Living body physiological index monitoring device, system, method and data processing device

Abstract

The application discloses a living body physiological index monitoring device, system, method and data processing device, and belongs to the technical field of biomedical engineering and automatic detection. The device comprises a containing module, the containing module comprising an initial area for containing a living body and a monitoring area connected to the initial area; a guiding module configured to utilize the spontaneous behavior tropism of the living body to guide the living body to automatically move from the initial area to the monitoring area; and a monitoring module configured to collect sensing data of the living body located in the monitoring area, the sensing data being capable of determining a physiological index of the living body. The application can realize living body monitoring of the living body without interfering with the normal physiological state of the living body, and obtain real data which can be used for analyzing the physiological index of the living body.

Description

Technical Field

[0001] This application relates to the fields of biomedical engineering and automated detection technology. Specifically, it relates to a device, system, method and data processing device for monitoring physiological indicators of live animals, which is particularly suitable for dynamic quantitative analysis of physiological reactions (such as allergic reactions, inflammatory reactions, changes in vascular permeability, etc.) of the body surface of live animals (such as mice) under non-stress conditions. Background Technology

[0002] In life science and medical research, monitoring physiological indicators in live animals is a key method for elucidating disease mechanisms and evaluating drug efficacy. For example, the Miles Assay (Evans blue exudation test), widely used in immunology and pharmacology research, involves injecting Evans blue dye into the tail vein. When a local allergic reaction occurs, vascular permeability increases, causing the dye bound to albumin to leak out and accumulate in the skin tissue, forming blue spots.

[0003] However, existing monitoring technologies face insurmountable technological gaps, hindering the in-depth development of research in this field.

[0004] On the one hand, existing technologies mainly rely on endpoint detection. For example, chemical extraction methods require euthanizing animals, cutting off skin for dye extraction, and spectrophotometric quantification. While this method is precise, it can only obtain endpoint data at a single time point, completely losing information about the complete dynamic process of the reaction from its occurrence, development, to its decline. This makes it impossible to provide high temporal resolution data support for pharmacokinetic / pharmacodynamic studies. Furthermore, euthanizing animals may induce stress responses or physiological changes, potentially distorting the endpoint data.

[0005] On the other hand, traditional methods for in vivo monitoring require anesthesia or physical restraint (such as binding) of animals, followed by visual observation or photography with a regular camera and manual scoring. However, this process itself induces a strong stress response in the animals. Studies have shown that physical restraint leads to a rapid increase in cortisol, increased heart rate, fluctuations in body temperature, and changes in peripheral hemodynamics in mice. These fluctuations in physiological parameters directly interfere with vascular permeability, a core detection indicator, resulting in severely distorted experimental data.

[0006] In summary, the long-standing but unresolved technical challenge in this field lies in how to conduct live animal monitoring without interfering with the normal physiological state of the animal, and obtain non-distorted data that can be used to analyze the physiological indicators of the animal. Summary of the Invention

[0007] In view of this, embodiments of this application provide a live animal physiological indicator monitoring device, system, method, and data processing device to achieve live animal monitoring without interfering with the normal physiological state of the live animal, and obtain non-distorted data that can be used to analyze the physiological indicators of the live animal.

[0008] In a first aspect, embodiments of this application provide a device for monitoring physiological indicators of live animals, comprising: A containment module, the containment module including an initial area for containing live animals and a monitoring area connected to the initial area; A guidance module, configured to utilize the spontaneous behavioral tendency of the live animal to guide it to automatically move from the initial area to the monitoring area; and A monitoring module is configured to collect sensor data of the live animal located within the monitoring area, the sensor data being capable of determining the physiological indicators of the live animal.

[0009] The technical solutions provided in the embodiments of this application have at least the following technical effects: By setting up a containment module that includes an initial area and a monitoring area, and using a guidance module to drive live animals to automatically move from the initial area to the monitoring area based on their spontaneous behavioral tendencies, this method differs from existing methods of physical restraint, anesthesia, and animal sacrifice. It employs a stress-free guidance method that does not interfere with the normal physiological state of the live animal, guiding it to move autonomously to the designated monitoring area—a fixed region. This achieves stress-free localization of the live animal. In this stress-free state, the monitoring module can collect sensor data from the live animal within the monitoring area, which exhibits no physical resistance or physiological fluctuations. The data can accurately reflect the actual condition of live animals under normal physiological homeostasis. Therefore, it is possible to determine the true physiological indicators of live animals based on sensor data, avoiding the distortion of physiological indicators caused by forced methods such as physical restraint, anesthesia, and animal sacrifice. It can realize the live monitoring of live animals without interfering with their normal physiological state, and obtain true data that can be used to analyze the physiological indicators of live animals. Furthermore, it can be adapted to continuous monitoring of live animals, obtaining continuous sensor data of live animals, which is conducive to the continuous monitoring of the physiological indicators of live animals.

[0010] Secondly, embodiments of this application provide a live animal physiological indicator monitoring system, comprising: The live animal physiological indicator monitoring device according to any embodiment of the first aspect; and A data processing device electrically connected to the monitoring module, the data processing device being configured to determine the physiological indicators of the live animal based on the sensor data.

[0011] The technical solutions provided in the embodiments of this application have at least the following technical effects: By setting up a data processing device electrically connected to the monitoring module, after the monitoring module collects sensor data of live animals located in the monitoring area, the data processing device can determine the physiological indicators of the live animals based on the sensor data, realize the automatic determination of the physiological indicators of live animals, and adapt to the automatic continuous monitoring of the physiological indicators of live animals.

[0012] Thirdly, embodiments of this application provide a method for monitoring physiological indicators of live animals, applied to a data processing device, the method comprising: The data processing device acquires sensor data; wherein, the sensor data is external sensor data collected from live animals used in biological experiments and capable of determining the physiological indicators of the live animals. The data processing device determines the physiological response characteristic data of the live animal based on the sensor data; wherein, the physiological response characteristic data is used to reflect the characteristics of the monitored physiological response produced by the live animal in the biological experiment. The data processing device determines the physiological indicators of the live animal based on the physiological response characteristic data.

[0013] Fourthly, embodiments of this application provide a data processing apparatus, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the data processing apparatus performs the method described in the second aspect.

[0014] It is understood that the beneficial effects of the third and fourth aspects mentioned above can be found in the relevant descriptions in the first or second aspects mentioned above, and will not be repeated here. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A schematic diagram of the structure of the live animal physiological indicator monitoring device provided in the embodiments of this application; Figure 2 A flowchart illustrating the method for monitoring physiological indicators of live animals provided in the embodiments of this application; Figure 3 This is a schematic diagram of the structure of the data processing apparatus provided in the embodiments of this application; Figure 4 This is a schematic diagram of the overall structure of the automatic monitoring system for allergic reactions on the mouse skin surface provided in the embodiments of this application; Figure 5 This is a schematic diagram of the thermal avoidance positioning principle in the embodiments of this application; Figure 6 This is a flowchart illustrating the image processing algorithm logic in the embodiments of this application; Figure 7 This is a physical image of the mouse surface allergy monitoring device in the embodiments of this application; Figure 8 This is a diagram illustrating the system setup and experimental operation scenario in the embodiments of this application; Figure 9 This is one of the machine vision automatic recognition and quantization results in the embodiments of this application; Figure 10 The second image shows the automatic recognition and quantization results of machine vision in the embodiments of this application.

[0017] The attached figures are labeled as follows: 10. Live animal physiological indicator monitoring device; 11. Containment module; 12. Guiding module; 13. Monitoring module; 1101. Initial area; 1102. Monitoring area; 121. Heating element; 11011. Narrowing area; 111. First side wall; 112. Second side wall; 131. Background enhancement part; 1103. Retention area. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. However, it should be understood that the specific embodiments of this invention are only for explaining the invention and are not intended to limit the scope of protection of this invention.

[0019] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0020] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0021] References to "one embodiment" or "some embodiments" in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized.

[0022] In related technologies, the monitoring of physiological indicators of live animals either sacrifices the physiological authenticity of the live animals in order to facilitate data collection (by anesthetizing or physically restraining the live animals for forced restraint), or sacrifices the possibility of continuous monitoring in order to obtain more authentic data than forced restraint (endpoint method of euthanizing animals to obtain endpoint data). However, the process of euthanizing animals may still cause stress responses or changes in physiological state of the animals, which in turn leads to distortion of endpoint data.

[0023] In view of the above problems, this application provides a live animal physiological indicator monitoring device, system, method and data processing device, which can realize live monitoring without interfering with the normal physiological state of the live animal and obtain non-distorted data that can be used to analyze the physiological indicators of the live animal.

[0024] Below, we will first introduce the live animal physiological indicator monitoring device provided in the embodiments of this application.

[0025] Please see Figure 1 The live animal physiological indicator monitoring device 10 provided in this application embodiment includes a housing module 11, a guiding module 12 and a monitoring module 13.

[0026] The containment module 11 includes an initial area 1101 for containing live animals and a monitoring area 1102 connected to the initial area 1101. The initial area 1101 is used to receive the live animals placed inside, and the monitoring area 1102 is used to limit the range of movement of the live animals during monitoring, and is spatially connected to the initial area 1101. It can be understood that the containment module 11 is a structure that provides space for the live animals to move around and defines the spatial boundaries, and can be a shell structure, box structure, frame structure, fence structure, drawer structure, etc., but is not limited to these. Figure 1 The example shown is a case where the housing module 11 is a shell structure that is closed at the bottom and all four sides and open at the top. Of course, in some cases the top can also be covered or closed.

[0027] The guidance module 12 is configured to utilize the spontaneous behavioral tendency of the live animal to guide it to automatically move from the initial area 1101 to the monitoring area 1102. This can be understood as stimulating the animal's inherent instinctive behavioral response (such as seeking advantage and avoiding harm), causing it to actively and voluntarily move to the monitoring area 1102. The guidance module 12 is a functional component that utilizes the animal's spontaneous behavioral tendency for stress-free localization. Unlike methods that restrict animal activity through external force such as mechanical clamps, restraints, or drug anesthesia, this non-stress-based guidance method ensures that the live animal is not physically restrained during movement and remains in a physiological homeostasis.

[0028] For example, this spontaneous behavioral tropism can be thermotaxis based on temperature differences, phototaxis based on light differences, or chemotaxis based on odor gradients. This non-invasive guidance and positioning mechanism fundamentally eliminates stress responses caused by forced restraint (such as elevated cortisol, abnormal heart rate, and hemodynamic changes), allowing the live animal to remain in a natural physiological homeostasis during monitoring, thereby improving the authenticity and reliability of the acquired sensor data. Simultaneously, because the animal voluntarily remains in the monitoring area 1102, its posture within the monitoring area 1102 is more natural and relaxed, facilitating the acquisition of a more complete surface imaging field.

[0029] The monitoring module 13 is configured to collect sensor data from live animals located in the monitoring area 1102. The sensor data can determine the physiological indicators of the live animals. Since the animals are guided to the monitoring area 1102 and stay there in a stress-free state, the sensor data collected by the monitoring module 13 can truly and accurately reflect the actual condition of the live animals under normal physiological homeostasis.

