600 for training a livestock animal tissue recognition and measurement model. Process 600 may be executed by any of the systems, or system components, disclosed herein. In many cases, process 600 is executed by computing platform 430 using, for example, information received from database 435, ultrasound device 450, measurement device 410, livestock evaluation device 440, and / or user device 415.

[0065] Initially, in step 605, a set of images of animal tissue for a plurality of living livestock animals may be received. In some embodiments, the images may be obtained via non-invasive scans of the animal tissue and may be, for example, ultrasound images like ultrasound images 201-205, hyperspectral images, and / or multispectral images. When images are received, they may include multiple views, or scans, of one or more animals and, in many cases, may include a plurality (e.g. ,000-100, 000, 000) of images of a plurality (e.g., 100-1 ,000,000) of animals and / or animal tissue. Each image may be associated with various identifying characteristics such as an identifier for the animal shown in the image, a location (e.g., along first, second, or third scan paths 110, 120, or 130) on the animal from which an image was taken, an age of the animal, a breed of the animal, and / or a date, time, and / or location at which the ultrasound image was taken. In some embodiments, image identifying characteristics may be associated with the imagesas metadata or other encoded data as, for example, a code and / or identification number tagged to, or otherwise associated with, the animal. In some embodiments, the images received in step 605 may show portions of an animal that show musculature, bone, striations, and / or IMF. In some embodiments, the images received in step 605 may be a series (e.g., 2-100) of images of a region (e.g., ribeye, back, flank, etc.) of the same animal that may collectively analyzed to determine properties of the region scanned using the series of images. For example, a series (e.g., 2-100) images of different regions of a cow’s ribeye area may be received in step 605 and stitched together to generate a larger image the cow’s ribeye area in a manner similar to generating a panoramic image. In some embodiments, an identifier for the livestock animal whose tissue is shown in one or more of the images may be received in step 605. The identifier may identify, for example, a code, name, owner, and / or caregiver for the livestock animal and / or associate an image with same. In some embodiments, the measurement data may include measurements made and / or images taken of a particular animal over time (e.g., every year, every five years, 1 month before scheduled slaughter, etc.).

[0066] In step 610, a set of post-slaughter measurement data samples for each of the livestock animals for which images were received in step 605 may be received. The measurement data may be ground truth data measured from the animals associated with the ultrasound images of step 605 following slaughter such as slice shear test results, Warner-Bratzler shear test results, and / or physical measurements of the animal and / or parts of the animal. Exemplary physical measurements include, but are not limited to, measurements of back fat, IMF, yield, yield by category of meat type, REA, and / or internal fat (e.g., KPH fat). On some occasions, measurement data received in step 610 may be measurements of an animal prior to slaughter and may be obtained from, for example, analysis of ultrasound images, scales, calipers, and the like. In some embodiments, an identifier for the livestock animal whose tissue is being measured may be received in step 610. The identifier may identify, for example, a code, name, owner, and / or caregiver for the livestock animal and / or associate an ultrasound image with same.

[0067] Optionally, in step 615, additional information about one or more of the livestock animals for which ultrasound image(s) were received in step 605 and / or sets of ultrasound images, measurement data, and / or identification information for a particular animal may be received. Additional information includes, but is not limitedto, animal breed, age, weight, sex, maturity (i.e., physiological (as opposed to chronological) age), information received from database 435, USDA grade for the animal and / or cuts of meat from the animal. Animal maturity may be determined using, for example, bone characteristics, cartilage ossification, and / or color and / or texture of ribeye muscle. At times, information used to determine animal maturity may include a scan (e.g., ultrasound, X-ray, CT, and / or MRI) of an animal’s vertebra or vertebrae.

[0068] In step 620, the set of pre-slaughter images of step 605 and the corresponding post-slaughter measurement data may be divided into a training set and a testing set. The training set may be input into a machine learning architecture that may be resident on, for example, computer platform 430 so that a first version of a livestock animal tissue recognition and measurement model may be generated (step 625). The training set may be used in step 625 to build and / or train a model that correlates ultrasound image characteristics with measured data for the animal corresponding to the ultrasound images so that, for example, the model may be trained to recognize characteristics of ultrasound images associated with various types of animal tissue (e.g., adipose tissue, muscle, skin, bone, etc.) and measure characteristics of the recognized tissue including, but not limited to, width, length, area, mass, weight, and volume. For example, the livestock animal tissue recognition and measurement model may be trained via execution of step 625 to recognize characteristics of ultrasound images associated with back fat, IMF, mass, and / or REA. In some embodiments, the model may be trained and / or tuned using animal characteristics such as age, gender, weight, and / or breed.

