An experimental animal monitoring system and method

Abstract

The application relates to the technical field of animal monitoring, and particularly discloses an experimental animal monitoring system and method, which comprises the following modules: an acquisition module for acquiring basic animal data, monitoring stage data, monitoring stage labels, monitoring target data and animal monitoring data of an experimental animal to be monitored, and determining real-time monitoring data; a monitoring module for determining monitoring abnormal data; an analysis module for acquiring historical abnormal data and determining reason abnormal data of each abnormal reason; an abnormal module for determining a candidate abnormal reason set and an abnormal reason label of each real-time monitoring label of the experimental animal to be monitored; and a determination module for determining real-time animal monitoring data of the experimental animal to be monitored. The candidate reasons and core causes of abnormal animals can be accurately positioned, the integration of abnormal monitoring, reason tracing and real-time data optimization is realized, the monitoring accuracy and systematicness are improved, different monitoring requirements are adapted to, and the real-time animal monitoring data is ensured to be comprehensive and reliable.

Description

Technical Field

[0001] This invention relates to the field of animal monitoring technology, and in particular to a laboratory animal monitoring system and method. Background Technology

[0002] Laboratory animal monitoring is a crucial step in ensuring the scientific validity, reliability, and animal welfare of experimental data. Early monitoring relied primarily on manual observation and piecemeal data collection using simple instruments, which was cumbersome, inefficient, and only allowed for preliminary identification of animal abnormalities, failing to provide phased integration and management of monitoring data. With technological advancements, simplified monitoring systems have emerged; however, traditional systems lack a systematic division of the feeding and experimental stages, cannot link historical abnormal data, and cannot classify and archive the causes of abnormalities. This leads to difficulties in tracing the source of abnormalities, poor data correlation, and insufficient monitoring accuracy and systematicity, failing to meet the practical needs of modern life science experiments for refined monitoring, data traceability, and anomaly traceability. Therefore, there is an urgent need for a system that can integrate multiple types of data and achieve anomaly tracing and systematic monitoring.

[0003] Therefore, this invention proposes a laboratory animal monitoring system and method. Summary of the Invention

[0004] This invention provides a laboratory animal monitoring system and method. It acquires basic animal data, monitoring stage data, monitoring stage tags, monitoring target data, and animal monitoring data of the laboratory animals to be monitored. It determines the real-time monitoring data, abnormal monitoring data, and historical abnormal data of the laboratory animals to be monitored. It also identifies the causal abnormal data for each abnormality, determines the set of candidate abnormal causes and abnormal cause tags for each laboratory animal with an abnormality in the real-time monitoring abnormal data, and determines the real-time animal monitoring data of the laboratory animals to be monitored. This system can accurately locate the candidate causes and core triggers of abnormal animals, achieving integrated abnormality monitoring, cause tracing, and real-time data optimization. It improves monitoring accuracy and systematicity, adapts to different monitoring needs, and ensures comprehensive and reliable real-time animal monitoring data.

[0005] This invention provides a laboratory animal monitoring system, comprising: Acquisition Module: Acquires basic animal data, monitoring stage data, and monitoring stage tags of the experimental animals to be monitored; acquires monitoring target data and animal monitoring data of the experimental animals to be monitored; and determines real-time monitoring data of the experimental animals to be monitored based on the monitoring stage tags of the experimental animals to be monitored. Monitoring module: Based on the real-time monitoring data, monitoring stage tags, and monitoring target data of the experimental animals to be monitored, the module monitors the experimental animals to be monitored and identifies abnormal monitoring data. Analysis module: Based on the monitoring target data, monitoring stage tags, and monitoring abnormal data of the experimental animals to be monitored, historical abnormal data of the experimental animals to be monitored is obtained, the historical abnormal data is analyzed, and the cause of each abnormality is determined. Anomaly Module: Based on the monitoring anomaly data of the experimental animals to be monitored and the cause anomaly data of all anomalies, determine the set of candidate anomaly causes and anomaly cause labels for each experimental animal whose real-time monitoring label is abnormal in the monitoring anomaly data of the experimental animals to be monitored; Determining the module: Based on the monitoring stage label of the experimental animal to be monitored, the abnormal cause label of all experimental animals with abnormal real-time monitoring labels in the abnormal monitoring data, and the cause abnormal data of all abnormal causes, the real-time animal monitoring data of the experimental animal to be monitored is determined.

[0006] Preferably, an experimental animal monitoring system includes an acquisition module comprising: Basic animal data unit: Acquire basic animal data of the experimental animals to be monitored. The basic animal data includes at least the animal name, multiple experimental animals, and the animal number and sex of each experimental animal. Monitoring phase data unit: Acquire monitoring phase data of the experimental animals to be monitored. The monitoring phase data includes feeding phase data and experimental phase data. The feeding phase data includes a set of feeding sub-phases, the feeding objectives of each feeding sub-phase in the set of feeding sub-phases, and the feeding monitoring cycle. The experimental phase data includes a set of experimental sub-phases, the experimental objectives of each experimental sub-phase in the set of experimental sub-phases, and the experimental monitoring cycle. Monitoring phase labeling unit: Based on the monitoring phase data of the experimental animals to be monitored, determine the monitoring phase label of the experimental animals to be monitored; Monitoring target data unit: Based on the basic animal data of the experimental animals to be monitored, the monitoring target data of the experimental animals to be monitored is obtained. The monitoring target data includes the feeding monitoring vector and feeding range matrix of each feeding sub-stage in the feeding sub-stage set, and the experimental monitoring vector and experimental range matrix of each experimental sub-stage in the experimental sub-stage set. Animal monitoring data unit: Based on the monitoring stage tags and monitoring target data of the experimental animals to be monitored, animal monitoring data of the experimental animals to be monitored is obtained. The animal monitoring data includes animal monitoring sub-data of multiple experimental animals. First animal monitoring sub-data unit: If the monitoring stage label belongs to the set of feeding sub-stages, the animal monitoring sub-data includes animal monitoring value vectors for multiple feeding monitoring cycles of multiple feeding sub-stages, and monitoring trend diagrams for each monitoring feature in the animal monitoring value vectors for all feeding monitoring cycles of each feeding sub-stage. The second animal monitoring sub-data unit: If the monitoring stage label belongs to the experimental sub-stage set, the animal monitoring sub-data includes animal monitoring value vectors for multiple feeding monitoring cycles of multiple feeding sub-stages and multiple experimental monitoring cycles of multiple experimental sub-stages, monitoring trend diagrams of monitoring features in the animal monitoring value vectors of all feeding monitoring cycles of each feeding sub-stage, and monitoring trend diagrams of each monitoring feature in the animal monitoring value vectors of all experimental monitoring cycles of each experimental sub-stage.

[0007] Preferably, in an experimental animal monitoring system, the acquisition module further includes: Real-time monitoring vector unit: Based on the monitoring stage label of the experimental animal to be monitored, the corresponding feeding monitoring vector of the feeding sub-stage or the experimental monitoring vector of the experimental sub-stage in the monitoring target data is used to monitor each experimental animal number in the basic animal data of the experimental animal to be monitored in real time, and determine the real-time monitoring vector of each experimental animal in the basic animal data of the experimental animal to be monitored. Real-time monitoring data unit: Based on the real-time monitoring vectors of all experimental animals in the basic animal data of the experimental animals to be monitored, the real-time monitoring data of the experimental animals to be monitored is determined.

[0008] Preferably, a laboratory animal monitoring system includes a monitoring module comprising: Comparison Unit: The real-time monitoring vector of each experimental animal in the real-time monitoring data of the experimental animal to be monitored is compared with the feeding range matrix or experimental range matrix of the corresponding feeding sub-stage in the monitoring target data of the experimental animal to be monitored. If the real-time monitoring value of each monitoring feature in the real-time monitoring vector is within the range of the corresponding monitoring feature in the experimental range matrix, the real-time monitoring label of the experimental animal is determined to be normal. If the real-time monitoring value of any monitoring feature in the real-time monitoring vector is not within the monitoring range of the corresponding monitoring feature in the experimental range matrix, the real-time monitoring label of the experimental animal is determined to be abnormal. Based on all monitoring features that are not within the monitoring range, the abnormal monitoring feature vector of the experimental animal is determined. Based on the real-time monitoring values ​​of all monitoring features that are not within the monitoring range, the abnormal monitoring feature value vector of the experimental animal is determined. Based on the monitoring range of all monitoring features that are not within the monitoring range, the abnormal monitoring range matrix of the experimental animal is determined. The abnormal data monitoring unit determines the abnormal monitoring data of the experimental animals to be monitored based on the abnormal monitoring feature vector, abnormal monitoring feature value vector, and abnormal monitoring range matrix of all experimental animals with abnormal real-time monitoring tags.