[0030] It is understood that the sensor data is collected by the monitoring module 13, and can be various types of sensor data capable of determining the physiological indicators of live animals. For example, it can be one or more of image data, thermal radiation data, sound wave data, electromagnetic reflection data, depth data, and spectral data, but is not limited to these. The monitoring module 13 is a corresponding sensing device capable of collecting corresponding sensor data, such as one or more of a camera, infrared thermal imaging device, sound sensor, radar (UWB, continuous wave radar), and lidar. The specific settings can be configured according to the biological experiments conducted on the live animals and the physiological responses of the live animals to be monitored, and are not limited to a single method here.

[0031] It is understandable that by setting up functional partitions for the initial area 1101 and the monitoring area 1102, and utilizing the guidance module 12 for stress-free localization based on the animal's spontaneous behavior, the monitoring module 13 can collect sensor data when the animal is in a physiological homeostasis. The entire device constitutes a collaborative working architecture for stress-free behavioral guidance and fixed-point sensor data acquisition, fundamentally avoiding data distortion caused by coercive methods such as anesthesia and restraint. This partitioning design of the initial area 1101 and the monitoring area 1102 spatially decouples the animal introduction process from the data acquisition process, avoiding the impact of operational interference on monitoring stability.

[0032] As can be seen from the above, the live animal physiological indicator monitoring device 10 provided in this application embodiment, by setting up a receiving module 11 including an initial area 1101 and a monitoring area 1102, and using a guidance module 12 to drive the live animal to automatically move from the initial area 1101 to the monitoring area 1102 based on its spontaneous behavioral tendency, is different from the existing methods of physical restraint, anesthesia, and animal sacrifice. It adopts a stress-free guidance method that does not interfere with the normal physiological state of the live animal, guiding the live animal to move autonomously to the fixed area of ​​the set monitoring area 1102, thereby achieving stress-free positioning of the live animal. In this stress-free state, the monitoring module 13 can collect data located in the monitoring area 1102. The sensor data of live animals within 102 without physical resistance or physiological fluctuations can accurately reflect the actual condition of live animals under normal physiological homeostasis. Therefore, the true physiological indicators of live animals can be determined based on the sensor data, avoiding the distortion of physiological indicators caused by forced methods such as physical restraint, anesthesia, and animal sacrifice. It can realize the live monitoring of live animals without interfering with their normal physiological state, and obtain true data that can be used to analyze the physiological indicators of live animals. Furthermore, it can be adapted to continuous monitoring of live animals, obtaining continuous sensor data of live animals, which is conducive to the continuous monitoring of the physiological indicators of live animals.

[0033] In some embodiments of this application, the guiding module 12 is configured to form a stimulus gradient field along the initial region 1101 to the monitoring region 1102, the stimulus gradient field being used to drive the live animal to move automatically from the initial region 1101 to the monitoring region 1102.

[0034] It can be understood that the stimulus gradient field is a spatial field in which the intensity of a certain physical or chemical stimulus changes continuously or stepwise along the direction from the initial region 1101 to the monitoring region 1102. Live animals, out of instinct, will spontaneously move from the initial region 1101, where the stimulus is stronger, to the monitoring region 1102, where the stimulus is weaker. The stimulus gradient field utilizes the perception ability and spontaneous behavioral tendency of live animals to generate the driving force for directional movement. Unlike traditional mechanical restraint or external forced thrust, the stimulus gradient field internalizes the positioning dynamic into the animal's own avoidance or pursuit of benefit behavior, enabling the animal to autonomously find and remain in the monitoring region 1102, where the environmental parameters are suitable, after entering the initial region 1101.

[0035] It should be understood that the stimulus gradient field can be a temperature gradient field, a light intensity gradient field, a humidity gradient field, an odor concentration gradient field, or a combination thereof, as long as it can stimulate spontaneous avoidance or orientation behavior responses in live animals.

[0036] By creating a stimulus gradient field and utilizing the animal's natural avoidance response to adverse stimuli, precise spatial dispersal without physical contact is achieved. This transforms the animal's instinctive escape behavior into deterministic and reproducible spatial movement. The dispersal process itself does not introduce physical stress, unlike traditional physical restraint or anesthesia-based positioning.

[0037] In some embodiments, the stimulus gradient field is a temperature gradient field, with the temperature of the initial region 1101 being higher than that of the monitoring region 1102. Thus, by utilizing the temperature sensitivity of the living animal, it is driven to spontaneously move from the relatively warm initial region 1101 to the cooler monitoring region 1102, and can remain in the monitoring region 1102.

[0038] It should be noted that the temperature of the initial area 1101 is a temperature that will not cause a stress response to the live animal; it is simply higher than the temperature of the monitoring area 1102, in order to guide the animal to move spontaneously to the monitoring area 1102.

[0039] Optionally, the temperature range of the initial region 1101 is 50°C to 70°C, and the temperature range of the monitoring region 1102 is 20°C to 30°C.

[0040] The selection of this specific parameter range is based on a comprehensive consideration of experimental animal behavior and physiological safety boundaries. Specifically, for common small laboratory animals (such as mice and rats), the upper limit of their thermal comfort zone is usually around 30-35℃. When the ambient temperature rises to the range of 50-70℃, it significantly exceeds the animal's thermal comfort range, which is sufficient to trigger a thermal avoidance reflex in a short time, forcing the animal to quickly leave the area; at the same time, this temperature range is strictly controlled below the threshold that would cause skin burns or irreversible thermal damage, and because the animal stays in the area for a very short time (usually only a few seconds), it will not cause substantial physiological harm or stress-induced pathological changes.

[0041] In contrast, the 20-30℃ range set in monitoring area 1102 covers the thermoneutral zone or suitable living temperature range for most experimental animals. This allows animals to achieve thermal comfort within monitoring area 1102 after escaping from initial area 1101, thus voluntarily and quietly remaining there, providing a stable posture basis for subsequent detection by monitoring module 13. If the temperature in initial area 1101 is too low (e.g., below 40℃), escape behavior may not be effectively triggered, leading to positioning failure or excessive time consumption; if the temperature is too high (e.g., above 80℃), there are safety hazards and it may trigger excessive panic reactions, affecting the accuracy of sensor data. Similarly, if the temperature in monitoring area 1102 deviates from the suitable range, animals may exhibit restlessness or attempt to move again, affecting the quality of sensor data, especially image data. Therefore, the combination of 50-70℃ and 20-30℃ constitutes the optimal parameter window that balances positioning efficiency, animal welfare, and data quality.

[0042] For example, the initial temperature of region 1101 is set to 65°C, and the temperature of monitoring region 1102 is the same as the ambient temperature, which is about 25°C.

[0043] It should be emphasized that the above temperature values ​​are only preferred examples for specific experimental subjects. In practical applications, those skilled in the art can adaptively adjust the temperature settings of the initial area 1101 and the monitoring area 1102 according to the species, body size, coat thickness and heat tolerance characteristics of the live animal to be monitored. The core principle is always to maintain a sufficient temperature difference to drive behavior, while ensuring that the absolute temperature of each area is within a safe and ethical range.

[0044] Optionally, the guiding module 12 includes a heating element 121, which is disposed in the initial region 1101 and is used to heat the initial region 1101 to generate a local high temperature in the initial region 1101 that is higher than the ambient temperature, so that the temperature of the initial region 1101 is higher than that of the monitoring region 1102, thereby forming a temperature gradient.

[0045] Optionally, the monitoring area 1102 can be at ambient temperature or room temperature, which helps to provide a comfortable living environment for the live animals and improve their stability during monitoring. Optionally, a heat sink, a semiconductor cooling pad, or natural heat dissipation from the environment can be installed below the monitoring area 1102 to form a stable temperature step or a continuous temperature gradient zone between the initial area 1101 and the monitoring area 1102.

[0046] Optionally, the heating element 121 can be a resistance wire heater, a PTC (positive temperature coefficient) ceramic heater, a silicone heating element, a heating film, or an infrared heating module.

[0047] For example, the heating element 121 is a PTC heating plate, which has automatic temperature control characteristics and is safe and reliable.

[0048] Optionally, the heating element 121 is a heating plate, that is, a plate-shaped heating element. The heating plate is set at the bottom of the initial region 1101. It can be set on the inner or outer surface of the bottom of the initial region 1101, or it can be directly used as the bottom of the initial region 1101.

[0049] In this way, the bottom of the initial area 1101 is heated by heat conduction, allowing the soles of the feet of a living animal standing on it to directly feel the heat stimulus. Heating the soles of the feet directly from the bottom is a direct and effective method of heat-induced repulsion. The plate-like structure can fit into the bottom of the initial area 1101, forming a uniform heated surface, so that the temperature of the soles of the animal's feet is basically the same regardless of where the animal stands in the initial area 1101, thus achieving consistency in the repulsion behavior.

[0050] The live animal commonly used in biological experiments is the mouse. Placing the heat source at the bottom of the initial area 1101 is a specific design that takes advantage of the physiological characteristic that rodents' feet are highly sensitive to temperature, rather than ordinary space heating.

[0051] For example, the heating plate may be sized to match the bottom dimensions of the initial region 1101 and be fitted to the bottom outer surface of the initial region 1101.

[0052] It should be noted that the heating element 121 can be disposed not only at the bottom of the initial region 1101, but also on the side or top to heat the initial region 1101.

[0053] Optionally, the surface of the heating plate can be covered with a visible marking layer that is different from the background color (such as using a high-temperature resistant orange or red coating, tape, or indicator light as a visual warning). This not only helps with quick alignment during operation, but also helps the experimenter identify high-temperature areas and prevents accidental contact with high-temperature areas.

[0054] Optionally, the guide module 12 also includes a thermostat electrically connected to the heating element 121.

[0055] It can be understood that a thermostat is an electronic device used to control the temperature of a heating element. For example, a thermostat can automatically adjust the heating power by detecting the actual temperature of the heating element and comparing it with the set temperature, so that the heating element is maintained within the set temperature range.

[0056] By introducing temperature control, the intensity of the temperature gradient field can be adjusted and kept constant, avoiding scalding animals due to excessively high temperatures or failing to drive them away due to excessively low temperatures, which is conducive to achieving consistency and reproducibility of experimental conditions.

[0057] For example, the temperature controller uses a PID (proportional-integral-derivative) control algorithm to collect the surface temperature of the heating element in real time through a thermocouple or thermistor, compare it with the preset target temperature, and then output a control signal to adjust the power supply of the heating element.

[0058] Optionally, the thermostat can be a digital display thermostat with an LED display and button input, allowing the user to directly set the target temperature value. For example, the thermostat can be set to a target temperature of 65°C with a temperature control accuracy of ±1°C.

[0059] In other embodiments, the stimulus gradient field may also be a light intensity gradient field, where the light intensity of the initial region 1101 is greater than the light intensity of the monitoring region 1102.

[0060] It can be understood that the light intensity gradient field is a spatial field in which the light intensity decreases along the direction from the initial region 1101 to the monitoring region 1102. By utilizing the instinct of nocturnal animals such as mice to seek darkness and avoid light, they are moved from the brighter initial region 1101 to the weaker monitoring region 1102.

[0061] For example, a high-brightness light source (such as an LED array) is arranged above the initial area 1101, while only a weak light or complete shading is retained above the monitoring area 1102 to create a contrast between light and dark.

[0062] Light control offers a fast response, no thermal inertia, and causes no sensory stimulation to animals, making it particularly suitable for experimental subjects that are not sensitive to temperature. It transforms the tendency to seek darkness from an observational indicator into an active means of control.

[0063] For example, the initial area 1101 has a light intensity of 1500-2500 lux, and the monitoring area 1102 has a light intensity of 30-100 lux.