[0069] In step 630, the first version of the livestock animal tissue recognition and measurement model generated via execution of step 625 may be tested using the testing set of images and post-slaughter measurement data. Results of the testing may be evaluated (step 635) to iteratively update, tune, and / or modify the livestock animal tissue recognition and measurement model to, for example, improve accuracy and / or reduce processing time, thereby generating a second, or updated, version of the livestock animal tissue recognition and measurement model (step 640). Optionally, in step 645, the second, or updated, version of the livestock animal tissue recognition and measurement model may be provided to a user device like user device 410, a livestock animal evaluation device like livestock animal evaluation device 440, and / or a computer like computing platform 430 and / or stored in amemory and / or database like memory 506, ROM 508, storage device 510, and / or database 435.

[0070] FIG. 7 is a flowchart showing an exemplary process 700 for using a measuring or predicting anatomical features of livestock animals using a livestock animal tissue recognition and measurement model such as the livestock animal tissue recognition and measurement model generated in process 600. Process 700 may be executed by any of the systems, or system components, disclosed herein. In many cases, process 700 is executed by a software program running on livestock quality evaluation device 440 and / or user device 415.

[0071] Initially, one or more images and, optionally, an identifier, for a livestock animal may be received (step 705). The images may be obtained by, for example, ultrasound device 450 and / or measurement device 525 when operated along, for example, first, second, and / or third scan paths 110, 120, and / or 130. Optionally, an image received in step 705 may be analyzed for image quality and / or clarity (e.g., lack of noise and / or sufficient contrast) and an indication (e.g., message, tone, flashing light, etc.) that the image is, or is not, of quality may be provided to the user so that he or she may obtain one or more new images if necessary. In some embodiments, the images received in step 705 may correspond to different areas of an animal that are scanned and, at times, these images may be fit together to form a larger composite image of the animal and / or a portion thereof. For example, four different images of the ribeye area of an animal may be received in step 705 and stitched together to form a composite image of the ribeye area that is larger than each of the individual images. In some embodiments, the images received in step 705 may be a series (e.g., 2-40) of images of an area of the same animal (e.g., 10 images that collectively show a foot-long length of the animal along a scan path). The series of images may be generated by, for example, moving a scanner, imager, and / or ultrasound transducer along a scan path of an animal to image tissue underneath with a series of images that may be stitched together to form a larger image of the animal along, for example, the ribeye, back, or flank of the animal.

[0072] Optionally, in step 710, additional information about the livestock animal may be received. Additional information may include, for example, identification for the animal, age, breed, owner, where the animal was raised, weight, one or more measurements for the animal obtained from measurement device 410, and / or gender. Optionally, in step 715, user inputs and / or preferences may bereceived. User inputs and / or preferences may include, for example, preferences for how results of execution of process 700 are displayed, how input for livestock animal tissue recognition and measurement model should be weighted or prioritized and / or where information is stored. In some embodiments, user inputs and / or preferences may include information regarding equipment used to obtain the images (e.g., model number or unique identifier) and location where the images were taken. In some embodiments, a user preference may be an acceptable range of scores and / or measurements for each trait of an animal and / or the animal overall and / or a sorting category. Additionally, or alternatively, user preferences may be characteristics for sorting animals into one or more categories according to, for example, one or more criteria. In some embodiments, additional information received in step 710 may include animal maturity and / or scans or other data that may be used to determine animal maturity and / or physiological age. Additionally, or alternatively, user preferences may be preferences for feed types or how long to feed a particular animal prior to an event such as slaughter and / or treatment for a medical condition.

