Intelligent ward body temperature monitoring method, system, equipment and medium
The intelligent ward temperature monitoring system uses thermal imaging and user location data to identify hot areas and perform personalized anomaly detection, solving the problems of low frequency, large error, and cross-infection risk of traditional ward temperature monitoring, and achieving continuous, accurate temperature monitoring and rapid response.
Patent Information
- Application Number
- CN202511654659.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-12
- Publication Date
- 2026-02-13
AI Technical Summary
Traditional methods of temperature monitoring in wards rely on manual inspections, which result in low monitoring frequency, discontinuous data, easy omissions or misrecording, and the risk of cross-infection. They are difficult to meet the needs of modern smart wards for real-time monitoring, accuracy, and safety.
By acquiring thermodynamic images of the ward and user location data, hot areas are identified and coordinate information is extracted. Combined with basic user information, personalized abnormal body temperature discrimination rules are retrieved from a preset rule base to detect abnormalities in the body temperature data sequence and generate alarm signals.
It enables non-contact, continuous body temperature monitoring, improves the accuracy of judgment, reduces the risk of missed and false reports, and optimizes the efficiency of ward management.
Smart Images
Figure CN121512466A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of medical monitoring technology, and in particular relates to a method, system, device and medium for intelligent ward temperature monitoring. Background Technology
[0002] With the continuous development of medical monitoring technology, especially the integration and application of Internet of Things and artificial intelligence technologies, intelligent body temperature monitoring technology based on thermal imaging has emerged. This technology achieves large-area, continuous body temperature screening in a non-contact manner, and has the characteristics of high efficiency, real-time operation and low risk of infection.
[0003] Currently, common methods of temperature monitoring in wards include: relying on medical staff to manually measure body temperature periodically using electronic thermometers, or deploying a small number of fixed temperature measurement points for localized monitoring. Traditional techniques typically involve manual patrols, with nurses using handheld ear or forehead thermometers to measure patients' temperatures and manually entering the data into the system. This process is cumbersome and reliant on manpower. Current traditional methods suffer from low monitoring frequency, discontinuous data, a high risk of missed or incorrect readings, high human resource costs, and the potential to increase the risk of cross-infection in environments with high rates of infectious diseases. These methods fail to meet the demands of modern smart wards for real-time monitoring, accuracy, and security. Summary of the Invention
[0004] Therefore, it is necessary to provide a smart ward temperature monitoring method, system, equipment, and medium that can solve the above problems.
[0005] Firstly, this application provides a method for intelligent ward temperature monitoring, including:
[0006] Acquire thermodynamic images of the ward, user location data, and basic user information, including identity information, age information, and basic medical history information;
[0007] Identify thermal regions in a thermodynamic image and extract the coordinate information of each thermal region in the thermodynamic image;
[0008] The coordinate information of each thermal area image is matched with the user's location data to obtain the user's thermal image;
[0009] Extract the thermal center temperature value from the user's thermal image to form a body temperature data sequence;
[0010] Based on the user's basic information, the corresponding abnormal body temperature judgment rule is retrieved from the preset abnormal body temperature judgment rule library;
[0011] Anomaly detection was performed on the body temperature data sequence using body temperature anomaly discrimination rules to obtain body temperature detection results;
[0012] If the body temperature detection result is abnormal, an abnormal body temperature alarm signal will be generated by combining the body temperature detection result, user location data and user basic information.
[0013] In one embodiment, identifying thermal region images in a thermodynamic image and extracting the coordinate information of each thermal region image in the thermodynamic image includes:
[0014] Based on the range of thermal radiation intensity corresponding to human body temperature, a thermal area screening threshold is set.
[0015] Based on the thermal region screening threshold, the thermodynamic image is segmented into pixels to obtain the set of thermal region pixels;
[0016] Connectivity analysis is performed on the set of hot region pixels to group adjacent hot region pixels into the same hot region, forming multiple hot region images;
[0017] Using the preset coordinate system of the thermodynamic image as a reference, traverse all pixels of each thermal region image and extract the horizontal and vertical coordinate values of the pixels in the thermodynamic image.
[0018] Based on the x-coordinate and y-coordinate values, calculate the minimum x-coordinate, maximum x-coordinate, minimum y-coordinate, and maximum y-coordinate corresponding to each thermal region image;
[0019] The minimum and maximum x-coordinates, minimum and maximum y-coordinates are combined to form the coordinate range of the hot region as coordinate information.
[0020] In one embodiment, the coordinate information of each thermal region image is matched with user location data to obtain a user thermal image, including:
[0021] Extract the location coordinates of each user from the user location data to obtain the user coordinate points;
[0022] By determining whether the user's coordinates fall within the coordinate range of the hot zone, a point-to-region matching result is generated.
[0023] The conflict scenarios in the matching results of identification points and regions are obtained, and the conflict scenario identification results are obtained. The conflict scenarios include a single user coordinate point matching multiple hot area images, multiple user coordinate points matching a single hot area image, and multiple user coordinate points matching multiple hot area images within a preset coordinate range threshold.
[0024] Based on the conflict scene identification results, the point and region matching results of the conflict scene are identified. Conflict processing is performed according to preset conflict processing rules to obtain the matching results after conflict processing:
[0025] Based on the conflict scene identification results, the matching results of points and regions that did not identify conflict scenes or the matching results after conflict processing are used to determine the user's thermal image.
[0026] In one embodiment, a pre-defined body temperature abnormality discrimination rule library stores body temperature abnormality discrimination rules corresponding to different age ranges, different disease types, and different combinations of age ranges and different disease types, and each body temperature abnormality discrimination rule is associated with a unique user feature combination tag.