[0064] In other embodiments, the stimulus gradient field is an odor concentration gradient field, where the odor concentration in the initial region 1101 is greater than the odor concentration in the monitoring region 1102.

[0065] It can be understood that the odor concentration gradient field is a spatial field in which the concentration of a specific odor decreases along the direction from the initial region 1101 to the monitoring region 1102. By utilizing the animal's instinct to avoid certain specific odors, it is made to move from the initial region 1101, where the odor concentration is high, to the monitoring region 1102, where the odor concentration is low.

[0066] For example, an odor generating device (such as a slow-release device containing odors that mice avoid, such as menthol) is set in the initial area 1101, and a vent is set in the monitoring area 1102 to introduce fresh air, thereby creating an odor concentration difference.

[0067] For example, the odorant could be menthol, the initial concentration in zone 1101 could be maintained at a level that would be perceptible to animals but harmless, and the monitoring zone 1102 could be fresh air.

[0068] In some embodiments of this application, the monitoring area 1102 is configured as a channel structure that restricts the turning of a live animal, the width of which is less than the body length of the live animal.

[0069] As can be understood, body length refers to the length of a living animal's body. For example, for rodents, body length is the straight-line distance from the tip of the nose to the base of the tail. A channel structure that is narrower than its body length means that the animal cannot turn around within the channel, forcing its body axis to align with the channel's direction. This improves the stability and consistency of the sensor data collected by the monitoring module 13.

[0070] The channel structure is a long, narrow tubular or trough-shaped channel with carefully designed lateral dimensions. This ensures that when a live animal enters the area, its body longitudinal axis can only align along the direction of the channel, preventing it from turning around or making large lateral turns. This physical constraint is not intended to restrict the animal's normal physiological activities, but rather to reduce the impact of random animal turning on the acquisition of sensor data, especially to eliminate changes in the viewing angle of image data acquisition caused by random animal turning.

[0071] For example, when the sensing data includes image data, during continuous monitoring, if the animal can turn freely, the monitoring module 13 will alternately capture images of different parts such as the head, back, and side. This not only causes drastic changes in the position and shape of the physiological response area to be tested in the image, but also causes differences in light reflection due to changes in the curvature of the body surface, seriously interfering with the feature extraction and quantization accuracy of subsequent image processing algorithms. By limiting the channel width to less than the body length, this embodiment enforces the standardization of imaging posture, ensuring that each frame of real-time image data originates from the same anatomical perspective, laying a solid hardware foundation for the comparability of time-series data and the robustness of the algorithm. The image acquisition unit can continuously acquire images of the animal's body surface from a fixed angle, avoiding changes in image perspective and deformation of feature areas caused by the animal turning, significantly improving the accuracy and repeatability of subsequent image analysis and feature quantization. It should be understood that the specific width value of the channel should be adaptively set according to the species, size, and flexibility of the target live animal, as long as it can achieve the function of restricting turning without affecting the animal's normal breathing and blood circulation. This embodiment does not impose an absolute limitation on this.

[0072] For example, the cross-section of the channel structure is rectangular or trapezoidal, and the minimum distance (width) between its inner walls is designed according to the body size of the target animal.

[0073] For example, for adult BALB / c mice (body length approximately 8-10cm), the channel width is designed to be 5cm and the height is designed to be 5cm.

[0074] In some embodiments, the activity area of ​​the initial region 1101 is larger than that of the monitoring region 1102. The larger space of the initial region 1101 reduces the initial stress on the animal, while the smaller space of the monitoring region 1102 keeps the animal within the monitoring field of vision. This contrast in space size naturally compresses the animal's activity range. This aligns with the behavior of some animals, such as rodents, who prefer to enter confined spaces for a sense of security, facilitating the entry of live animals into the monitoring region 1102.

[0075] For example, the initial area 1101 is a square or circular area with a relatively large area; the monitoring area 1102 is a narrow passage with a relatively small area. Animals can move and explore freely in the initial area 1101, but their range of movement is restricted after entering the monitoring area 1102.

[0076] In some embodiments, the monitoring area 1102 is located on one side of the initial area 1101, and the initial area 1101 and the monitoring area 1102 are arranged linearly, which facilitates the discovery and entry of live animals into the monitoring area 1102.

[0077] In other embodiments, the two can also adopt an L-shaped, T-shaped or other arbitrarily connected topology, as long as the live animal can be smoothly transferred from the initial area 1101 to the monitoring area 1102.

[0078] Optionally, a narrowed region 11011 is formed on the side of the initial region 1101 facing the monitoring region 1102, and the width of the narrowed region 11011 gradually decreases along the direction from the initial region 1101 to the monitoring region 1102.

[0079] It can be understood that the narrowing region 11011 is the transition region between the initial region 1101 and the monitoring region 1102. Its width smoothly transitions from the larger width of the initial region 1101 to the smaller width of the monitoring region 1102, and the whole is funnel-shaped.

[0080] The gradually narrowing passage conforms to the animal's habit of exploring narrow spaces in its natural environment. The guidance process is smooth and unobstructed, avoiding the animal getting stuck or hesitating due to sudden changes in width, which is conducive to guiding live animals into the monitoring area 1102.

[0081] Optionally, the narrowed area 11011 is formed by the inwardly inclined extension of the sidewall of the receiving module 11, guiding the animal naturally from the wide initial area 1101 into the narrow monitoring area 1102.

[0082] like Figure 1 As shown, from a top-down view of the accommodating module 11, the sidewalls of the initial region 1101 converge toward the center to form a narrowed region 11011, which eventually connects to the narrow monitoring region 1102.

[0083] Specifically, the inner wall of the initial region 1101 includes a first sidewall 111 and a second sidewall 112 arranged opposite to each other. The first sidewall 111 and the second sidewall 112 extend obliquely into the monitoring region 1102, forming a narrowing region 11011 between the first sidewall 111 and the second sidewall 112. The two relatively oblique sidewalls together constitute a symmetrical funnel-shaped guide channel. This improves the guiding effect.

[0084] For example, the first sidewall and the second sidewall are at an angle of 45 degrees to the length direction of the receiving module, and extend from the rear wall of the initial region 1101 to the entrance of the monitoring region 1102.

[0085] In some embodiments, the containing module 11 further includes a retention area 1103, which is located on the side of the monitoring area 1102 opposite to the initial area 1101, and is used to contain the live animal after the monitoring area 1102 has collected sensor data. After the live animal has had its sensor data collected by the monitoring module 13 within the monitoring area 1102, it can continue to move forward to the retention area 1103.

[0086] Optionally, a removable partition (such as a baffle or fence) can be installed between the monitoring area 1102 and the detention area 1103 to restrict the live animal from moving to the detention area 1103 before the sensor data is collected. After the data is collected, the partition can be removed to allow the live animal to enter the detention area 1103.

[0087] The activity area of ​​the detention area 1103 can be larger than that of the monitoring area 1102, which is beneficial for processing live animals after the sensor data is collected.

[0088] Of course, in some other embodiments, the retention area 1103 may not be provided.

[0089] In some embodiments of this application, the monitoring module 13 includes an image acquisition unit, which is used to acquire image data of live animals located in the monitoring area 1102, and the sensing data includes image data.

[0090] As can be understood, an image acquisition unit is an electronic device capable of converting optical images into digital image signals. Image data can capture subtle color and morphological changes on the animal's body surface non-contactly and with high spatiotemporal resolution, making it an ideal data type for monitoring physiological responses on the body surface.

[0091] Optionally, the image acquisition unit can be a CMOS camera or a CCD camera. For example, the image acquisition unit is a Logitech C920 HD USB camera, supporting a maximum resolution of 1920×1080 and a frame rate of 30fps. Alternatively, the image acquisition unit can be an HD USB camera connected to a data processing device via a USB cable to transmit acquired image data in real time.

[0092] In some embodiments, the sensing data is real-time sensing data, and the image data is real-time image data.

[0093] Real-time sensor data is understood to be sensor data that is continuously acquired and transmitted in real time at a certain frequency, as opposed to static data acquired in a single instance. Real-time image data is a sequence of image frames acquired continuously in the form of a video stream.

[0094] Real-time data is beneficial for recording the entire dynamic process of physiological responses in live animals, from occurrence and development to regression, achieving a leap from point measurement to line monitoring, which existing endpoint methods or static photography cannot achieve. This leads to the application of real-time video stream analysis technology in this field.

[0095] For example, the image acquisition unit continuously acquires images at a fixed frame rate (e.g., 5 fps) and transmits each frame to the data processing device for processing in real time. For example, the real-time acquisition frame rate is 5 fps, which can be adjusted to 1-30 fps as needed.

[0096] In some embodiments, the image acquisition unit is used to acquire image data of the body surface of a live animal located in the monitoring area 1102. The image data can determine the physiological response characteristics of the body surface of the live animal, and thus determine the physiological indicators of the live animal.

[0097] It can be understood that physiological response feature data are data extracted from image data that directly characterize the physiological response to be monitored. For example, when the physiological response to be monitored is Evans blue exudation, the physiological response feature data could be a set of pixels in the blue area of ​​the image.

[0098] The method involves collecting images of the body surface of live animals without contact, and the physiological indicators of live animals can be determined from the images of the body surface. It is easy to collect sensor data and facilitates the analysis of physiological indicators.

[0099] By introducing physiological response feature data as an intermediate layer, the complex image analysis task is modularized. First, features directly related to the response are extracted from the image, and then the final physiological indicators are calculated based on these features. This improves the system's flexibility and scalability, making it easier to replace the corresponding feature extraction module for different physiological response types.

[0100] Of course, in some other embodiments, the image data may also be non-surface image data, such as X-ray images, the choice of which is related to the biological experimental response to be performed and the physiological indicators to be monitored.

[0101] In some embodiments, the image acquisition unit is positioned above the monitoring area 1102 for acquiring image data from a top-down perspective.

[0102] A top-down perspective is advantageous because the observation plane is parallel to the animal's back, minimizing perspective distortion. This facilitates the acquisition of overall image data and makes it easier to accurately calculate the area of ​​the reaction zone.

[0103] For example, the image acquisition unit is fixed directly above the monitoring area 1102 by a bracket, tape or other fixing structure, with the lens optical axis pointing vertically downward and the field of view covering the bottom of the entire monitoring area 1102.

[0104] For example, the image acquisition unit is 20cm away from the bottom of the monitoring area 1102, and the field of view covers an area of ​​about 15cm×10cm.

[0105] In some embodiments, the monitoring module 13 further includes a background enhancement portion 131, which is disposed in the monitoring area 1102. The background enhancement portion 131 has a color that forms an optical contrast with the body color of the live animal and the characteristic color of the physiological response to be monitored. The image acquisition unit can acquire image data against the color background formed by the background enhancement portion 131.

[0106] It can be understood that the background enhancement portion 131 is a portion with a specific color set within the imaging field of view of the monitoring area 1102, used to provide a high-contrast background for the image data. This maximizes the signal-to-noise ratio of the target signal to the background noise at the imaging source, providing a standardized and easily segmented data foundation for subsequent image processing algorithms. Optical contrast refers to the distance between two or more colors in the color space; the greater the distance, the higher the contrast, and the easier it is for image processing algorithms to distinguish them.