[0073] Data received in steps 705, 710, and / or 715 may then be input into the livestock animal tissue recognition and measurement model (step 720) of process 600 and one or more outputs may be received (step 725) and provided to the user a user device (e.g., user device 415) and / or a livestock animal evaluation device (livestock animal evaluation device 440) via, for example, a display device like display device incorporated into user interface 515 and / or a display device provided by user device 415 like user interface 512, which may be embodied as a display device (step 930). Additionally, or alternatively, execution of step 730 may include storage of the output in a database and / or memory (e.g., database 435, ROM 508, storage device 510, and / or memory 506). The output of step 725 may be, for example, one or more measurements of the livestock animal made using the image(s) received in step 705. For example, if the images received in step 705 is five ultrasound images taken along scan path 130, an output of step 730 may be recognizing back fat within the images and / or measurement of a thickness and / or volume of back fat shown in the images.

[0074] FIG. 8 is a flowchart showing an exemplary process 800 for training a livestock animal assessment model. Process 800 may be executed by any of the systems, or system components, disclosed herein. In many cases, process 800 is executed by computing platform 430 using, for example, information received fromdatabase 435, ultrasound device 450, measurement device 410, livestock evaluation device 440, and / or user device 415.

[0075] Initially, in step 805, a set of measurement data samples for a plurality of living livestock animals may be received. In some embodiments, the measurement data may be obtained via non-invasive scans of the animal tissue and may be, for example, ultrasound images like ultrasound images 201-205, hyperspectral images, multispectral images, oximetry information, and / or gas chromograph measurements. When images are received, they may include multiple views, or scans, of one or more animals and, in many cases, may include a plurality (e.g., 1 ,000-100,000,000) of images of a plurality (e.g., 100-1 ,000,000) of animals and / or animal tissue. Each piece of measurement data and / or image may be associated with various identifying characteristics such as an identifier for the animal shown in the image, a location (e.g., along first, second, or third scan paths 110, 120, or 130) on the animal from which an image was taken, an age of the animal, a breed of the animal, and / or a date, time, and / or location at which the measurement was taken. In some embodiments, image identifying characteristics may be associated with the images as metadata or other encoded data as, for example, a code and / or identification number tagged to, or otherwise associated with, the animal. In some embodiments, the ultrasound images received in step 805 may show portions of an animal that show musculature, fat, back fat, bone, striations, and / or IMF. In some embodiments, the images received in step 805 may be a series (e.g., 2-100) of images of a region (e.g., ribeye, back, flank, etc.) of the same animal that may collectively analyzed to determine properties of the region scanned using the series of images. For example, a series (e.g., 2-100) images of different regions of a cow’s ribeye area may be received in step 805 and stitched together to generate a larger image the cow’s ribeye area in a manner similar to generating a panoramic image. In some embodiments, an identifier for the livestock animal whose tissue is shown in one or more of the measurements and / or images may be received in step 805. The identifier may identify, for example, a code, name, owner, and / or caregiver for the livestock animal and / or associate an image and / or measurement data sample with same. In some embodiments, the measurement data may include measurements made and / or images taken of a particular animal over time (e.g., every year, every five years, 1 month before scheduled slaughter, etc.). In some embodiments, the measurement data received in step 805 may be generated via process 700.

[0076] In step 810, a set of post-slaughter measurement data samples for each of the livestock animals for which measurements and / or images (e.g., ultrasound images, hyperspectral images, multispectral images, oximetry information, gas chromograph measurements blood and / or tissue samples and / or analysis of same) were received in step 805 may be received. The measurement data may be ground truth data measured from the animals associated with the ultrasound images of step 805 following slaughter such as slice shear test results, Warner-Bratzler shear test results, and / or physical measurements of the animal and / or parts of the animal. Exemplary physical measurements include, but are not limited to, measurements of back fat, IMF, yield, yield by category of meat type, REA, and / or internal fat (e.g., KPH fat). On some occasions, measurement data received in step 810 may be measurements of an animal prior to slaughter and may be obtained from, for example, analysis of ultrasound images, scales, calipers, and the like. In some embodiments, an identifier for the livestock animal whose tissue is being measured may be received in step 810. The identifier may identify, for example, a code, name, owner, and / or caregiver for the livestock animal and / or associate an ultrasound image with same.

[0077] Optionally, in step 815, additional information about one or more of the livestock animals for which measurement data samples were received in step 805 and / or sets of ultrasound images, measurement data, and / or identification information for a particular animal may be received. Additional information includes, but is not limited to, animal breed, age, weight, sex, maturity (i.e., physiological (as opposed to chronological) age), information received from database 435, USDA grade for the animal and / or cuts of meat from the animal. Animal maturity may be determined using, for example, bone characteristics, cartilage ossification, and / or color and / or texture of ribeye muscle. At times, information used to determine animal maturity may include a scan (e.g., ultrasound, X-ray, CT, and / or MRI) of an animal’s vertebra or vertebrae.