[0027] Based on user information, the system retrieves the corresponding abnormal body temperature criteria from a pre-defined rule base, including:
[0028] Based on each user's basic information, the age information is divided into intervals to obtain age interval identifiers, and the basic medical history information is classified by disease type to obtain disease type identifiers;
[0029] The identity identifiers of each user's basic information are associated with the corresponding age range identifiers and disease type identifiers to form user feature combination tags;
[0030] Based on user characteristic combination tags, the abnormal body temperature discrimination rules are obtained by matching them in the preset abnormal body temperature discrimination rule library.
[0031] In one embodiment, an abnormal body temperature detection rule is used to detect anomalies in the body temperature data sequence, using the following formula:
[0032]
[0033] Where T = (t1, t2, ..., t n ) represents a body temperature data sequence, t i Let L represent the body temperature value measured in the sequence at the i-th time, and let θ represent the user feature combination label. L The threshold parameter for judging abnormal body temperature is the combination tag L of the user feature in the preset abnormal body temperature judgment rule base. I is an indicator function, which takes the value of 1 when the condition is met and 0 otherwise. f(T, L) is an abnormality detection function, which outputs a value of 1 to indicate abnormal body temperature and an output value of 0 to indicate normal body temperature.
[0034] In one embodiment, if the body temperature detection result is abnormal, a body temperature abnormality alarm signal is generated by combining the body temperature detection result, user location data, and user basic information, including:
[0035] Extract abnormal body temperature values, the onset time of abnormal body temperature, and the duration of abnormal body temperature from the body temperature detection results;
[0036] Extract the location coordinates and the ward area identifier where the location coordinates are located from the user's location data;
[0037] Based on the preset alarm level determination rules, abnormal body temperature values, age information, and basic medical history information are correlated and matched to determine the corresponding alarm level;
[0038] The system calls a preset alarm signal template and generates an abnormal body temperature alarm signal based on the abnormal body temperature value, the start time of the abnormal body temperature, the duration of the abnormal body temperature, the location coordinates, the ward area identifier, the identity identifier, the age range identifier, the underlying disease type identifier, and the alarm level.
[0039] In one embodiment, the method further includes:
[0040] Obtain body temperature detection results under the same user feature combination label. The same user feature combination label is a combination of user feature labels that contain the same age range identifier and disease type identifier.
[0041] Based on the body temperature detection results under the same user characteristic combination tags, analyze the differences in body temperature characteristics under the same user characteristic combination tags;
[0042] Based on the differences in body temperature characteristics, the discrimination parameters of the corresponding body temperature abnormality discrimination rules in the preset body temperature abnormality discrimination rule library are adjusted to obtain the updated body temperature abnormality discrimination rules. The updated body temperature abnormality discrimination rules are used for subsequent anomaly detection of body temperature data sequences.
[0043] Secondly, this application also provides an intelligent ward temperature monitoring system, comprising:
[0044] The monitoring data acquisition module is used to acquire thermodynamic images of the ward, user location data, and basic user information, including identity information, age information, and basic medical history information.
[0045] The thermal region identification module is used to identify thermal region images in a thermodynamic image and extract the coordinate information of each thermal region image in the thermodynamic image.
[0046] The region-user matching module is used to match the coordinate information of each thermal region image with user location data to obtain the user's thermal image;
[0047] The body temperature data extraction module is used to extract the thermal center temperature value of the user's thermal image and form a body temperature data sequence;
[0048] The discrimination rule matching module is used to retrieve the corresponding abnormal body temperature discrimination rule from the preset abnormal body temperature discrimination rule library based on the user's basic information;
[0049] The abnormal body temperature detection module is used to perform abnormal detection on the body temperature data sequence using abnormal body temperature discrimination rules, and obtain the body temperature detection result.
[0050] The alarm signal generation module is used to generate an abnormal body temperature alarm signal by combining the body temperature detection result, user location data, and user basic information if the body temperature detection result is abnormal.
[0051] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-mentioned intelligent ward temperature monitoring method.
[0052] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described intelligent ward temperature monitoring method.
[0053] The aforementioned intelligent ward temperature monitoring method, system, equipment, and medium achieve non-contact temperature data acquisition by acquiring thermodynamic images of the ward, user positioning data, and user basic information. It obtains user thermal images by identifying thermal regions in the thermodynamic images and extracting their coordinate information, matching them with user positioning data to establish a correlation between temperature data and specific users. By extracting the thermal center temperature value from the user's thermal image, it forms a temperature data sequence, enabling continuous and real-time temperature monitoring. Based on user basic information, it retrieves personalized temperature anomaly detection rules from a preset rule base to detect anomalies in the data sequence, improving accuracy and reducing the risk of missed and false alarms. When an abnormal detection result is detected, it generates an alarm signal by combining the detection result, positioning data, and basic information, enabling timely response and optimizing ward management efficiency. Attached Figure Description
[0054] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying 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.
[0055] Figure 1 This is a flowchart of a smart ward temperature monitoring method according to the present invention;
[0056] Figure 2 This is a structural diagram of an intelligent ward temperature monitoring system according to the present invention;
[0057] Figure 3 This is a structural diagram of an intelligent ward temperature monitoring system, as shown in one embodiment of the present invention. Detailed Implementation
[0058] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.