[0009] Preferably, an experimental animal monitoring system includes an analysis module comprising: Historical Abnormal Data Unit: Based on the monitoring target data of the experimental animals to be monitored, the monitoring stage label, the feeding target of the feeding sub-stage or the experimental target of the experimental sub-stage corresponding to the monitoring stage label, and all abnormal monitoring feature vectors in the monitoring abnormal data, historical abnormal data of the experimental animals to be monitored is obtained. Among them, the historical abnormal data includes animal abnormal data of multiple abnormal animals in multiple historical monitoring. The animal abnormal data includes the cause of abnormality, the abnormal time interval, the historical stage monitoring data of multiple feeding sub-stages or experimental sub-stages with normal historical monitoring labels within the abnormal time interval, and the historical abnormal feature vectors, historical abnormal feature value vectors, and historical abnormal feature range matrices of multiple feeding monitoring cycles or experimental monitoring cycles with abnormal historical monitoring labels. The historical stage monitoring data includes the historical trend map of each historical feature in the historical monitoring feature value vector of all feeding monitoring cycles or experimental monitoring cycles. Cause-based abnormal data unit: Based on the causes of abnormality of all abnormal animals in the historical abnormal data of the experimental animals to be monitored, the historical abnormal data is divided and the cause-based abnormal data of each abnormal cause is determined. The cause-based abnormal data includes the animal abnormal data of multiple abnormal animals.

[0010] Preferably, an experimental animal monitoring system includes an anomaly module comprising: Anomaly Consistency Value Unit: Based on the anomaly monitoring feature vector, anomaly monitoring feature value vector, and anomaly monitoring range matrix of each experimental animal with an abnormal real-time monitoring tag in the monitoring anomaly data of the experimental animals to be monitored, as well as the cause anomaly data of each anomaly, the anomaly consistency value of each experimental animal with an abnormal real-time monitoring tag and each anomaly cause in the monitoring anomaly data of the experimental animals to be monitored is calculated. Cause set unit: Extract the abnormal causes corresponding to each experimental animal with an abnormal real-time monitoring tag and all non-zero abnormal consistency values ​​in the abnormal monitoring data of the experimental animals to be monitored, and determine the candidate abnormal cause set for each experimental animal with an abnormal real-time monitoring tag in the abnormal monitoring data of the experimental animals to be monitored. Time interval unit: Sort the lengths of the abnormal time intervals in the animal abnormal data of all abnormal animals for each abnormal cause in ascending order, and select the abnormal time interval corresponding to the upper quartile of the sorted data as the cause time interval of the abnormal cause. Interval Trend Unit: Based on the set of candidate abnormal causes for each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored, and all feeding sub-stages and experimental sub-stages with normal historical monitoring labels within the cause time interval of each abnormal cause in the candidate cause set, all monitoring trend charts in the animal monitoring sub-data of the experimental animals in the animal monitoring data are extracted to determine the interval trend data of each abnormal cause in the candidate cause set for each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored; The latency consistency value unit: The interval trend data of each abnormal cause in the candidate cause set of each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored, and the historical stage monitoring data of all feeding sub-stages or experimental sub-stages with normal historical monitoring labels in the abnormal time interval of all abnormal animals in the cause abnormal data are input into the abnormal latency consistency evaluation model to determine the latency consistency value of each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored and each abnormal cause in the candidate abnormal cause set; Abnormal cause label unit: Based on the abnormality value and latent consistency value of each abnormal cause in the monitoring abnormality data of the experimental animals to be monitored and the abnormality label of each experimental animal with an abnormal real-time monitoring label, the abnormal cause label of each experimental animal with an abnormal real-time monitoring label in the candidate abnormal cause set is determined.

[0011] Preferably, a laboratory animal monitoring system includes a determination module comprising: Tag anomaly data unit: Based on the anomaly cause tags of all experimental animals with abnormal real-time monitoring tags in the monitoring anomaly data of the experimental animals to be monitored, the tag anomaly data of each anomaly cause tag of the experimental animals to be monitored is determined. The tag anomaly data includes anomaly monitoring feature vectors, anomaly monitoring feature value vectors and anomaly monitoring range matrix of multiple experimental animals with abnormal real-time monitoring tags. Abnormal evolution and development data unit: Based on the monitoring stage label of the experimental animal to be monitored, the cause abnormal data of the abnormal cause corresponding to each abnormal cause label, the historical abnormal feature vector, historical abnormal feature value vector, and historical abnormal feature range matrix of all feeding monitoring cycles or experimental monitoring cycles with abnormal historical monitoring labels in the animal abnormal data of all abnormal animals, the abnormal evolution and development data of each abnormal cause label of the experimental animal to be monitored are extracted to determine the abnormal evolution and development data of each abnormal cause label of the experimental animal to be monitored. Real-time animal monitoring data unit: Based on the tagged abnormal data and abnormal evolution data of all abnormal cause tags of the experimental animals to be monitored, the real-time animal monitoring data of the experimental animals to be monitored is determined.

[0012] This invention provides a method for monitoring laboratory animals, used to execute any one of the laboratory animal monitoring systems in Examples 1 to 7, comprising: S1: Obtain basic animal data, monitoring stage data, and monitoring stage tags of the experimental animals to be monitored; obtain monitoring target data and animal monitoring data of the experimental animals to be monitored; and determine the real-time monitoring data of the experimental animals to be monitored based on the monitoring stage tags of the experimental animals to be monitored. S2: Based on the real-time monitoring data, monitoring stage tags, and monitoring target data of the experimental animals to be monitored, monitor the experimental animals to be monitored and identify abnormal monitoring data; S3: Based on the monitoring target data, monitoring stage labels, and monitoring abnormal data of the experimental animals to be monitored, obtain historical abnormal data of the experimental animals to be monitored, analyze the historical abnormal data, and determine the cause of each abnormality. S4: Based on the monitoring abnormal data of the experimental animals to be monitored and the cause abnormal data of all abnormal causes, determine the candidate abnormal cause set and abnormal cause label of each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored; S5: Based on the monitoring stage label of the experimental animal to be monitored, the abnormal cause label of all experimental animals with abnormal real-time monitoring labels in the abnormal monitoring data, and the cause abnormal data of all abnormal causes, determine the real-time animal monitoring data of the experimental animal to be monitored.

[0013] The beneficial effects of this invention compared to existing technologies are as follows: It acquires basic animal data, monitoring stage data, monitoring stage tags, monitoring target data, and animal monitoring data of the experimental animals to be monitored; determines real-time monitoring data, abnormal monitoring data, and historical abnormal data of the experimental animals to be monitored; identifies the causal abnormal data for each abnormal cause; determines the set of candidate abnormal causes and abnormal cause tags for each experimental animal with an abnormal real-time monitoring tag in the abnormal monitoring data of the experimental animals to be monitored; and determines the real-time animal monitoring data of the experimental animals to be monitored. This allows for precise location of candidate causes and core triggers for abnormal animals, achieving integrated abnormal monitoring, cause tracing, and real-time data optimization, improving monitoring accuracy and systematicness, adapting to different monitoring needs, and ensuring comprehensive and reliable real-time animal monitoring data.

[0014] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in this application.

[0015] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0016] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of an experimental animal monitoring system according to an embodiment of the present invention; Figure 2 This is a flowchart of an experimental animal monitoring method according to an embodiment of the present invention. Detailed Implementation

[0017] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0018] Example 1: This invention provides a laboratory animal monitoring system, with reference to Figure 1 ,include: Acquisition Module: Acquires basic animal data, monitoring stage data, and monitoring stage tags of the experimental animals to be monitored; acquires monitoring target data and animal monitoring data of the experimental animals to be monitored; and determines real-time monitoring data of the experimental animals to be monitored based on the monitoring stage tags of the experimental animals to be monitored. Monitoring module: Based on the real-time monitoring data, monitoring stage tags, and monitoring target data of the experimental animals to be monitored, the module monitors the experimental animals to be monitored and identifies abnormal monitoring data. Analysis module: Based on the monitoring target data, monitoring stage tags, and monitoring abnormal data of the experimental animals to be monitored, historical abnormal data of the experimental animals to be monitored is obtained, the historical abnormal data is analyzed, and the cause of each abnormality is determined. Anomaly Module: Based on the monitoring anomaly data of the experimental animals to be monitored and the cause anomaly data of all anomalies, determine the set of candidate anomaly causes and anomaly cause labels for each experimental animal whose real-time monitoring label is abnormal in the monitoring anomaly data of the experimental animals to be monitored; Determining the module: Based on the monitoring stage label of the experimental animal to be monitored, the abnormal cause label of all experimental animals with abnormal real-time monitoring labels in the abnormal monitoring data, and the cause abnormal data of all abnormal causes, the real-time animal monitoring data of the experimental animal to be monitored is determined.