[0107] The background enhancement component 131 uses a reverse selection method based on the surface color of the live animal and the characteristic colors of the physiological responses to be monitored. For example, the selection principle is based on the complementary color theory of the color wheel or the distribution rules of the color space, so that the background color can be maximized to distinguish it from the foreground target (animal surface and response area) in terms of chromaticity characteristics.

[0108] This setup helps create high contrast at the physical level of imaging, simplifying the originally complex image segmentation problem into simple threshold segmentation, which greatly reduces the computational overhead and development difficulty of the algorithm.

[0109] For example, when the live animal is a white mouse and the characteristic color of the physiological response to be monitored is blue, the background enhancement portion 131 is selected as green, as green can be well distinguished from both white and blue on the color wheel. When the live animal is a black mouse and the characteristic color of the physiological response to be monitored is red, the background enhancement portion 131 is selected as cyan.

[0110] It should be understood that the specific color or spectral reflectance characteristics of the background enhancement portion 131 should be adapted to the species and physiological response characteristics of the animal being tested, and this embodiment does not make specific limitations here.

[0111] Optionally, the background enhancement portion 131 is disposed at the bottom and / or side of the monitoring area 1102, that is, it can be disposed at the bottom of the monitoring area 1102 (which can be the inner or outer surface of the bottom), or at the side of the monitoring area 1102 (which can be the inner or outer surface of the side), or it can be disposed at both the bottom and side of the monitoring area 1102.

[0112] For example, in the case of image acquisition using a top-down view, the background enhancement portion 131 is laid on the side of the monitoring area 1102, which can make the live animals stand out in the acquired image data.

[0113] Optionally, the background enhancement portion 131 uses a matte surface as the background, such as a frosted surface. It can be understood that a matte surface is one that diffusely reflects light, has low gloss, and does not produce obvious specular highlights. The matte surface causes uniform diffuse reflection in the background, preventing highlights from being misidentified as target features by image processing algorithms, thus improving the robustness and accuracy of foreground segmentation.

[0114] Optionally, the background enhancement part 131 is a background plate, that is, a plate-like structure or a sheet-like structure.

[0115] For example, the background enhancement portion 131 is made of a sheet material with a frosted or matte finish. The matte surface allows light emitted from the lighting source to be evenly scattered across the background, avoiding the formation of localized highlight areas.

[0116] For example, the background enhancement portion 131 is a green frosted PVC board with a surface roughness Ra value of 1.6-3.2μm and a 60° gloss of less than 10GU.

[0117] In some embodiments, the live animal physiological indicator monitoring device 10 further includes a light-shielding portion that covers the monitoring area 1102.

[0118] The light-shielding part can be a cover or covering made of opaque material that encloses the monitoring area 1102 of the housing module 11, used to isolate the monitoring area 1102 from ambient light. For example, the light-shielding part can be made of materials with high light absorption rates, such as black velvet, matte coating, or dark opaque plastic, and can be attached to or cover the outside of the monitoring area 1102 of the housing module 11.

[0119] This setup serves two purposes. First, from an optical imaging perspective, the light-blocking portion effectively blocks stray light from the external environment from entering the monitoring area 1102, eliminating background noise fluctuations caused by laboratory lighting flicker, shadows from personnel movement, or changes in natural light, thus creating a pure darkroom environment for image acquisition. Second, from an animal behavior perspective, most experimental rodents (such as mice and rats) naturally exhibit negative phototaxis or dark-taxis. A dark environment provides them with a sense of security and reduces their desire to explore, making them more willing to stay in that area, further reducing their stress levels and improving the authenticity and reliability of the sensor data. Therefore, the dark environment created by the light-blocking portion not only serves image quality but also acts as an invisible behavioral guide, working in synergy with the guidance module to encourage live animals to remain more quietly and for longer periods within the monitoring area 1102, further reducing motion artifacts caused by agitation.

[0120] For example, the light-shielding part is made of black light-absorbing velvet and is fitted over the monitoring area 1102 of the housing module 11 to form a detachable flexible darkroom. For example, the light-shielding part is a double layer of black velvet with a light transmittance of less than 0.1% and covers the outside of the monitoring area 1102 of the housing module 11.

[0121] In some embodiments, the monitoring module 13 further includes an illumination source for providing illumination for the image acquisition unit to acquire image data.

[0122] It can be understood that the lighting source is a light-emitting device that provides active illumination, used to provide a stable and controllable light environment for image acquisition inside the dark chamber formed by the light-shielding part.

[0123] Active and constant illumination enables images acquired at different times and in different batches to have direct and comparable brightness, color, and contrast, which is beneficial for the automatic quantification of image data acquisition and subsequent image data analysis and processing.

[0124] For example, the lighting source is a high color rendering index (CRI) LED light strip, installed inside the housing module 11 or on the inner wall of the light-shielding part, providing uniform, flicker-free lighting with constant spectral characteristics. For instance, the lighting source is an LED white light strip with a color temperature of 5500K (daylight color), and the illuminance is uniformly distributed at the bottom of the monitoring area 1102, at approximately 500 lux, with no flicker.

[0125] In the automated quantitative analysis of in vivo physiological indicators, the stability of illumination conditions is beneficial to improving the effectiveness of algorithm thresholds. This effectively reduces the possibility that pixel grayscale values ​​or color components in the acquired image data may change due to time-varying illumination brightness or color temperature.

[0126] The shading component, in conjunction with the lighting source, creates a closed and controlled artificial light environment, isolating external light interference. This ensures that throughout the entire continuous monitoring period, the image data acquired by the image acquisition unit reflects only the true changes in the physiological characteristics of the live animal's body surface, rather than disturbances caused by ambient light.

[0127] In some embodiments, the live animal is a white mouse, the physiological response to be monitored is an allergic reaction based on Evans blue dye, the characteristic color is blue, and the background enhancement area is green.

[0128] It is understandable that Evans blue dye is a commonly used indicator of vascular permeability, which binds to plasma albumin after being injected into blood vessels. When an allergic reaction occurs, vascular permeability increases, and the Evans blue-albumin complex leaks into the interstitial space, forming blue spots on the skin surface.

[0129] In this embodiment, the system configuration was optimized for an Evans blue allergy model in white BALB / c mice. The mouse body surface was white, the exudative spots to be monitored were blue, and the background enhancement area was selected as green, which forms a strong contrast with both white and blue. This color combination utilizes the principle of complementary colors, so that in the HSV color space, the hue of the green background can be clearly distinguished from the white mouse (low saturation) and the blue spots (high saturation).

[0130] According to the principle of complementary colors, green and red (the base color of blood vessels and pink tissue in mouse skin) are complementary colors and are the farthest apart in the color space, which can maximize the separation between the animal subject and the background. At the same time, green and blue also have significant differences on the color wheel, so that the indigo staining features still maintain a clear recognizability against a green background.

[0131] This dual-color high-contrast design allows the data processing module to quickly and accurately remove the background and generate a foreground mask from complex surface textures using background chromaticity features, and then precisely identify subtle physiological response areas within the foreground using characteristic chromaticity features. If the background color is inappropriately chosen (e.g., using a white or gray background), the animal edges will blend into the background, or physiological features will be submerged, causing the algorithm to be unable to distinguish between the target and noise, or even to fail completely. Therefore, the predetermined color of the background enhancement portion is the bridge connecting physical imaging and digital algorithms; its optical contrast characteristics directly determine the sensitivity and accuracy of the system's quantitative analysis. In practical applications, those skilled in the art should determine the optimal background color through color space analysis based on the specific animal coat color and the spectral characteristics of the physiological response colorant to achieve the best segmentation effect.

[0132] For example, the color value of the green background enhancement portion is RGB(0, 180, 0), the mouse strain is BALB / c albino mouse, and Evans blue dye is injected via tail vein at a dose of 100 μL / mouse (1% concentration).

[0133] The following describes the live animal physiological indicator monitoring system provided in the embodiments of this application.

[0134] The live animal physiological indicator monitoring system provided in this application includes a live animal physiological indicator monitoring device 10 as described in any of the foregoing embodiments, and a data processing device. The data processing device is electrically connected to the monitoring module 13, and is configured to determine the physiological indicators of the live animal based on sensor data.

[0135] It can be understood that the data processing device is an electronic device with data processing capabilities, used to receive sensor data collected by the monitoring module 13, execute analysis algorithms, and output physiological indicators. The electrical connection between the data processing device and the monitoring module 13 can be a wired connection (such as a USB data cable) or a wireless connection (such as WiFi or Bluetooth).

[0136] For example, the data processing device can transform the unstructured sensor data collected by the monitoring module 13 into structured quantitative physiological indicators. This process is fully automated, which not only avoids the subjective bias and efficiency bottlenecks caused by manual scoring, but also enables time-series analysis of continuously collected sensor data, thereby generating a complete curve reflecting the dynamic evolution of physiological responses, providing high temporal resolution data support for pharmacological research or pathological analysis.

[0137] Optionally, the data processing device can be a personal computer, an embedded processing board (such as the NVIDIA Jetson series), or a cloud server.

[0138] For example, the data processing device is a computer with dedicated analysis software installed. The computer connects to the image acquisition unit via a USB cable to receive real-time image data. Separating the hardware for data acquisition and data processing facilitates system deployment and independent upgrades.

[0139] Of course, in some other embodiments, the data processing device can also be integrated with the monitoring module 13, that is, integrated into the monitoring module 13.

[0140] In some embodiments of this application, the data processing device is configured to: acquire sensor data collected by the monitoring module; determine physiological response characteristic data of a live animal based on the sensor data; wherein the physiological response characteristic data is used to reflect the characteristics of the physiological response to be monitored produced by the live animal in the biological experiment; and determine the physiological indicators of the live animal based on the physiological response characteristic data.

[0141] In this embodiment, the data processing device implements an automated analysis process from sensor data to physiological indicators through software programs. Using physiological response characteristic data as an intermediate layer makes the data processing process modular and configurable. The core functional framework of the data processing device is presented, realizing the automatic conversion from sensor data to physiological indicators.

[0142] In some embodiments, the sensing data is image data. Determining physiological response characteristic data of a live animal based on the sensing data includes: separating a foreground image containing the live animal from the image data; identifying a physiological response characteristic region from the foreground image based on the chromaticity characteristics of the response characteristic color; wherein the response characteristic color is the characteristic color of the physiological response to be monitored produced by the live animal, and the physiological response characteristic region is an image of the region containing the response characteristic color. Determining physiological indicators of the live animal based on the physiological response characteristic data includes: determining the physiological indicators of the live animal based on the physiological response characteristic region.

[0143] It can be understood that the foreground image is the image of the live animal's body area after removing the background from the original image data. The physiological response feature area is the local area in the foreground image that displays the characteristic color of the response, that is, the direct representation of the physiological response to be monitored in the image.

[0144] This paper presents an automated image analysis algorithm framework based on color information, suitable for the quantification of color responses on body surfaces. It introduces colorimetric analysis into the field of physiological monitoring of live animals, utilizing the color changes of the response itself for label-free detection.

[0145] Optionally, the image data is acquired against a colored background formed by the background enhancement portion, which has a color that provides optical contrast with both the surface color of the live animal and the characteristic color of the physiological response to be monitored. Separating the foreground image containing the live animal from the image data includes: separating the foreground image containing the live animal from the image data based on the background chromaticity features of the background enhancement portion.

[0146] In this embodiment, since the color of the background enhancement portion is known and forms a high contrast with the foreground target, the data processing device can quickly and accurately separate the foreground image by identifying and removing pixels with background color characteristics in the image.