[0059] In one embodiment, such as Figure 1 As shown, a smart ward temperature monitoring method is provided. This embodiment illustrates the method's application to a terminal, but it can also be applied to a server or a system including both a terminal and a server, and implemented through interaction between the terminal and the server. The implementation environment includes a thermal imaging camera deployed in the ward, a user positioning device (such as an RFID reader), a backend server storing a pre-set rule base for abnormal temperature detection, and a smart terminal held by medical staff. Application scenarios include: when non-contact, continuous, and accurate monitoring of patient temperature is required to reduce the risk of cross-infection and improve management efficiency, the thermal imaging camera acquires thermodynamic images, the positioning device obtains user positioning data, and both, along with pre-stored user basic information, are synchronously uploaded to the server. After thermal area identification, coordinate matching, temperature extraction, and personalized abnormality detection, the server pushes an abnormality alarm signal to the medical staff terminal for rapid response. In this embodiment, the method includes the following steps:
[0060] S01, acquire the thermodynamic images of the ward, user location data, and basic user information, including identity information, age information, and basic medical history information.
[0061] Thermodynamic images can be captured using thermal imaging devices (such as infrared cameras) to reflect visualized data of temperature distribution. User location data can be obtained in real time through wireless positioning technologies (such as RFID, UWB, or Bluetooth beacons). User basic information includes identification information (such as patient ID or electronic tags), age information, and basic medical history information (such as chronic disease records), which can be automatically collected through IoT devices or retrieved from pre-stored medical databases to provide basic data support for subsequent body temperature monitoring.
[0062] S02, identify the thermal region images in the thermodynamic image and extract the coordinate information of each thermal region image in the thermodynamic image.
[0063] Specifically, a thermal region screening threshold (e.g., a temperature range of 35.5℃-40.0℃) can be set based on the range of thermal radiation intensity corresponding to human body temperature. The thermodynamic image is segmented at the pixel level using image processing algorithms to filter out a set of thermal region pixels that meet the threshold range. Spatially adjacent thermal pixels are clustered into independent thermal region images (e.g., high-temperature areas such as the forehead and neck of the human body) using connected component analysis technology. Each thermal region image represents a potential user heat source. Based on the preset coordinate system of the thermodynamic image, all pixels of each thermal region image are traversed. The thermal region coordinate range (e.g., a rectangular bounding box) is generated by calculating the extreme coordinates (minimum / maximum horizontal and vertical coordinates) of the pixel points. The thermal radiation signal is transformed into structured data with spatial location information, providing a basis for subsequent user matching.
[0064] S03, match the coordinate information of each thermal area image with the user's location data to obtain the user's thermal image.
[0065] The coordinate information of the thermal area image is the range of thermal area coordinates (such as the coordinate range of a rectangle) extracted through image processing. By judging whether the user's coordinates fall within the range of thermal area coordinates in the user's location data, and in scenarios such as multiple users overlapping or areas intersecting, a conflict handling mechanism (such as the nearest neighbor algorithm or priority rules) can be used to handle conflicts, so as to associate thermal radiation data with specific users and form a user thermal image.
[0066] S04, extract the thermal center temperature value of the user's thermal image to form a body temperature data sequence.
[0067] The thermal center temperature value is a representative body temperature data point (such as the highest temperature point, regional average temperature, or weighted thermonuclear temperature) determined through thermal imaging analysis. The body temperature data sequence is a collection of continuous temperature measurements arranged in chronological order. The thermal center temperature value can be automatically calculated using temperature extraction algorithms (such as extreme value detection based on pixel value statistics or regional temperature field fitting algorithms). A data caching mechanism integrates temperature values from different time points into a time-series data sequence by timestamp, enabling continuous monitoring of body temperature changes. In implementation, temperature calibration parameters (such as ambient temperature compensation) can be used to correct the original thermal radiation value, and data smoothing processing (such as moving average filtering) can be used to eliminate instantaneous fluctuations, forming a stable and reliable body temperature data sequence, providing a data foundation for subsequent anomaly detection.
[0068] S05, based on the user's basic information, retrieve the corresponding abnormal body temperature judgment rule from the preset abnormal body temperature judgment rule library.
[0069] The system includes a pre-stored rule base for abnormal body temperature detection, containing rules associated with different combinations of user characteristic tags. By extracting and tagging user information (e.g., mapping age to age range identifiers, and classifying medical history into disease type identifiers), and associating these identifiers with user identity tags to form unique user characteristic combination tags, the system can retrieve corresponding abnormal body temperature detection rules from the rule base through a tag matching mechanism, enabling dynamic invocation of personalized detection rules. In implementation, the rule base can be stored in a database or configuration file, and efficient rule matching can be achieved through a query interface. This ensures that different physiological characteristics (e.g., the elderly, cardiovascular disease patients) are treated with different abnormal body temperature judgment standards, improving monitoring accuracy.
[0070] S06, use the body temperature abnormality discrimination rule to perform abnormality detection on the body temperature data sequence and obtain the body temperature detection result.
[0071] The abnormal body temperature discrimination rule is a predefined detection logic (including threshold comparison, statistical model or machine learning algorithm) based on user feature combination labels (such as age range and disease type). The body temperature data sequence is a set of user thermal core temperature values arranged in chronological order. The data sequence can be analyzed in real time through anomaly detection functions (such as indicator functions based on preset threshold parameters, sliding window analysis or time-series pattern recognition algorithms) to calculate abnormal indicators (such as temperature deviation, duration or trend) and output binary or graded detection results (such as normal, abnormal or alarm level) to achieve automated and personalized abnormal body temperature judgment.
[0072] S07. If the body temperature detection result is abnormal, then combine the body temperature detection result, user location data and user basic information to generate an abnormal body temperature alarm signal.