[0019] In this embodiment, all basic and real-time monitoring data are collected and filtered for the entire monitoring system. Basic animal data consists of the inherent background information of the experimental animals to be monitored, including animal name, individual number, sex, and strain. This data is obtained through prior manual registration and supplier qualification verification, and is used to distinguish individuals and identify monitoring targets. Monitoring stage data comprises relevant planning data for each stage of experimental animal monitoring, including the goals and monitoring cycles for each sub-stage of feeding and experimentation. This data is obtained according to feeding guidelines and experimental protocols, and is used to define the monitoring stage divisions. Monitoring stage labels are stage identifiers determined based on monitoring stage data and the current growth or experimental progress of the experimental animals, used to mark the specific stage of current monitoring. Animal monitoring data is real-time animal monitoring data collected throughout the experiment, acquired in real-time through various monitoring devices, covering physiological, behavioral, and other relevant indicators. Monitoring target data is the core monitoring standard data preset for the experiment, including monitoring vectors and range matrices for each stage. This data is obtained according to the preset experimental research goals, and is used to define the monitoring focus and normal range.

[0020] In this embodiment, by comparing real-time monitoring data with the normal range of the corresponding stage in the monitoring target data, and combining the stage characteristics corresponding to the monitoring stage label, it is determined whether the monitoring data of each experimental animal deviates from the normal range. If a deviation exists, it is identified as abnormal monitoring data. The abnormal monitoring data includes relevant abnormal information of all experimental animals whose real-time monitoring labels are abnormal.

[0021] In this embodiment, historical anomaly data is retrieved and categorized by cause to form cause-based anomaly data. Monitoring target data clarifies the current monitoring standards and research objectives, used to filter historical anomaly data matching the current monitoring scenario and avoid interference from irrelevant historical data. Monitoring stage tags mark the current specific stage, used to filter historical anomaly data consistent with the current stage type, ensuring the relevance of historical data. Monitoring anomaly data contains information such as the currently occurring anomaly characteristics, used to further accurately match historical data with similar anomaly characteristics. Based on these three types of data, historical anomaly data matching the current scenario and anomaly characteristics from past monitoring is retrieved from the system's historical data storage module. This historical anomaly data covers relevant information on all abnormal animals from multiple historical monitoring events. Analyzing the historical anomaly data involves classifying and organizing the retrieved historical anomaly data, identifying the specific cause of each anomaly, and then grouping all historical anomaly data with the same cause together. The grouped historical anomaly data corresponding to each anomaly cause is the cause-based anomaly data for that anomaly cause.

[0022] In this embodiment, by comparing the monitoring abnormality data of each abnormal experimental animal with the causal abnormality data of all abnormal causes, all abnormal causes similar to the current abnormal manifestation are screened out. These similar abnormal causes are integrated to form a candidate abnormality cause set for the abnormal animal. The candidate abnormality cause set contains all possible causes of the current abnormality. Based on this, by comprehensively comparing the latency consistency value of the current abnormality with the historical abnormality data corresponding to each candidate cause, the core abnormality cause of the abnormal animal is selected from the candidate cause set according to the abnormality consistency value and the latency consistency value, and a corresponding identifier is marked for the core cause. This identifier is the abnormal cause label.

[0023] The beneficial effects of the above technologies are as follows: They acquire basic animal data, monitoring phase data, monitoring phase tags, monitoring target data, and animal monitoring data of the experimental animals to be monitored; determine real-time monitoring data, abnormal monitoring data, and historical abnormal data of the experimental animals to be monitored; identify the causal abnormal data for each abnormal cause; determine the set of candidate abnormal causes and abnormal cause tags for each experimental animal with an abnormal real-time monitoring tag in the abnormal monitoring data of the experimental animals to be monitored; and determine the real-time animal monitoring data of the experimental animals to be monitored. This allows for precise location of candidate causes and core triggers for abnormal animals, achieving integrated abnormal monitoring, cause tracing, and real-time data optimization, improving monitoring accuracy and systematicness, adapting to different monitoring needs, and ensuring comprehensive and reliable real-time animal monitoring data.

[0024] Example 2: Based on Example 1, an experimental animal monitoring system includes an acquisition module comprising: Basic animal data unit: Acquire basic animal data of the experimental animals to be monitored. The basic animal data includes at least the animal name, multiple experimental animals, and the animal number and sex of each experimental animal. Monitoring phase data unit: Acquire monitoring phase data of the experimental animals to be monitored. The monitoring phase data includes feeding phase data and experimental phase data. The feeding phase data includes a set of feeding sub-phases, the feeding objectives of each feeding sub-phase in the set of feeding sub-phases, and the feeding monitoring cycle. The experimental phase data includes a set of experimental sub-phases, the experimental objectives of each experimental sub-phase in the set of experimental sub-phases, and the experimental monitoring cycle. Monitoring phase labeling unit: Based on the monitoring phase data of the experimental animals to be monitored, determine the monitoring phase label of the experimental animals to be monitored; Monitoring target data unit: Based on the basic animal data of the experimental animals to be monitored, the monitoring target data of the experimental animals to be monitored is obtained. The monitoring target data includes the feeding monitoring vector and feeding range matrix of each feeding sub-stage in the feeding sub-stage set, and the experimental monitoring vector and experimental range matrix of each experimental sub-stage in the experimental sub-stage set. Animal monitoring data unit: Based on the monitoring stage tags and monitoring target data of the experimental animals to be monitored, animal monitoring data of the experimental animals to be monitored is obtained. The animal monitoring data includes animal monitoring sub-data of multiple experimental animals. First animal monitoring sub-data unit: If the monitoring stage label belongs to the set of feeding sub-stages, the animal monitoring sub-data includes animal monitoring value vectors for multiple feeding monitoring cycles of multiple feeding sub-stages, and monitoring trend diagrams for each monitoring feature in the animal monitoring value vectors for all feeding monitoring cycles of each feeding sub-stage. The second animal monitoring sub-data unit: If the monitoring stage label belongs to the experimental sub-stage set, the animal monitoring sub-data includes animal monitoring value vectors for multiple feeding monitoring cycles of multiple feeding sub-stages and multiple experimental monitoring cycles of multiple experimental sub-stages, monitoring trend diagrams of monitoring features in the animal monitoring value vectors of all feeding monitoring cycles of each feeding sub-stage, and monitoring trend diagrams of each monitoring feature in the animal monitoring value vectors of all experimental monitoring cycles of each experimental sub-stage.

[0025] In this embodiment, basic identification and biological information of the experimental animals to be monitored are collected to clarify the monitoring targets and achieve individual differentiation for the entire monitoring system. The basic animal data includes the animal name (specific breed name of the experimental animal to be monitored, such as mouse, rat, New Zealand rabbit, etc.), multiple experimental animals (referring to all groups of experimental animals to be monitored), and an animal number (a unique identifier assigned to each animal to distinguish different individuals). Sex is a basic biological attribute of each animal. This basic animal data is obtained primarily before the experimental animals are introduced and raised by verifying the qualification certificates provided by the animal supplier, combined with artificial individual registration, to number and record the sex information of each animal, while simultaneously confirming the animal name.

[0026] In this embodiment, the stages of the entire monitoring process for the experimental animals to be monitored are clearly defined, along with the key monitoring points and monitoring time requirements for each stage. The monitoring stage data are mainly divided into two categories: feeding stage data and experimental stage data. Feeding stage data refers to the relevant monitoring planning data of the experimental animals during the feeding and breeding stage before they enter the formal experiment. The set of feeding sub-stages is a set of multiple sub-stages that divide the entire feeding stage according to the growth pattern or feeding needs of the experimental animals, such as the weaning period, growth period, and adulthood. The feeding objective of each feeding sub-stage is the breeding purpose that needs to be achieved in that sub-stage. For example, during the weaning period, the goal is to ensure that the animals are successfully weaned and adapt to independent feeding; during the growth period, the goal is to ensure that the animals gain weight normally and meet the physical standards. The feeding monitoring cycle is the time interval and total duration of monitoring within each feeding sub-stage, such as monitoring once a week for four weeks. Experimental phase data refers to the relevant monitoring and planning data after the laboratory animals enter the formal experiment. The set of experimental sub-phases is a series of subdivided phases according to the steps of the experimental protocol, such as the modeling phase, drug administration phase, and observation phase. The experimental objective of each sub-phase is the experimental goal to be achieved in that sub-phase. For example, the modeling phase involves constructing a disease model that meets the experimental requirements; the drug administration phase involves administering the drug and observing the initial response. The experimental monitoring cycle is the time interval and total duration of monitoring within each experimental sub-phase, such as monitoring once a day for two weeks. These monitoring phase data are obtained in advance based on the laboratory animal husbandry standards and specific experimental protocols, combined with past husbandry and experimental experience, clearly defining the criteria, objectives, and monitoring cycles for each sub-phase.

[0027] In this embodiment, the specific monitoring sub-stage that the experimental animal to be monitored is currently in is marked, achieving precise positioning of the monitoring stage. The determination of the monitoring stage label is based on the monitoring stage data obtained by the monitoring stage data unit. Specifically, it combines the set of feeding sub-stages and the set of experimental sub-stages in the monitoring stage data to first identify all the subdivided sub-stages included in the entire monitoring process. Then, by judging the current growth status or experimental progress of the experimental animal to be monitored in real time, it is determined which specific sub-stage it is in. For example, it is determined that the experimental animal is currently in the growth period of the feeding sub-stage or in the drug administration stage of the experimental sub-stage. Then, a corresponding identifier is generated as the monitoring stage label.

