Animal feature status

Abstract

A point-of-care PON instrument (200) for acquiring an indicator of physiological characteristics of an animal, the instrument (200) comprising: an insertable rotor cartridge (202) for sample detection comprising a plurality of reagent chambers (202a) for receiving a sample of animal body fluid, each reagent chamber (202a) containing a reagent configured to react with the sample of animal body fluid; a rotor configured to drive rotation of the rotor cartridge (202) to agitate reactants in the plurality of reagent chambers (202a); an analyzer (220) comprising an optical analysis device and a processor (206) configured to optically analyze the reactants of each reagent chamber (202) after reaction of the reagents and the animal body fluid sample to generate biomarker data; a transmitter (208) configured to communicate with a remote biomarker recognition system (400), where the analyzer (220) is configured to generate biomarker data in a particular format that supports the transmitter (208) to transmit the biomarker data to the biomarker recognition system (400) for analysis and to generate animal physiological feature status results.

Description

[0001] This disclosure relates to an instant detection instrument for acquiring animal physiological characteristic indicators, an animal physiological characteristic recognition system, and a computer implementation method for determining the physiological characteristic state of an animal.

[0002] Animals may undergo physiological changes throughout their lives, which can be cyclical or permanent. Studying an animal's physiological characteristics is of great importance to scientists researching that animal. In the case of domesticated animals, monitoring these physiological changes may be necessary. Understanding these physiological changes or current physiological characteristics can guide breeders (such as farmers) on how to move animals between different habitats, adjust feeding methods, change their function (e.g., from breeding or egg-laying animals to slaughter-ready animals), or otherwise monitor their development and their suitability for mating.

[0003] In a specific example of animal physiological changes discussed in this article, silvering is a complex series of behavioral, developmental, and physiological changes that represent the adaptation of fish from freshwater to saltwater or marine environments. Also known as Pallas's juvenile silvering, it occurs in salmonid fish, specifically those belonging to the family Salmonidae, such as Atlantic salmon. As a physiological process, silvering enables anadromous spawning fish to migrate from freshwater to saltwater, survive in constantly changing osmotic pressure environments, and maintain fluid homeostasis.

[0004] In the wild, fish undergo a transition from juvenile to adult when they first migrate to the sea. In salmon farms, farmed fish are transferred from hatcheries to marine growth areas. Farmers need to determine the optimal time to transfer farmed fish from freshwater to saltwater. If transferred too early, the fish will not be able to fully adapt to the saltwater environment and will be unable to survive and reproduce; if transferred too late, the fish will no longer be adapted to the freshwater environment. Fish may be ready to be transferred to saltwater, but if they are not transferred, they will revert to freshwater physiological states. Therefore, fish need to be transferred within the optimal time window.

[0005] Improper timing of fish rearing can hinder growth and lead to death due to unsuitable environments. Failure to reach the required size is uneconomical and wasteful for fish farmers, and can also pose hygiene risks (live and dead fish may come into close contact). With increasing public concern about animal welfare in aquaculture, minimizing unnecessary loss of life before fish are ready for food and other production purposes is crucial.

[0006] Animal transfer is not limited to aquaculture. For example, ornamental fish and other pets, mollusks and other aquatic species, other pets, and livestock may also be transferred under different conditions.

[0007] Typically, experts in this field analyze whether fish can successfully transition from freshwater to saltwater environments by testing samples from their gills. However, this analytical method often causes fatal harm to the selected test fish.

[0008] Other physiological characteristics of fish are also worth monitoring—for example, non-anaggregative spawning fish (i.e., those that do not migrate from freshwater rivers to the ocean and return to spawn) may experience physiological changes due to stress, so it may be necessary to assess the health status of non-anaggregative spawning fish.

[0009] Generally, farmers, as well as veterinarians and feed producers, will benefit from efficient and reliable animal health benchmarking methods, such as aquatic animal health assessments, and from gaining biological insights—for example, indicating the condition of juvenile fish or gonadal maturity.

[0010] Statement of Invention Content The disclosed techniques are intended to alleviate, eliminate, mitigate, or cure various problems known in the art. While the invention is defined by the appended claims, the summary section lists various aspects of the disclosed techniques, including the claimed techniques, and provides examples of preferred embodiments and descriptions of possible technical effects.

[0011] The first aspect of the disclosed technology relates to a point-of-care testing (PON) instrument for acquiring animal physiological indicators, the PON instrument comprising: an insertable rotor cartridge for sample determination, the cartridge containing multiple reagent chambers for receiving animal body fluid samples, each reagent chamber containing a reagent configured to react with the animal body fluid sample; a rotor configured to drive the rotor cartridge to rotate to agitate the reactants in the multiple reagent chambers; an analyzer including an optical analysis device and a processor, the analyzer being configured to perform optical analysis on the reactants in each reagent chamber after the reagents react with the animal body fluid sample to generate biomarker data; and a transmitter configured to communicate with a remote biomarker identification system, wherein the analyzer is configured to generate biomarker data in a specific format that allows the transmitter to transmit the biomarker data to the biomarker identification system for analysis and to generate animal physiological state results.

[0012] Point-of-care testing (POC) instruments can be used to obtain indicators of physiological characteristics in domestic animals. Animal body fluid samples can be body fluid samples from domestic animals.

[0013] Domestic animals can be one of the following: farmed terrestrial animals (including farmed arboreal animals), farmed aquatic or semi-aquatic animals, pets, or animals used in scientific research. For example, domestic animals can be fish.

[0014] The format of biomarker data may allow for processing of biomarker data within the laboratory information system of a biomarker identification system.

[0015] The format of biomarker data may conform to the Layer 7 (HL7) basic standard for health information exchange.

[0016] The format of biomarker data may conform to the Level 7 Rapid Healthcare Interoperability Resource (FHIR) standard for the exchange of health information.

[0017] Biomarker data may be formatted in accordance with the FHIR standard and encoded as JavaScript Object Notation (JSON) data.

[0018] Another aspect of this disclosure provides an animal physiological characteristic identification system, comprising: a point-of-care testing instrument as described above; a biomarker identification system located remotely from the point-of-care testing instrument, comprising: a receiver configured to receive biomarker data from the point-of-care testing instrument; a memory containing a machine learning model, wherein the model is a classification or regression model configured to generate an animal physiological characteristic state based on biomarkers detected in the biomarker data; and a transmitter configured to transmit the animal physiological characteristic state to a user computing device; and a user computing device comprising: a receiver configured to receive signals from the biomarker identification system; and a display configured to provide results generated by the biomarker identification system analyzing the received biomarker data to identify the presence of biomarkers in a sample, the results indicating the animal physiological characteristic state.

[0019] The model can be configured to identify the presence of biomarkers in a sample based on biomarker data and to assign animal state indicators to the biomarker data, wherein the model is configured to aggregate two or more animal state indicators to generate a comprehensive animal physiological state.

[0020] The model may contain an extreme gradient boosting machine learning algorithm set consisting of decision trees, each of which is configured to receive biomarker data parsed from a first data format to a second data format as input, and the set of decision trees is configured to generate a score based on which the determination of the animal's physiological state is based.

[0021] The biomarker identification system also includes a laboratory information system, which is configured to parse the received biomarker data from a first data format and convert it into a second data format suitable for input to the model, and to use the parsed biomarker data as input to the model. The laboratory information system is also configured to generate animal physiological characteristic status reports for transmission to the user's computing device.

[0022] The animal physiological characteristic identification system may further include at least one point-of-care testing instrument according to any one of the preceding claims, wherein the laboratory information system is configured to parse the received biomarker data from each point-of-care testing instrument into a format suitable for input into the model, and to transmit the parsed biomarker data from each point-of-care testing instrument to the model.

[0023] Animal physiological characteristics may include the fish’s ability to survive in saltwater (e.g., calculated probabilities).

[0024] Animal body fluid samples may include any of the following: blood, whole blood, plasma, serum, mucus, feces, and ascites.

[0025] The model can be configured to determine one or more animal state indicators based on biomarkers in a sample, and wherein the model is configured to generate animal physiological state based on one or more animal state indicators, wherein the one or more animal state indicators include one of the following indicators: health state, stress state, gonadal state, metabolic state, and silvering state.

[0026] One indicator of animal condition is chronic stress. In fish, biomarkers in the sample may include phosphorus, sodium, calcium, growth hormone, and chloride. It's worth noting that these biomarkers may apply to non-fish, but are particularly relevant to fish. Therefore, one indicator of animal condition is chronic stress, and biomarkers in the sample may include phosphorus, sodium, calcium, growth hormone, and chloride.

[0027] Another aspect of this disclosure provides a computer-implemented method for determining the physiological state of an animal, the method being executed at a biomarker identification system, comprising: receiving biomarker data from an animal body fluid sample acquired from a point-of-care testing instrument, wherein the biomarker data is in a first data format; parsing the biomarker data into a second format suitable for input into a classification or regression model at a server remotely to the point-of-care testing instrument; inputting the parsed biomarker data into a model at the server remotely to the point-of-care testing instrument, the model being configured to process the parsed biomarker data to obtain multiple animal state indicators; aggregating the multiple animal state indicators into a single animal physiological state; and outputting the animal physiological state through the model.

[0028] Another aspect of this disclosure provides a non-transitory computer-readable medium containing computer code that, when loaded from memory and executed by one or more processors or processing circuits, enables a biomarker identification system to perform the methods described above.

[0029] The disclosed aspects and embodiments may be combined with each other in any suitable manner that will be obvious to those skilled in the art.