[0073] The abnormal body temperature alarm signal is a structured alarm output integrating multi-source data, including dynamic detection results such as abnormal body temperature value, onset time of abnormal body temperature, and duration of abnormal body temperature; location coordinates and ward area identifiers from user location data; and identity identifiers, age range identifiers, and underlying disease type identifiers from user basic information. This multi-dimensional information can be correlated and integrated using data fusion technology (such as information aggregation algorithms or event triggering mechanisms). The alarm level is determined based on preset alarm level judgment rules (such as dynamically classifying alarm levels according to body temperature deviation, user risk level, or clinical priority). A composite signal containing timestamps, location information, user characteristics, and alarm level is generated by calling preset alarm signal templates (such as standardized data formats or configurable message structures), achieving automated and precise abnormal response. In implementation, input data can be processed in real time using software modules (such as alarm generators or message queues), and alarm signals can be pushed to medical staff terminals or monitoring systems through communication interfaces (such as network protocols or API calls) to ensure timely intervention and optimized ward management.
[0074] The aforementioned intelligent ward temperature monitoring method reduces the risk of contact infection by acquiring thermodynamic images of the ward, user location data, and basic user information. It identifies thermal areas in the thermodynamic images and extracts their coordinate information, matching this with user location data to obtain user thermal images. This association of heat sources with specific users eliminates the tediousness and errors of manual intervention. The method extracts the thermal center temperature value from the user's thermal images to form a temperature data sequence, supporting continuous real-time monitoring and overcoming the drawback of low frequency. Based on user basic information (such as age and medical history), it retrieves personalized abnormal temperature judgment rules from a preset rule base for anomaly detection, adapting the detection logic to different physiological characteristics, improving judgment accuracy, and reducing false alarms and missed alarms. When the detection result is abnormal, it combines the detection result, location data, and basic information to generate an alarm signal, enabling rapid response and optimizing ward management efficiency.
[0075] In one embodiment, identifying thermal region images in a thermodynamic image and extracting the coordinate information of each thermal region image in the thermodynamic image includes:
[0076] S11, based on the range of thermal radiation intensity corresponding to human body temperature, sets the thermal area screening threshold;
[0077] S12, Perform pixel segmentation on the thermodynamic image based on the thermal region screening threshold to obtain a set of thermal region pixels;
[0078] S13, perform connected component analysis on the set of hot region pixels, classify adjacent hot region pixels into the same hot region, and form multiple hot region images;
[0079] S14, using the preset coordinate system of the thermodynamic image as a reference, traverse all pixels of each thermal region image and extract the horizontal and vertical coordinate values of the pixels in the thermodynamic image.
[0080] S15, Calculate the minimum x-coordinate, maximum x-coordinate, minimum y-coordinate, and maximum y-coordinate corresponding to each thermal region image based on the x-coordinate and y-coordinate values;
[0081] S16 combines the minimum and maximum x-coordinates, minimum and maximum y-coordinates to form the coordinate range of the hot region as coordinate information.
[0082] For example, a thermal region screening threshold can be set based on the range of thermal radiation intensity corresponding to human body temperature. This threshold is used to distinguish human heat sources from background noise. The thermodynamic image is segmented at the pixel level according to this threshold, and a set of thermal region pixels that meet the threshold range is selected by image processing algorithms to achieve preliminary heat source localization. Connectivity analysis is performed on the set of thermal region pixels, and clustering technology is used to group spatially adjacent thermal pixels into the same thermal region, forming multiple independent thermal region images (such as the forehead or neck region of a human body). Each thermal region image represents a potential user heat source. Based on the preset coordinate system of the thermodynamic image, all pixels of each thermal region image are traversed, and the horizontal and vertical coordinate values of each pixel in the image are extracted to obtain the original position data. Based on these coordinate values, the minimum horizontal coordinate, maximum horizontal coordinate, minimum vertical coordinate, and maximum vertical coordinate corresponding to each thermal region image are determined by an extreme value calculation algorithm to quantify the boundary of the thermal region. These extreme coordinates are combined to form the thermal region coordinate range (such as a rectangular bounding box), which serves as structured coordinate information and provides a spatial reference for subsequent user matching.
[0083] In one embodiment, the coordinate information of each thermal region image is matched with user location data to obtain a user thermal image, including:
[0084] S21, Extract the location coordinates of each user from the user location data to obtain the user coordinate points;
[0085] S22, by determining whether the user's coordinates fall within the coordinate range of the hot area, a point-to-area matching result is generated;
[0086] S23, identify conflict scenarios in the matching results of points and regions, and obtain conflict scenario identification results. Conflict scenarios include a single user coordinate point matching multiple hot area images, multiple user coordinate points matching a single hot area image, and multiple user coordinate points matching multiple hot area images within a preset coordinate range threshold.
[0087] S24, Based on the conflict scene identification results, identify the point and region matching results of the conflict scene, perform conflict processing according to the preset conflict processing rules, and obtain the matching results after conflict processing:
[0088] S25. Based on the point and region matching results of the conflict scene identification results that did not identify the conflict scene or the matching results after conflict processing, determine the user's thermal image.
[0089] Specifically, the system extracts the location coordinates of each user from user location data (such as two-dimensional or three-dimensional coordinate points obtained through RFID or UWB technology) to obtain precise user coordinate points. Using a spatial relationship judgment algorithm, it checks whether each user coordinate point falls within the thermal region coordinate range of the thermal region image (such as a rectangular bounding box), generating preliminary point-to-region matching results. It identifies conflict scenarios in the matching results, including single user coordinate points matching multiple thermal region images (possibly due to overlapping or reflection of heat sources), multiple user coordinate points matching a single thermal region image (such as multiple people approaching each other), or multiple user coordinate points matching multiple thermal region images within a preset coordinate range threshold (such as dense areas), obtaining conflict scenario identification results. For identified conflict scenarios, it dynamically processes them by calling preset conflict handling rules (such as algorithms based on nearest neighbor distance, user priority, or thermal intensity weighting), such as selecting the nearest thermal center region when there is a single-point multi-region conflict, or allocating based on positioning accuracy when there is a multi-point single-region conflict, obtaining matching results after conflict handling. Finally, it combines the matching results of unidentified conflict scenarios and the matching results after conflict handling to determine the thermal image corresponding to each user, achieving accurate and automated association between thermal radiation data and specific users.