[0028] In this embodiment, based on basic animal data, the monitoring target data is clearly defined to ensure that the monitoring work does not deviate from the target. The monitoring target data is standard data used to clarify the monitoring focus. The feeding monitoring vector of the feeding sub-stage is the set of features that need to be monitored in the feeding sub-stage, such as weight, food intake, body temperature, etc. The feeding range matrix is ​​the normal value range corresponding to each feeding monitoring feature. The experimental monitoring vector of the experimental sub-stage is the set of features that need to be monitored in the experimental sub-stage, such as blood glucose, tumor volume, inflammation, etc. The experimental range matrix is ​​the normal value range corresponding to each experimental monitoring feature.

[0029] In this embodiment, by combining monitoring stage tags, the specific sub-stage in which the experimental animal is currently located is determined. Whether it is the feeding sub-stage or the experimental sub-stage, relevant data for the corresponding stage are collected specifically based on the monitoring target data, thereby obtaining animal monitoring data and monitoring target data. Among them, the animal monitoring data is the actual monitoring data of all experimental animals collected in the current stage, and the animal monitoring sub-data of the experimental animals refers to the monitoring data corresponding to each experimental animal, ensuring that individual data is independent and traceable.

[0030] In this embodiment, when the monitoring stage label determined by the monitoring stage label unit belongs to the set of feeding sub-stages, that is, when the experimental animal is currently in a specific feeding sub-stage, the content of the animal monitoring sub-data will be specifically adapted to the needs of the feeding stage. The animal monitoring value vector for multiple feeding monitoring cycles across multiple feeding sub-stages refers to the specific set of values ​​for each monitoring feature collected each time, according to the feeding monitoring cycle of that sub-stage. For example, if a feeding sub-stage has a monitoring cycle of once a week for three weeks, then there will be animal monitoring value vectors for three monitoring cycles, each vector containing all feature values ​​from that monitoring. The monitoring trend graph for each monitoring feature in the animal monitoring value vectors of all feeding monitoring cycles for each feeding sub-stage is a graphical representation of the change of that feature with the monitoring cycle, generated by organizing the values ​​of the same monitoring feature from all monitoring cycles within that feeding sub-stage, such as a trend graph of weight gain with each week. This data is acquired by collecting the monitoring feature values ​​of the experimental animal within each feeding monitoring cycle using corresponding monitoring equipment, organizing them into animal monitoring value vectors, and then using data processing tools to convert the multi-cycle values ​​of the same feature into monitoring trend graphs.

[0031] In this embodiment, when the monitoring stage label determined by the monitoring stage label unit belongs to the set of experimental sub-stages, it means that the experimental animal is currently in a specific experimental sub-stage. At this time, the animal monitoring sub-data needs to take into account the monitoring data of both the feeding stage and the experimental stage, because the state of the experimental animal in the experimental stage will be affected by the state of the feeding stage in the previous stage. Specifically, the animal monitoring sub-data includes animal monitoring value vectors for multiple feeding monitoring cycles of multiple feeding sub-stages. This part is all the monitoring data collected in the previous feeding stage. It also includes animal monitoring value vectors for multiple experimental monitoring cycles of multiple experimental sub-stages. This part is the real-time monitoring data collected in each sub-stage and each monitoring cycle of the current experimental stage. In addition, it also includes monitoring trend charts of monitoring features in the animal monitoring value vectors of all feeding monitoring cycles of each feeding sub-stage, used to intuitively show the changing patterns of each feature in the feeding stage, and monitoring trend charts of each monitoring feature in the animal monitoring value vectors of all experimental monitoring cycles of each experimental sub-stage, used to intuitively show the changing patterns of each feature in the experimental stage. The data is obtained by first retrieving the monitoring value vector and trend chart stored in the early feeding stage, and then collecting the monitoring values ​​of each experimental sub-stage and each monitoring cycle in real time through the monitoring equipment in the experimental stage, organizing them into the monitoring value vector of the experimental stage, and then generating the monitoring trend chart of the experimental stage.

[0032] The beneficial effects of the above technologies are: obtaining basic animal data, monitoring stage data, and monitoring stage tags of the experimental animals to be monitored; obtaining monitoring target data and animal monitoring data of the experimental animals to be monitored; adapting to the monitoring needs of different sub-stages; and improving the relevance, completeness, and intuitiveness of the monitoring data.

[0033] Example 3: Based on Example 2, an experimental animal monitoring system, including an acquisition module, further includes: Real-time monitoring vector unit: Based on the monitoring stage label of the experimental animal to be monitored, the corresponding feeding monitoring vector of the feeding sub-stage or the experimental monitoring vector of the experimental sub-stage in the monitoring target data is used to monitor each experimental animal number in the basic animal data of the experimental animal to be monitored in real time, and determine the real-time monitoring vector of each experimental animal in the basic animal data of the experimental animal to be monitored. Real-time monitoring data unit: Based on the real-time monitoring vectors of all experimental animals in the basic animal data of the experimental animals to be monitored, the real-time monitoring data of the experimental animals to be monitored is determined.

[0034] In this embodiment, based on the monitoring stage label of the experimental animal to be monitored, the corresponding sub-stage is found in the monitoring target data, and then the corresponding feeding monitoring vector or experimental monitoring vector is found to clarify the specific characteristics that need to be monitored. Then, based on these clear monitoring characteristics, the experimental animal corresponding to each animal number in the basic animal data is monitored in real time. Real-time values ​​of each animal on these monitoring characteristics are collected through the corresponding monitoring equipment, and these real-time values ​​are integrated in sequence to form a real-time monitoring vector unique to each experimental animal.

[0035] In this embodiment, the real-time monitoring vectors of all experimental animals in the basic animal data are collected, and these vectors are integrated and processed to form a complete dataset, which is the real-time monitoring data of the experimental animals to be monitored.

[0036] The beneficial effects of the above technologies are as follows: Based on the monitoring stage tags of the experimental animals to be monitored, the real-time monitoring data of the experimental animals to be monitored can be determined, which can realize accurate real-time monitoring of each experimental animal and improve the accuracy, pertinence and efficiency of real-time monitoring.

[0037] Example 4: Based on Example 3, an experimental animal monitoring system includes a monitoring module comprising: Comparison Unit: The real-time monitoring vector of each experimental animal in the real-time monitoring data of the experimental animal to be monitored is compared with the feeding range matrix or experimental range matrix of the corresponding feeding sub-stage in the monitoring target data of the experimental animal to be monitored. If the real-time monitoring value of each monitoring feature in the real-time monitoring vector is within the range of the corresponding monitoring feature in the experimental range matrix, the real-time monitoring label of the experimental animal is determined to be normal. If the real-time monitoring value of any monitoring feature in the real-time monitoring vector is not within the monitoring range of the corresponding monitoring feature in the experimental range matrix, the real-time monitoring label of the experimental animal is determined to be abnormal. Based on all monitoring features that are not within the monitoring range, the abnormal monitoring feature vector of the experimental animal is determined. Based on the real-time monitoring values ​​of all monitoring features that are not within the monitoring range, the abnormal monitoring feature value vector of the experimental animal is determined. Based on the monitoring range of all monitoring features that are not within the monitoring range, the abnormal monitoring range matrix of the experimental animal is determined. The abnormal data monitoring unit determines the abnormal monitoring data of the experimental animals to be monitored based on the abnormal monitoring feature vector, abnormal monitoring feature value vector, and abnormal monitoring range matrix of all experimental animals with abnormal real-time monitoring tags.

[0038] In this embodiment, the real-time monitoring vector of each experimental animal is compared feature-by-feature with the range matrix of the corresponding stage. Each real-time monitoring value in the real-time monitoring vector is checked to see if it falls within the normal range of the corresponding feature in the range matrix. If the real-time monitoring values ​​of all monitoring features are within their corresponding normal ranges, the experimental animal is assigned a real-time monitoring label indicating normality. If the real-time monitoring value of any monitoring feature exceeds its corresponding normal range, the experimental animal is assigned a real-time monitoring label indicating abnormality. After determining that an experimental animal is abnormal, the comparison unit filters out all monitoring features that exceed the normal range and combines them to form the abnormal monitoring feature vector for that experimental animal. Then, the real-time monitoring values ​​corresponding to these abnormal monitoring features are organized and combined to form the abnormal monitoring feature value vector for that experimental animal. Simultaneously, the normal monitoring ranges corresponding to these abnormal monitoring features in the range matrix are extracted and combined to form the abnormal monitoring range matrix for that experimental animal.

[0039] In this embodiment, all experimental animals whose real-time monitoring tags are marked as abnormal, as well as the abnormal monitoring feature vector, abnormal monitoring feature value vector and abnormal monitoring range matrix corresponding to each abnormal experimental animal, are summarized to form a complete dataset covering all abnormal individuals. This dataset is the monitoring abnormal data of the experimental animals to be monitored.