[0030] Brief description of the attached figures Some embodiments of the disclosed technology will be described below with reference to the accompanying drawings, which are merely examples, wherein: Figure 1A The fish farming enclosure environment is shown; Figure 1B Different fish farming enclosure environments are shown; Figure 2 A block diagram of an exemplary PON (point-of-care detection) instrument according to an embodiment is shown; Figure 3 An exemplary rotary chuck and instant detection instrument according to an embodiment are shown; Figure 4 An exemplary biomarker identification system according to an embodiment is shown, the system including a server; Figure 5 An example of a decision tree for a model according to an embodiment is shown; Figure 6 An exemplary workflow of a biomarker identification system is illustrated, which, according to an embodiment, begins with three exemplary point-of-care testing instruments; Figure 7A An exemplary timeframe for analysis according to an embodiment is shown; Figure 7B An exemplary timeframe is shown for analysis based on an example, where the analysis results are fed back to a point-of-care detection instrument. Figure 8 A method for training a model according to an embodiment is shown; Figure 9A An exemplary data stream is shown in a fish physiological feature recognition system according to an embodiment; Figure 9B An exemplary data stream is shown, which passes through a fish physiological characteristic recognition system, and, according to the example, the results are fed back to a real-time detection instrument; Figure 10 An exemplary anatomical surface is shown in the model; Figure 11 An example user interface display containing an exemplary decomposition configuration file is shown; and Figure 12 The correlation between an exemplary machine learning model used to analyze the state of juvenile fish and the PCR gold standard is shown. Detailed Implementation

[0031] This disclosure uses an example animal—a fish—as an example, but as stated above, this disclosure generally relates to animals, and more specifically to domestic animals. Fish represents any suitable animal. The following discussion of “fish” can be extended to “animal” or “domestic animal”, and may also apply to other exemplary animals, such as farmed terrestrial animals, aquatic or semi-aquatic animals, pets, or animals being studied because of their physiological characteristics.

[0032] Examples of farmed terrestrial animals include cattle, sheep, alpacas, etc. The term "farming" does not only refer to raising animals for the purpose of obtaining animal by-products / meat, but may also include farm-raised animals or animals kept in captivity for recreational or other purposes. "Farming" here means animals that are not wild, raised for profit / as part of a commercial activity (i.e., not pets), whose characteristics will be determined as described herein. Pets are another example of domesticated animals. For the purposes of this application, "domestication" also includes animals under scientific observation / research, such as birds whose migratory behavior is being monitored. Animals, including wild animals, whose physiological characteristics can be determined as described herein. However, it is foreseeable that the importance of this disclosure lies in monitoring animals that are owned or being studied (and therefore these animals are likely to be tagged or identifiable in the wild).

[0033] Figure 1A An exemplary animal is shown, namely, fish 102 in a fishpond environment 100. Fish 102 is likely anadromous fish, meaning they migrate from freshwater to seawater to spawn. In a non-wild environment, fish do not migrate freely out of instinct. Fish breeders, such as fish farmers, need to simulate this migration process by transferring fish 102 to saltwater to ensure the fish 102 develops properly. Figure 1A Fish 102, which are currently living in freshwater, are shown in a stage of their life cycle that may be in preparation for migration to saltwater (which may be seawater or an artificially prepared equivalent saltwater). Figure 1B The second fishpond environment 104 is shown. Figure 1B The fish 106 depicted is larger than the fish 102 in the first fishpond environment 100. Compared to the fish 102 in the first fishpond environment 100, Figure 1BA fish 106, possibly at a later stage of development, is shown. The second fishpond environment 104 may be larger than the first fishpond environment 100 to accommodate larger fish. The second fishpond environment 104 may contain seawater or other brackish water, making the environment suitable for fish 106 that can survive and reproduce in seawater or brackish water. For example, the second fishpond environment 104 may be located in the sea. At the appropriate stage of development of fish 102, when they are anadromous fish or have otherwise evolved to live in saltwater, they can be transferred from the first fishpond environment 100 to the second fishpond environment 104. For example, this might involve placing fish 102 into a marine-based fishpond.

[0034] In this way, the life cycle of wild anadromous spawning fish 102 and 106 can be replicated in an artificial environment. Determining the optimal time to transfer fish 102 is extremely useful for its breeders, helping to maintain the fish's health in a suitable environment appropriate for its life stage. Whether to transfer fish 102 to brackish water can be identified using indicators of brackish water survivability, which measure the likelihood or suitability of fish 102 to survive in brackish water. Fish 102 capable of surviving in brackish water may exhibit salinity osmotic regulation capabilities, meaning they can extract oxygen without dying from electrolyte imbalance.

[0035] One method for determining survival rate is to have a veterinarian or other expert personally assess fish 102. Samples can be collected from the location of fish 102 and evaluated. The samples are then transported to a laboratory. Common methods for assessing survival rate include polymerase chain reaction (PCR) testing of gill samples. This invention eliminates the need for a veterinarian or other expert to travel to the location of the fish (or other example animal), and also eliminates the need for any evaluation in a real laboratory setting. This invention can be used directly by fish farmers or other non-fish physiologists or other non-experts. Using this invention will not result in the death of the fish (or other example animal).

[0036] The Point-of-Sight (PON) instrument 200 can be used to collect fluid samples; in this example, a fluid sample is collected from fish 102 in the first fishpond environment 100. Figure 2 and Figure 3 An example of a point-of-care testing instrument 200 is shown. Figure 2 The functional block diagram of the point-of-care testing instrument 200 is shown. Figure 3 An example view showing the instantaneous detection instrument 200 and the rotor box 202 side by side is presented.

[0037] Point-of-Demand (PON) instruments 200 are commonly used in fields such as food safety, environmental monitoring, and veterinary medicine, where the ability to rapidly detect and identify pathogens or contaminants is crucial. Point-of-Demand (POD) testing refers to testing conducted in the field, such as at or near a fish farm, for example, in a fishpond environment 100. POD samples can be obtained using the PON instrument 200, which is configured to provide rapid and accurate test results at the location where testing is needed, without sending samples to a laboratory for analysis.

[0038] Samples may be blood samples, scale samples, fish skin scrapings, gill biopsies or fin biopsies, etc. Collecting scale samples may include performing mucus smears. Samples may be mucus samples or contain mucus. Samples may be fecal samples or contain feces. Samples may be ascites samples or other body fluid samples that are present or accumulated due to health conditions. In this application and the accompanying drawings, "blood sample" may be interchanged with another sample from fish 102 or other animals, unless the description specifically refers to blood. Blood samples may be whole blood samples or plasma samples, for example, serum samples. Samples may be collected in a lethal or non-lethal manner using needles or other sampling mechanisms, such as collecting saliva or mucus by swab, extracting urine or feces, or homogenizing organs or larval stages.

[0039] Biomarkers were identified in animal body fluid samples. In this example, biomarkers were identified in fish body fluid samples. Biomarkers are biomolecules that can be identified in a sample as indicators of a process or condition (such as disease). Point-of-care testing instrument 200 is configured to allow the isolation and in vitro measurement of certain blood characteristics from blood samples. Point-of-care testing instrument 200 includes analyzer 220 configured for analyzing animal (e.g., fish) body fluid samples, and analyzer 220 may include, for example, a blood testing device. A blood sample testing device is an analytical device used to analyze specific substances (such as proteins, enzymes, electrolytes, or antibodies) in a blood sample. Blood sample testing devices typically use a small amount of blood and employ various techniques (such as photometric detection, hematological, immunoassay, enzyme detection, or molecular diagnostics) for analysis. In the point-of-care testing instrument 200 of this application, analyzer 220 is configured for optical evaluation of animal body fluid samples.

[0040] The point-of-care testing instrument 200 can be configured to receive a rotor cartridge 202. The rotor cartridge 202 may contain multiple reagent chambers 202a. Each reagent chamber 202a may contain a chemical reagent and may contain a diluent. Each reagent chamber 202a can be configured to receive animal body fluids and can be pre-calibrated for a specific chemical test (before receiving any animal body fluids). The reagent material may be lyophilized reagent beads, for example, these beads have undergone appropriate stabilization treatment for point-of-care testing.

[0041] Rotor cartridge 202 may be for single use, i.e., only for one round of analysis. Rotor cartridge 202 may be pre-calibrated for each test to ensure accuracy and consistency. Rotor cartridge 202 may contain a reserved area (e.g., located in the center) for receiving animal body fluid samples, which are dispensed into reagent chamber 202a as rotor cartridge 202 rotates. The animal body fluid sample then reacts with the reagents in each reagent chamber 202a. In the example using fish as samples, only 90-120 μL of fish body fluid sample needs to be added to rotor cartridge 202. Of course, more sample can be added, but it is not necessary to collect a lethal amount of sample from the animal.

[0042] This reaction causes a change in the light absorption properties of the fluid in reagent chamber 202a, which may be due to a change in solution color, or a change in fluid opacity or translucency, or a change in fluid optical properties. Based on the changes in light absorption / optical properties, the concentration of a specific biomarker in the sample can be determined, as described below.

[0043] Example biomarkers include: potassium (K+) biomarkers; amylase (AMY) biomarkers; albumin (ALB) biomarkers; cholesterol (CHOL) biomarkers; alkaline phosphatase (ALT) biomarkers; sodium (NA+) biomarkers; chloride (Cl-) biomarkers; magnesium (MG) biomarkers; carbon dioxide (CO2) biomarkers; calcium (Ca) biomarkers; and phosphorus (P) biomarkers. The biomarkers detected can be determined based on the type of physiological characteristics of the animal being studied. For example, blood analytes such as phosphorus, sodium, calcium, growth hormone, and chloride are suitable candidate indicators for assessing chronic stress.

[0044] The point-of-care testing instrument 200 may include a housing 200a for accommodating the analyzer 220 of the point-of-care testing instrument 200 and for housing the rotor cartridge 202 for analysis. Once the rotor cartridge 202 is loaded with a sample, it can be inserted into an opening 222 in the housing 200a of the point-of-care testing instrument. The opening 222 may be a slot in the housing 200a, which may be covered or uncovered, or it may be a drawer-type slot. Alternatively, the housing 200a may be opened to accommodate the rotor cartridge 202, or it may even be disassembled into multiple parts after the rotor cartridge 202 is inserted for reconstruction around the rotor cartridge 202. The housing 200a may accommodate any component of the point-of-care testing instrument 200 described herein.

[0045] Inserting the rotor cartridge 202 into the opening 222 automatically aligns the compartment 202a with the components in the analyzer 220 used for sample analysis, and automatically aligns the rotor cartridge 202 with the rotor so that the rotor cartridge 202 can rotate.

[0046] The analyzer 220 may include a rotor configured to drive a rotor cartridge to rotate, and may include a motor for driving the rotor, and a power source (e.g., a battery) configured to supply power to the motor.

