Medical image department resource scheduling method and device and computer equipment

Abstract

The invention relates to a medical image department resource scheduling method and device and computer equipment. The method comprises the following steps: acquiring historical department equipment operation data and historical department physiological data of a medical image department; performing real-time virtual state modeling on the medical image department according to the historical department equipment operation data and the historical department physiological data to obtain department real-time state data; according to the department real-time state data and the current department physiological data of the medical image department, carrying out physiological rhythm analysis on the medical image department to obtain department rhythm analysis data; inputting the department real-time state data and the department rhythm analysis data into a resource scheduling algorithm of the medical image department to obtain each resource scheduling parameter group; and according to the department equipment health data of the medical image department, optimizing each resource scheduling parameter group to obtain a target department resource scheduling parameter group. By adopting the method, the operation efficiency of the medical image department can be effectively improved.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent scheduling, and particularly to a method, device, and computer equipment for scheduling resources in a medical imaging department. Background Art

[0002] In the management of traditional imaging departments, the scheduling of department resources mainly relies on manual arrangements. For example, staff manually arrange equipment usage time, assign examining doctors, and notify abnormal objects according to examination applications. Due to the lack of systematic information integration, there are often information silos between different systems (such as registration systems, reporting systems, and imaging storage systems), making it difficult to synchronize the status of abnormal objects and equipment usage in a timely manner, resulting in low overall department operation efficiency. Summary of the Invention

[0003] Based on this, in view of the above technical problems, it is necessary to provide a method, device, and computer equipment for scheduling resources in a medical imaging department that can effectively improve the operation efficiency of the medical imaging department.

[0004] In a first aspect, the present application provides a method for scheduling resources in a medical imaging department, including:

[0005] Obtaining historical department equipment operation data and historical department physiological data of the medical imaging department;

[0006] Based on the historical department equipment operation data and the historical department physiological data, performing real-time virtual state modeling on the medical imaging department to obtain department real-time state data;

[0007] Based on the department real-time state data and the current department physiological data of the medical imaging department, performing physiological rhythm analysis on the medical imaging department to obtain department rhythm analysis data;

[0008] Inputting the department real-time state data and the department rhythm analysis data into a resource scheduling algorithm for the medical imaging department to obtain each resource scheduling parameter group;

[0009] Optimizing each resource scheduling parameter group according to the department equipment health data of the medical imaging department to obtain a target department resource scheduling parameter group.

[0010] In a second aspect, the present application further provides a device for scheduling resources in a medical imaging department, including:

[0011] A data acquisition module, configured to obtain historical department equipment operation data and historical department physiological data of the medical imaging department;

[0012] A status virtual module, configured to perform real-time virtual status modeling on the medical imaging department according to the historical department equipment operation data and the historical department physiological data, so as to obtain department real-time status data;

[0013] A rhythm analysis module, configured to perform physiological rhythm analysis on the medical imaging department according to the department real-time status data and the current department physiological data of the medical imaging department, so as to obtain department rhythm analysis data;

[0014] A scheduling analysis module, configured to input the department real-time status data and the department rhythm analysis data into a resource scheduling algorithm for the medical imaging department, so as to obtain each resource scheduling parameter group;

[0015] A scheduling optimization module, configured to optimize each of the resource scheduling parameter groups according to the department equipment health data of the medical imaging department, so as to obtain a target department resource scheduling parameter group.

[0016] In a third aspect, the present application further provides a computer device, including a memory and a processor, where the memory stores a computer program, and when the processor executes the computer program, any step of a method for resource scheduling of a medical imaging department is implemented.

[0017] The above-mentioned method, device and computer device for resource scheduling of a medical imaging department systematically collect and fuse the historical equipment operation data and historical department physiological data of the medical imaging department, construct a real-time virtual status model for the department, so that the operation status of the department can be grasped in real time without interfering with the actual work, and physiological rhythm analysis is performed in combination with the current department physiological data, so as to accurately depict the rhythm change and cycle law of the department operation; on this basis, the constructed department real-time status data and rhythm analysis results are input into a resource scheduling algorithm to generate multiple groups of resource scheduling parameters, and each parameter group is further optimized in multiple dimensions in combination with the equipment health data of the medical imaging department, so as to screen out the optimal resource scheduling plan. It can realize the intelligent dynamic configuration of resources such as personnel, equipment, and inspection time periods, effectively improve the operation efficiency of the medical imaging department, enhance the forward-looking and scientific nature of overall operation and maintenance, and ultimately significantly improve the coordination ability, service quality and abnormal object medical experience of the department operation. Description of the Drawings