[0090] In one embodiment, a pre-defined body temperature abnormality discrimination rule library stores body temperature abnormality discrimination rules corresponding to different age ranges, different disease types, and different combinations of age ranges and different disease types, and each body temperature abnormality discrimination rule is associated with a unique user feature combination tag.
[0091] Based on user information, the system retrieves the corresponding abnormal body temperature criteria from a pre-defined rule base, including:
[0092] S31. Based on the basic information of each user, the age information is divided into intervals to obtain age interval identifiers, and the basic medical history information is classified into disease types to obtain disease type identifiers.
[0093] S32, associate the identity identifier of each user's basic information with the corresponding age range identifier and disease type identifier to form a user feature combination label;
[0094] S33, based on user characteristic combination tags, matches the abnormal body temperature discrimination rules obtained from the preset abnormal body temperature discrimination rule library.
[0095] For example, age information can be grouped based on each user's basic information (e.g., divided into ranges such as 0-3 years, 4-12 years, 13-65 years, and over 65 years) to generate age range identifiers. Basic medical history information can be categorized through a disease classification system (e.g., divided into cardiovascular, metabolic, or immune diseases) to obtain disease type identifiers. The user's identity identifier (e.g., a unique patient ID) is associated and bound with the aforementioned age range identifier and disease type identifier to form a unique user feature combination label, which serves as the index key for rule base retrieval. Based on the user feature combination label, a query and matching operation is performed in a preset abnormal body temperature discrimination rule base (stored in a database table or structured configuration file, where each rule is associated with specific label parameters) to obtain the corresponding abnormal body temperature discrimination rule (e.g., threshold comparison rule or statistical model). This enables dynamic invocation of discrimination rules based on user-personalized characteristics (e.g., age and medical history). Through automated label generation and rule matching mechanisms, manual intervention is eliminated, increasing consistency and response efficiency.
[0096] In one embodiment, S41, anomaly detection is performed on the body temperature data sequence using a body temperature anomaly discrimination rule, employing the following formula:
[0097]
[0098] Where T = (t1, t2, ..., t n ) represents a body temperature data sequence, t i Let L represent the body temperature value measured in the sequence at the i-th time, and let θ represent the user feature combination label. L The threshold parameter for judging abnormal body temperature is the combination tag L of the user feature in the preset abnormal body temperature judgment rule base. I is an indicator function, which takes the value of 1 when the condition is met and 0 otherwise. f(T, L) is an abnormality detection function, which outputs a value of 1 to indicate abnormal body temperature and an output value of 0 to indicate normal body temperature.
[0099] Specifically, automatic discrimination can be achieved through the anomaly detection function f(T, L), where T represents the body temperature data sequence arranged in chronological order (e.g., t1, t2, ..., t...). nEach t i θ represents the body temperature value measured in the i-th measurement, L is the user feature combination label (a unique identifier generated based on age range and disease type), and θ L The function iterates through all t values in the body temperature data sequence T to set the threshold parameter corresponding to L in the predefined body temperature abnormality discrimination rule base (e.g., setting a threshold of 37.5℃ for elderly cardiovascular patients); i The value is checked using the indicator function I (which outputs 1 if the condition is true, and 0 otherwise) to see if any t exists. i More than θ L The situation, i.e., calculation The Boolean result is mapped to a binary output: if the condition is true, f(T,L) = 1 indicates abnormal body temperature, otherwise f(T,L) = 0 indicates normal body temperature; this achieves real-time performance, repeatability, and adaptability to personalized thresholds, improving monitoring accuracy and reducing false alarms.
[0100] In one embodiment, if the body temperature detection result is abnormal, a body temperature abnormality alarm signal is generated by combining the body temperature detection result, user location data, and user basic information, including:
[0101] S51, extract the abnormal body temperature value, the start time of the abnormal body temperature, and the duration of the abnormal body temperature from the body temperature detection results;
[0102] S52, extract the location coordinates and the ward area identifier where the location coordinates are located from the user's location data;
[0103] S53, based on the preset alarm level determination rules, associates and matches abnormal body temperature values, age information and basic medical history information to determine the corresponding alarm level;
[0104] S54 calls the preset alarm signal template and generates an abnormal body temperature alarm signal based on the abnormal body temperature value, the start time of the abnormal body temperature, the duration of the abnormal body temperature, the location coordinates, the ward area identifier, the identity identifier, the age range identifier, the underlying disease type identifier, and the alarm level.
[0105] For example, when a body temperature detection result is determined to be abnormal, key dynamic parameters can be extracted from the result, including the specific value of the abnormal temperature, the start time of the abnormal temperature, and the duration of the abnormal state. The user's precise location coordinates and the identifier of their ward area (such as ward number or ward code) can be obtained from user location data to achieve precise spatial positioning of the abnormal event. Based on preset alarm level determination rules (using multi-dimensional correlation matching to comprehensively analyze abnormal body temperature values with user age information and underlying medical history information, such as triggering a level one alarm when an elderly patient with a history of cardiovascular disease has a temperature exceeding 38°C), the corresponding alarm level is dynamically determined. A preset alarm signal template is invoked, and structured data filling technology is used to integrate multi-source information such as abnormal body temperature value, start time of the abnormal temperature, duration of the abnormal temperature, location coordinates, ward area identifier, identity information, age range identifier, underlying disease type identifier, and alarm level to generate a standardized abnormal body temperature alarm signal. This signal can be pushed to medical staff terminals in real time via a communication interface, enabling rapid positioning and tiered response to abnormal situations and optimizing ward management efficiency.