[0147] By utilizing known background color information, the complexity of the foreground separation algorithm is greatly simplified, avoiding complex object detection or semantic segmentation models.

[0148] Specifically, this processing employs a strictly time-sequential two-step chromaticity separation strategy. First, using the known background chromaticity information provided by the background enhancement component, the live animal is completely separated from the complex environment, forming a clean foreground image. Then, only within this foreground image, the chromaticity features specific to the physiological responses being monitored are used for secondary identification and quantification. This sequence of background removal followed by feature extraction is irreversible. Its core purpose is to eliminate interference through a two-stage filtering mechanism: the first stage eliminates noise from non-target areas such as the environmental background and the walls of the monitoring area, significantly narrowing the search range of subsequent algorithms and eliminating artifact interference; the second stage accurately locates weak physiological signals within the clean animal surface area. Attempting to directly identify physiological features in the original image, or performing feature extraction without complete background separation, is highly susceptible to false positives due to background texture, uneven lighting, or mixed colors in the animal's fur, leading to distorted quantization results. Therefore, this two-step method is not only an arrangement of the algorithm flow but also a key logical architecture ensuring high robustness of the system under uncontrolled biological surface imaging conditions.

[0149] Optionally, based on the background chromaticity features of the background enhancement portion, the foreground image containing the live animal is separated from the image data, including: converting the image data to a first color space; generating a live animal region mask using a background chromaticity threshold pre-calibrated according to the color of the background enhancement portion; and separating the foreground image from the image data based on the live animal region mask.

[0150] It can be understood that the first color space is a suitable color space for background color segmentation, such as the HSV (Hue, Saturation, Lightness) color space. The background color threshold is the numerical range used to distinguish background pixels from foreground pixels in the first color space, and this threshold is pre-calibrated based on the color of the background enhancement portion. The live animal region mask is a binary image with the same size as the original image, where the foreground pixel value is 1 (white) and the background pixel value is 0 (black).

[0151] For example, the data processing device first converts the acquired RGB format image data to the HSV color space. In the HSV space, the hue and saturation range of the background pixels are pre-defined as background color chromaticity thresholds based on the color of the background enhancement portion (such as green). Each pixel in the image is iterated over; if its HSV value falls within the background color chromaticity threshold range, it is marked as background (0) in the mask; otherwise, it is marked as foreground (1). Finally, the mask is bitwise ANDed with the original image to obtain a foreground image containing only the live animal.

[0152] The HSV color space was chosen for background segmentation because it is more robust to changes in lighting and has a high degree of separation of hue components. This allows it to separate color information (green background) from brightness information, stably separating a specific color background from any color foreground target.

[0153] In the foreground separation stage, the HSV (Hue, Saturation, Brightness) color space is chosen as the processing basis instead of the original RGB space. This is because the HSV space effectively decouples color information from brightness information, where the hue component H is relatively insensitive to changes in light intensity. In actual monitoring, although the monitoring area is provided with constant illumination, the curved reflection of the live animal's surface, hair occlusion, and slight movements still cause dynamic fluctuations in local brightness. If segmentation is performed using only a fixed grayscale threshold in the RGB space, these brightness changes will be misjudged as color differences, resulting in broken or holed mask edges. In the HSV space, as long as the color tone of the background enhancement part remains stable, even with a certain range of brightness variations, its hue value can still be maintained within the preset threshold range, thus generating a continuous, complete, and smooth-edged animal region mask. After denoising using morphological opening and closing operations, this mask is used as a binary template applied to the original image, which can accurately extract the foreground image containing only the live animal, eliminating pixel interference from the green background or other environmental elements.

[0154] The HSV color space is chosen as the primary color space because it is insensitive to changes in illumination. Based on the known color of the background enhancement portion (e.g., green), the system sets corresponding hue, saturation, and brightness threshold ranges in the HSV space to generate a binarized background mask. By inverting and morphologically filtering this mask, a foreground mask accurately covering the live animal can be obtained. This allows the background pixels in the original image to be zeroed out or replaced, retaining only the animal's surface information. This separation strategy based on background chromaticity features leverages the optical contrast advantage provided by the hardware design, exhibiting higher robustness and lower computational cost compared to traditional edge detection or general semantic segmentation algorithms. Furthermore, the process of separating the foreground image from the image data based on the background chromaticity features of the background enhancement portion transforms physical-level optical enhancement into digital-level signal extraction.

[0155] Optionally, the first color space can also be the Lab color space or the YCbCr color space. For example, when the background enhancement area is green, the pre-defined background color thresholds in the HSV space are: Hue range 40-80, Saturation range 50-255. By having the experimenter select a background area in the software interface, the program automatically calculates the average HSV value and standard deviation of that area, and generates the threshold range accordingly.

[0156] Optionally, based on the chromaticity features of the reaction feature color, the physiological reaction feature region is identified from the foreground image, including: converting the foreground image to a second color space; extracting specific channel components in the second color space; and using a chromaticity threshold of the reaction feature color pre-calibrated according to the reaction feature color to identify the physiological reaction feature region based on the specific channel components.

[0157] It can be understood that the second color space is a color space suitable for feature color recognition and segmentation, such as the Lab color space (CIE L). a b The specific channel component is the channel in the second color space that is most relevant to the color dimension of the reaction feature color. For example, when the reaction feature color is blue, the specific channel is the b channel (blue-yellow axis) in the Lab color space. The reaction feature color threshold is a numerical threshold used to distinguish the reaction feature region from the normal skin region on the specific channel in the second color space. This threshold is pre-calibrated based on the reaction feature color.

[0158] For example, the data processing device converts the separated foreground image from the RGB color space to the Lab color space. The Lab space consists of a luminance channel (L) and two color-opposite channels (a: green-red axis, b: blue-yellow axis). When the characteristic color of the physiological response to be monitored is blue, the b channel component is extracted. Since blue has a negative value in the b channel, by setting a threshold for the b channel (e.g., b < -5), pixels below this threshold are identified as physiological response characteristic regions.

[0159] Choosing the b channel of the Lab color space for blue recognition is a precise selection based on the principles of color science. The Lab color space is designed so that color differences are approximately proportional to the color differences perceived by the human eye, and the b channel specifically encodes blue-yellow axis information, which can purely separate the blue bleed area, unaffected by changes in brightness or the red undertone of the skin.

[0160] The Lab color space is designed based on the uniformity of human visual perception, with its b-channel specifically encoding color contrast information from blue to yellow. Physiological response features exhibiting blue hues, such as Evans blue exudation, show a significant negative response in the b-channel, creating a stark contrast with normal animal skin color (usually pinkish-white or pale yellow, with a positive or near-zero b-value). In contrast, identifying blue in the HSV color space is difficult because blue may be adjacent to certain dark shadows or hair reflections on the color wheel, and saturation is easily affected by lighting, making it hard to set a stable segmentation threshold. The b-channel components of the Lab space are highly specific for blue-yellow differences, capable of purifying weak blue pathological signals from complex surface textures. By using a preset negative b-channel threshold (e.g., b < -5, the specific value can be pre-calibrated according to a standard color chart) to perform threshold segmentation on the foreground image, the system can highly sensitively identify all suspected physiological response areas, effectively overcoming the missed detection problem caused by low contrast in traditional methods. The key to this step is that its processing object is limited to the foreground image generated in the previous step, which reduces the false triggering of feature recognition by residual background noise and improves the accuracy of quantization results.

[0161] Optionally, the second color space can also be HSV or RGB. For example, when the characteristic color of the reaction is blue, the b-channel threshold is set to b < -5 (this can be dynamically adjusted according to the actual image quality). The threshold calibration method is as follows: select known blue reaction region samples in the software, analyze their b-channel distribution, and automatically determine the optimal threshold.

[0162] It should be understood that although the embodiments of this application have been described in detail using HSV space for background separation and Lab space for feature recognition as examples, this is not the only limitation of this application. In other embodiments, if the spectral characteristics of the background color and the feature color change, other suitable color space combinations such as YCrCb and Luv can be selected, or a semantic segmentation model based on machine learning can be used to replace threshold segmentation. As long as they follow the principle of separating the foreground based on background chromaticity and recognizing the target based on feature chromaticity, they should all be considered within the scope of protection of this application.

[0163] In some embodiments, determining physiological indicators of a live animal based on a physiological response feature region includes: determining feature parameters of the physiological response feature region; and determining physiological indicators based on the feature parameters of the physiological response feature region.

[0164] It can be understood that feature parameters are quantifiable attribute values ​​extracted from the physiological response feature regions, used to characterize the size, intensity, or morphology of the region.

[0165] A computable mapping relationship from image pixels to physiological indicators was established. Through the intermediate layer of feature parameters, the original image region attributes can be flexibly transformed into different types of physiological indicators.

[0166] Optionally, the data processing device traverses the identified physiological response feature regions (i.e., pixels marked as response regions in the binary mask), counts their feature parameters, and then converts the feature parameters into indicators with physiological meaning according to a preset mapping relationship.

[0167] Optionally, the feature parameters of the physiological response feature region include the pixel area of ​​the physiological response feature region, and the physiological index includes the reaction area of ​​the physiological response to be monitored, wherein the reaction area is determined based on the pixel area; and / or, the feature parameters of the physiological response feature region include the average color depth of the physiological response feature region, and the physiological index includes the reaction depth of the physiological response to be monitored, wherein the reaction depth is determined based on the average color depth.

[0168] Pixel area refers to the total number of pixels contained in a physiological response feature region. Response area refers to the index obtained by converting pixel area into actual physical area, used to reflect the spatial diffusion range of the physiological response. Response area can be directly represented by pixel area (i.e., the two are equal), or the corresponding response area can be determined by mapping pixel area, or the corresponding response area can be calculated from pixel area (which can be set according to actual needs, such as multiplying the pixel area by a constant).

[0169] Average color depth refers to the average absolute value of all pixels within a physiological reaction characteristic region on a specific color channel (such as the b channel in Lab). It reflects the intensity of the physiological reaction, indicating the concentration of chromogenic substances per unit area or the severity of the reaction, such as the density of dye leakage in a vascular permeability experiment. Reaction depth is a reaction intensity index determined based on average color depth. Reaction depth can be directly characterized by average color depth (i.e., the two are equal), or it can be determined by mapping the average color depth. Alternatively, the reaction depth can be calculated from the average color depth (which can be set according to actual needs, such as multiplying the average color depth by a constant).

[0170] By replacing the subjective, semi-quantitative assessment of traditional manual scoring with objective pixel-level measurements, standardized and automated quantitative analysis of experiments such as the Miles Assay has been achieved. Area and depth indicators can capture subtle changes that are indistinguishable to the human eye, thus improving detection sensitivity.

[0171] Optionally, the reaction area is obtained by statistically analyzing the total number of white pixels in the masked image of the physiological reaction feature region. The reaction depth is obtained by calculating the average absolute value of the b-channel within the physiological reaction feature region. Area and color depth quantify the reaction from two dimensions—range and intensity—respectively, and their combined use allows for a more comprehensive evaluation of the reaction's severity.

[0172] When the physiological reaction to be monitored is an allergic reaction based on Evans blue dye, the characteristic color is blue, the reaction area is the number of blue pixels in the physiological reaction characteristic area, and the reaction depth is the average color depth of the blue pixels in the physiological reaction characteristic area (reflecting the dye concentration).