[0047] The real-time testing instrument 200 may include a user interface 204, which may include physical buttons, a touchscreen, a camera, or a microphone, for receiving user input and enabling the real-time testing instrument 200 to perform functions. The real-time testing instrument 200 may include a display 212 for displaying information to the user; this display may be a touchscreen and therefore can also serve as the user interface 204. The user interface 204 enables the user to power on / off the real-time testing instrument 200 and / or instruct the real-time testing instrument to perform its functions. The real-time testing instrument 200 enables the user to rotate the rotor cartridge 202, and / or the analyzer 220 may include a sensor for sensing that the rotor cartridge 202 has been inserted into the analyzer 220 and automatically rotating the rotor cartridge 202 upon insertion.

[0048] The analyzer 220 can be configured to rotate the rotor at a sufficiently high speed to ensure that the sample is evenly distributed in the reagent chamber 202a, thereby ensuring uniform reaction and accurate measurement.

[0049] The analyzer 220 of the point-of-care testing instrument 200 may include an optical analysis device. This optical analysis device may include a light source, a wavelength selector, and a photodetector. The optical analysis device may be configured to perform absorption spectroscopy, turbidimetry, or absorption spectrophotometry.

[0050] As the rotor cartridge 202 rotates, the light source can be configured to illuminate each reagent chamber 202a at one or more specific wavelengths, and the photodetector can be configured to measure the amount of light not absorbed by the reacted sample—that is, any light that is transmitted, reflected, or otherwise not absorbed. Different biomarkers that may be present in the sample absorb light in different ways (i.e., absorb specific wavelengths), thus giving each biomarker a unique absorption “fingerprint.” Therefore, the transmission spectrum will show whether the biomarker is present.

[0051] The photodetector can be, for example, a photodiode detector or an array of photodiode detectors. The photodetector can also be a photovoltaic cell.

[0052] Examples of wavelength selectors include prisms, diffraction gratings, and filters. In turbidimetry, a wavelength selector can be a filter located between the light source and the rotor box to limit the wavelength of light reaching the sample.

[0053] A wavelength selector can be configured to control a specific wavelength for each test. Since different biomarker molecules absorb light at different wavelengths, using the correct wavelength is crucial for accurate measurements. The point-of-care testing instrument 200 can be configured to detect more than one different wavelength of light (at the photodetector) to enable it to identify different biomarkers in the sample. The point-of-care testing instrument 200 can be configured to detect light at wavelengths such as 340 nm, 405 nm, 450 nm, 505 nm, 546 nm, 600 nm, 630 nm, and / or 850 nm. The light source can be visible light or infrared light to provide these desired wavelengths to be detected. The point-of-care testing instrument 200 may contain more than one number of light sources to provide the desired wavelength range.

[0054] The point-of-care testing instrument 200 may include a processor or processing circuitry 206 and a memory 210. Once the photodetector detects unabsorbed light in the reagent chamber 202a, the point-of-care testing instrument 200 can be configured to compare the absorption of the wavelength at the processor 206 with previous data (possibly test data). Previous data can be retrieved from remote storage (such as cloud storage) or from the memory 210 of the point-of-care testing instrument 200.

[0055] The memory 210 is configured to store test data, which may include sample data previously acquired by the point-of-care testing instrument 200, and / or may include sample data previously acquired by different sampling instruments. This prior sample data may include test data, for example, data that may come from a control group of fish 102 or other animals. This prior sample data can be used to calibrate the point-of-care testing instrument 200, and / or can be used as reference data for comparison with sample data acquired by the point-of-care testing instrument 200. The point-of-care testing instrument 200 may be configured to retrieve test data or other prior sample data from cloud storage or other remote storage. The point-of-care testing instrument 200 may include a transmitter and a receiver or transceiver 208 so that the point-of-care testing instrument 200 can transmit and receive data—for example, transmit data to or receive data from cloud storage.

[0056] The point-of-care testing instrument 200 can be configured to generate biomarker data at a processor or processing circuit 206 based on detected light. The biomarker data generated by the point-of-care testing instrument 200 may include the concentration of a specific biomarker in the sample. The biomarker data generated by the point-of-care testing instrument 200 may include indicators of the concentration of more than one biomarker in the sample. The processor or processing circuit 206 can generate biomarker data for remote analysis.

[0057] In absorption spectroscopy, the transmission spectrum on the detector shows the location of absorption in the sample. Biomarker data may be generated by the transmission spectrum of the processor 206 of the point-of-care testing instrument 200. Absorption at different wavelengths indicates the presence / absence / abundance of different biomarkers. Therefore, the content of each biomarker in the sample can be calculated based on the absorption at different wavelengths. The processor 206 or processing circuitry may be configured to identify substances present in the sample. For example, the processor 206 or processing circuitry may be configured to identify proteins, enzymes, electrolytes, or antibodies in the input sample.

[0058] Memory 210 may contain code 214, which includes an operating system application 216 for the point-of-care testing instrument 200. In the example where the animal is a fish, memory 210 may also contain a fish analysis client application 218. Operating system 216 may be configured to run the basic functions of the point-of-care testing instrument 200. The role of operating system 216 is to execute fish analysis client application 218. Once a sample is received, fish analysis client application 218 enables analyzer 220 to analyze the sample in rotor cartridge 202. Fish analysis client application 218 enables processor 206 to process the analysis results of analyzer 220. Fish analysis client application 218 may be configured to send a request to remote biomarker identification system 400 to perform fish analysis services based on data generated by analyzer 220. Another suitable animal analysis service performed by the animal analysis client application may have the above-described features.

[0059] Multiple samples can be acquired in the field using multiple point-of-care testing instruments 200—for example, a group of users may be in the field and want to analyze fish 102 or other animals in bulk. Multiple rotor cartridges can be used consecutively and inserted into the same point-of-care testing instrument 200 for rapid and efficient processing of multiple samples. The processor 206 can sort the analysis results from the analyzer 220 to distinguish between samples / rotor cartridges 202 (and samples).

[0060] Transmitter or transceiver 208 may be configured to transmit digital signals, and receiver or transceiver 208 may be configured to receive digital signals. Processor or processing circuitry 206 may be configured to process the received digital signals. Transmitter or transceiver 208 may be configured to transmit a signal containing information about the input blood / mucus sample collected from processor or processing circuitry 206.

[0061] Transmitter or transceiver 208 may be configured to send a request to biomarker identification system 400, which may include a remote platform for analyzing biomarker data. Transmitter or transceiver 208 may be configured to send a request to biomarker identification system 400 to obtain animal physiological status / reports, such as obtaining results / reports containing information about the status of juvenile fish.

[0062] The real-time testing instrument 200 may include an input / output interface or I / O 222 for data input and data output. Figure 2 An example input (user interface 204) and an example output (display 212) are shown. I / O 222 may include any suitable mechanism for inputting data to and outputting data from the instantaneous detection instrument 200, such as via a keyboard (possibly a touchscreen or physical buttons), a microphone, or a printer.

[0063] Display 212 can be configured to visually display information to a user. The displayed information may be instructions for use of the point-of-care testing instrument 200, or other useful indicators such as battery life, date, time, etc. In one example, the point-of-care testing instrument 200 is configured to receive results from a biomarker identification system that has transmitted readings via transmitter 208, and display 612 can be configured to display the animal's physiological status or reports to the user. The point-of-care testing instrument display 212 can also be configured to display other results generated by the instrument.

[0064] An example of a point-of-care testing instrument that can be used to acquire samples and identify biomarkers in those samples is the Seamaty SMT-120VP. The Seamaty SMT-120VP can simultaneously measure multiple analytes by analyzing light detected at different wavelengths. The device typically prints out the optical analysis results of the sample. This form of biomarker information may be understandable to veterinarians or other relevant professionals, but is (typically) not available to farmers or pet owners. In this disclosure, biomarker data is transmitted for remote analysis, allowing the user computing device 703a to report farmer-friendly animal physiological indicators and provide additional information and a simple presentation for farmers or other animal owners or other non-professionals.

[0065] Remote server 438 can be configured for biomarker identification—the figure shows an example biomarker identification system 400. Server 438 can be a cloud server (i.e., a virtual server running in a cloud computing environment, rather than a physical server). Figure 4 In this context, server 438 is labeled "Machine Learning Model Server" because it is configured to run classification or regression models that utilize biomarkers in the samples to characterize the animal from which the samples were taken, such as fish 102. The following will combine... Figure 5 A more detailed description of the model is provided.

[0066] The point-of-care testing instrument 200 can be configured to transmit biomarker data to a biomarker identification system 400 located remotely from the instrument 200. Figure 4Example server 438 and example components of the biomarker identification system 400 are shown. "Remote" means that the biomarker identification system 400 and the point-of-care testing instrument 200 are not part of the same physical device (e.g., not within enclosure 200a). The biomarker identification system 400 is not a laboratory in the traditional sense—the point-of-care testing instrument 200 does not output blood samples (or other samples) for delivery to a physical laboratory. The point-of-care testing instrument 200 can be configured to transmit biomarker data as a digital signal to the biomarker identification system 400.

[0067] Biomarker identification system 400 may include receiver 408 for receiving biomarker data. The received data may be forwarded to a laboratory information system (LIS) 418 of the biomarker identification system 400. The LIS 418 may be configured to parse the data from its initial format upon arrival at the biomarker identification system 400 into a format suitable for input into a classification / regression model. The LIS 418 may be configured to run scripts that force the input biomarker data into a functional data frame for input into the model. The functional data frame may be stored in a cloud database in .json or JSON format. The functional data frame may also be stored in the cloud database in BLOB (Binary Large Object) format.

[0068] It is beneficial to send data from the point-of-care testing instrument 200 for analysis and evaluation. While the point-of-care testing instrument 200 may include onboard processing circuitry configurable for identifying biomarkers in a sample, transmitting biomarker data from the point-of-care testing instrument 200 to a remote server 438 offers numerous benefits, including but not limited to: 1) Combining multiple biomarkers to enhance and refine diagnosis 2) Incorporate the results of multiple real-time detection instruments into the diagnostic algorithm. 3) Assisted diagnostic services, easy implementation of new and / or updated algorithms and new features, and the introduction of new biological conditions.