[0078] In step 820, the set of pre-slaughter measurement data samples of step 805 and the corresponding post-slaughter measurement data may be divided into a training set and a testing set. The training set may be input into a machine learning architecture that may be resident on, for example, computer platform 430 so that a first version of a livestock animal assessment model may be generated (step 825). The training set may be used in step 825 to build and / or train a model thatcorrelates ultrasound image characteristics with measured data for the animal corresponding to the ultrasound images so that, for example, the model may be trained to recognize characteristics of ultrasound images associated with various measured characteristics of animals. For example, the model may be trained via execution of step 825 to recognize characteristics of ultrasound images associated with animal tenderness as measured by, for example, slice shear force and / or Warner-Bratzler shear test results. Additionally, or alternatively, the model may be trained via execution of step 825 to recognize characteristics of ultrasound images associated with measured and / or determined values for a corresponding animal’s yield, USDA grade, backfat, IMF, and / or REA. In some embodiments, the model may be trained and / or tuned using animal characteristics such as age, gender, weight, and / or breed.

[0079] In some embodiments, data received in steps 805, 810, and / or 815 may correspond to different points in time over an animal’s life span. For example, ultrasound images for an animal may be collected over the animal’s life span on, for example, a periodic (e.g., annual, semi-annual, or every five years) basis and / or an as-needed basis (e.g., when the animal is receiving veterinary care, is being transferred to another location, etc.) and these images may be correlated with measurement data for the animal following slaughter so that, for example, the model may learn how image data over the animal’s life span may correlate to, for example, yield and / or quality metrics (e.g., IMF, BF, REA, USDA grade, etc.) upon slaughter.

[0080] In step 830, the first version of the livestock animal assessment model generated via execution of step 825 may be tested using the testing set of ultrasound images and measurement data. Results of the testing may be evaluated (step 835) to iteratively update, tune, and / or modify the livestock animal assessment model to, for example, improve accuracy and / or reduce processing time, thereby generating a second, or updated, version of the livestock animal assessment model (step 840). Optionally, in step 845, the second, or updated, version of the livestock animal assessment model may be provided to a user device like user device 410, a livestock animal evaluation device like livestock animal evaluation device 440, and / or a computer like computing platform 430 and / or stored in a memory and / or database like memory 506, ROM 508, storage device 510, and / or database 435.

[0081] FIG. 9A provides a graph 901 that plots a predicted tenderness score for a plurality of cuts of meat as a function of a Warner-Bratzler shear force testmeasurement in Newtons (N) for each respective cut of meat. The data plotted on graph 901 is a possible output of execution of step(s) 830 and / or 835. The tenderness score of graph 901 is a prediction made by the first and / or second version of the livestock animal assessment model based upon analysis of the ultrasound images and optional other information received and input into the livestock animal assessment model. As may be seen in graph 901 , the predicted tenderness scores are well correlated with the measured tenderness (represented as Warner-Bratzler shear force test measurements) for each of the of cuts of meat with a R2value of 0.5063.

[0082] FIG. 9B provides a graph 902 that plots a predicted tenderness score for a plurality of cuts of meat as a function of slice shear force test measurement in kilograms-force (KgF) for each respective cut of meat. The data plotted on graph 902 is a possible output of execution of step(s) 830 and / or 835. As may be seen in graph 902, the predicted tenderness scores are well correlated with the measured tenderness (represented as slice shear force test measurements) for each of the of cuts of meat with a R2value of 0.5519.

[0083] FIG. 10 is a block diagram depicting steps of process 1000 and use of the second, or updated, version of the livestock animal assessment model when installed on a user device like user device 415 and / or a livestock quality evaluation device like livestock quality evaluation device 440. Data like the data received in steps 805, 810, and / or 815 is provided in block 1010 and is split into a set of testing data 1015 and a set of training data 1020 when, for example, step 820 is executed. The first version of the livestock animal assessment model 1030 is then developed as part of model creation 1025 as, for example, described above with regard to step 825. The first version of the livestock animal assessment model 1030 is then tested using testing data 1015 as, for example, described above with regard to step 830 and evaluated (see e.g., step 835) with evaluation results used to generate a second, or updated, version of the livestock animal assessment model 1045 (see e.g., step 840), which is communicated to livestock quality evaluation device 440 (see e.g., step 845). Production data 1040 is then input into model 1045 and prediction results 1050 are prepared as shown in, for example, FIG. 11 and described below.