[0040] The beneficial effects of the above technologies are as follows: based on the real-time monitoring data, monitoring stage tags, and monitoring target data of the experimental animals to be monitored, the monitoring of the experimental animals to be monitored can identify abnormal data, thereby achieving the refinement and structuring of abnormal information and improving the accuracy and depth of abnormal identification.

[0041] Example 5: Based on Example 4, an experimental animal monitoring system includes an analysis module, comprising: Historical Abnormal Data Unit: Based on the monitoring target data of the experimental animals to be monitored, the monitoring stage label, the feeding target of the feeding sub-stage or the experimental target of the experimental sub-stage corresponding to the monitoring stage label, and all abnormal monitoring feature vectors in the monitoring abnormal data, historical abnormal data of the experimental animals to be monitored is obtained. Among them, the historical abnormal data includes animal abnormal data of multiple abnormal animals in multiple historical monitoring. The animal abnormal data includes the cause of abnormality, the abnormal time interval, the historical stage monitoring data of multiple feeding sub-stages or experimental sub-stages with normal historical monitoring labels within the abnormal time interval, and the historical abnormal feature vectors, historical abnormal feature value vectors, and historical abnormal feature range matrices of multiple feeding monitoring cycles or experimental monitoring cycles with abnormal historical monitoring labels. The historical stage monitoring data includes the historical trend map of each historical feature in the historical monitoring feature value vector of all feeding monitoring cycles or experimental monitoring cycles. Cause-based abnormal data unit: Based on the causes of abnormality of all abnormal animals in the historical abnormal data of the experimental animals to be monitored, the historical abnormal data is divided and the cause-based abnormal data of each abnormal cause is determined. The cause-based abnormal data includes the animal abnormal data of multiple abnormal animals.

[0042] In this embodiment, the core function of the historical anomaly data unit is to retrieve historical anomaly-related data that highly matches the current monitoring scenario and current anomaly characteristics. The monitoring target data clarifies the core objectives and various monitoring standards for the current experimental animal monitoring. The monitoring stage label marks the specific breeding or experimental sub-stage the experimental animal is currently in. The breeding or experimental objectives corresponding to the monitoring stage label clarify the core breeding or experimental purpose of the current stage. All anomaly monitoring feature vectors in the monitoring anomaly data clarify the specific monitoring characteristics of the current experimental animal's anomaly. When acquiring historical anomaly data, the historical anomaly data unit uses these four elements as filtering criteria to retrieve qualified historical monitoring data from the system's historical data storage module. This ensures that the retrieved historical anomaly data matches the current monitoring scenario and current anomaly situation, avoiding irrelevant historical data from interfering with subsequent analysis. The historical anomaly data includes animal anomaly data for multiple experimental animals that have shown anomalies in multiple historical monitoring sessions. Here, "multiple historical monitoring sessions" refers to past experimental animal monitoring work similar to the current monitoring objectives and stage types. "Multiple abnormal animals" refers to all experimental animals marked as abnormal in these historical monitoring sessions. Each abnormal animal's abnormality data contains several core pieces of information. The cause of the abnormality refers to the specific triggers for the animal's abnormality during historical monitoring, such as excessive environmental parameters, drug reactions, or the animal's own health problems. The abnormality time interval refers to the specific time period during which the animal exhibited abnormality, covering the complete cycle from the incubation period before the abnormality occurred to its recovery or the end of the experiment. The historical monitoring data for multiple feeding or experimental sub-stages labeled as normal within the abnormality time interval refers to the monitoring data from those feeding or experimental sub-stages when the animal was in a normal state before and after the abnormality occurred. This data is used to compare and analyze the state changes before and after the abnormality, helping to trace the triggers for the abnormality. The historical stage monitoring data also includes historical trend graphs for each historical feature in the historical monitoring feature value vector for all feeding or experimental monitoring cycles. These trend graphs allow for a visual view of the changing patterns of each monitoring feature during the normal stages. The historical abnormal feature vector, representing multiple feeding or experimental monitoring cycles with abnormal historical monitoring tags, refers to the set of specific monitoring features that indicate an abnormality in each abnormal monitoring cycle within the abnormal time interval. The historical abnormal feature value vector represents the actual monitoring values ​​of these abnormal features within the corresponding cycle, and the historical abnormal feature range matrix represents the normal monitoring range corresponding to these abnormal features. These three data items collectively preserve the specific details of historical abnormalities. This unit acquires historical abnormal data automatically from the historical data storage module, combining monitoring target data, monitoring stage tags, stage targets, and the current abnormal feature vector, filtering, retrieving, and organizing the data without requiring manual searching and filtering.

[0043] In this embodiment, for all historical abnormal data obtained by the historical abnormal data unit, based on the abnormal causes of all abnormal animals in the historical abnormal data, the animal abnormal data of all experimental animals with the same abnormal cause are all collected together. All the related animal abnormal data corresponding to each abnormal cause after collection constitute the cause abnormal data of that abnormal cause.

[0044] The beneficial effects of the above technologies are as follows: Based on the monitoring target data, monitoring stage tags, and monitoring abnormal data of the experimental animals to be monitored, historical abnormal data of the experimental animals to be monitored can be obtained. By analyzing the historical abnormal data, the cause of each abnormality can be determined. This can achieve accurate and structured classification of historical abnormal data, providing targeted and clearly classified historical data support for the judgment of abnormal causes, and improving the efficiency and accuracy of abnormal cause location.

[0045] Example 6: Based on Example 5, an experimental animal monitoring system, including an anomaly module, comprises: Anomaly Consistency Value Unit: Based on the anomaly monitoring feature vector, anomaly monitoring feature value vector, and anomaly monitoring range matrix of each experimental animal with an abnormal real-time monitoring tag in the monitoring anomaly data of the experimental animals to be monitored, as well as the cause anomaly data of each anomaly, the anomaly consistency value of each experimental animal with an abnormal real-time monitoring tag and each anomaly cause in the monitoring anomaly data of the experimental animals to be monitored is calculated. Cause set unit: Extract the abnormal causes corresponding to each experimental animal with an abnormal real-time monitoring tag and all non-zero abnormal consistency values ​​in the abnormal monitoring data of the experimental animals to be monitored, and determine the candidate abnormal cause set for each experimental animal with an abnormal real-time monitoring tag in the abnormal monitoring data of the experimental animals to be monitored. Time interval unit: Sort the lengths of the abnormal time intervals in the animal abnormal data of all abnormal animals for each abnormal cause in ascending order, and select the abnormal time interval corresponding to the upper quartile of the sorted data as the cause time interval of the abnormal cause. Interval Trend Unit: Based on the set of candidate abnormal causes for each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored, and all feeding sub-stages and experimental sub-stages with normal historical monitoring labels within the cause time interval of each abnormal cause in the candidate cause set, all monitoring trend charts in the animal monitoring sub-data of the experimental animals in the animal monitoring data are extracted to determine the interval trend data of each abnormal cause in the candidate cause set for each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored; The latency consistency value unit: The interval trend data of each abnormal cause in the candidate cause set of each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored, and the historical stage monitoring data of all feeding sub-stages or experimental sub-stages with normal historical monitoring labels in the abnormal time interval of all abnormal animals in the cause abnormal data are input into the abnormal latency consistency evaluation model to determine the latency consistency value of each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored and each abnormal cause in the candidate abnormal cause set; Abnormal cause label unit: Based on the abnormality value and latent consistency value of each abnormal cause in the monitoring abnormality data of the experimental animals to be monitored and the abnormality label of each experimental animal with an abnormal real-time monitoring label, the abnormal cause label of each experimental animal with an abnormal real-time monitoring label in the candidate abnormal cause set is determined.