[0069] Advantageously, sending biomarkers from the point-of-care testing instrument 200 in a first data format and then parsing the data into a second (different) data format for evaluation can facilitate the easy implementation of ancillary diagnostic services, new and / or updated algorithms and functions, and the realization of new biological conditions. The point-of-care testing instrument 200 is configured to provide network connectivity and functional parsing of blood sample results directly to the database of a remote server 438 and current biomarker identification services. The device firmware of the point-of-care testing instrument 200 can be configured to support this function, namely, redirecting data to a laboratory information system endpoint compatible with HL7 (Layer 7 of Health Information Exchange) and FHIR (Layer 7 of Health Information Exchange Rapid Healthcare Interoperability Resource) data messages (payloads), as described below.

[0070] Metadata can be included in biomarker data. For example, biomarker data obtained from a specific animal body fluid sample (such as a fish body fluid sample) can be combined with metadata so that the laboratory information system 418 can identify from the metadata the time of sample acquisition, the location of sample acquisition, and the environmental conditions at the time of sample acquisition (such as ambient temperature or salinity). The processor 206 can be configured to include metadata in the biomarker data, for example, indicating which rotor cartridge 202 the data came from. Including metadata helps the laboratory information system 418 process the biomarker data arriving at the biomarker identification system 400.

[0071] When server 438 is a cloud-based server, its components may also be cloud-based. Biomarker recognition system 400 may include infrastructure that supports cloud-based server 438, such as processor 406.

[0072] Receiver or transceiver 408 may be configured to receive requests from point-of-care testing instrument 200 to perform animal analysis services. Processor or processing circuitry 406 may be configured to execute a computer program product containing computer code 420 to enable server 438 to provide animal analysis services, which may include retrieving biomarker data received from point-of-care testing instrument 200, parsing the data at laboratory information system 418, inputting the parsed data into a model, and running the model to generate animal physiological state.

[0073] Figure 4 Operating system application 414 is shown, which can be configured to run the basic functions of server 438. The role of operating system 414 is to execute animal analysis service application 416. Animal analysis service application 416 can be configured to run models and analyze data from animal body fluid samples.

[0074] Animal analysis service application 416 can be configured to be executed by a processor or processing circuitry 406 to perform the animal analysis service described herein.

[0075] Server 438 may include an input / output (I / O) interface 420 for inputting and / or outputting data to and from server 438. For example, server 438 may include a user interface 404 configurable to allow a user to control the operation of server 438. For example, a user can trigger animal analysis through user interface 404. Server 438 may include a display 412. For example, display 412 and user interface 404 may be provided by the same touchscreen.

[0076] The model can be accessed by the Animal Analysis Service application 416 through the model service, rather than being embedded in application 416. An API (Application Programming Interface) can be implemented to access the model.

[0077] The API can be remote, meaning it is neither located at the point-of-care testing instrument 200 (where the request for animal analysis by the biomarker identification system is initiated) nor at the server 438. The API can be a Web API. Requests for animal analysis can be formatted as HTTP request messages; however, a suitable request format is a JSON (JavaScript Object Notation) data frame. The JSON data frame can contain biomarker data already parsed by the laboratory information system 418. The JSON data exchange format is considered suitable for this purpose due to its lightweight, human-readable text characteristics and the ease of handling biomarker data transfer.

[0078] Figure 5 An example decision tree for characterizing fish 102 in the example of Figure 1 is shown. This model may incorporate a machine learning (ML) classification algorithm. The ML classification algorithm can use a decision tree to generate fish state indices based on input biomarker data indicating the levels of biomarkers in a fish's body fluid sample. Fish state indices are indicators of whether the fish from which the sample originated meet specific physiological criteria. Multiple fish state indices can be considered together (e.g., reviewed or summarized as a group) to derive the physiological state of the fish. Of course, if the animal is not fish 102, it can also be used... Figure 5 The decision tree shown is used to represent the animals mentioned above.

[0079] To determine a machine learning (ML) score (the score output by the model) that can serve as a basis for a fish's physiological characteristic state (in this example, although more generally, an "animal" characteristic state), the model can be set to "regression" mode, using a regression tree as appropriate. Alternatively, the model can be set to classification mode to predict binary conditions, which is consistent with... Figure 5 The values ​​shown on the middle leaves are different.

[0080] The model can be configured to automatically fine-tune based on the parsed biomarker data input into it. Decision trees can be augmented to generate a set of decision trees where training on each new instance focuses on previously mismodeled instances.

[0081] Depending on the animal characteristics being assessed (such as juvenile status, health status, stress status, gonadal status, and metabolic status), there may be two or more animal status indicators that can be used to indicate the outcome of the animal characteristic assessment. In the case of two animal status indicators, one may be positive and the other negative: an example of one indicator is "poor health" and the other is "good health." Another example of another indicator is "fertile" and the other is "infertile."

[0082] Various animal states can be identified based on biomarkers found in animal body fluid samples. The model can be configured to generate a score, compare it to a threshold, and assign an animal state index based on that score. To comprehensively understand the physiological characteristics of animals, multiple animal state indices can be developed to reflect different physiological features.

[0083] Animal status indicators may not be limited to binary (e.g., positive or negative)—for example, a model may be configured to generate a predictive animal status indicator. Taking fish as an example, an exemplary predictive fish status indicator is "likely to juvenile within [time period]". When a fish status indicator is predictive, the model may be configured to determine a score (as described above) and further determine whether that score falls within a predetermined range of values. This range of values ​​may be provided during model training and / or input during actual use of the model, such as by a user through a user interface. Based on this range of values, the model may determine a time window in which a score might rise or fall to enter a different range of values, thus predicting when the fish status will change. For example, the model may determine a score of 96 out of 100 (as an example only). The model may be trained to recognize that a score of 96 out of 100 means the sampled fish is currently capable of surviving in seawater. This could be the fish status indicator—"capable of surviving in seawater"—which may be provided to the user. However, the model may (e.g., further) determine that, since the score is between 80 and 100 (as an example only), the fish will juvenile within 2 weeks. Therefore, the model may provide a predictive indicator of fish status, such as "will lose its ability to survive in seawater in 2 weeks".

[0084] In addition to providing the overall physiological characteristics of the animal, the system can also provide animal status indicators, or only animal status indicators without providing the overall physiological characteristics. When a user only wants to analyze the status of a specific animal, the physiological characteristics can include only one animal status indicator, meaning they are the same indicator.

[0085] The model may use one or more decision trees to derive the physiological characteristics of the animal. Figure 5 An example machine learning (ML) algorithm decision tree, labeled Tree 1, is shown. Biomarkers present in the samples (and in the biomarker data) are evaluated to construct the decision tree. The first biomarker is evaluated in step 502. In this example, the first biomarker is potassium. The biomarker data is analyzed, and nodes are split according to information gain, as shown in the figure. These values ​​are for illustrative purposes only. After step 502, the decision tree reaches a leaf node in step 504, a node that is no longer split. Following the branch of step 502, another biomarker (decision node) is evaluated in step 506—in this example, magnesium. This is followed by another leaf node 508 and another branch 510. Another biomarker is evaluated at decision node 510—in this example, total carbon dioxide. After step 510, two leaf nodes are reached—steps 512 and 514. This is just an example illustrating how a single decision tree contributes to the model.

[0086] Figure 5 The units labeled in the example decision tree are normalized (i.e., all variables are scaled according to their respective (maximum - minimum) / standard deviation), rather than using absolute biomarker values ​​obtained from fish body fluid samples. For example, scaling of biomarker data might be performed at laboratory information system 418.

[0087] Figure 5 The example tree shown may be one of the ensembles contained in the model.

[0088] The model is configured to generate animal physiological state based on the results on the leaf, specifically by generating an ML_score and interpreting the meaning of the score (a numerical score) based on the animal's health status.

[0089] In some embodiments, the machine learning (ML) algorithm includes an implementation of the XGBoost machine learning algorithm, a supervised machine learning algorithm that uses a gradient boosting framework to implement an optimized distributed gradient boosting machine learning algorithm. XGBoost stands for Extreme Gradient Boosting, and it is a scalable distributed gradient boosting decision tree (GBDT) machine learning library that uses supervised machine learning algorithms to provide parallel tree boosting to achieve multiple decision trees, ensemble learning, and gradient boosting.

[0090] In some embodiments of the disclosed technology, the XGBoost algorithm is configured using the second gradient of the loss function, in addition to the first gradient, based on the Taylor expansion of the loss function of the test data used as training data. In some embodiments of the XGBoost algorithm, Taylor expansions of other loss functions may be used in certain cases; for example, in binary classification (such as a salinity status binary classification indicating whether fish can survive in saltwater), logistic loss may be used.

[0091] Alternatives to the XGBoost algorithm or similar gradient boosting machine learning (ML) algorithms can be random forests, variance clustering, or any implementation of a machine learning algorithm that uses deep learning.

[0092] Gradient boosting takes a set of labeled (also called targets) training instances as input and builds a model designed to correctly predict the label (target) of each training example based on other known unlabeled information about the samples (i.e., the features of the instances). Once trained, the model should be able to accurately label future data with unknown labels. One or more decision trees can be used to minimize a suitable loss function.

[0093] To prevent overfitting, the XGBoost algorithm transforms the loss function into an objective function that includes regularization terms. These regularization terms add penalty terms to the loss function to prevent the addition of new decision leaf nodes to the model. These penalty terms are proportional to the size of the leaf node weights (see example). Figure 5 Without regularization, gradient boosting models can quickly become bloated and overfit to noise in the training data, causing the algorithm to perform poorly when dealing with new test data, such as biomarker data.

[0094] XGBoost is an example of a machine learning (ML) algorithm that includes an exhaustive list of six or seven hyperparameters, which are automatically selected. Each hyperparameter can be tuned, with values ​​ranging from zero to infinity. These hyperparameters provide specifications for the machine learning model, such as tree depth, number of trees, learning rate, loss reduction, sample size, etc. These hyperparameters affect how the final machine learning model interprets new data during training and when making predictions.

[0095] XGBoost performs parallel processing, making its models train faster than algorithms that do not. The model may automatically trigger retraining when new test data is received. Graphics processing units (GPUs) can be used to compute to solve parallel problems; in this context, GPU algorithms can be used for gradient boosting. The training dataset can be divided into groups or quantiles of equal size. This improves the accuracy of the training algorithm, simplifies the tree-building algorithm, and makes the implementation more efficient.