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 It is an application environment diagram of a medical imaging department resource scheduling method in an embodiment;

[0020] Figure 2 It is a schematic flowchart of a medical imaging department resource scheduling method in an embodiment;

[0021] Figure 3 It is a schematic flowchart of a method for obtaining the first type of department rhythm analysis data in an embodiment;

[0022] Figure 4 It is a schematic flowchart of a method for obtaining the second type of department rhythm analysis data in an embodiment;

[0023] Figure 5 It is a schematic flowchart of a method for obtaining a rhythm synchronous subpopulation in an embodiment;

[0024] Figure 6 It is a schematic flowchart of a method for obtaining the third type of department rhythm analysis data in an embodiment;

[0025] Figure 7 It is a schematic flowchart of a method for obtaining the first type of resource scheduling parameter group in an embodiment;

[0026] Figure 8 It is a schematic flowchart of a method for obtaining the second type of resource scheduling parameter group in an embodiment;

[0027] Figure 9 It is a schematic flowchart of a method for obtaining a target department resource scheduling parameter group in an embodiment;

[0028] Figure 10 It is a structural block diagram of a medical imaging department resource scheduling device in an embodiment;

[0029] Figure 11 It is an internal structure diagram of a computer device in an embodiment. Detailed implementation manners

[0030] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0031] A medical imaging department resource scheduling method provided by an embodiment of the present application can be applied as follows Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or can be placed on the cloud or other network servers. Among them, the server 104 can be implemented by an independent server or a server cluster composed of multiple servers.

[0032] In an exemplary embodiment, as Figure 2 shown, a method for scheduling resources in a medical imaging department is provided. Taking the server in Figure 1 as an example, the following steps 202 to 210 are included. Among them:

[0033] Step 202, obtain the historical department equipment operation data and historical department physiological data of the medical imaging department.

[0034] Among them, the medical imaging department can be a department in a hospital that is specifically responsible for various medical imaging examination tasks, mainly including imaging examination equipment such as CT, MRI, X-ray, ultrasound, and their operating personnel, diagnostic doctors and other component units.

[0035] Among them, the historical department equipment operation data can be the operation records of imaging equipment in the medical imaging department in the past period of time, including information such as equipment power-on and power-off times, usage frequency, task completion quantity, failure times, operation and maintenance logs, usage duration, etc., for evaluating the load level and usage trend of the equipment.

[0036] Among them, the historical department physiological data can be historical behavioral data related to the operation of the department, such as the number of past abnormal object visits, daily peak visit distribution, examination type structure, medical staff scheduling information, examination time length, etc.

[0037] Specifically, through the hospital information system (HIS), radiology information system (RIS), and equipment management platform, collect the operation data of each equipment in the medical imaging department in the past period of time as the historical department equipment operation data, including equipment startup frequency, operation duration, equipment load, failure records, usage time period distribution, etc.; at the same time, extract the historical department physiological data of the medical imaging department from the hospital big data platform, such as abnormal object arrival frequency, abnormal object type distribution in different time periods, examination item types, medical staff scheduling data, etc.

[0038] Step 204, according to the historical department equipment operation data and historical department physiological data, perform real-time virtual state modeling on the medical imaging department to obtain department real-time state data.

[0039] Among them, real-time virtual state modeling can be to integrate historical device data and physiological data, and use artificial intelligence modeling methods to construct a digital twin model of a medical imaging department, enabling it to dynamically simulate the operation state of the department under real-time data input, such as device load, task allocation, personnel working state, etc.

[0040] Among them, the real-time state data of the department can be a data set obtained from real-time virtual state modeling that can reflect the overall operation state of the medical imaging department at the current moment, usually including key indicators such as device operation state, waiting situation of abnormal objects, schedule execution situation, task processing rate, etc.

[0041] Specifically, perform feature engineering processing on historical department device operation data and historical department physiological data, including time series reconstruction, key indicator extraction (such as device utilization rate, average queuing time, peak load, etc.) and normalization operations. Further select modeling algorithms for time series and dynamic feature learning, such as long short-term memory network (LSTM) or variational autoencoder (VAE), to train the historical department device operation data and historical department physiological data after feature engineering processing, and establish a digital twin model that reflects the operation characteristics of the medical imaging department. This digital twin model can simulate dynamic changes such as device operation state, personnel workload, and task distribution, and perform immediate reasoning and state output after accessing real-time data streams (such as real-time queuing situation, inspection task allocation, current device state), and finally obtain the real-time state data of the department that reflects the usage situation and operation state of various resources in the medical imaging department at the current time point.