[0084] FIG. 11 is a flowchart showing an exemplary process 1100 for assessing livestock animals to, for example, predict livestock quality, yield, futuregrowth, and / or value using a livestock animal assessment model such as the livestock animal assessment model generated in process 800. Process 1100 may be executed by any of the systems, or system components, disclosed herein. In many cases, process 1100 is executed by a software program running on livestock quality evaluation device 440 and / or user device 415.

[0085] Initially, one or more ultrasound images and, optionally, an identifier, for a livestock animal may be received (step 1105). The ultrasound images may be obtained by, for example, ultrasound device 450 and / or measurement device 525 when operated along, for example, first, second, and / or third scan paths 110, 120, and / or 130. Optionally, an image received in step 1105 may be analyzed for image quality and / or clarity (e.g., lack of noise and / or sufficient contrast) and an indication (e.g., message, tone, flashing light, etc.) that the image is, or is not, of quality may be provided to the user so that he or she may obtain one or more new images if necessary. In some embodiments, the ultrasound images received in step 1105 may correspond to different areas of an animal that are scanned and, at times, these images may be fit together to form a larger composite image of the animal and / or a portion thereof. For example, four different images of the ribeye area of an animal may be received in step 1105 and stitched together to form a composite image of the ribeye area that is larger than each of the individual images. In some embodiments, the images received in step 1105 may be a series (e.g., 2-40) of images of an area of the same animal (e.g., 10 images that collectively show a foot-long length of the animal along a scan path). The series of images may be generated by, for example, moving a scanner, imager, and / or ultrasound transducer along a scan path of an animal to image tissue underneath with a series of images that may be stitched together to form a larger image of the animal along, for example, the ribeye, back, or flank of the animal.

[0086] Optionally, in step 1110, additional information about the livestock animal may be received. Additional information may include, for example, identification for the animal, age, breed, owner, where the animal was raised, weight, one or more measurements for the animal obtained from measurement device 410, and / or gender. Optionally, in step 1115, user inputs and / or preferences may be received. User inputs and / or preferences may include, for example, preferences for how results of execution of process 1100 are displayed, how input for livestock animal assessment model should be weighted or prioritized and / or where informationis stored. In some embodiments, user inputs and / or preferences may include information regarding equipment used to obtain the ultrasound images (e.g., model number or unique identifier) and location where the images were taken. In some embodiments, a user preference may be an acceptable range of scores and / or measurements for each trait of an animal and / or the animal overall and / or a sorting category. Additionally, or alternatively, user preferences may be characteristics for sorting animals into one or more categories according to, for example, one or more criteria. In some embodiments, additional information received in step 1110 may include animal maturity and / or scans or other data that may be used to determine animal maturity and / or physiological age. Additionally, or alternatively, user preferences may be preferences for feed types or how long to feed a particular animal prior to an event such as slaughter and / or treatment for a medical condition.

[0087] Data received in steps 1105, 1110, and / or 1115 may then be input into the livestock animal assessment model (step 1120) of process 800 and one or more outputs may be received (step 1125) and provided to the user, a user device (e.g., user device 415) and / or a livestock animal evaluation device (livestock animal evaluation device 440) via, for example, a display device like display device incorporated into user interface 515 and / or a display device provided by user device 415 like user interface 512, which may be embodied as a display device (step 1130). Additionally, or alternatively, execution of step 1130 may include storage of the output in a database and / or memory (e.g., database 435, ROM 508, storage device 510, and / or memory 506). FIG. 12 provides a rendering of an exemplary graphic user interface 1200 that shows an exemplary manner in which step 1130 may be executed. Graphic user interface 1200 includes an optional first ultrasound image 1235, an optional second ultrasound image 1240, an optional third ultrasound image 1245, an IMF banner 1210 that includes an IMF percentage (in this case, 11%) as well as an IMF score (in this case, 6), a BF banner 1215 that includes an BF thickness measurement (in this case, 0.547inches) as well as a BF score (in this case, 5), a tenderness banner 1220 that includes a tenderness score (in this case,1), a yield estimate banner 1225 that includes a yield estimate score (in this case, 3), a sorting category (in this case, 3) and a banner of icons 1230. Optional first ultrasound image 1235, second ultrasound image 1240, and / or third ultrasound image 1245 may provide image of, for example, IMF, BF, and / or rib depth. Icons within banner of icons 1230 may provide options for a user to interact with GUI 1200to, for example, select settings, adjust measurements, or otherwise interact with GUI 1200 and / or one or more systems and / or devices disclosed herein.