[0096] This model generates a score (or ML score) based on the values ​​of the leaf nodes of the decision tree. This score may be a numerical value based on the animal's physiological state. For example, the score might be a number between 1 and 100. The model can be configured to compare this score to a threshold and generate the animal's physiological state based on the comparison result. The model can also be configured to output a score "x". This score can be compared to a threshold "y", which is used to determine whether the animal meets physiological standards—see below. Figure 6 Any example evaluation shown and described may have a threshold “y” to determine which of the two evaluation results the sample conforms to.

[0097] In one example, the state of juvenile eels is being evaluated, and a score is used to determine whether the fish from the sample source can survive in seawater. The machine learning (ML) score might be "93", and the threshold might be "50", where a score above 50 indicates that the fish can migrate to seawater and survive. For example, the model might be configured to determine that a score of 93 exceeds 50 and generate a fish physiological state that reflects the fish's ability to migrate and survive.

[0098] The threshold (“y”) can be a probability cutoff value between 0 and 100%, such as 50% as described above. For the specific physiological state being evaluated, the threshold can be a predetermined and fixed value. Alternatively, the threshold can be a continuously drifting threshold that changes with model training and retraining. For example, the threshold can be adjusted based on new biomarker data. By using the model, the threshold can be adjusted to improve the accuracy of the ML_score.

[0099] The threshold can be adjusted based on one or more configured scripts used to determine the optimal threshold for the physiological state being assessed, such as using ROC curves. These scripts can be automated, and the threshold can be changed at any given time, provided biomarker data is available.

[0100] An animal's physiological state may be a composite result based on multiple animal state indicators—that is, a single result is generated by aggregating multiple states to depict the overall health of the animal. An animal's physiological state may be based on the aggregation of values ​​from one or more (e.g., each) leaf nodes of a machine learning (ML) algorithm decision tree.

[0101] This model can be configured to output comments based on machine learning (ML) scores. These scores are not the final physiological state of the animal, but rather an evaluation of an individual animal characteristic that can influence the overall physiological state. The comments can explain the meaning of the machine learning scores to non-experts. For example, as mentioned above, a score of "93" might generate comments like "positive result" or "surviving in seawater" when the threshold is "50," but based on multiple machine learning scores, the overall physiological state of the fish might be output as "healthy fish."

[0102] The generation of animal physiological state may occur outside the model. For example, the model may be configured to output a score to another part of the biomarker recognition system 400—processor 406 and / or laboratory information system 418—to generate animal physiological state based on the score.

[0103] To enable the model to evaluate biomarkers in samples, biomarker data is sent from point-of-care testing instrument 200 to server 438. The biomarker data can be input into processor 206, which records and processes the data at point-of-care testing instrument 200. The data can be appropriately structured for input into the model. Alternatively, the data can be appropriately structured for parsing by a laboratory information system before being input into the model. Examples of data formats include those conforming to the Layer 7 (HL7) Foundation Standard for Health Information Exchange and those conforming to the HL7 Rapid Healthcare Interoperability Resource (FHIR) standard. A suitable data format example is JSON (JavaScript Object Notation) for data transmission.

[0104] The HL7 main standard is an international data standard specifically designed for the transmission of clinical data. This data format is flexible, allowing communication between different systems. In the context of this application, the HL7 main standard includes: Version 2.x Messaging Standard—Interoperability Specification for Health and Medical Transactions; Version 3 Messaging Standard—Interoperability Specification for Health and Medical Transactions; Clinical Document Architecture (CDA) — A clinical document exchange model based on HL7 V3; Continuous Care Documentation (CCD)—A US Medical Summary Exchange Standard Based on CDA; Structured Product Label (SPL) – Published information accompanying pharmaceutical products based on HL7 version 3; Clinical Context Objects Working Group (CCOW) — Interoperability specifications for visual integration of user applications.

[0105] Source: https: / / en.wikipedia.org / wiki / Health_Level_7 Version 2 uses a non-XML encoded syntax based on paragraphs / lines and single-character delimiters. This data format is compatible with laboratory information systems.

[0106] Version 3 includes XML encoding syntax.

[0107] The following is an example of an actual HL7 message transmitted from the point-of-care testing instrument 200 to the laboratory information system API, serving as an example of HL7 formatted data: MSH|^~\&|SMT|SMT- 120VP|||20230513220842||ORU^R01|9|P|2.3.1|A|||0||ASCII|||V1.00.01.09| PID|9|4|||^||A|||Delphinidae||||||||1||||||||||||| OBR|9||2|121004686|||20230513212647|||||1^^||BloodSerum|||71^922542^019 5922542^Bloodgasplusparameters|0^0^0^1||||||||||||||||||||||||||| OBX|1|NM|BUN / CREA|BUN / CREA|150.158|||N|||F||||||| OBX|2|NM|Ca|Ca|2.69|mmol / L||N|||F||||||| OBX|3|NM|GLU|GLU|6.09|mmol / L||N|||F||||||| OBX|4|NM|LAC|LAC|12.12|mmol / L||N|||F||||||| OBX|5|NM|Cl|Cl|140.9|mmol / L||N|||F||||||| OBX|6|NM|BUN|BUN|3.47|mmol / L||N|||F||||||| OBX|7|NM|tCO2|tCO2|7.7|mmol / L||N|||F||||||| OBX|8|NM|Na|Na|152.3|mmol / L||N|||F||||||| OBX|9|NM|Mg|Mg|1.35|mmol / L||N|||F||||||| OBX|10|NM|Crea|Crea|23.1|umol / L||N|||F||||||| OBX|11|NM|PHOS|PHOS||mmol / L||N|>6.00||F||||||| OBX|12|NM|K|K|6.09|mmol / L||N|||F||||||| OBX|13|NM|pH|pH|6.86|||N|||F||||||| This example includes an exemplary biomarker (such as Ca) and indicates that the sample is a serum sample. The concentration of the detected biomarker is expressed in millimoles per liter (mmol / L).

[0108] FHIR format is a type of HL7 format that provides a secure way to transfer data between computer systems, regardless of how that data is stored in those systems.

[0109] An example data format for biomarker data transmitted from the point-of-care testing instrument 200 is FHIR encoded as JSON. This format allows data to be transmitted to the laboratory information system 418. One advantage of using FHIR encoded as JSON is that messages can be sent via HTTP clients and used within traditional application programming interfaces (APIs).

[0110] Figure 6 The system architecture of the Biomarker Identification System 400 was demonstrated. One or more point-of-care testing instruments, 200i, 200ii, and 200iii, were used in the field. Figure 6 The example shows three samples taken from animal 102.

[0111] In the example of fish as the animal, in a fish farm environment, the number of fish 102 that can be represented by a small sample could be as high as 100,000 or even 200,000. Reaching a population level of 100,000 means that multiple point-of-care testing instruments 200 can be used at the location of the fish to collect a sufficient number of samples to represent the population level. In this disclosure, biomarker data is aggregated at the biomarker identification system 400. Other data can also be provided to the biomarker identification system 400 if different assay devices are used on the samples. For example, point-of-care testing instruments can collect data to indicate biomarkers through methods such as PCR testing, rather than observing absorption and collecting biomarker data. Once the data is collected, the biomarker identification system 400 is device-independent—in other words, the system 400 does not care about the instrument used. The laboratory information system 418 can be configured to receive data in different formats and consistently parse the data; as long as the data is consistently parsed for input into the model, the biomarker identification system 400 is device-independent.

[0112] In this way, the point-of-care testing instrument 200 can be developed / improved, while the biomarker identification system 400 will remain a useful tool.

[0113] Therefore, point-of-care testing instruments 200i, 200ii, and 200iii may represent different types of measuring devices, and not just point-of-care testing instrument 200 as described herein.

[0114] Once the biomarker content is identified from samples from instruments 200i, 200ii, and 200iii, the biomarker information can be sent to the biomarker index data storage 402. The biomarker index data storage 402 can be located in network-connected storage or distributed cloud storage. The biomarker index data storage 402 can be configured to provide input data to the server 438 of the biomarker identification system for analysis.

[0115] Server 438 can be configured to perform one or more evaluations on input biomarker information. Figure 6 A series of example assessments and results are presented, which are passed to output analysis 436 to generate a classification and output. This classification can be based on the physiological state of fish across multiple assessments. Each assessment generates a fish state index. The fish physiological state can be the result of one of the assessments (i.e., a fish state index output from the analysis) or it can be a combination of results from multiple assessments.

[0116] Figure 6The images, from left to right, illustrate assessments performed on silvering (or juvenile salmonization) status 604, health status 606, stress status 610, reproductive and sexual health status (or gonadal analysis) 612, and metabolic status 614. Another possible assessment is a general physiological analysis, where general physiological status indicates biological functions such as circulation, respiration, and fluid balance.

[0117] Each evaluation can be performed using decision trees from a set of decision trees that make up the model.

[0118] Animal health status is one example of an indicator of animal condition, reflecting whether an animal is free from disease, infection, viruses, bacteria, and parasites. Animal health status may be based on the immune response to pathogens and the inflammatory response of the mucosa and intestinal barrier, which can be indicated by biomarkers in the sample.

[0119] With advancements in modeling and testing techniques, we have selected biomarkers to identify changes in animal health over time. Currently, biomarkers are ranked according to their usefulness in indicating specific animal health conditions, and some of the top-ranked biomarkers have been evaluated. For example, to identify stress, biomarkers such as phosphorus, sodium, calcium, growth hormone, chloride, cortisol, uric acid, magnesium, glucose, globulin, total protein, and albumin can be evaluated because inventors currently consider these biomarkers to be useful for identifying animal stress.

[0120] Taking fish as an example, the stress state of fish 102 may represent whether fish 102 has the ability to adapt to and survive the processing procedures in aquaculture (including crowding within or between farms, pumping, sorting, grading, and transportation). The stress state of fish 102 may be based on the presence of cortisol, adrenaline, or noradrenaline, as well as any set of relevant biomarkers that have predictive value for animal stress response.