[0042] Step 206, perform physiological rhythm analysis on the medical imaging department according to the real-time state data of the department and the current department physiological data of the medical imaging department to obtain department rhythm analysis data.

[0043] Among them, physiological rhythm analysis can be based on real-time state data and current physiological data, and apply time series analysis and rhythm modeling methods to extract and analyze the operation rhythm, periodic changes and load fluctuation rules of the department in the time dimension.

[0044] Among them, the department rhythm analysis data can be the output result of physiological rhythm analysis, specifically manifested as a data structure that reflects the rhythm changes, trend predictions and load patterns of the department in different time periods and different operation states.

[0045] Specifically, the real-time status data of the department is matched and fused with the current physiological data of the department, where the current physiological data of the department includes key indicators such as the real-time abnormal object flow, inspection type distribution, medical staff scheduling information, abnormal object waiting duration, task response time, etc. By using rhythm modeling methods such as Fourier transform, wavelet analysis, or recurrent neural networks (such as GRU, LSTM), time series analysis and processing are performed on the fused data set to extract periodic features, trend changes, and sudden abnormal signals in the data, and identify the operation rhythm characteristics of the department in different time periods, such as peak periods, flat peak periods, and potential overload periods; combining the rhythm changes with the current state evolution trend, further evaluate dynamic behaviors such as department load fluctuation rules, resource utilization peak shift, inspection task accumulation, etc., and finally form department rhythm analysis data.

[0046] Step 208, input the real-time status data of the department and the department rhythm analysis data into the resource scheduling algorithm for the medical imaging department to obtain each resource scheduling parameter group.

[0047] Among them, the resource scheduling algorithm can be an algorithm model that automatically generates an optimal resource allocation plan according to the department operation status and rhythm analysis results, and usually includes intelligent optimization methods such as genetic algorithms, ant colony algorithms, and particle swarm algorithms.

[0048] Among them, the resource scheduling parameter group can be a candidate resource allocation plan output by the resource scheduling algorithm, including a combination of multiple parameters such as specific equipment usage time, personnel scheduling arrangement, and task priority strategy.

[0049] Specifically, the real-time status data of the department and the department rhythm analysis data are subjected to standardization and feature encoding processing to ensure that they enter the resource scheduling algorithm model in a unified input format; further, based on the actual operation requirements and resource constraint conditions of the department (such as the total number of devices, personnel shifts, inspection task types, etc.), a resource scheduling optimization objective function is constructed, covering multiple indicators such as task completion efficiency, personnel load balance, maximum equipment utilization rate, and minimum abnormal object waiting time, and an algorithm model suitable for multi-objective optimization problems is selected, such as genetic algorithm, ant colony optimization algorithm, or particle swarm optimization algorithm. Then, using the standardized and feature-encoded real-time status data of the department and the department rhythm analysis data as the initial input, search and iterate the scheduling parameter space to generate multiple groups of feasible resource scheduling parameter groups, and these parameter groups include equipment usage plans, personnel scheduling plans, inspection item allocation strategies, etc. for each time period.

[0050] Step 210, optimize each resource scheduling parameter group according to the department equipment health data of the medical imaging department to obtain the target department resource scheduling parameter group.

[0051] Among them, the department equipment health data can be comprehensive evaluation data on the current operating status of imaging equipment, covering equipment maintenance cycles, operating stability, current fault status, performance degradation indicators, etc., and is used to determine whether the equipment is in good condition to avoid equipment overload or failure caused by improper scheduling.

[0052] Among them, the target department resource scheduling parameter group can be the optimal scheduling plan obtained by optimizing and screening based on multiple scheduling parameter groups in combination with equipment health data. It can not only meet the current operating needs of the department but also ensure equipment safety and efficient resource use, and is the final recommended configuration result for actual operation.