[0106] In one embodiment, the method further includes:
[0107] S61, Obtain the body temperature detection results under the same user feature combination label, where the same user feature combination label is a user feature combination label that contains the same age range identifier and disease type identifier;
[0108] S62, Based on the body temperature detection results under the same user characteristic combination tags, analyze the differences in body temperature characteristics of the same user characteristic combination tags;
[0109] S63, based on the differences in body temperature characteristics, adjust the discrimination parameters of the corresponding body temperature abnormality discrimination rules in the preset body temperature abnormality discrimination rule library to obtain the updated body temperature abnormality discrimination rules. The updated body temperature abnormality discrimination rules are used for subsequent abnormal detection of body temperature data sequences.
[0110] Specifically, by acquiring historical body temperature detection results under similar user characteristic combination labels (i.e., user groups containing the same age range and disease type labels), a data query mechanism is used to retrieve continuous time-period body temperature data sequences and their abnormality markers from the storage system; based on statistical analysis (such as calculating the mean, standard deviation, or abnormal frequency distribution of body temperature), the differences in body temperature characteristics among similar user groups are quantitatively assessed to identify group-specific patterns (such as the trend of lower baseline body temperature in elderly cardiovascular patients); based on the analyzed differences in body temperature characteristics, parameter optimization algorithms (such as sliding window adjustment or machine learning-based threshold calibration) can be used to dynamically adjust the discrimination parameters (such as correcting the threshold θ) of the corresponding rules in the preset body temperature abnormality discrimination rule base. LTo reduce the false alarm rate, updated body temperature abnormality discrimination rules are generated, and the rule base is iterated in real time through database write operations. The updated rules are immediately applied to the anomaly detection of subsequent body temperature data sequences, forming a closed-loop optimization system to improve the system's adaptability and monitoring accuracy.
[0111] The aforementioned intelligent ward temperature monitoring method utilizes thermodynamic images of the ward, user location data, and basic user information. Thermal imaging technology avoids physical contact, reducing the risk of infection. Simultaneously, user location data is matched with the coordinates of thermal regions in the thermodynamic images. Combined with connected component analysis and conflict resolution mechanisms, heat sources are associated with specific users, eliminating the tediousness and errors of manual intervention. The thermal center temperature value of the user's thermal image is extracted to form a temperature data sequence, supporting continuous real-time monitoring and overcoming the low frequency limitation of traditional methods. Based on user basic information (such as age range and disease type), personalized discrimination rules are retrieved from a pre-set abnormal temperature discrimination rule library. Formulated abnormal detection adapts to different physiological characteristics, improving discrimination accuracy and reducing false alarms and missed alarms. When an abnormality is detected, a tiered alarm signal is generated by combining the temperature value, location coordinates, and basic information to achieve rapid response.
[0112] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.
[0113] Based on the same inventive concept, this application also provides an intelligent ward temperature monitoring system for implementing the intelligent ward temperature monitoring method described above. The solution provided by this system is similar to the solution described in the above method; therefore, the specific limitations of one or more intelligent ward temperature monitoring system embodiments provided below can be found in the limitations of the intelligent ward temperature monitoring method described above, and will not be repeated here.
[0114] In one exemplary embodiment, such as Figure 2 As shown, an intelligent ward temperature monitoring system is provided, including:
[0115] The monitoring data acquisition module 101 is used to acquire thermodynamic images of the ward, user location data, and user basic information, including identity information, age information, and basic medical history information.
[0116] The thermal region identification module 102 is used to identify thermal region images in a thermodynamic image and extract the coordinate information of each thermal region image in the thermodynamic image.
[0117] The region-user matching module 103 is used to match the coordinate information of each thermal region image with user positioning data to obtain a user thermal image;
[0118] The body temperature data extraction module 104 is used to extract the thermal center temperature value of the user's thermal image and form a body temperature data sequence.
[0119] The discrimination rule matching module 105 is used to retrieve the corresponding abnormal body temperature discrimination rule from the preset abnormal body temperature discrimination rule library based on the user's basic information;
[0120] The abnormal body temperature detection module 106 is used to perform abnormal detection on the body temperature data sequence using an abnormal body temperature discrimination rule, and obtain the body temperature detection result.
[0121] The alarm signal generation module 107 is used to generate an abnormal body temperature alarm signal by combining the body temperature detection result, user location data and user basic information if the body temperature detection result is abnormal.
[0122] In one embodiment, the thermal area identification module 102 is further configured to:
[0123] Based on the range of thermal radiation intensity corresponding to human body temperature, a thermal area screening threshold is set.
[0124] Based on the thermal region screening threshold, the thermodynamic image is segmented into pixels to obtain the set of thermal region pixels;
[0125] Connectivity analysis is performed on the set of hot region pixels to group adjacent hot region pixels into the same hot region, forming multiple hot region images;
[0126] Using the preset coordinate system of the thermodynamic image as a reference, traverse all pixels of each thermal region image and extract the horizontal and vertical coordinate values of the pixels in the thermodynamic image.