[0121] Gonadal status may represent maturity. Metabolic status may represent growth performance, the ability to convert feed into growth, and the quality of aquatic organisms in terms of economic value—status may include color, texture, and biochemical composition. Taking fish as an example, gonadal status can be identified using one or more biomarkers similar to stress (e.g., [specific biomarkers]) or other different biomarkers. It may be possible to identify gonadal status in fish from some biomarkers and not from others; these possible biomarkers can then be ranked to determine which biomarkers are most useful (or provide the best indication) of gonadal status. This may be applicable to any fish health status being studied.

[0122] This model can be configured to automatically perform all possible evaluations it has been trained on once parsed biomarker data is input into the model. The model can perform all types of evaluations it is capable of performing on the input biomarker data and return an output indicating that no meaningful result can be obtained from the input biomarker data if no meaningful result is found. For example, the model might determine that the biomarker required to determine stress state is missing, therefore a stress state conclusion cannot be drawn, and thus no stress state analysis is performed. Alternatively, it might perform a stress state analysis in the absence of the required biomarker and return a result of "no conclusion."

[0123] Alternatively, a biomarker analysis service request from the point-of-care testing instrument 200 may include a request for one or more specific assessments. For example, a user may not want to receive analysis results related to animal stress, but only be concerned with gonadal status. The request may only include a request for gonadal status analysis, and the model may be configured to generate only a gonadal status score, thereby deriving the animal's physiological status solely based on the gonadal status score. The model may derive the gonadal status score using only a single decision tree.

[0124] Each assessment may yield a binary result, providing one of two possible indicators of animal condition. For example, Figure 6 The silvering status analysis results in the sample 604 show that the seawater is either surviving (616) or non-surviving (618).

[0125] When assessing health status, a health status analysis 606 is performed, and the result may be poor health 620 or good health 622—both are indicators of animal condition. When assessing stress status, a stress status analysis 610 is performed, and the result may be high stress 624 or low stress 626. When assessing sexual and reproductive health, a gonadal status analysis 612 is performed, and the result may be infertility or non-fertility 628 or fertility 630. Some fish spawn, while others produce live fry—for example, depending on the type of fish 102 being analyzed, a gonadal status analysis may conclude that fish 102 is currently pregnant.

[0126] When assessing metabolic status, a metabolic status analysis is performed 614, with results that may be healthy 632 or unhealthy 634. Biomarkers used in the model to determine metabolic status may include any combination of indicators such as total cholesterol, low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, triglycerides, aspartate aminotransferase (AST), alanine aminotransferase (ALT), blood glucose levels, and vitamin levels, as well as any combination of biomarkers with predictive value associated with metabolic dysfunction in animals.

[0127] The model is configured to analyze the evaluation results of the input biomarker information at output analysis 636. Output analysis 636 may include a compilation of results from multiple decision trees. The results of the output analysis include animal physiological status. Animal physiological status may include one or more animal status indicators. For example, animal physiological status may include each animal status indicator determined by animal analysis server 438. Animal physiological status may include a summary of two or more animal status indicators. Animal physiological status may be an overall conclusion based on output analysis 636, such as "survivably transfer from freshwater to saltwater." Animal physiological status may include multiple animal status indicators, such as "healthy and fertile, but unsuitable for transfer to saltwater." This is because samples can be evaluated multiple times, and multiple biomarkers can be identified and classified to draw conclusions at output analysis 636, such as... Figure 6 As shown.

[0128] The information to be presented to the user may be in the form of an animal analysis report, which includes the animal's physiological characteristics and status. The biomarker identification system 400 may include a processor or processing circuit 406 configured to process the results of the output analysis 436 into a report (which may be in a user-friendly format, such as text, numbers, or charts).

[0129] In addition to providing an animal physiological status as a feasible conclusion regarding whether fish 102 can survive in saline (e.g.), the animal physiological status may also include an indicator of salinity osmotic regulation, which, for example, allows a user to infer whether fish 102 can survive in saline. The nature of the animal physiological status may depend on the preferences, skills, or knowledge of the intended end-user; for example, a fish farmer might prefer a "yes" or "no" statement, indicating whether fish 102 can be transferred to saline, while a scientist or veterinarian studying fish 102 might prefer to see indicators such as salinity osmotic regulation or other more detailed indicators.

[0130] The biomarker identification system 400 may include a mobile application, such as a webpage viewable in a web browser, for displaying animal analysis reports. The animal analysis reports are transmitted to a user computing device that provides a client portal user interface or dashboard through which users (such as farmers) can access the animal's physiological status based on data from a point-of-care testing instrument (from which biomarker data is sent). The user computing device 703a can be any suitable device with network access capabilities, and the user interface can be web-based to provide the animal's physiological status. The user computing device 703a can be portable, such as a smartphone or tablet; this is particularly useful for users in the field who possess point-of-care testing instruments 200i, 200ii, 200iii who want to obtain feedback from server 438 as quickly and efficiently as possible, and access the client portal user interface or dashboard through their user computing device 703a when needed. They can take action as needed based on the output analysis 636, such as relocating fish 102.

[0131] Animal analysis reports can be sent to multiple computing devices, provided that the customer portal user interface or dashboard is accessible through that device, allowing users not currently near the point-of-care testing instrument 200 to view the results, or allowing users who wish to study the results outside the animal's location to do so. The results of output analysis 636, or a report based thereon, can be sent as a file (e.g., as an email attachment) to another computing device via the customer portal user interface or dashboard. One advantage of this disclosure is that the animal's physiological status can be received as needed via the user computing device 703a for rapid action based on the results, but sending the results to a laboratory or other location remote from the point of need can also be beneficial, for example, for data processing or research.

[0132] The animal's physiological characteristics can be presented in the form of a report or as part of a report. This report can be a summary of the output analysis results 636, and may include text (e.g., "Animal is not viable"), numerical tables, charts, graphs, etc. The report can be accessed electronically; for example, users can choose different possible formats for the output analysis results 636.

[0133] The report may be actionable, meaning the results of output analysis 636 may be presented to the user in a way that includes interactive elements. To facilitate user presentation, supplementary information about the animal's physiological characteristics may be provided; for example, one or more external links may be provided alongside or as part of the report, enabling end-users to access additional information beyond the report. Information generated as part of output analysis 636 may be transferred to a data store and accessed by the user through links associated with or included in the report.

[0134] The report may include additional information beyond Output Analysis 636, such as educational information on how to address problems identified by the Biomarker Recognition System 400. The report may contain one or more external links to other content, such as web pages or videos, that may contain information explaining Output Analysis 436 or the report itself, such as a video detailing how to reduce animal stress. The report may also include other content, such as images or videos, to supplement Output Analysis 636.

[0135] In one example, the display 212 of the point-of-care testing instrument 200 may be configured to display the animal's physiological state to a user. The point-of-care testing instrument 200 may include a computer application configured to receive and display the animal's physiological state. Alternatively, a computer application on the user's computing device may be configured accordingly, or, as described above, a dashboard / user interface may be provided in a browser.

[0136] The computer application may include a user navigation dashboard to allow users to view animal physiological condition status, multiple animal physiological condition statuses, historical physiological condition statuses, or select a desired format to display animal physiological condition statuses (e.g., a table). Through this dashboard, users can view output analysis 636 (not summarized into animal physiological condition statuses) or reports based on it.

[0137] The real-time monitoring instrument 200 or the user computing device 703a can be configured to store previous output analysis 636 or reports based on that analysis 636, and the user can access the previous output analysis 636 or reports based on that analysis 636 through a computer application. In this way, the user can track the animal's progress from the past to the present.

[0138] The time required from inserting the rotor cartridge into the instantaneous detection instrument 200 to receiving the animal's physiological status or other output analysis reports may be less than 24 hours. Figure 7A This diagram illustrates the turnaround time from collecting animal body fluid samples to receiving the animal's physiological state or other outputs (e.g., data in a different form than the animal's physiological state).

[0139] Compared to other available solutions at the time of this application, the turnaround time of this invention can be significantly reduced. For example, the time required from inserting the rotor cartridge 202 into the point-of-care testing instrument 200 (step 701) to receiving the animal's physiological status or other output analysis report (step 703) may be less than 12 hours, less than 6 hours, or even less than 1 hour. It is estimated that the time required from inserting the rotor cartridge 202 into the point-of-care testing instrument 200 (step 701) to receiving the animal's physiological status or other output analysis report (step 703) may be as low as 15 minutes, 12 minutes, or even less. The turnaround time is limited by the time required for the chemical reaction between the sample and reagents and the optical analysis performed by the analyzer 220. The transmission and processing time of biomarker data, the analysis time of the server 438 (step 702), and the time to provide output (such as animal physiological status) are all extremely short. Currently, the analyzer 220 can perform optical analysis on a sample and generate biomarker data in approximately 12 minutes. This time may be further reduced in the future, while biomarker recognition systems will only slightly increase the time required from inserting a sample to receiving results from the model.

[0140] Step 701 is performed at the location of the animal, and the user computing device 703a is expected to accompany the user to the required location so that they can receive the results when needed (in this case, at the location of the animal).

[0141] Figure 7B This demonstrates an example turnaround process where results are fed back to point-of-care testing instrument 200, and the remaining processes are similar. Figure 7A The same as shown and described.

[0142] Figure 8 A method for training a model is described 800. Figure 8 A training method 800, including sampling 804, is demonstrated. Sampling can be performed using the aforementioned instantaneous testing instrument 200. Samples used to train the model can be obtained in a different manner than those taken when actually evaluating the model. It is not necessary to use the instantaneous testing instrument 200 to generate training data—for example, to generate sufficient test data, whole-animal sampling might be necessary, and the instantaneous testing instrument 200 might not provide a sufficiently large amount of data.

[0143] Method 800 includes input 802 empirical animal state information. Empirical animal state information may be related to various characteristics of the animal, such as empirical juvenile status information (information indicating whether the fish is undergoing or has undergone the juvenile process), empirical health status information (information indicating whether the fish 102 is healthy), empirical stress status information (information indicating whether the fish 102 is under stress), or empirical reproductive status information (information indicating whether the fish 102 is capable of reproduction).

[0144] Empirical animal state information (any example of a state) may come from testing on a group of experimental animals. Public data can be used as input data; however, this model is trained using raw test results.