[0053] Specifically, call the current equipment health data of the medical imaging department, including equipment operation duration, maintenance cycle, failure rate, performance degradation indicators, and real-time fault warning information, etc., to construct an equipment health assessment model to quantify the current availability and risk level of each device. Conduct correlation analysis between each resource scheduling parameter group and the equipment health status, identify scheduling plans that may lead to high-load operation or exceed the equipment tolerance threshold, and accordingly set the safety boundary conditions of the scheduling parameters. Use a weighted multi-objective function to comprehensively score each parameter group in dimensions such as operation efficiency, safety, and impact on equipment life, and screen out the target department resource scheduling parameter group that maximizes resource utilization and optimizes scheduling efficiency while ensuring equipment health through algorithm optimization means (such as weighted linear combination, Pareto frontier analysis, etc.). Among them, the target department resource scheduling parameter group can be used for scientific research management, shift management, teaching management, system management, personnel management, equipment management, data management, meeting management, ledger management, performance management, quality control management, and system management in the medical imaging department.

[0054] In the above-mentioned method for scheduling resources in a medical imaging department, by systematically collecting and integrating the historical equipment operation data and historical department physiological data of the medical imaging department, a real-time virtual state model for the department is constructed, enabling real-time monitoring of the department's operating status without interfering with actual work, and conducting physiological rhythm analysis in combination with the current department physiological data, thereby accurately depicting the rhythm changes and periodic laws of the department's operation; on this basis, using the constructed department real-time state data and rhythm analysis results, input them into the resource scheduling algorithm to generate multiple groups of resource scheduling parameters, and further conduct multi-dimensional optimization of each parameter group in combination with the equipment health data of the medical imaging department, so as to screen out the optimal resource scheduling plan. It can realize the intelligent dynamic configuration of resources such as personnel, equipment, and inspection time periods, effectively improve the operation efficiency of the medical imaging department, enhance the foresight and scientific nature of overall operation and maintenance, and ultimately significantly improve the collaborative ability, service quality, and abnormal object visit experience of the department.

[0055] In an exemplary embodiment, such asFigure 3 As shown, according to the real-time status data of the department and the current physiological data of the medical imaging department, physiological rhythm analysis is carried out on the medical imaging department to obtain department rhythm analysis data, including steps 302 to 306. Among them:

[0056] Step 302, perform physiological decomposition on the current physiological data of the department to obtain the physiological data of each department member.

[0057] Among them, physiological decomposition can be a process of analyzing the current physiological data of the medical imaging department. According to the identity, responsibilities, and behavior trajectories of individual personnel, the overall department data is disassembled into separate individual data sets. By identifying the work schedules, work intensities, task records, etc. of each department member, the collective behavior data is restored to the individual work status, forming an independent data basis for the personnel dimension.

[0058] Among them, the physiological data of department members can be an individualized work status data set formed for each department member after physiological decomposition, usually including time-correlated indicators such as working hours, task frequencies, response delays, rest intervals, and operation behavior trajectories.

[0059] Specifically, basic data such as personnel work schedule information, work task assignment, task completion status, abnormal object reception records, working hours, and rest intervals are extracted from the current physiological data of the medical imaging department. Then, the data is classified and managed according to personnel roles (such as radiologists, technicians, nursing staff, etc.), and individualized decomposition is carried out through a unique personnel identifier. After individualized decomposition, the behavior data of each department member within the current time period is integrated, and their current workload, operation frequency, response time, task density, and short-term physiological rhythm change trends are calculated to obtain the structured physiological data of each department member.

[0060] Step 304, perform periodic time series signal processing on the physiological data of each department member to obtain individual rhythm characteristic information.

[0061] Among them, periodic time series signal processing can be a technical process for analyzing the time series information in the physiological data of department members, used to identify the periodicity, trend, and volatility in the data. Through methods such as Fourier transform, wavelet analysis, autocorrelation function, and recurrent neural network, the original time series is converted into a signal structure with periodic characteristics, and the physiological rhythm laws of individuals at different time scales are extracted.

[0062] Among them, the individual rhythm characteristic information can be rhythmic indexes extracted from the data of each person through the processing of periodic time series signals, usually including rhythm period, intensity, stability, active interval, fluctuation range, etc., which depict the regular behavior patterns presented by personnel in their daily work and are used to evaluate the work rhythm and physiological load status of individuals.