[0127] Based on the x-coordinate and y-coordinate values, calculate the minimum x-coordinate, maximum x-coordinate, minimum y-coordinate, and maximum y-coordinate corresponding to each thermal region image;
[0128] The minimum and maximum x-coordinates, minimum and maximum y-coordinates are combined to form the coordinate range of the hot region as coordinate information.
[0129] In one embodiment, the region-user matching module 103 is further configured to:
[0130] Extract the location coordinates of each user from the user location data to obtain the user coordinate points;
[0131] By determining whether the user's coordinates fall within the coordinate range of the hot zone, a point-to-region matching result is generated.
[0132] The conflict scenarios in the matching results of identification points and regions are obtained, and the conflict scenario identification results are obtained. The conflict scenarios include a single user coordinate point matching multiple hot area images, multiple user coordinate points matching a single hot area image, and multiple user coordinate points matching multiple hot area images within a preset coordinate range threshold.
[0133] Based on the conflict scene identification results, the point and region matching results of the conflict scene are identified. Conflict processing is performed according to preset conflict processing rules to obtain the matching results after conflict processing:
[0134] Based on the conflict scene identification results, the matching results of points and regions that did not identify conflict scenes or the matching results after conflict processing are used to determine the user's thermal image.
[0135] In one embodiment, the discrimination rule matching module 105 has a preset body temperature abnormality discrimination rule library that stores body temperature abnormality discrimination rules corresponding to different age ranges, different disease types, and different combinations of age ranges and different disease types, and each body temperature abnormality discrimination rule is associated with a unique user feature combination tag.
[0136] Based on user information, the system retrieves the corresponding abnormal body temperature criteria from a pre-defined rule base, including:
[0137] Based on each user's basic information, the age information is divided into intervals to obtain age interval identifiers, and the basic medical history information is classified by disease type to obtain disease type identifiers;
[0138] The identity identifiers of each user's basic information are associated with the corresponding age range identifiers and disease type identifiers to form user feature combination tags;
[0139] Based on user characteristic combination tags, the abnormal body temperature discrimination rules are obtained by matching them in the preset abnormal body temperature discrimination rule library.
[0140] In one embodiment, the abnormal body temperature detection module 106 uses the following formula to perform abnormal detection on the body temperature data sequence using an abnormal body temperature discrimination rule:
[0141]
[0142] Where T = (t1, t2, ..., t n) represents a body temperature data sequence, t i Let L represent the body temperature value measured in the sequence at the i-th time, and let θ represent the user feature combination label. L The threshold parameter for judging abnormal body temperature is the combination tag L of the user feature in the preset abnormal body temperature judgment rule base. I is an indicator function, which takes the value of 1 when the condition is met and 0 otherwise. f(T, L) is an abnormality detection function, which outputs a value of 1 to indicate abnormal body temperature and an output value of 0 to indicate normal body temperature.
[0143] In one embodiment, the alarm signal generation module 107 is further configured to:
[0144] Extract abnormal body temperature values, the onset time of abnormal body temperature, and the duration of abnormal body temperature from the body temperature detection results;
[0145] Extract the location coordinates and the ward area identifier where the location coordinates are located from the user's location data;
[0146] Based on the preset alarm level determination rules, abnormal body temperature values, age information, and basic medical history information are correlated and matched to determine the corresponding alarm level;
[0147] The system calls a preset alarm signal template and generates an abnormal body temperature alarm signal based on the abnormal body temperature value, the start time of the abnormal body temperature, the duration of the abnormal body temperature, the location coordinates, the ward area identifier, the identity identifier, the age range identifier, the underlying disease type identifier, and the alarm level.
[0148] In one embodiment, such as Figure 3 As shown, it also includes a feedback optimization module 108, used for:
[0149] Obtain body temperature detection results under the same user feature combination label. The same user feature combination label is a combination of user feature labels that contain the same age range identifier and disease type identifier.
[0150] Based on the body temperature detection results under the same user characteristic combination tags, analyze the differences in body temperature characteristics under the same user characteristic combination tags;
[0151] Based on the differences in body temperature characteristics, the discrimination parameters of the corresponding body temperature abnormality discrimination rules in the preset body temperature abnormality discrimination rule library are adjusted to obtain the updated body temperature abnormality discrimination rules. The updated body temperature abnormality discrimination rules are used for subsequent anomaly detection of body temperature data sequences.
[0152] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the intelligent ward temperature monitoring method as described above.
[0153] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.
[0154] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.
[0155] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.
Claims
1. A method for monitoring body temperature in an intelligent ward, characterized in that, The method includes: The system acquires thermodynamic images of the ward, user location data, and basic user information, including identity information, age information, and basic medical history information. Identify thermal region images in the thermodynamic image and extract the coordinate information of each thermal region image in the thermodynamic image; The coordinate information of each of the thermal area images is matched with the user positioning data to obtain the user thermal image; Extract the thermal center temperature value from the user's thermal image to form a body temperature data sequence; Based on the user's basic information, the corresponding abnormal body temperature judgment rule is retrieved from the preset abnormal body temperature judgment rule library; The body temperature data sequence is subjected to anomaly detection using the aforementioned body temperature anomaly discrimination rule to obtain body temperature detection results; If the body temperature detection result is abnormal, an abnormal body temperature alarm signal is generated by combining the body temperature detection result, the user location data, and the user's basic information.