[0173] In some embodiments, the data processing device is further configured to generate a dynamic curve of the physiological response to be monitored changing over time based on the physiological indicators corresponding to multiple frames of image data.

[0174] By connecting discrete indicator points into a time series curve, the dynamic process of the reaction is fully presented. This achieves a leap from point measurement to line monitoring, providing time-dimensional information that traditional endpoint methods cannot obtain.

[0175] For example, the data processing device performs the above analysis process on each frame of image acquired in real time to obtain the physiological indicators (such as reaction area) corresponding to each frame. In one implementation, the indicator values ​​of each frame are connected to form a curve with time as the horizontal axis and the physiological indicator as the vertical axis, thus generating a dynamic curve of reaction changing over time. In another implementation, the indicator values ​​of each frame are connected to form a curve with time as the horizontal axis and the weighted fusion value of multiple physiological indicators corresponding to each frame of image data as the vertical axis, thus generating a dynamic curve of reaction changing over time.

[0176] Optionally, based on the physiological indicators corresponding to multiple frames of image data, a dynamic curve showing the change of the monitored physiological response over time is generated, including: determining the severity score of the response corresponding to each frame of image data based on the physiological indicators corresponding to each frame of image data; wherein, the severity score of the response is used to reflect the severity of the monitored physiological response corresponding to the current frame of image data; and performing time-series change analysis on the severity scores of the response corresponding to multiple frames of image data to generate a dynamic curve showing the change of the severity score of the response over time.

[0177] It is understandable that the reaction severity score is a comprehensive quantitative indicator. For example, it can be obtained by weighting and fusing multiple individual indicators (such as area and depth) corresponding to each frame of image data, and is used to more comprehensively assess the overall severity of the reaction in each frame of image.

[0178] For example, the formula for calculating the severity score is: Score = α × Area + β × Intensity, where α and β are preset weighting coefficients that can be adjusted by the user according to specific experimental needs. The data processing device calculates the Score value for each frame and then generates a dynamic curve with time on the horizontal axis and the Score value on the vertical axis.

[0179] By combining area and depth metrics into a weighted score, a more comprehensive reflection of the overall severity of the reaction is achieved. A configurable comprehensive evaluation framework is provided, allowing users to adjust the weights according to specific experimental needs.

[0180] For example, the weighting coefficients are set to α = 0.5 and β = 0.5 by default, and users can adjust them through the software interface.

[0181] The data processing device can not only output static indicators from a single frame image, but also perform time-series tracking and analysis of continuously acquired multi-frame image data. By recording the physiological indicator values ​​at each moment, the system can automatically generate dynamic change curves reflecting the entire process of physiological response occurrence, development, peak, and decline. This time-series information is of irreplaceable value for pharmacokinetic studies, drug onset time assessment, and allergic reaction process analysis, overcoming the limitation of traditional endpoint methods that can only obtain data at a single time point. It should be understood that the generation frequency of dynamic change curves depends on the image acquisition frame rate and processing speed, and can be flexibly configured according to experimental needs in practical applications.

[0182] The following describes the method for monitoring physiological indicators of live animals provided in the embodiments of this application.

[0183] The method for monitoring physiological indicators of live animals provided in this application is applied to a data processing device, which is the executing entity of this method. Please refer to [link / reference]. Figure 2 The method includes the following steps S101 to S103.

[0184] Step S101: The data processing device acquires sensor data; wherein, the sensor data is external sensor data collected from live animals used in biological experiments and capable of determining the physiological indicators of the live animals.

[0185] It is understood that the sensor data can be external measurement data from the surface of a live animal that reflects its physiological state, collected by the monitoring module of the aforementioned live animal physiological indicator monitoring device or system.

[0186] In this embodiment, after the live animal is kept in the monitoring area in a stress-free manner by the guidance module, the monitoring module collects sensor data from its body surface. Taking image data as an example, the image acquisition unit in the monitoring module continuously captures images of the live animal's body surface at a fixed frame rate against a color background formed by the background enhancement portion and in a constant light environment provided by the illumination source.

[0187] The key difference between the sensor data obtained in this step and traditional methods is that the data acquisition is carried out when the animal is in a stress-free physiological homeostasis, which fundamentally achieves the non-distortion of the data.

[0188] In step S102, the data processing device determines the physiological response characteristic data of the live animal based on the sensor data; wherein, the physiological response characteristic data is used to reflect the characteristics of the physiological response to be monitored produced by the live animal in the biological experiment.

[0189] Physiological response feature data can be understood as a subset of data extracted from raw sensor data that directly characterizes the physiological response to be monitored. For example, when the sensor data is image data, the physiological response feature data could be pixel regions in the image that display the characteristic colors of the response.

[0190] In this embodiment of the application, taking image data as an example, the data processing device performs color space conversion and chromaticity threshold segmentation on the image data obtained in step S101, and extracts the physiological reaction feature region as physiological reaction feature data.

[0191] Step S103: The data processing device determines the physiological indicators of the live animal based on the physiological response characteristic data.

[0192] In this embodiment of the application, the data processing device performs feature parameter statistics (such as pixel area and average color depth) on the physiological reaction feature region determined in step S102, and converts the feature parameters into quantitative indicators with clear physiological meaning (such as reaction area and reaction depth) according to the preset mapping relationship.

[0193] The above method provides a hardware-independent core data processing workflow, enabling fully automated analysis from raw sensor data to physiological indicators. This standardized data processing workflow facilitates automated and continuous quantitative analysis of physiological indicators in live animals.

[0194] In some embodiments of this application, the sensing data is image data. Determining physiological response characteristic data of a live animal based on the sensing data includes: separating a foreground image containing the live animal from the image data; identifying a physiological response characteristic region from the foreground image based on the chromaticity characteristics of the response characteristic color; wherein, the response characteristic color is the characteristic color of the physiological response to be monitored produced by the live animal, and the physiological response characteristic region is an image of the region containing the response characteristic color. Determining physiological indicators of the live animal based on the physiological response characteristic data includes: determining the physiological indicators of the live animal based on the physiological response characteristic region.

[0195] In some embodiments of this application, the image data is acquired against a colored background formed by a background enhancement portion, the background enhancement portion having a color that forms optical contrast with the surface color of the live animal and the characteristic color of the physiological response to be monitored; separating the foreground image containing the live animal from the image data includes: separating the foreground image containing the live animal from the image data based on the background chromaticity features of the background enhancement portion.

[0196] In some embodiments of this application, separating a foreground image containing a live animal from image data based on the background chromaticity features of the background enhancement portion includes: converting the image data to a first color space; generating a live animal region mask using a background chromaticity threshold pre-calibrated according to the color of the background enhancement portion; and separating the foreground image from the image data based on the live animal region mask.

[0197] In some embodiments of this application, identifying physiological response feature regions from a foreground image based on the chromaticity features of the response feature color includes: converting the foreground image to a second color space; extracting specific channel components in the second color space; and identifying physiological response feature regions based on the specific channel components using a chromaticity threshold of the response feature color pre-calibrated according to the response feature color.

[0198] In some embodiments of this application, determining physiological indicators of a live animal based on physiological response feature regions includes: determining feature parameters of the physiological response feature regions; and determining physiological indicators based on the feature parameters of the physiological response feature regions.

[0199] In some embodiments of this application, the feature parameters of the physiological reaction feature region include the pixel area of ​​the physiological reaction feature region, and the physiological index includes the reaction area of ​​the physiological reaction to be monitored, wherein the reaction area is determined based on the pixel area; and / or, the feature parameters of the physiological reaction feature region include the average color depth of the physiological reaction feature region, and the physiological index includes the reaction depth of the physiological reaction to be monitored, wherein the reaction depth is determined based on the average color depth.

[0200] In some embodiments of this application, the method further includes: generating a dynamic curve of the physiological response to be monitored changing over time based on the physiological indicators corresponding to multiple frames of image data.

[0201] In some embodiments of this application, a dynamic curve of the physiological response to be monitored changing over time is generated based on the physiological indicators corresponding to multiple frames of image data, including: determining the severity score of the response corresponding to each frame of image data based on the physiological indicators corresponding to each frame of image data; wherein the severity score of the response is used to reflect the severity of the physiological response to be monitored corresponding to the current frame of image data; and performing time-series change analysis on the severity scores of the response corresponding to multiple frames of image data to generate a dynamic curve of the severity score of the response changing over time.

[0202] It is understood that the specific implementation principles and technical effects of each step in the above method implementation are the same as the corresponding data processing configuration in the aforementioned system implementation, and will not be repeated here.

[0203] The data processing apparatus provided in the embodiments of this application is described below.

[0204] The data processing apparatus provided in this application includes a memory 401, a processor 400, and a computer program 402 stored in the memory 401 and executable on the processor 400. When the processor 400 executes the computer program, it causes the data processing apparatus to perform any of the methods described in the above method embodiments.

[0205] Computer program 402 can be divided into one or more modules / units. One or more modules / units are stored in memory 401 and executed by processor 400 to complete the method of the embodiments of this application. One or more modules / units can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in a data processing device.

[0206] It should be understood that the embodiments of this application do not limit the specific form of the data processing device. In different application scenarios, the data processing device can be a general computing device such as a personal computer, laptop, or workstation, or it can be a dedicated embedded controller, industrial tablet computer, or even a remote server or cloud computing node connected to the image acquisition unit via a network.

[0207] The data processing apparatus may include, but is not limited to, a processor 400 and a memory 401. Those skilled in the art will understand that... Figure 3 This is merely an example of a data processing device and does not constitute a limitation on the data processing device. It may include more or fewer components than shown, or combine certain components, or different components. For example, a data processing device may also include input / output devices, network access devices, buses, etc.

[0208] The processor 400 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0209] Memory 401 can be an internal storage unit of the data processing device, such as a hard disk or RAM. Memory 401 can also be an external storage device of the data processing device, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, memory 401 can include both internal and external storage units. Memory 401 is used to store computer programs and other programs and data required by the data processing device. Memory 401 can also be used to temporarily store data that has been output or will be output.

[0210] It should be noted that, for the sake of convenience and brevity, the structure of the above-mentioned data processing device can also be referred to the specific description of the structure in the method and system embodiments, which will not be repeated here.

[0211] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0212] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for various specific applications, but such implementations should not be considered beyond the scope of this application.

[0213] The following description is based on specific embodiments.

[0214] Example 1 This embodiment provides a specific application scenario for in situ continuous monitoring of mouse allergic reactions based on the aforementioned system and method, aiming to verify the technical effectiveness of this application in solving the problem of dynamic quantification of in vivo physiological indicators through real experimental cases.

[0215] In this application scenario, the live experimental animal is set as a mouse, and the physiological response to be monitored is an allergic reaction based on Evans blue dye exudation. The predetermined color of the background enhancement portion is configured as green. Specifically, when an allergic reaction occurs, Evans blue dye leaks from blood vessels into the skin tissue and forms blue spots. The area and color depth of the exudation directly characterize the degree of change in vascular permeability and the severity of the allergic reaction. Green is chosen as the color of the background enhancement portion based on the principle of complementary colors in color science: green and the pinkish-white skin (reddish / yellowish tone) commonly found on the mouse body surface, as well as the blue Evans blue exudation spots, maintain a significant distance on the color wheel. This maximizes the optical contrast between the animal subject and the background, and between physiological features and normal skin at the imaging end, providing an ideal physical basis for high-precision automatic segmentation of the subsequent data processing device. It should be understood that although this embodiment uses mice and Evans blue as examples, the monitoring logic of this invention is also applicable to other physiological response models with surface coloration characteristics, such as inflammatory erythema and fluorescent marker tracing. Only the color of the background enhancement portion needs to be adjusted accordingly to maintain a high contrast relationship.