[0145] Empirical animal state information can be obtained from test data using typical animal analysis methods, such as PCR testing. Models can be trained to generate a score and / or state of animal physiological characteristics that is representative of the sample compared to empirical animal state information.

[0146] Empirical animal status information may include indications of biomarkers present in the test sample, as well as the physiological characteristics of the animals that possess these biomarkers. For example, empirical animal status information may describe which biomarkers are present in animals under prolonged stress, and their relative levels.

[0147] Method 800 also includes collecting samples 804 and acquiring biomarker data 806. Collecting samples 804 and acquiring biomarker data 806 can be accomplished by point-of-care testing instrument 200 or other suitable analytical tools (because at this stage, the model is being trained, and the acquisition of biomarker data is not necessarily aimed at the efficiency of the required points).

[0148] The training data is generated based on the obtained sample biomarkers 806 and the input empirical animal state information 808. Based on the empirical information, the meaning of the biomarker data can be interpreted, and it can be appropriately designated as an animal state indicator. In step 808, the model can be trained to identify the presence of example biomarkers exceeding or falling below example threshold amounts as indicators of example animal state.

[0149] Once the training data is generated, it is fed into the model in step 810. The model is then trained—see step 812. As new biomarker data is provided to the model, it can be continuously trained during use.

[0150] The training protocol undergoes thousands of iterations and selects the best model, for example, by ranking it using ROC-AUC.

[0151] Figure 9A The data flow between the example point-of-care testing instrument 200 and the example machine learning (ML) server 438 for acquiring the physiological state of an animal is illustrated. As shown, the data flow 900 includes the steps of acquiring sample 901 described above. The point-of-care testing instrument 200 receives the sample as input and is configured to analyze the sample 904 to identify biomarkers 906 in the sample.

[0152] like Figure 9A As shown, the point-of-care testing instrument 200 can be configured to indicate certain analytical results 916 (labeled as field results, i.e., not yet sent to the machine learning (ML) server 438, but calculated on the point-of-care testing instrument 200) solely based on the analysis performed by the point-of-care testing instrument 200, without first requesting animal analysis services. These results can be displayed on the display 212, for example. These results could be detection wavelength results from a photodetector, or data processed by the processor 206 after the analyzer 220 analyzes the sample, displaying spectra, etc. Any data not yet entered into the model can be made available to the user of the point-of-care testing instrument 200 via the display 212.

[0153] In step 906, biomarker data from point-of-care testing instrument 200 may be transmitted to machine learning (ML) server 438. In step 908, this biomarker data may be used as input to the model. As described above, the biomarker data sent from point-of-care testing instrument 200 may be processed to adopt a format suitable for input to the model before being input into the model. For example, the biomarker data may be parsed at laboratory information system (LIS) 418.

[0154] In step 910, a model can be run on the input data to determine the physiological characteristics of the animal.

[0155] In step 914, the animal's physiological characteristics (or output data in any suitable format, such as a set of animal status indicators) are transmitted to the user computing device 703a for the user to receive. In the example above, as... Figure 9B As shown, the results are transmitted to the real-time detection instrument 200 itself—this can serve as an alternative to, or supplement to, the user computing device 703a (which can supplement the results sent to the user computing device) if the real-time detection instrument has the necessary computing power. In any case, the user computing device 703a is expected to be used to access the results at a real-time location on-site, thereby realizing the benefit of taking action based on the results at a real-time location on-site. The steps occurring at the real-time location on-site are as follows: Figure 9A As shown in the dashed box. Figure 9B In the example, the fact that the results are received by the instant detection instrument 200 used to transmit biomarker data to the server 438 implies the benefit of obtaining results on demand.

[0156] To date, animal classification models have been described as generating a state of animal physiological characteristics. This state of animal physiological characteristics can also be referred to as a machine learning (ML) prediction, which serves as the output of the model.

[0157] Figure 10An example of a sample's cross-section is shown. A user interface can be provided, containing one or more input boxes for entering biomarker values ​​as feature variables for the model. Users can enter biomarker values, for example, by typing, or the input boxes can be automatically filled based on data received from server 438.

[0158] In this specific example, input biomarkers include potassium, sodium, chloride, total carbon dioxide, calcium, magnesium, and phosphate. These values ​​may change during model training, for example. Biomarkers may be adjusted to include more than seven biomarkers (in this example) when training the model. The selection of biomarkers can be determined based on one or more physiological characteristics explored by the point-of-care testing instrument 200 and the farmer's needs.

[0159] exist Figure 10 In the diagram, the intercept is represented by a dashed line and labeled "xgboost". This intercept represents the model's default prediction when all input variables are set to zero. In the example shown, the XGBoost algorithm is used for prediction.

[0160] The dissected plot shows the exact contribution of each biomarker to the machine learning (ML) prediction. The plot shows the model's prediction as "92.6". Based on this value, or the machine learning score, the machine learning prediction can be determined as "seawater viability". For example, a threshold for seawater viability can be input into the model during training, or the model can be updated over time when scientists determine an appropriate threshold. For instance, the model can be trained based on thresholds found in test data that indicate animal health. Currently, the scores for seawater viability are as follows: Score > 80: Seawater is suitable and abundant. Rating 40-80: Marine aquaculture is feasible, but faces difficulties. Score < 40: Not suitable for marine aquaculture Therefore, as shown in this example, a predicted value of 92.6 indicates that the sample originated from a viable and thriving saltwater fish.

[0161] according to Figure 10 The information provided can be used to calculate and determine the survival time of a specific fish species in seawater (i.e., how long is the window of opportunity before desalination). Once a machine learning (ML) model is trained with sufficient empirical data describing fish biomarkers (such as age, water temperature, salinity, and pH), the model can predict the developmental state and future trajectory of the fish. Combined with ML predictions, this can determine (i) the fish's health status (e.g., desalination or stress) and (ii) the effective duration of that status (e.g., the time before desalination). Figure 10In the example, the window period is approximately 3 weeks.

[0162] The following are examples of the contents of this disclosure: 1. A point-of-care (PON) instrument for acquiring physiological characteristic indicators of fish, the instrument comprising: An insertable rotor cartridge for sample testing includes multiple reagent chambers for receiving fish body fluid samples, each containing a reagent configured to react with the fish body fluid sample. A rotor configured to drive a rotor cartridge to rotate in order to stir the reactants in multiple reagent chambers; An analyzer, comprising an optical analysis device and a processor, is configured to perform optical analysis on the reactants in each reagent chamber after the reagents have reacted with a fish body fluid sample, in order to generate biomarker data. A transmitter configured to communicate with a remote biomarker identification system, wherein the analyzer is configured to generate biomarker data in a format that allows the transmitter to transmit biomarker data to the biomarker identification system for analysis; A receiver configured to receive signals from a biomarker recognition system; and A display configured to show the results generated by a biomarker identification system after analyzing received biomarker data to identify the presence of biomarkers in a sample, the results indicating the physiological state of the fish.

[0163] 2. The point-of-care testing instrument according to claim 1, wherein the format of the biomarker data allows the data to be processed in the laboratory information system of the biomarker identification system.

[0164] 3. The point-of-care testing instrument according to item 1 or 2, wherein the format of the biomarker data conforms to the Layer 7 (HL7) basic standard for health information exchange.

[0165] 4. The point-of-care testing instrument according to item 3, wherein the format of the biomarker data conforms to the HL7 Rapid Healthcare Interoperability Resource (FHIR) standard.

[0166] 5. The point-of-care testing instrument according to item 4, wherein the biomarker data is formatted in accordance with the FHIR standard and encoded using JavaScript Object Notation (JSON).

[0167] 6. A fish physiological characteristic recognition system, comprising: The point-of-care testing instrument described in any of the above entries; and A biomarker identification system that operates remotely from point-of-care testing instruments, comprising: A receiver configured to receive biomarker data from a point-of-care testing instrument; The memory includes a machine learning model, wherein the model is a classification model configured to generate physiological characteristic states of fish based on biomarkers detected in biomarker data; and The transmitter is configured to transmit the physiological characteristics of fish to a real-time detection instrument.

[0168] 7. The fish physiological characteristic identification system according to item 6, wherein the classification model is configured to identify the presence of biomarkers in a sample based on biomarker data and to assign fish state indicators to the biomarker data, wherein the classification model is configured to summarize two or more fish state indicators to generate a comprehensive fish physiological characteristic state.

[0169] 8. The fish physiological characteristic identification system according to claim 6 or 7, wherein the classification model includes the XGBoost machine learning algorithm, which consists of a set of decision trees, wherein each decision tree is configured to receive biomarker data parsed from a first data format to a second data format as input, and wherein the set of decision trees is configured to generate a score based on which the determination of the physiological characteristic state of the fish is based.

[0170] 9. A fish physiological characteristic identification system according to any one of items 6, 7, or 8, wherein the biomarker identification system further includes a laboratory information system configured to parse received biomarker data from a first data format and convert it into a second data format suitable for input to a classification model, and to provide the parsed biomarker data as input to the classification model. The laboratory information system is further configured to generate a report on the physiological characteristic status of the fish, and to transmit the report to a point-of-care testing instrument.

[0171] 10. The fish physiological characteristic identification system according to claim 9, further comprising at least one point-of-care testing instrument as described in any of the preceding claims, wherein the laboratory information system is configured to parse biomarker data from each point-of-care testing instrument into a format suitable for input into a classification model, and to transmit the parsed biomarker data from each point-of-care testing instrument to the classification model.

[0172] 11. The instant detection instrument according to any one of claims 1 to 5 or the fish physiological characteristic identification system according to any one of claims 6 to 10, wherein the fish physiological characteristic status includes the fish's ability to survive in seawater.

[0173] 12. The instant detection instrument according to any one of the preceding items, or the fish physiological characteristic identification system according to any one of the preceding items, wherein the fish body fluid sample includes any one of the following body fluids: blood, whole blood, plasma, serum, mucus, feces, and ascites.

[0174] 13. A fish physiological characteristic identification system according to any one of the preceding claims, wherein the classification model is configured to determine one or more fish state indicators based on biomarkers in a sample, and is configured to generate fish physiological characteristic states based on the one or more fish state indicators, wherein the one or more fish state indicators include at least any one of the following indicators: health status, stress status, gonadal status, metabolic status, and silvering status.