[0063] Specifically, convert the key physiological behavior indexes (such as task completion frequency, response time, continuous working duration, rest interval, etc.) in the structured physiological data of each department staff into a time series form, and perform standardization processing in chronological order to ensure the consistency and comparability of the data. Use periodic analysis methods, such as fast Fourier transform (FFT) to identify frequency domain characteristics, wavelet transform to detect multi-scale rhythm changes, or use autocorrelation function to analyze periodic repeatability, to extract the rhythm characteristics of individuals in a specific time dimension; for non-linear or complex behavior patterns, recurrent neural networks such as LSTM and GRU can be further introduced to model and predict long sequences, identify potential rhythm fluctuation trends and fatigue accumulation patterns, and finally output the rhythm characteristic information of each department staff, including rhythm period, intensity, fluctuation range and stability, etc.

[0064] Step 306, perform physiological rhythm aggregation processing on each individual rhythm characteristic information to obtain department rhythm analysis data.

[0065] Among them, physiological rhythm aggregation processing can be a process of integrating, classifying and analyzing the rhythm characteristics of multiple individuals. Through methods such as cluster analysis, statistical modeling, and weight assignment, the rhythm characteristics of all personnel are fused into one, and macroscopic laws such as the overall rhythm trend, operation cycle, and synchrony of the department are identified.

[0066] Specifically, vectorize the individual rhythm characteristic information (such as rhythm period, activity intensity, fatigue index, response delay, etc.) of each department staff, construct a group rhythm characteristic matrix, and use methods such as cluster analysis (such as K-means, DBSCAN) or principal component analysis (PCA) to classify and reduce the dimension of the individual rhythm patterns in the group rhythm characteristic matrix, identify the similarities and differences between the rhythms of different personnel, and calculate the rhythm synchrony data to quantify the degree of collaborative operation between groups. On this basis, through a weighted aggregation model, according to the importance of the roles of department staff and the proportion of work tasks, further comprehensive analysis is carried out on each rhythm characteristic in the group rhythm characteristic matrix to form the overall rhythm pattern data of the medical imaging department at the current time as the department rhythm analysis data, which includes information such as the average rhythm period, fluctuation trend, and peak load distribution of the department.

[0067] In this embodiment, by performing refined individual decomposition on the physiological data of the current department, the individual physiological data of the department staff is extracted. Then, combined with the periodic time series signal processing technology, the rhythm characteristic information of each person is accurately identified. Furthermore, through rhythm aggregation modeling, the overall rhythm analysis data at the department level is formed. This not only improves the dynamic perception ability of the working state and rhythm law of the staff, but also effectively reveals the collaborative mode and load difference among groups. Finally, it provides a scientific basis for resource scheduling, shift optimization, and rhythm intervention, significantly enhancing the operation efficiency, rhythm adaptability, and personnel health management level of the medical imaging department.

[0068] In an exemplary embodiment, as Figure 4 shown, perform physiological rhythm aggregation processing on the individual rhythm characteristic information to obtain department rhythm analysis data, including steps 402 to 406. Among them:

[0069] Step 402, map each individual rhythm characteristic information into an individual rhythm time series vector respectively.

[0070] Among them, the individual rhythm time series vector can be a time series vector formed by encoding the rhythm characteristic information of a certain department staff according to the time dimension, usually including the physiological rhythm changes of this person within a specific period, such as the dynamic changes of work activity, fatigue accumulation, task response rate, etc. on the time axis.

[0071] Specifically, structurally organize the individual rhythm characteristic information of each department staff, extract the key time-related indicators therein, such as rhythm cycle, active period, work load fluctuation, response delay trend, etc.; align and encode these indicators with a unified time axis (such as in hours or minutes) to form a vector representation with time series characteristics, that is, each individual rhythm characteristic information is mapped into a group of initial rhythm time series vectors, which sequentially reflect the rhythm performance of the individual at each time point on the time axis, such as peak intensity, fatigue accumulation degree, or efficiency change trend. To enhance comparability, it is also necessary to perform normalization processing on each group of initial rhythm time series vectors to ensure that the data between different individuals is within the same scale. The final individual rhythm time series vector formed retains the behavioral rhythm characteristics of the individual in the time dimension.

[0072] Step 404, perform rhythm synchronization clustering processing according to the similarity information between the individual rhythm time series vectors to obtain each rhythm synchronization sub-group.

[0073] Among them, the similarity information can be a measurement result of the relative proximity calculated between multiple individual rhythm time series vectors, used to measure the similarity degree of the rhythm patterns between different people. Commonly used calculation methods include dynamic time warping (DTW), Euclidean distance, Manhattan distance, or cosine similarity, etc. The similarity information is presented in the form of a matrix or score.