[0088] In some embodiments, execution of process 1100 not only evaluates the overall quality of livestock animals based on measurements back fat, ribeye area, and IMF, but also has the capability to analyze and consider additional traits such as weight, sex, age, and breed when evaluating the quality of an animal and / or predicting features (e.g., tenderness) of an animal. This comprehensive analysis provides users with a holistic view of livestock animals, allowing for more nuanced decision-making in areas such as breeding programs, health management, and market readiness for each animal and / or group of animals. Additionally, or alternatively, an output of execution of process 1100 may simply be a category for the animal and / or a prediction of a degree of tenderness of meat from the animal.

[0089] In some embodiments, execution of one or more processes described herein may yield generation of one or more comprehensive quality score(s) for each animal, which may serve as a quantitative representation of the livestock's overall quality, incorporating various factors, including, but not limited to, tenderness, muscle density, fat distribution, weight, sex, age, and breed. For example, if results indicate that the animal is of high quality, it may assist the user in determining that it is ready for slaughter. Alternatively, if results indicate that the animal does not have enough fat, the feeding regimen for the animal may change. At times, this comprehensive quality score may be correlated to one or more USDA classifications for meat. At times, these determinations may be facilitated by one or more user inputs received in step 1115 of process 1100. For example, a user may have a particular number (e.g., 2-10) different categories into which it wants to sort animals and the user input received in step 1115 may provide criteria for the different categories so that an output of execution of process 1100 is the name (or other identifier) of the category into which the animal fits. Continuing with this example, a user may enter preferences and / or metrics by which to categorize animal quality on a scale of 1-5 with 1 being low quality and 5 being high quality so that an output of process 1100 may be assigning the animal a quality score (e.g., 1-5) that may assist the user in making future decisions about the animal such as what and how long to continue to feed the animal, when to slaughter the animal, and / or how to sort the animal within a livestock facility.

[0090] FIG. 13 provides a rendering of a first graphic user interface 1300 that is an exemplary manner in which step 1130 may be executed. First graphic user interface 1300 includes an identification table 1310, a current USDA measures table 1315, a predicted value table 1320, a grade table 1325, a scan status table 1330, a feed recommendation table 1335, and a tenderness score table 1340. Identification table 1310 includes information about an animal, in this case, an angus steer with an identifying number RA15AG00FY1 (referred to herein as “animal FY1”) that is fifteen months old, 750 pounds, in good health, and has been range fed. Current USDA measures in current USDA measures table 1315 for animal FY1 are a BF value of 0.35 inches, a REA of 10 square inches, and IMF of 3.0%. Predicted values for animal FY1 provided by predicted value table 1320 may be populated via execution of process 1100 and include a predicted BF value of 0.5 inches, a predicted REA of 14 square inches, and a predicted IMF value of 3.5%. Grade table 1325 provides a yield grade for animal FY1 and a quality grade of select plus. Scan status table 1330 provides an indication of the status of ultrasound scans, which may indicate whether or not the scan is of a sufficient quality to execute process 1100. Scan status table 1330 indicates that the BF, REA, and IMF scans were all acceptable. The feed recommendation value provided by feed recommendation value table 1335 indicates that is recommended to provide grow feed from Lot A for animal FY1 . The tenderness score provided by tenderness score table 1340 provides a value of 1 , or high tenderness.