[0046] In this embodiment, based on the abnormal monitoring feature vector, abnormal monitoring feature value vector, and abnormal monitoring range matrix of each experimental animal with an abnormal real-time monitoring tag in the monitoring abnormal data of the experimental animals to be monitored, as well as the cause abnormal data of each abnormal cause, the abnormal consistency value of each experimental animal with an abnormal real-time monitoring tag and each abnormal cause in the monitoring abnormal data of the experimental animals to be monitored is calculated. The calculation formula is expressed as follows: in, This represents the anomalous consistency value between the j-th experimental animal and the m-th cause of abnormality in the monitoring abnormality data of the experimental animals under monitoring. This represents the abnormal monitoring feature vector of the j-th experimental animal in the abnormal monitoring data of the experimental animals to be monitored. This represents the k-th monitoring feature in the abnormal monitoring feature vector of the j-th experimental animal in the monitoring abnormality data. This represents the k-th real-time monitoring value in the abnormal monitoring feature value vector of the j-th experimental animal in the monitoring abnormal data of the experimental animal to be monitored. This represents the lower limit of the real-time range and the upper limit of the real-time range in the first row and k column of the abnormal monitoring range matrix for the j-th experimental animal in the monitoring abnormal data. This represents the number of monitoring features in the abnormal monitoring feature vector of the j-th experimental animal in the monitoring abnormality data of the experimental animals to be monitored. This represents the first out-of-range direction value of the k-th real-time monitoring value in the abnormal monitoring feature value vector of the j-th experimental animal in the monitoring abnormal data. This represents the out-of-range value of the k-th real-time monitoring value in the abnormal monitoring feature value vector of the j-th experimental animal in the monitoring abnormal data. This represents the earliest historical monitoring tag within the abnormal time interval of the p-th abnormal animal in the abnormal data of the m-th abnormal cause, which is the historical abnormal feature vector of the abnormal feeding monitoring cycle or experimental monitoring cycle. The nth historical feature of the historical abnormal feature vector represents the earliest historical monitoring tag within the abnormal time interval of the p-th abnormal animal in the abnormal data of the m-th abnormal cause. This tag is the tag of the abnormal feeding monitoring cycle or experimental monitoring cycle. This represents the nth historical monitoring value in the vector of historical abnormal feature values ​​for the p-th abnormal animal within the abnormal time interval of the m-th abnormal cause data. mpN2 represents the lower limit of the historical range in the first row and nth column of the abnormal monitoring range matrix for the p-th abnormal animal within the abnormal time interval of the abnormal time interval, with the earliest historical monitoring tag being the abnormal feeding monitoring cycle or experimental monitoring cycle. mpN3 represents the number of historical features in the historical abnormal feature vector for the p-th abnormal animal within the abnormal time interval of the m-th abnormal cause's abnormal data. The second out-of-range direction value is the nth historical monitoring value in the historical abnormal feature value vector of the pth abnormal animal within the abnormal time interval of the mth abnormal cause data. This value is the earliest historical monitoring tag within the abnormal feeding monitoring cycle or experimental monitoring cycle. This represents the out-of-range value of the nth historical monitoring value in the historical abnormal feature value vector within the abnormal time interval of the pth abnormal animal in the abnormal data of the mth abnormal cause. The earliest historical monitoring tag within the abnormal feeding monitoring cycle or experimental monitoring cycle is the out-of-range value. This represents the k-th real-time monitoring value in the abnormal monitoring feature value vector of the j-th experimental animal in the abnormal monitoring data of the experimental animals to be monitored. Based on the abnormal data of the p-th abnormal animal in the cause of the m-th abnormality, the first indicator function is the n-th historical monitoring value in the historical abnormal feature value vector within the abnormal time interval, with the earliest historical monitoring tag being the abnormal feeding monitoring cycle or experimental monitoring cycle. Let represent all monitoring features of the abnormal monitoring feature vector of the j-th experimental animal in the abnormal monitoring data of the experimental animals to be monitored, and let represent all historical features of the historical abnormal feature vector of the p-th abnormal animal in the abnormal animal abnormal data of the m-th abnormal cause, whose earliest historical monitoring label is the abnormal feeding monitoring cycle or experimental monitoring cycle within the abnormal time interval. , This represents the empty set.

[0047] In this embodiment, , .

[0048] In this embodiment, the others in the first indicator function are... In this embodiment, the cause set unit processes the abnormality consistency values ​​corresponding to each abnormal experimental animal and all abnormal causes. An abnormality consistency value of zero indicates that the abnormal details of the current abnormal animal do not match the historical abnormal details of the cause at all, meaning the cause cannot be the trigger for the current abnormality and can be directly excluded. All non-zero abnormality consistency values ​​for each abnormal animal are extracted, and the corresponding abnormal causes are found. These abnormal causes are then integrated to form the candidate abnormality cause set for that abnormal animal. Each abnormal animal has its own exclusive candidate abnormality cause set, and each abnormal cause in the set has a certain degree of similarity to the current abnormality.

[0049] In this embodiment, the time interval unit determines a representative typical abnormal time interval for each abnormal cause. The cause-related abnormal data for each abnormal cause includes animal abnormality data from multiple historical animals that exhibited abnormalities due to that cause. Each historical abnormal animal's animal abnormality data contains a corresponding abnormal time interval, which represents the complete time period of the animal's abnormality. The length of the time interval for different animals exhibiting abnormalities due to that cause may vary; some durations are long, and some are short. This unit first collects the abnormal time intervals of all historical abnormal animals corresponding to each abnormal cause, extracts the length of each time interval, and then sorts these time interval lengths in ascending order. After sorting, the abnormal time interval corresponding to the upper quartile position in the sorted results is selected and determined as the cause-related time interval for that abnormal cause. Selecting the time interval corresponding to the upper quartile is to avoid interference from extreme values; neither extremely long-duration cases nor extremely short-duration special cases are selected, ensuring that the determined cause-related time interval is representative and can reflect the typical duration of the abnormality caused by that abnormal cause.

[0050] In this embodiment, the interval trend unit extracts monitoring trend maps within a specific time range for each abnormal animal corresponding to a candidate abnormal cause, forming interval trend data. The processing basis of this unit includes the set of candidate abnormal causes for each abnormal animal, and all feeding and experimental sub-stages with historical monitoring labels of "normal" within the cause time interval corresponding to each abnormal cause in the candidate set. The cause time interval is the typical abnormal duration of the abnormal cause. Sub-stages with historical monitoring labels of "normal" within it refer to feeding or experimental sub-stages in which animals that historically exhibited abnormalities due to this cause were in a normal state before the abnormality occurred. The trend changes of these normal stages contain latent characteristics before the abnormality occurred. Based on the range of these normal sub-stages, the unit extracts all corresponding monitoring trend maps from the animal monitoring sub-data of the abnormal animal in the animal monitoring data. These extracted monitoring trend maps constitute the interval trend data for the abnormal animal corresponding to this candidate abnormal cause. The monitoring trend maps can intuitively reflect the changing patterns of monitoring characteristics over time, and the interval trend data is a collection of these trend maps, covering the characteristic changes of the normal stages before the abnormality occurred.

[0051] In this embodiment, the latency consistency value unit calculates the latency similarity between each abnormal animal and each cause in the candidate abnormality cause set, i.e., the latency consistency value, to achieve a quantitative assessment of the latent characteristics of abnormalities. The input data for this unit includes two parts: one part is the interval trend data for each abnormal animal corresponding to each candidate abnormality cause, reflecting the monitoring characteristic change trend of the current abnormal animal in the normal phase before the abnormality occurred; the other part is the historical stage monitoring data of all historical abnormal animals within the abnormal time interval in the animal abnormality data of each candidate abnormality cause, specifically the historical stage monitoring data of all feeding or experimental sub-stages where the historical monitoring label was normal during the abnormal time interval. This part reflects the monitoring characteristic change trend and related data of animals that historically became abnormal due to this cause in the normal phase before the abnormality occurred. This unit inputs both parts of data into the abnormality latency consistency assessment model, and the model compares and analyzes the two sets of trend data and related monitoring data to quantify their similarity.

[0052] In this embodiment, the anomaly latency consistency assessment model quantifies the similarity between the trend data of abnormal individuals in the monitored experimental animals during the anomaly latency stage and the trend data of all abnormal animals under the corresponding candidate anomaly causes during the same latency stage. It then outputs a latency consistency value that measures the degree of correlation between the two. The model uses Dynamic Time Warping (DTW) as its core algorithm. First, it preprocesses the input interval trend data of the current abnormal experimental animals and the historical monitoring data of historical abnormal animals, converting both types of data into time series corresponding to each monitoring feature. Then, it standardizes the series to eliminate differences in the dimensions and numerical amplitudes of different monitoring features, ensuring the fairness of subsequent comparisons. Next, it constructs a distance matrix between the two sets of time series using the DTW algorithm, calculates the single-point distance between every two data points in the series, and then uses dynamic programming to find... The optimal regularization path satisfies boundary constraints, monotonicity constraints, and continuity constraints. This path allows for flexible scaling of the time axis, adapting to situations where the trend changes of two sets of sequences have different lengths and misaligned rhythms, thus avoiding the limitations of traditional point-by-point comparisons. Finally, the minimum cumulative distance between the two sets of sequences is calculated. The smaller the minimum cumulative distance, the higher the overall similarity of the two trends. Then, the minimum cumulative distance is normalized and mapped to a value in the range of zero to one. Combined with the preset weights of different monitoring features, the similarity results of each feature are weighted and fused to finally obtain a quantified latency consistency value. The closer this value is to one, the higher the similarity between the latency trend of the current abnormal experimental animal and the latency trend of the historical abnormal animal corresponding to the candidate abnormal cause, and the higher the probability that the abnormal cause has caused the current abnormality in the experimental animal.

[0053] In this embodiment, by comprehensively analyzing the magnitudes of the abnormal consistency value and the latent consistency value, the candidate abnormal cause with the highest overall consistency value is selected first and identified as the core abnormal cause of the abnormal animal. This core abnormal cause is then labeled with a corresponding identifier, which is the abnormal cause tag. If multiple candidate causes have similarities, the most likely abnormal cause is determined by combining the overall consistency value of the two candidates and tagged accordingly. The overall consistency value can be calculated using a weighted average.