[0216] The automatic monitoring system for allergic reactions on mouse skin provided in this embodiment (a specific example of a live animal physiological indicator monitoring system) has the following structure: Figure 4 . Figure 7 This embodiment demonstrates the overall physical structure of the mouse surface allergic reaction monitoring device (a specific example of a live animal physiological indicator monitoring device). Figure 8 The system setup and experimental operation scenarios were demonstrated.

[0217] Hardware configuration The housing module is constructed from a transparent acrylic sheet, forming an overall square box-like structure. The interior is further defined by transparent acrylic sheets, delineating the initial and monitoring areas. The monitoring area is primarily a channel made of a transparent acrylic square tube, measuring 30cm long, 5cm wide, and 5cm high, serving as the confined space for the mouse. The width of this channel (5cm) is less than the body length of an adult mouse (approximately 8-10cm), creating a channel structure that restricts turning. Once the mouse enters the channel, its body axis remains aligned with the channel direction, preventing free turning and ensuring that the camera always captures images of the mouse's side or back from above.

[0218] Light-shielding components and lighting source: Channel structure (monitoring area, i.e.) Figure 4 The observation area is covered with a double layer of black velvet (i.e., Figure 4The passageway incorporates a black light-blocking cloth, forming a removable, flexible darkroom to isolate stray light interference. This black cloth also serves as a light shield, utilizing the mouse's instinct to seek darkness and avoid light, allowing it to naturally move towards the darker observation area within the passageway, further aiding in positioning. An integrated LED white light strip provides uniform, flicker-free, constant illumination to ensure the consistency and repeatability of acquired images.

[0219] Guiding Module (Temperature Gradient Field): A 10cm × 5cm PTC heating plate is attached to the bottom of one end of the channel structure (initial area). An external temperature controller sets its surface temperature to a constant 65℃, forming a high-temperature driving zone (i.e., Figure 4 The heating plate is covered with orange heat-resistant tape as a visual identifier. The channel structure is a room-temperature zone, i.e., the monitoring area, with a temperature consistent with the ambient temperature (approximately 22-25℃). A 20cm long area at the other end of the channel structure is a room-temperature placement zone, also with a temperature consistent with the ambient temperature (approximately 22-25℃). The high temperature of 65℃ is higher than the mice's thermal comfort zone, effectively stimulating their heat-avoidance behavior and driving them to quickly leave and move towards the channel structure; while the room-temperature zone provides a comfortable environment for the mice to stay. This temperature gradient field design achieves stress-free, spontaneous localization in the mice. Figure 5 A schematic diagram of the thermal avoidance positioning principle is shown.

[0220] Monitoring Module: A green frosted PVC board is laid at the bottom of the ambient temperature observation area to serve as a background enhancement. A Logitech C920 high-definition USB camera is fixed directly above the channel structure, with the lens pointing vertically downwards and the focus adjusted to clearly cover the entire monitoring area. The green background board (background enhancement) creates high optical contrast with the fur color of the white mice and the blue Evans blue exudate spots to be monitored, providing excellent conditions for subsequent image processing.

[0221] Data processing device (i.e.) Figure 4 (Data processing terminal): The camera is connected to a laptop computer with dedicated analysis software installed via a USB cable. This software is the specific implementation of the data processing device.

[0222] System working principle Mice injected with Evans blue dye via their tail veins and sensitized to their ears were placed in a high-temperature repulsion zone. Upon contact of their paws with a 65°C orange heating plate, the mice, driven by their heat-avoidance instinct, immediately turned and ran towards the monitoring area, eventually settling spontaneously in the area covered by a green background. Throughout this process, the mice were not physically restrained or anesthetized, behaved autonomously, and maintained a stable physiological state. A camera continuously captured real-time images of the mice's backs or ears under stable artificial light and transmitted the data to a data processing device.

[0223] Image processing algorithms and data output Please see Figure 6 The software within the data processing device executes the following algorithm steps in real time: Foreground image separation (background culling): Acquire a single frame of RGB image from the video stream. Convert the image from the RGB color space to the HSV color space. Generate a background mask by pre-calibrating the HSV thresholds (e.g., Hue: 40-80, Saturation: 50-255, Value: 50-255) based on the predetermined color of the green background. Obtain a binary mask of the mouse region through an inversion operation, and use morphological opening operations to remove minor noise, thereby obtaining a clean mouse foreground image.

[0224] Physiological Response Feature Recognition and Quantification: The mouse foreground image region was converted from RGB to Lab color space. In Lab space, the b-channel represents the color range from blue to yellow. Since Evans blue bleed spots are blue, they have a significant negative response on the b-channel. A pre-calibrated b-channel threshold (e.g., b < -5 or dynamically adjusted according to image quality) was determined based on the feature color to be monitored (blue). Pixels below this threshold were identified as blue bleed areas (i.e., regions of interest, ROIs). The total number of pixels identified as blue bleed areas was calculated and denoted as the diffusion area. The average of the absolute values ​​of the b-channel values ​​of these blue pixels (or converted to a blue depth index) was calculated and denoted as the blue intensity.

[0225] Dynamic curve generation: The software records the area and depth indicators at a preset frequency (e.g., once per second) and calculates the severity index (Score = α × Area + β × Intensity, where α and β are preset weighting coefficients) as needed. Finally, the software generates and displays dynamic curves showing the changes of the above indicators over time until the reaction subsides, achieving a complete record of the entire allergic reaction process. Figure 9 and Figure 10 This demonstrates the automatic identification and quantification results of the blue exudate area on the mouse body surface using this system. Figure 9 and Figure 10 The yellow-marked areas represent the automatically identified regions of interest (ROIs) for indigo staining. The area is obtained by counting the total number of white pixels in the mask image and is used to characterize the reaction range. The depth is calculated by averaging the color channels within the blue area; higher saturation indicates a deeper blue and a stronger reaction. The Severity Score is a weighted fusion of area and depth to comprehensively quantify the degree of reaction; the formula is Score = α × Area + β × Intensity.

[0226] Comparative example: To verify the technical effectiveness of this invention, the traditional method of taking photos and manually scoring after anesthesia and restraint was used as a comparison. Traditional methods typically require anesthetizing or physically fixing the mice before acquiring images under external lighting conditions, and then relying on human observation or scoring. In contrast, this invention uses the system described in Example 1, which achieves automatic monitoring and quantitative analysis of surface reactions through animal self-localization, background enhancement imaging, and automatic machine vision recognition.

[0227] stress level Traditional anesthesia or physical restraint methods often interfere with the normal behavior of animals and may further affect their physiological performance during testing. In contrast, this invention utilizes the spontaneous behavioral tropism of laboratory animals to achieve localization, eliminating the need for forced restraint. This facilitates monitoring with minimal human intervention, thereby improving the accuracy and reliability of test results.

[0228] Data accuracy and variability Traditional manual observation or scoring methods rely heavily on operator experience and subjective judgment, which can lead to variations between different operators and between different observations. In contrast, this invention automatically identifies and quantifies the target region under relatively uniform imaging conditions based on a preset image processing workflow. This reduces the impact of human factors on the results and helps improve the consistency, repeatability, and automation level of the detection results.

[0229] Dynamic information The comparative group could only obtain static data at the moment of anesthesia and restraint, making it impossible to determine whether the response was in the rising or declining phase. In contrast, the system of this invention can successfully record the complete dynamic process of each mouse's response from its onset, enhancement to its decline, providing rich pharmacodynamic information including the time dimension.

[0230] Automation level and detection efficiency This system employs a fully automated design, eliminating the need for human intervention throughout the entire process from animal localization and image acquisition to feature recognition and quantification. The complete detection process for a single mouse (from placement in the channel to outputting quantitative indicators) takes approximately 10 seconds, while traditional methods (including animal restraint, manual observation, photography, and manual scoring) typically take more than ten minutes. Furthermore, this system supports batch continuous operation, allowing for the continuous monitoring of multiple mice, significantly improving experimental throughput and data output efficiency.

[0231] In summary, this embodiment successfully achieves in-situ continuous monitoring with high temporal resolution, without anesthesia or restraint, by organically integrating stress-free heat avoidance localization, green screen optical enhancement, and dual-color machine vision algorithms into a mouse allergic reaction monitoring scenario. Compared with traditional manual scoring or anesthesia photography methods, this application not only eliminates subjective bias and stress artifacts but also improves the detection efficiency of a single mouse to the second level and supports batch continuous operation, providing an innovative technical means for immunological and pharmacological research that combines data authenticity, richness, and high throughput. It should be understood that the specific temperature values, color space thresholds, and scoring formulas mentioned in this embodiment are only preferred examples for this specific application scenario. In actual applications, they can be adaptively adjusted according to experimental needs and animal strain characteristics without departing from the scope of protection of this application.

[0232] Those skilled in the art will understand that the specific parameters given in Example 1, such as the HSV threshold (Hue 40-80) and b-channel threshold (b<-5), are preferred examples for specific experimental conditions (green background, specific camera model, fixed lighting). In practical applications, the above thresholds can be pre-calibrated and adjusted according to different background colors, light intensities, animal breed coat colors, and other factors. The data processing module should provide a user interface, allowing users to calibrate the algorithm parameters before starting the experiment by selecting a background area and a standard color chart, to ensure the system's adaptability and stability under different environments. This calibration process is a conventional technique in the art.

[0233] Compared with the prior art, the embodiments of this application have the following beneficial effects: This invention pioneers a "behavior-imaging" collaborative localization mechanism, fundamentally eliminating stress interference. Unlike traditional physical restraint or anesthesia, this invention is the first to propose utilizing spontaneous animal behavioral tendencies (such as heat avoidance) as a "virtual restraint" method. This not only achieves stress-free localization, but more importantly, it precisely couples the animal's "spontaneous behavior" with the "monitoring field of view," ensuring that each data acquisition is conducted under conditions of physiological homeostasis and consistent imaging conditions. This design fundamentally solves the problem of physiological index fluctuations caused by animal restraint, laying the physical foundation for obtaining real and reliable dynamic monitoring data. The guidance module of this invention is not an independent functional unit, but rather a prerequisite and guarantee for the entire monitoring chain.

[0234] The "background-feature" optical enhancement design overcomes the challenges of image recognition. This invention significantly enhances the distinguishability of the animal subject from the target features (such as blue spots) at the imaging source by setting a high-contrast background (e.g., a green screen) in the monitoring area, utilizing optical principles. This design eliminates the need for subsequent image processing algorithms to handle complex and varied natural backgrounds; simple thresholding within a stable color space is sufficient to obtain high-precision foreground and target features, significantly reducing algorithm complexity and hardware requirements, and improving system robustness and repeatability. The background enhancement component is not an optional accessory but a key component that works collaboratively with the image acquisition unit and data processing unit to form an efficient and stable "optical-algorithm" detection pathway.