[0175] 14. The point-of-care testing instrument or the fish physiological characteristic identification system according to any one of the foregoing claims, wherein the fish condition indicators are chronic stress indicators, and the biomarkers in the sample include phosphorus, sodium, calcium, growth hormone, and chloride.

[0176] 15. A computer-implemented method for determining the physiological state of fish, the method comprising, at a biomarker recognition system: Receive biomarker data from fish body fluid samples obtained from a point-of-care testing instrument, wherein the biomarker data is in a first data format, and parse the biomarker data into a second format suitable for input into a classification model on a server located away from the point-of-care testing instrument; The parsed biomarker data is input into a classification model on a server located away from the point-of-care testing instrument. The classification model is configured to process the parsed biomarker data to obtain multiple fish status indicators. Multiple fish status indicators are summarized into a single fish physiological characteristic status; and The classification model outputs the physiological characteristics of the fish.

[0177] 16. A non-transitory computer-readable medium containing computer code that, when loaded from memory and executed by one or more processors or processing circuits, enables a biomarker identification system to perform the method described in claim 15.

[0178] When the disclosed technology is described in the form of block diagrams and / or flowcharts using accompanying drawings, it should be understood that multiple entities in the drawings (e.g., various blocks in the block diagram) and combinations of entities in the drawings can be implemented by computer program instructions that can be stored in a computer-readable storage medium and loaded onto a computer or other programmable data processing apparatus. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, and / or other programmable data processing apparatus to generate a machine such that the instructions, which execute via the processor of the computer and / or other programmable data processing apparatus, create means for implementing the functions / actions specified in the block diagrams and / or flowcharts.

[0179] In some implementations, according to certain aspects of this disclosure, the execution order of the functions or steps marked in the boxes may differ from the order described in the operating instructions. For example, depending on the function / action involved, two steps in consecutively displayed boxes may actually be executed substantially simultaneously, or the boxes may sometimes be executed in reverse order. Furthermore, according to certain aspects of this disclosure, the functions or steps marked in the boxes may be executed cyclically and continuously.

[0180] The description of the exemplary embodiments provided herein is for illustrative purposes only. This description is not intended to be exhaustive, nor is it intended to limit the exemplary embodiments to the precise forms disclosed. Modifications and variations can be made based on the foregoing teachings, or by practicing various alternatives to the provided embodiments. The examples discussed herein are intended to explain the principles and nature of various exemplary embodiments and their practical application, enabling those skilled in the art to utilize these exemplary embodiments in various ways and with various modifications for a particular purpose. Features of the embodiments described herein can be combined in all possible combinations of methods, apparatus, modules, systems, and computer program products. It should be understood that the exemplary embodiments provided herein can be practiced in any combination of their respective forms.

[0181] It should be noted that the word "comprising" does not necessarily exclude the presence of other elements, features, functions, or steps besides those listed, and the modifiers "a" or "a plurality of" preceding an element do not exclude the presence of multiple such elements, features, functions, or steps. It should also be noted that any reference numerals in the drawings do not limit the scope of the claims, exemplary embodiments can be implemented at least in part by hardware and software, and multiple "apparatus," "units," or "devices" may be represented by the same hardware item.

[0182] The various exemplary embodiments described herein are described in the general context of methods and may involve elements, functions, steps, or processes, one, more, or all of which may be implemented in one aspect by a computer program product embodied in a computer-readable medium, including computer-executable instructions such as program code that are executed by a computer in a networked environment.

[0183] Exemplary aspects of this disclosure have been disclosed in the accompanying drawings and description. However, many variations and modifications can be made to these aspects, all of which fall within the scope of the appended claims. Therefore, this disclosure should be considered illustrative rather than restrictive in supporting the scope of the claims, and the scope of the claims is not limited to the specific examples of the foregoing aspects and embodiments. This disclosure is exemplified by the various aspects and embodiments described above, and the scope of this disclosure is defined by the following claims.

Claims

1. A point-of-care testing instrument for acquiring physiological characteristic indicators of animals, the point-of-care testing instrument comprising: An insertable rotor cartridge for sample testing includes multiple reagent chambers for receiving animal body fluid samples, each containing reagents configured to react with the animal body fluid samples. A rotor configured to drive a rotor cartridge to rotate in order to stir the reactants in multiple reagent chambers; An analyzer comprising an optical analysis device and a processor, the analyzer being configured to perform optical analysis on the reactants in each reagent chamber after the reagents have reacted with the animal body fluid sample, in order to generate biomarker data; as well as A transmitter configured to communicate with a remote biomarker identification system, wherein the analyzer is configured to generate biomarker data in a specific format that allows the transmitter to transmit the biomarker data to the biomarker identification system for analysis and generation of animal physiological state results.

2. The instantaneous detection instrument according to claim 1, wherein the instrument is used to acquire physiological characteristic indicators of domestic animals, and wherein the animal body fluid sample is a body fluid sample collected from a domestic animal.

3. The instant detection instrument according to claim 2, wherein the domesticated animal is one of the following: artificially bred terrestrial animals, artificially bred aquatic animals or artificially bred semi-aquatic animals, pets or animals used for scientific research, wherein the artificially bred terrestrial animals include artificially bred arboreal animals.

4. The instant detection instrument according to claim 3, wherein the domesticated animal is a fish.

5. The point-of-care testing instrument according to any one of the preceding claims, wherein, The format of the biomarker data is such that it can be processed in the laboratory information system (LIS) of the biomarker identification system.

6. The point-of-care testing instrument according to any one of the preceding claims, wherein the format of the biomarker data conforms to the Layer 7 (HL7) basic standard for health information exchange.

7. The instant detection instrument according to claim 6, wherein, The format of the biomarker data conforms to the Level 7 Rapid Healthcare Interoperability Resource (FHIR) standard for Health Information Exchange.

8. The point-of-care testing instrument of claim 7, wherein the biomarker data is formatted in accordance with the Rapid Healthcare Interoperability Resource Standard and is encoded as JavaScript object representation data.

9. An animal physiological characteristic recognition system, comprising: The instantaneous detection instrument as described in any of the preceding claims; A biomarker identification system remotely deployed with the aforementioned point-of-care testing instrument, the biomarker identification system comprising: A receiver configured to receive biomarker data from the point-of-care testing instrument; A memory, comprising a machine learning model, wherein the model is a classification model or a regression model, the classification model or the regression model being configured to generate animal physiological state characteristics based on biomarkers detected in the biomarker data; and A transmitter configured to transmit the physiological state of the animal to a user computing device; and User computing device, the user computing including: A receiver, configured to receive signals from the biomarker identification system; and A display configured to show the results generated by the biomarker identification system, which analyzes the received biomarker data to identify the presence of biomarkers in a sample and generates the results indicating the physiological state of the animal.

10. The animal physiological characteristic recognition system according to claim 9, wherein, The model is configured to identify the presence of biomarkers in a sample based on the biomarker data and to assign animal state indicators to the biomarker data. The model is configured to aggregate two or more animal state indicators to generate a comprehensive animal physiological state.

11. The animal physiological characteristic recognition system according to claim 9 or claim 10, wherein the model includes an extreme gradient boosting machine learning algorithm, the extreme gradient boosting machine learning algorithm comprising a set of decision trees, wherein each decision tree is configured to receive biomarker data parsed from a first data format to a second data format as input, and wherein the set of decision trees is configured to generate a score, wherein the determination of the animal physiological characteristic state is based on the score.

12. The animal physiological characteristic recognition system according to any one of claims 9, 10, or 11, wherein, The biomarker identification system also includes a laboratory information system (LIS), wherein the LIS is configured to parse the received biomarker data from a first data format and convert it into a second data format suitable for input to the model, and to provide the parsed biomarker data as input to the classification model. The LIS is also configured to generate animal physiological characteristic status reports for transmission to the user computing device.

13. The animal physiological characteristic recognition system of claim 12, further comprising at least one or more point-of-care testing instruments as described in any of the preceding claims, wherein the laboratory information system is configured to parse biomarker data received from each point-of-care testing instrument into a format suitable for input to the model, and to transmit the parsed biomarker data from each point-of-care testing instrument to the model.

14. The instant detection instrument according to any one of claims 1 to 8 or the animal physiological characteristic identification system according to any one of claims 9 to 13, wherein the animal is a fish, and the animal physiological characteristic status includes the fish's ability to survive in saltwater.

15. The point-of-care testing instrument according to any one of the preceding claims, or the animal physiological characteristic identification system according to any one of the preceding claims, wherein the animal body fluid sample includes any one of the following body fluids: blood, whole blood, plasma, serum, mucus, feces, and ascites.

16. The animal physiological characteristic recognition system according to any one of the preceding claims, wherein the model is configured to determine one or more animal state indicators based on biomarkers in the sample, and is configured to generate an animal physiological characteristic state based on the one or more animal state indicators, wherein the one or more animal state indicators include at least one of the following indicators: health status, stress status, gonadal status, metabolic status, and silvering status.

17. The instant detection instrument according to any one of the preceding claims, or the animal physiological characteristic recognition system according to any one of the preceding claims, wherein, One of the animal condition indicators is an indicator of chronic stress, wherein the animal is a fish, and the biomarkers in the sample include phosphorus, sodium, calcium, growth hormone, and chloride.

18. A computer-implemented method for determining the physiological state of an animal, the method being executed at a biomarker identification system, comprising the following steps: Receive biomarker data from an automated body fluid sample obtained from a point-of-care testing instrument, wherein the biomarker data is in a first data format; On a server remotely deployed with the aforementioned point-of-care testing instrument, biomarker data is parsed and converted into a second format suitable for input into a classification or regression model; On the remotely deployed server, the parsed biomarker data is input into the model, which is configured to process the parsed biomarker data to obtain multiple animal state indicators. The multiple animal status indicators are summarized into a single animal physiological characteristic status; as well as The model outputs the physiological characteristics of the animal.

19. A non-transitory computer-readable medium comprising computer code that, when loaded from memory and executed by one or more processors or processing circuitry, enables a biomarker identification system to perform the method of claim 18.

Images