2. The method according to claim 1, characterized in that, The step of identifying thermal region images in the thermodynamic image and extracting the coordinate information of each thermal region image in the thermodynamic image includes: Based on the range of thermal radiation intensity corresponding to human body temperature, a thermal area screening threshold is set. Based on the thermal region screening threshold, the thermodynamic image is segmented into pixels to obtain a set of thermal region pixels; Perform connected component analysis on the set of hot region pixels to group adjacent hot region pixels into the same hot region, forming multiple hot region images; Using the preset coordinate system of the thermodynamic image as a reference, all pixels of each thermal region image are traversed, and the horizontal and vertical coordinate values of the pixels in the thermodynamic image are extracted. Based on the x-coordinate and y-coordinate values, calculate the minimum x-coordinate, maximum x-coordinate, minimum y-coordinate, and maximum y-coordinate corresponding to each of the thermal region images; The minimum x-coordinate, maximum x-coordinate, minimum y-coordinate, and maximum y-coordinate are combined to form the coordinate range of the hot region, which is used as the coordinate information.
3. The method according to claim 2, characterized in that, The step of matching the coordinate information of each of the thermal region images with the user positioning data to obtain a user thermal image includes: Extract the location coordinates of each user from the user location data to obtain the user coordinate points; By determining whether the user's coordinates fall within the coordinate range of the hot zone, a point-to-region matching result is generated. The conflict scenarios in the point and region matching results are identified to obtain conflict scenario identification results. The conflict scenarios include a single user coordinate point matching multiple hot area images, multiple user coordinate points matching a single hot area image, and multiple user coordinate points matching multiple hot area images within a preset coordinate range threshold. Based on the conflict scene identification results, the point and region matching results of the conflict scene are identified, and conflict processing is performed according to preset conflict processing rules to obtain the matching results after conflict processing: Based on the conflict scene identification results, the matching results of points and regions where no conflict scene was identified or the matching results after conflict processing, the user's thermal image is determined.
4. The method according to claim 1, characterized in that, The preset body temperature abnormality discrimination rule library stores body temperature abnormality discrimination rules corresponding to different age ranges, different disease types, and combinations of different age ranges and different disease types, and each body temperature abnormality discrimination rule is associated with a unique combination of user feature tags; Based on the aforementioned user information, the corresponding abnormal body temperature judgment rule is retrieved from the preset abnormal body temperature judgment rule library, including: For each of the aforementioned basic user information, the age information is divided into intervals to obtain age interval identifiers, and the basic medical history information is classified by disease type to obtain disease type identifiers; The identity identifiers of each user's basic information are associated with the corresponding age range identifiers and disease type identifiers to form the user feature combination tags; Based on the user feature combination tags, the body temperature abnormality discrimination rule is obtained by matching in the preset body temperature abnormality discrimination rule library.
5. The method according to claim 4, characterized in that, The abnormal body temperature data sequence is detected using the aforementioned abnormal body temperature discrimination rule, employing the following formula: Where T = (t1, t2, ..., t n ) represents a body temperature data sequence, t i Let L represent the body temperature value measured in the sequence at the i-th time, and let θ represent the user feature combination label. L The threshold parameter for judging abnormal body temperature is the combination tag L of the user feature in the preset abnormal body temperature judgment rule base. I is an indicator function, which takes the value of 1 when the condition is met and 0 otherwise. f(T, L) is an abnormality detection function, which outputs a value of 1 to indicate abnormal body temperature and an output value of 0 to indicate normal body temperature.
6. The method according to claim 1, characterized in that, If the body temperature detection result is abnormal, then, combining the body temperature detection result, the user location data, and the user's basic information, a body temperature abnormality alarm signal is generated, including: Extract the abnormal body temperature value, the onset time of the abnormal body temperature, and the duration of the abnormal body temperature from the body temperature detection results; Extract the location coordinates and the ward area identifier where the location coordinates are located from the user location data; Based on preset alarm level determination rules, the abnormal body temperature value, the age information and the basic medical history information are correlated and matched to determine the corresponding alarm level; The preset alarm signal template is invoked, and the abnormal body temperature alarm signal is generated based on the abnormal body temperature value, the start time of the abnormal body temperature, the duration of the abnormal body temperature, the location coordinates, the ward area identifier, the identity information, the age range identifier, the underlying disease type identifier, and the alarm level.
7. The method according to claim 4, characterized in that, The method further includes: Obtain body temperature detection results under similar user feature combination tags, wherein similar user feature combination tags are user feature combination tags that contain the same age range identifier and disease type identifier; Based on the body temperature detection results under the same user characteristic combination tags, analyze the differences in body temperature characteristics under the same user characteristic combination tags; Based on the differences in body temperature characteristics, the discrimination parameters of the corresponding body temperature abnormality discrimination rules in the preset body temperature abnormality discrimination rule library are adjusted to obtain updated body temperature abnormality discrimination rules. The updated body temperature abnormality discrimination rules are used for subsequent abnormal detection of the body temperature data sequence.
8. An intelligent ward temperature monitoring system, characterized in that, The system includes: The monitoring data acquisition module is used to acquire thermodynamic images of the ward, user location data, and user basic information, including identity information, age information, and basic medical history information. A thermal region identification module is used to identify thermal region images in the thermodynamic image and extract the coordinate information of each thermal region image in the thermodynamic image; The region-user matching module is used to match the coordinate information of each of the thermal region images with the user positioning data to obtain a user thermal image; The body temperature data extraction module is used to extract the thermal center temperature value of the user's thermal image to form a body temperature data sequence; The discrimination rule matching module is used to retrieve the corresponding abnormal body temperature discrimination rule from the preset abnormal body temperature discrimination rule library based on the user's basic information; A body temperature abnormality detection module is used to perform abnormality detection on the body temperature data sequence using the body temperature abnormality discrimination rule, and obtain body temperature detection results; An alarm signal generation module is used to generate an abnormal body temperature alarm signal by combining the body temperature detection result, the user location data, and the user's basic information if the body temperature detection result is abnormal.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.