[0054] The beneficial effects of the above technologies are as follows: Based on the monitoring abnormal data of the experimental animals to be monitored and the cause abnormal data of all abnormal causes, the set of candidate abnormal causes and abnormal cause labels of each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored can be determined. This can achieve accurate matching and scientific labeling of abnormal causes, and improve the comprehensiveness, accuracy and efficiency of abnormal cause determination.

[0055] Example 7: Based on Example 6, an experimental animal monitoring system, including a determination module, comprises: Tag anomaly data unit: Based on the anomaly cause tags of all experimental animals with abnormal real-time monitoring tags in the monitoring anomaly data of the experimental animals to be monitored, the tag anomaly data of each anomaly cause tag of the experimental animals to be monitored is determined. The tag anomaly data includes anomaly monitoring feature vectors, anomaly monitoring feature value vectors and anomaly monitoring range matrix of multiple experimental animals with abnormal real-time monitoring tags. Abnormal evolution and development data unit: Based on the monitoring stage label of the experimental animal to be monitored, the cause abnormal data of the abnormal cause corresponding to each abnormal cause label, the historical abnormal feature vector, historical abnormal feature value vector, and historical abnormal feature range matrix of all feeding monitoring cycles or experimental monitoring cycles with abnormal historical monitoring labels in the animal abnormal data of all abnormal animals, the abnormal evolution and development data of each abnormal cause label of the experimental animal to be monitored are extracted to determine the abnormal evolution and development data of each abnormal cause label of the experimental animal to be monitored. Real-time animal monitoring data unit: Based on the tagged abnormal data and abnormal evolution data of all abnormal cause tags of the experimental animals to be monitored, the real-time animal monitoring data of the experimental animals to be monitored is determined.

[0056] In this embodiment, the abnormal cause tags of all abnormal experimental animals are first identified, and different tag types are distinguished. Then, all abnormal experimental animals belonging to the same abnormal cause tag are screened out, and the abnormal monitoring feature vectors, abnormal monitoring feature value vectors, and abnormal monitoring range matrices of these animals are summarized. The summarized overall data is the tag abnormal data for that abnormal cause tag. Each abnormal cause tag corresponds to independent tag abnormal data.

[0057] In this embodiment, the abnormal evolution and development data unit extracts historical abnormal evolution-related data corresponding to each abnormal cause label, and sorts out the historical development patterns of similar abnormal causes. For each abnormal cause label, its corresponding abnormal cause is first matched, and then, from the abnormal data of the cause of the abnormal cause, relevant data for all feeding monitoring cycles or experimental monitoring cycles of animals with historical abnormal labels are extracted from the animal abnormal data of all historical abnormal animals. These data include historical abnormal feature vectors, historical abnormal feature value vectors, and historical abnormal feature range matrices. During extraction, the monitoring stage label is combined to filter historical abnormal cycle data that matches the current stage type, ensuring that historical data matches the current scenario. The extracted data completely records the characteristic changes, numerical fluctuations, and normal reference ranges of similar abnormal causes in each historical abnormal cycle. The integrated data is the abnormal evolution and development data of that abnormal cause label.

[0058] In this embodiment, the label anomaly data and anomaly evolution development data corresponding to all anomaly cause labels are correlated and integrated to form a complete dataset that includes both real-time anomaly details and historical evolution references. The resulting dataset is the real-time animal monitoring data of the experimental animals to be monitored.

[0059] The beneficial effects of the above technologies are as follows: Based on the monitoring stage tags of the experimental animals to be monitored, the abnormal cause tags of all experimental animals with abnormal real-time monitoring tags in the abnormal monitoring data, and the cause abnormal data of all abnormal causes, the real-time animal monitoring data of the experimental animals to be monitored can be determined. This can realize the structuring, traceability, and predictability of abnormal data, improve the depth and practical value of monitoring data, and provide comprehensive support for abnormal handling and trend analysis.

[0060] Example 8: This invention provides a method for monitoring laboratory animals, used to implement any one of the laboratory animal monitoring systems in Examples 1 to 7, with reference to... Figure 2 ,include: S1: Obtain basic animal data, monitoring stage data, and monitoring stage tags of the experimental animals to be monitored; obtain monitoring target data and animal monitoring data of the experimental animals to be monitored; and determine the real-time monitoring data of the experimental animals to be monitored based on the monitoring stage tags of the experimental animals to be monitored. S2: Based on the real-time monitoring data, monitoring stage tags, and monitoring target data of the experimental animals to be monitored, monitor the experimental animals to be monitored and identify abnormal monitoring data; S3: Based on the monitoring target data, monitoring stage labels, and monitoring abnormal data of the experimental animals to be monitored, obtain historical abnormal data of the experimental animals to be monitored, analyze the historical abnormal data, and determine the cause of each abnormality. S4: Based on the monitoring abnormal data of the experimental animals to be monitored and the cause abnormal data of all abnormal causes, determine the candidate abnormal cause set and abnormal cause label of each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored; S5: Based on the monitoring stage label of the experimental animal to be monitored, the abnormal cause label of all experimental animals with abnormal real-time monitoring labels in the abnormal monitoring data, and the cause abnormal data of all abnormal causes, determine the real-time animal monitoring data of the experimental animal to be monitored.

[0061] The beneficial effects of the above technologies are as follows: They acquire basic animal data, monitoring phase data, monitoring phase tags, monitoring target data, and animal monitoring data of the experimental animals to be monitored; determine real-time monitoring data, abnormal monitoring data, and historical abnormal data of the experimental animals to be monitored; identify the causal abnormal data for each abnormal cause; determine the set of candidate abnormal causes and abnormal cause tags for each experimental animal with an abnormal real-time monitoring tag in the abnormal monitoring data of the experimental animals to be monitored; and determine the real-time animal monitoring data of the experimental animals to be monitored. This allows for precise location of candidate causes and core triggers for abnormal animals, achieving integrated abnormal monitoring, cause tracing, and real-time data optimization, improving monitoring accuracy and systematicness, adapting to different monitoring needs, and ensuring comprehensive and reliable real-time animal monitoring data.

[0062] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A laboratory animal monitoring system, characterized in that, include: Acquisition Module: Acquires basic animal data, monitoring stage data, and monitoring stage tags of the experimental animals to be monitored; acquires monitoring target data and animal monitoring data of the experimental animals to be monitored; and determines real-time monitoring data of the experimental animals to be monitored based on the monitoring stage tags of the experimental animals to be monitored. Monitoring module: Based on the real-time monitoring data, monitoring stage tags, and monitoring target data of the experimental animals to be monitored, the module monitors the experimental animals to be monitored and identifies abnormal monitoring data. Analysis module: Based on the monitoring target data, monitoring stage tags, and monitoring abnormal data of the experimental animals to be monitored, historical abnormal data of the experimental animals to be monitored is obtained, the historical abnormal data is analyzed, and the cause of each abnormality is determined. Anomaly Module: Based on the monitoring anomaly data of the experimental animals to be monitored and the cause anomaly data of all anomalies, determine the set of candidate anomaly causes and anomaly cause labels for each experimental animal whose real-time monitoring label is abnormal in the monitoring anomaly data of the experimental animals to be monitored; Determining the module: Based on the monitoring stage label of the experimental animal to be monitored, the abnormal cause label of all experimental animals with abnormal real-time monitoring labels in the abnormal monitoring data, and the cause abnormal data of all abnormal causes, the real-time animal monitoring data of the experimental animal to be monitored is determined.

2. The experimental animal monitoring system according to claim 1, characterized in that, The acquisition module includes: Basic Animal Data Unit: Acquires basic animal data of the experimental animals to be monitored; Monitoring phase data unit: Acquire monitoring phase data of the experimental animals to be monitored, including feeding phase data and experimental phase data; Monitoring phase labeling unit: Based on the monitoring phase data of the experimental animals to be monitored, determine the monitoring phase label of the experimental animals to be monitored; Monitoring target data unit: Based on the basic animal data of the experimental animals to be monitored, the monitoring target data of the experimental animals to be monitored is obtained. The monitoring target data includes the feeding monitoring vector and feeding range matrix of each feeding sub-stage in the feeding sub-stage set, and the experimental monitoring vector and experimental range matrix of each experimental sub-stage in the experimental sub-stage set. Animal monitoring data unit: Based on the monitoring stage tags and monitoring target data of the experimental animals to be monitored, animal monitoring data of the experimental animals to be monitored is obtained. The animal monitoring data includes animal monitoring sub-data of multiple experimental animals.