[0235] This invention represents a leap from "point measurement" to "line monitoring," providing fully automated dynamic quantitative analysis. By combining the aforementioned "stress-free localization" with "high-contrast imaging," this invention achieves, for the first time, uninterrupted, high-temporal-resolution in vivo monitoring of the entire reaction process in experiments such as the Miles Assay. The system integrates fully automated algorithms from image acquisition, foreground extraction, feature recognition to index quantification, enabling real-time output of quantitative indicators such as reaction area and color depth, along with their dynamic change curves. This provides unprecedentedly rich data for pharmacological and immunological research, with technical effects far exceeding the simple summation of endpoint methods or manual scoring. This invention organically integrates three seemingly independent technical means—animal behavior control, optical imaging design, and machine vision analysis—creating a synergistic effect that collectively solves a long-standing comprehensive technical challenge in this field.

[0236] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, several improvements and modifications can be made without departing from the spirit and principle of the present invention, such as replacing thermal aversion with photoaversion or electroaversion, or replacing the detection target from allergic reactions with other physiological changes on the body surface (such as wound healing, tumor growth, etc.). These improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A device for monitoring physiological indicators of live animals, characterized in that, include: A containment module, the containment module including an initial area for containing live animals and a monitoring area connected to the initial area; A guidance module is configured to guide the live animal to move automatically from the initial area to the monitoring area by utilizing the spontaneous behavioral tendency of the live animal; as well as A monitoring module is configured to collect sensor data of the live animal located within the monitoring area, the sensor data being capable of determining the physiological indicators of the live animal.

2. The apparatus according to claim 1, characterized in that, The guiding module is configured to form a stimulus gradient field along the initial region to the monitoring region, the stimulus gradient field being used to drive the live animal to move automatically from the initial region to the monitoring region; Preferably, the stimulus gradient field is a temperature gradient field, and the temperature of the initial region is higher than the temperature of the monitoring region; Preferably, the guiding module includes a heating element disposed in the initial area for heating the initial area; more preferably, the guiding module further includes a temperature controller electrically connected to the heating element; even more preferably, the heating element is a heating plate disposed at the bottom of the initial area. Preferably, the temperature range of the initial area is 50℃~70℃, and the temperature range of the monitoring area is 20℃~30℃; Additionally, the stimulus gradient field is a light intensity gradient field, and the light intensity of the initial region is greater than the light intensity of the monitoring region; Additionally, the stimulus gradient field is an odor concentration gradient field, and the odor concentration in the initial region is greater than the odor concentration in the monitoring region.

3. The apparatus according to claim 1, characterized in that, The monitoring area is configured as a channel structure that restricts the live animal from turning around, and the width of the channel structure is less than the body length of the live animal; Preferably, the active area of ​​the initial region is larger than the active area of ​​the monitoring region; Preferably, the monitoring area is located on one side of the initial area, and the initial area and the monitoring area are linearly arranged; preferably, a narrowing area is formed on the side of the initial area facing the monitoring area, and the width of the narrowing area gradually decreases along the direction from the initial area to the monitoring area; more preferably, the inner sidewall of the initial area includes a first sidewall and a second sidewall disposed opposite to each other, the first sidewall and the second sidewall respectively extending obliquely to the monitoring area, and the narrowing area is formed between the first sidewall and the second sidewall. Preferably, the containing module further includes a retention area located on the side of the monitoring area opposite to the initial area, for containing a live animal after the sensor data is collected in the monitoring area.

4. The apparatus according to any one of claims 1 to 3, characterized in that, The monitoring module includes an image acquisition unit, which is used to acquire image data of the live animal located within the monitoring area, and the sensor data includes the image data; Preferably, the sensing data is real-time sensing data, and the image data is real-time image data; Preferably, the image acquisition unit is used to acquire image data of the body surface of the live animal located within the monitoring area. The image data can determine the physiological response characteristics of the body surface of the live animal, and thus determine the physiological indicators of the live animal. Preferably, the image acquisition unit is positioned above the monitoring area for acquiring image data from a top-down perspective; Preferably, the monitoring module further includes a background enhancement section, which is disposed in the monitoring area and has a color that forms an optical contrast with the surface color of the live animal and the characteristic color of the physiological response to be monitored; the image acquisition unit is able to acquire the image data against the color background formed by the background enhancement section. Preferably, the background enhancement portion is disposed at the bottom and / or side of the monitoring area; Preferably, the surface of the background enhancement portion serving as the background is a matte surface, and / or the background enhancement portion is a background panel; Preferably, the live animal physiological indicator monitoring device further includes a light-shielding part that covers the monitoring area; Preferably, the monitoring module further includes a lighting source, which provides illumination for the image acquisition unit to acquire the image data; Preferably, the live animal is a white mouse, the physiological reaction to be monitored is an allergic reaction based on Evans blue dye, the characteristic color is blue, and the color of the background enhancement portion is green.

5. A monitoring system for physiological indicators of live animals, characterized in that, include: The live animal physiological indicator monitoring device as described in any one of claims 1 to 4; as well as A data processing device electrically connected to the monitoring module, the data processing device being configured to determine the physiological indicators of the live animal based on the sensor data.

6. The system according to claim 5, characterized in that, The data processing device is configured to: Acquire the sensor data collected by the monitoring module; Physiological response characteristic data of the live animal are determined based on the sensor data; wherein, the physiological response characteristic data is used to reflect the characteristics of the physiological response to be monitored produced by the live animal in the biological experiment. The physiological indicators of the live animal are determined based on the physiological response characteristic data; Preferably, the sensing data is image data, and the step of determining the physiological response characteristics of the live animal based on the sensing data includes: Separate the foreground image containing the live animal from the image data; Based on the chromaticity features of the reaction feature color, a physiological reaction feature region is identified from the foreground image; wherein, the reaction feature color is the feature color of the physiological reaction to be monitored produced by the live animal, and the physiological reaction feature region is an image of the region containing the reaction feature color; The determination of physiological indicators of the live animal based on the physiological response characteristic data includes: The physiological indicators of the live animal are determined based on the physiological response characteristic regions. Preferably, the image data is acquired against a colored background formed by a background enhancement portion, the background enhancement portion having a color that provides optical contrast with both the surface color of the live animal and the characteristic color of the physiological response to be monitored; the step of separating the foreground image containing the live animal from the image data includes: Based on the background chromaticity features of the background enhancement portion, the foreground image containing the live animal is separated from the image data; Preferably, separating the foreground image containing the live animal from the image data based on the background chromaticity features of the background enhancement portion includes: Convert the image data to a first color space; A live animal region mask is generated using a background color chromaticity threshold pre-calibrated based on the color of the background enhancement portion; The foreground image is separated from the image data based on the live animal region mask; Preferably, the identification of physiological response feature regions from the foreground image based on the chromaticity features of the response feature colors includes: Convert the foreground image to a second color space; Extract specific channel components from the second color space; The physiological response feature region is identified based on the specific channel component by using a pre-calibrated chromaticity threshold of the response feature color; Preferably, determining the physiological indicators of the live animal based on the physiological response characteristic region includes: Determine the characteristic parameters of the physiological response region; The physiological indicators are determined based on the characteristic parameters of the physiological response feature region; More preferably, the feature parameters of the physiological response feature region include the pixel area of ​​the physiological response feature region, the physiological index includes the reaction area of ​​the physiological response to be monitored, and the reaction area is determined based on the pixel area; and / or, the feature parameters of the physiological response feature region include the average color depth of the physiological response feature region, the physiological index includes the reaction depth of the physiological response to be monitored, and the reaction depth is determined based on the average color depth; Preferably, the data processing device is further configured to: generate a dynamic curve of the physiological response to be monitored changing over time based on the physiological indicators corresponding to the multiple frames of image data; Preferably, generating a dynamic curve of the physiological response to be monitored changing over time based on the physiological indicators corresponding to multiple frames of the image data includes: The severity score of the reaction corresponding to each frame of image data is determined based on the physiological indicators corresponding to each frame of image data; wherein, the severity score of the reaction is used to reflect the severity of the physiological reaction to be monitored corresponding to the current frame of image data; A time-series change analysis is performed on the reaction severity scores corresponding to multiple frames of image data to generate a dynamic curve of the reaction severity scores changing over time.

7. A method for monitoring physiological indicators in live animals, characterized in that, Applied to a data processing apparatus, the method includes: The data processing device acquires sensor data; wherein, the sensor data is external sensor data collected from live animals used in biological experiments and capable of determining the physiological indicators of the live animals. The data processing device determines the physiological response characteristic data of the live animal based on the sensor data; wherein, the physiological response characteristic data is used to reflect the characteristics of the monitored physiological response produced by the live animal in the biological experiment. The data processing device determines the physiological indicators of the live animal based on the physiological response characteristic data.

8. The method according to claim 7, characterized in that, The sensing data is image data, and the step of determining the physiological response characteristics of the live animal based on the sensing data includes: Separate the foreground image containing the live animal from the image data; Based on the chromaticity features of the reaction feature color, a physiological reaction feature region is identified from the foreground image; wherein, the reaction feature color is the feature color of the physiological reaction to be monitored produced by the live animal, and the physiological reaction feature region is an image of the region containing the reaction feature color; The determination of physiological indicators of the live animal based on the physiological response characteristic data includes: The physiological indicators of the live animal are determined based on the physiological response characteristic regions. Preferably, the image data is acquired against a colored background formed by a background enhancement portion, the background enhancement portion having a color that provides optical contrast with both the surface color of the live animal and the characteristic color of the physiological response to be monitored; the step of separating the foreground image containing the live animal from the image data includes: Based on the background chromaticity features of the background enhancement portion, the foreground image containing the live animal is separated from the image data; Preferably, separating the foreground image containing the live animal from the image data based on the background chromaticity features of the background enhancement portion includes: Convert the image data to a first color space; A live animal region mask is generated using a background color chromaticity threshold pre-calibrated based on the color of the background enhancement portion; The foreground image is separated from the image data based on the live animal region mask; Preferably, the identification of physiological response feature regions from the foreground image based on the chromaticity features of the response feature colors includes: Convert the foreground image to a second color space; Extract specific channel components from the second color space; The physiological response feature region is identified based on the specific channel component by using a pre-calibrated chromaticity threshold of the response feature color; Preferably, determining the physiological indicators of the live animal based on the physiological response characteristic region includes: Determine the characteristic parameters of the physiological response region; The physiological indicators are determined based on the characteristic parameters of the physiological response feature region; More preferably, the feature parameters of the physiological response feature region include the pixel area of ​​the physiological response feature region, the physiological index includes the reaction area of ​​the physiological response to be monitored, and the reaction area is determined based on the pixel area; and / or, the feature parameters of the physiological response feature region include the average color depth of the physiological response feature region, the physiological index includes the reaction depth of the physiological response to be monitored, and the reaction depth is determined based on the average color depth.

9. The method according to claim 8, characterized in that, The method further includes: generating a dynamic curve of the physiological response to be monitored changing over time based on the physiological indicators corresponding to the multiple frames of image data; Preferably, generating a dynamic curve of the physiological response to be monitored changing over time based on the physiological indicators corresponding to multiple frames of the image data includes: The severity score of the reaction corresponding to each frame of image data is determined based on the physiological indicators corresponding to each frame of image data; wherein, the severity score of the reaction is used to reflect the severity of the physiological reaction to be monitored corresponding to the current frame of image data; A time-series change analysis is performed on the reaction severity scores corresponding to multiple frames of image data to generate a dynamic curve of the reaction severity scores changing over time.

10. A data processing apparatus, characterized in that, The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it causes the data processing apparatus to perform the method as described in any one of claims 7 to 9.

Images