[0091] FIG. 14 provides a rendering of a second graphic user interface 1400 that is an exemplary manner in which step 1130 may be executed. Second graphic user interface 1400 includes an identification table 1410, a current USDA measures table 1415, a predicted value table 1420, a scan status table 1430, a feed recommendation table 1435, a tenderness score table 1440, and a weight / feed table 1445. Identification table 1410 is similar to identification table 1310 includes information about an animal, in this case, an angus steer with an identifying number RA15G4D (referred to herein as “animal FY2”) that is fifteen months old. As indicated by weight / feed table 1445, animal FY2 is 800 pounds, has had zero days on feed to date and data regarding prior scans is not available / not applicable. Current USDA measures in current USDA measures table 1415 for animal FY2 are a BF value of 0.35 inches, a REA of 10 square inches, and IMF of 9.5%, a maturity grade of A, a yield grade of 3, and a quality grade of prime. Predicted USDA valuesfor animal FY2 provided by predicted value table 1420 may be populated via execution of process 1100 and include a predicted BF value of 0.5 inches, a predicted REA of 14 square inches, a predicted IMF value of 11%, a predicted maturity grade of A, a predicated yield grade of 3, and a predicted quality grade of prime. Scan status table 1430 provides an indication of the status of ultrasound scans, which may indicate whether or not the scan is of a sufficient quality to execute process 1100. Scan status table 1430 indicates that the BF, REA, IMF, and vertebrae scans were all acceptable. The feed recommendation value provided by feed recommendation value table 1435 indicates that is recommended to finish providing the animal FY2 with feed and that there are forty days until harvest. The tenderness score provided by tenderness score table 1440 provides a value of A, or high tenderness.

Claims

Claims:We claim:

1. A device comprising: a measurement device configured to generate a measurement of livestock animal tissue; a user interface configured to receive input from and / or display output to a user; a processor communicatively coupled to the measurement device, the user interface, and a memory, the processor being configured to execute one or more sets of instructions stored on the memory; the memory configured to store a livestock animal assessment model and a set of instructions thereon, which when executed by the processor cause the processor to: receive the image of livestock animal tissue from the measurement device; input the ultrasound image into the livestock animal assessment model ; receive an output from the livestock animal assessment model, the output including a prediction of a tenderness classification for the imaged livestock animal tissue; provide an indication of the output to the user interface; a power source configured to provide power to the measurement device, the user interface, the processor, and the memory; and a housing configured to house the measurement device, the user interface, the processor, the memory, and the power source., wherein the measurement is from a living livestock animal.

2. The device of claim 1 , further comprising: an animal identification device configured to identify a livestock animal associated with the livestock animal tissue.

3. The device of the claim 2, wherein the animal identification device is at least one of a camera, an optical scanner, and a radio-frequency identification device.

4. The device of any of the above claims wherein the device is portable.

5. The device of any of the above claims wherein the housing is configured to be handheld.

6. The device of any of the above claims, wherein the indication of the output is further provided to a livestock herd management system.

7. The device of any of the above claims, wherein the indication of the output includes at least one of a tenderness score, an estimated value of the livestock animal, a recommendation for the care and / or feeding of the livestock animal, and a sorting classification for the livestock animal.

8. The device of claim 7, wherein the sorting classification for the livestock animal is at least one of fender, average, and tough.

9. The device of any of the above claims, wherein the device is configured as a portable chute-side measurement device for evaluating and sorting livestock animal based on output from the livestock animal assessment model.

10. The device of any of the above claims, wherein the set of instructions, which when executed by the processor, further cause the processor to provide an instruction to a livestock sorting gate.11 .The device of claim 10, wherein the instruction is to open or stay closed.

12. The device of claim 10, wherein the livestock sorting gate is one of a plurality of livestock sorting gates, each livestock sorting gate of the plurality being associated with an output of the livestock animal assessment model.

13. The device of claim 10, wherein output of the livestock animal assessment model is a classification of the livestock and the livestock sorting gate is one of a plurality of livestock sorting gates, each livestock sorting gate of the plurality being associated with classification for livestock.

14. The of any of the above claims, further comprising: a transceiver for communicating the indication of the output to an external device, wherein the transceiver is resident within the housing.

15. The device of claim 14, wherein the transceiver is configured to communicate using a near-field communication protocol, Wi-Fi, a cellular communication protocol, and a communication port.

16. The device of any of the above claims, wherein the livestock animal assessment model is trained using ultrasound-shear force paired datasets.

17. The device of any of the above claims, wherein the livestock animal assessment model is trained using feedback from the device.