3. The experimental animal monitoring system according to claim 2, characterized in that, The acquisition module also includes: First animal monitoring sub-data unit: If the monitoring stage label belongs to the set of feeding sub-stages, the animal monitoring sub-data includes animal monitoring value vectors for multiple feeding monitoring cycles of multiple feeding sub-stages, and monitoring trend diagrams for each monitoring feature in the animal monitoring value vectors for all feeding monitoring cycles of each feeding sub-stage. The second animal monitoring sub-data unit: If the monitoring stage label belongs to the experimental sub-stage set, the animal monitoring sub-data includes animal monitoring value vectors for multiple feeding monitoring cycles of multiple feeding sub-stages and multiple experimental monitoring cycles of multiple experimental sub-stages, monitoring trend diagrams of monitoring features in the animal monitoring value vectors of all feeding monitoring cycles of each feeding sub-stage, and monitoring trend diagrams of each monitoring feature in the animal monitoring value vectors of all experimental monitoring cycles of each experimental sub-stage. Real-time monitoring vector unit: Based on the monitoring stage label of the experimental animal to be monitored, the corresponding feeding monitoring vector of the feeding sub-stage or the experimental monitoring vector of the experimental sub-stage in the monitoring target data is used to monitor each experimental animal number in the basic animal data of the experimental animal to be monitored in real time, and determine the real-time monitoring vector of each experimental animal in the basic animal data of the experimental animal to be monitored. Real-time monitoring data unit: Based on the real-time monitoring vectors of all experimental animals in the basic animal data of the experimental animals to be monitored, the real-time monitoring data of the experimental animals to be monitored is determined.

4. The experimental animal monitoring system according to claim 1, characterized in that, The monitoring module includes: Comparison Unit: The real-time monitoring vector of each experimental animal in the real-time monitoring data of the experimental animals to be monitored is compared with the feeding range matrix or experimental range matrix of the corresponding feeding sub-stage in the monitoring target data of the experimental animals to be monitored. Based on the comparison result, the real-time monitoring label is determined to be normal or abnormal. If the real-time monitoring label is abnormal, the abnormal monitoring feature vector of the experimental animal is determined based on all monitoring features that are not within the monitoring range. The abnormal monitoring feature value vector of the experimental animal is determined based on the real-time monitoring values ​​of all monitoring features that are not within the monitoring range. The abnormal monitoring range matrix of the experimental animal is determined based on the monitoring range of all monitoring features that are not within the monitoring range. The abnormal data monitoring unit determines the abnormal monitoring data of the experimental animals to be monitored based on the abnormal monitoring feature vector, abnormal monitoring feature value vector, and abnormal monitoring range matrix of all experimental animals with abnormal real-time monitoring tags.

5. The experimental animal monitoring system according to claim 1, characterized in that, The analysis module includes: Historical abnormal data unit: Based on the monitoring target data of the experimental animals to be monitored, the monitoring stage label, the feeding target of the feeding sub-stage or the experimental target of the experimental sub-stage corresponding to the monitoring stage label, and all abnormal monitoring feature vectors in the monitoring abnormal data, the historical abnormal data of the experimental animals to be monitored are obtained. Cause-based abnormal data unit: Based on the causes of abnormality of all abnormal animals in the historical abnormal data of the experimental animals to be monitored, the historical abnormal data is divided and the cause-based abnormal data of each abnormal cause is determined. The cause-based abnormal data includes the animal abnormal data of multiple abnormal animals.

6. The experimental animal monitoring system according to claim 5, characterized in that, in, Historical anomaly data includes animal anomaly data from multiple historical monitoring of multiple abnormal animals. Animal anomaly data includes the cause of the anomaly, the time interval of the anomaly, historical stage monitoring data of multiple feeding sub-stages or experimental sub-stages with the historical monitoring label as normal within the anomaly time interval, and historical anomaly feature vectors, historical anomaly feature value vectors, and historical anomaly feature range matrices of multiple feeding monitoring cycles or experimental monitoring cycles with the historical monitoring label as abnormal. Historical stage monitoring data includes historical trend graphs of each historical feature in the historical monitoring feature value vectors of all feeding monitoring cycles or experimental monitoring cycles.

7. The experimental animal monitoring system according to claim 1, characterized in that, The exception module includes: Anomaly Consistency Value Unit: Based on the anomaly monitoring feature vector, anomaly monitoring feature value vector, and anomaly monitoring range matrix of each experimental animal with an abnormal real-time monitoring tag in the monitoring anomaly data of the experimental animals to be monitored, as well as the cause anomaly data of each anomaly, the anomaly consistency value of each experimental animal with an abnormal real-time monitoring tag and each anomaly cause in the monitoring anomaly data of the experimental animals to be monitored is calculated. Cause set unit: Extract the abnormal causes corresponding to each experimental animal with an abnormal real-time monitoring tag and all non-zero abnormal consistency values ​​in the abnormal monitoring data of the experimental animals to be monitored, and determine the candidate abnormal cause set for each experimental animal with an abnormal real-time monitoring tag in the abnormal monitoring data of the experimental animals to be monitored. Time interval unit: Sort the lengths of the abnormal time intervals in the animal abnormal data of all abnormal animals for each abnormal cause in ascending order, and select the abnormal time interval corresponding to the upper quartile of the sorted data as the cause time interval of the abnormal cause. Interval Trend Unit: Based on the candidate abnormality cause set for each experimental animal with an abnormal real-time monitoring label in the monitoring abnormality data of the experimental animals to be monitored, and all feeding sub-stages and experimental sub-stages with normal historical monitoring labels within the cause time interval of each abnormality cause in the candidate cause set, all monitoring trend charts in the animal monitoring sub-data of the experimental animals in the animal monitoring data are extracted to determine the interval trend data of each abnormality cause in the candidate cause set for each experimental animal with an abnormal real-time monitoring label in the monitoring abnormality data of the experimental animals to be monitored.

8. The experimental animal monitoring system according to claim 7, characterized in that, The exception module also includes: The latency consistency value unit: The interval trend data of each abnormal cause in the candidate cause set of each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored, and the historical stage monitoring data of all feeding sub-stages or experimental sub-stages with normal historical monitoring labels in the abnormal time interval of all abnormal animals in the cause abnormal data are input into the abnormal latency consistency evaluation model to determine the latency consistency value of each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored and each abnormal cause in the candidate abnormal cause set; Abnormal cause label unit: Based on the abnormality value and latent consistency value of each abnormal cause in the monitoring abnormality data of the experimental animals to be monitored and the abnormality label of each experimental animal with an abnormal real-time monitoring label, the abnormal cause label of each experimental animal with an abnormal real-time monitoring label in the candidate abnormal cause set is determined.

9. The experimental animal monitoring system according to claim 1, characterized in that, The module to be determined includes: Tag anomaly data unit: Based on the anomaly cause tags of all experimental animals with abnormal real-time monitoring tags in the monitoring anomaly data of the experimental animals to be monitored, the tag anomaly data of each anomaly cause tag of the experimental animals to be monitored is determined. The tag anomaly data includes anomaly monitoring feature vectors, anomaly monitoring feature value vectors and anomaly monitoring range matrix of multiple experimental animals with abnormal real-time monitoring tags. Abnormal evolution and development data unit: Based on the monitoring stage label of the experimental animal to be monitored, the cause abnormal data of the abnormal cause corresponding to each abnormal cause label, the historical abnormal feature vector, historical abnormal feature value vector, and historical abnormal feature range matrix of all feeding monitoring cycles or experimental monitoring cycles with abnormal historical monitoring labels in the animal abnormal data of all abnormal animals, the abnormal evolution and development data of each abnormal cause label of the experimental animal to be monitored are extracted to determine the abnormal evolution and development data of each abnormal cause label of the experimental animal to be monitored. Real-time animal monitoring data unit: Based on the tagged abnormal data and abnormal evolution data of all abnormal cause tags of the experimental animals to be monitored, the real-time animal monitoring data of the experimental animals to be monitored is determined.

10. A method for monitoring laboratory animals, characterized in that, An experimental animal monitoring system for performing any one of claims 1 to 9, comprising: S1: Obtain basic animal data, monitoring stage data, and monitoring stage tags of the experimental animals to be monitored; obtain monitoring target data and animal monitoring data of the experimental animals to be monitored; and determine the real-time monitoring data of the experimental animals to be monitored based on the monitoring stage tags of the experimental animals to be monitored. S2: Based on the real-time monitoring data, monitoring stage tags, and monitoring target data of the experimental animals to be monitored, monitor the experimental animals to be monitored and identify abnormal monitoring data; S3: Based on the monitoring target data, monitoring stage labels, and monitoring abnormal data of the experimental animals to be monitored, obtain historical abnormal data of the experimental animals to be monitored, analyze the historical abnormal data, and determine the cause of each abnormality. S4: Based on the monitoring abnormal data of the experimental animals to be monitored and the cause abnormal data of all abnormal causes, determine the candidate abnormal cause set and abnormal cause label of each experimental animal with an abnormal real-time monitoring label in the monitoring abnormal data of the experimental animals to be monitored; S5: Based on the monitoring stage label of the experimental animal to be monitored, the abnormal cause label of all experimental animals with abnormal real-time monitoring labels in the abnormal monitoring data, and the cause abnormal data of all abnormal causes, determine the real-time animal monitoring data of the experimental animal to be monitored.

Images