18. The device of any of the above claims, further comprising: non-volatile memory configured to store at least one of the ultrasound image and the output from the livestock animal assessment model.

19. A method comprising using the device of any of claims 1-18 to predict a quality of a livestock animal.

20. A system comprising: the device of any of claims 1-18; and a feedback mechanism configured to provide feedback to the processor to update the livestock animal assessment model.

21. The system of claim 20, wherein the feedback mechanism is a measurement device configured to measure a characteristic of a slaughtered livestock animal.

22. The system of claim 20, further comprising: a computer system external to the device, wherein the feedback mechanism is configured to provide feedback to the computer system external to the device in addition to, or instead of, to the processor, the computer system external to the device being configured to update the livestock animal assessment model responsively to the feedback.

23. A method comprising: receiving, by a processor, an ultrasound image of livestock animal tissue; inputting, by the processor, the ultrasound image into a livestock animal assessment model ; receiving, by the processor, an output from the livestock animal assessment model, the output including a prediction of a tenderness classification for the imaged livestock animal tissue; and providing, by the processor, an indication of the output to a user interface.

24. The method claim 23, wherein the indication of the output includes at least one of a tenderness score, an estimated value of the livestock animal, a recommendation for the care and / or feeding of the livestock animal, and a sorting classification for the livestock animal.

25. The method of claim 23 or 24, further comprising: providing, by the processor, the indication of the output to a livestock herd management system.

26. The method of any of claims 23-25, further comprising: receiving, by the processor, an identification of a livestock animal associated with the ultrasound image; and associating, by the processor, the identification of the livestock animal with the output.

27. The method of any of claims 23-26, further comprising: providing, by the processor, at least one of the identification of the livestock animal and an association between the identification of the livestock animal with the output to a livestock herd management system.

28. The method of any of claims 23-27, wherein the method is performed chute-side while evaluating and sorting live cattle based the output from the livestock animal assessment model.

29. A livestock animal assessment model stored in a memory of a livestock evaluation device and executable by a processor of the device, the model being configured cause the processor to: receive a measurement and / or image of tissue from a livestock animal generated by a measurement device; process the measurement and / or image to extract one or more features indicative of livestock tissue characteristics; input the one or more features into the livestock animal assessment model; receive output from the livestock animal assessment model, the output including a set of predicted measurements for at least one of backfat thickness, ribeye area, intramuscular fat, and a tenderness classification; and generate an overall livestock quality score using the set of predicted measurements.

30. The livestock animal assessment model of claim 29, wherein the livestock animal assessment model is further configured to cause the processor configured to: receive user-defined threshold values for each type of the predicted measurements, wherein the generation of the overall livestock quality score is based upon the user-defined threshold values.31 .The livestock animal assessment model of claim 29 or 30, wherein the model is trained using a training dataset comprising, for each of a plurality of livestock animals: one or more pre-slaughter ultrasound images of tissue; post-slaughter values for at least one of a shear-force measurement, a backfat thickness measurement, a ribeye area measurement, and an intramuscular fat content for the tissue included in the one or more pre-slaughter ultrasound images; andan overall livestock quality score derived from the post-slaughter values according to a predetermined scoring function or classification rule.

32. The livestock animal assessment model of any of claims 29-31 , wherein the set of user-defined threshold values is configurable via a user interface of the livestock evaluation device to allow adjustment of how the predicted measurements are used to generate the overall livestock quality score.

33. The livestock animal assessment model of any of claims 29-32, wherein the set of predicted measurements further include one or more of internal fat percentage, estimated animal age, yield grade classification, and carcass value estimation.

34. A method for training a livestock animal assessment model comprising: receiving a plurality of training data samples, each sample comprising at least one pre-slaughter ultrasound image of tissue from a livestock animal and corresponding post-slaughter measurement data comprising at least one of a backfat thickness measurement, a ribeye area measurement, an intramuscular fat (IMF) measurement, and a tenderness measurement derived from a shearforce test; training the livestock animal assessment model using the pre-slaughter ultrasound images as input and the reference data and / or the overall livestock quality scores as output; and storing the trained livestock animal assessment model in a memory for use in a livestock evaluation device.

35. The method of claim 34, further comprising: computing an overall livestock quality score for each training sample based on one or more user-defined or standard scoring thresholds applied to the postslaughter measurement data.

Images