[0074] Among them, the rhythm synchronization clustering process can be a process of dividing personnel with similar rhythm patterns into several sub - groups by using the similarity information between individual rhythm time - series vectors. This process usually adopts unsupervised clustering algorithms (such as K - means, DBSCAN, spectral clustering, etc.) to discover the structural distribution of physiological rhythms within the medical imaging department and reveal which personnel have a high degree of consistency in work rhythm and physiological performance.

[0075] Among them, the rhythm synchronization sub - group can be several personnel sets formed in the clustering result. Each sub - group consists of personnel with similar rhythm patterns, and the members within the group show strong consistency in aspects such as work active periods, rhythm cycles, load change trends, etc.

[0076] Specifically, according to the preset time - series similarity measurement method, such as dynamic time warping (DTW) for dealing with the situation where the rhythm cycles are not completely aligned, or using Euclidean distance and cosine similarity to evaluate the overall trend similarity, calculate the pairwise similarity matrix between all individual rhythm time - series vectors. Taking this similarity matrix as the input and applying a clustering algorithm for group division, algorithms such as K - means algorithm (suitable for large amounts of data with a clear similarity distribution), hierarchical clustering (suitable for fine division of small samples), or spectral clustering (suitable for non - convex clustering structures) can be selected to automatically classify individual rhythm time - series vectors with similar rhythms into the same rhythm synchronization sub - group. Each rhythm synchronization sub - group represents a set of personnel with a high degree of consistency in aspects such as rhythm patterns, work peak periods, load changes, etc.

[0077] Step 406, calculate the rhythm analysis data of the department according to the rhythm peak overlap degree and physiological load balance degree between each rhythm synchronization sub - group.

[0078] Among them, the rhythm peak overlap degree can be the degree of overlap of the peak periods of their respective rhythm curves in the time dimension among all rhythm synchronization sub - groups. This index is used to judge whether multiple high - load groups reach physiological or work peaks simultaneously within the same time period.

[0079] Among them, the physiological load balance degree can be a quantitative evaluation of the rationality of the physiological work load distribution between each rhythm synchronization sub - group, usually calculated based on multiple dimensions such as task density, number of personnel, and rhythm fluctuation intensity. This index is used to measure whether there are problems such as uneven load and excessive work pressure in some groups among groups.

[0080] Specifically, the time series vectors of each rhythm synchronization subgroup are averaged to extract the typical rhythm pattern data of the subgroup, and the rhythm peak time period is identified with emphasis, that is, the time period with the highest work activity or the most concentrated physiological load, to obtain the rhythm peak data. Then, the time overlap degree of the rhythm peaks between different subgroups is calculated to determine whether multiple groups reach a high-load state simultaneously within the same time period, so as to identify potential resource concentration or overload risks, and the rhythm peak overlap degree is obtained. At the same time, parameters such as the number of personnel, task density, and rhythm fluctuation intensity of each rhythm synchronization subgroup are evaluated, and the physiological load balance degree is calculated to measure whether the resource allocation is reasonable and whether the work pressure of each group is balanced, and the physiological load balance degree is obtained. The rhythm peak overlap degree and the physiological load balance degree are fused into a multi-dimensional comprehensive index, and are quantified by using weighted average or scoring models, and the rhythm analysis data of the department is output.

[0081] In this embodiment, by mapping the individual rhythm characteristic information into time series vectors, the dynamic trajectory of the personnel rhythm changes is comprehensively retained, and rhythm synchronization clustering is performed based on the similarity between vectors, effectively identifying the personnel subgroups with similar rhythm patterns. Further, by combining the rhythm peak overlap degree and the physiological load balance degree between subgroups, refined and structured rhythm analysis data of the department is constructed. It not only improves the perception ability of the group cooperation efficiency and the rationality of the load distribution, but also provides strong data support for the early warning of high-risk rhythm conflicts, the collaborative optimization of shift scheduling, and the regulation of the operation rhythm, thus significantly enhancing the operation intelligence level and resource allocation scientificity of the medical imaging department.

[0082] In an exemplary embodiment, as Figure 5 shown, rhythm synchronization clustering processing is performed according to the similarity information between the individual rhythm time series vectors to obtain each rhythm synchronization subgroup, including steps 502 to 508. Among them:

[0083] Step 502, using the dynamic time warping distance algorithm, perform similarity processing on the individual rhythm time series vectors to obtain a similarity metric matrix.

[0084] Among them, the dynamic time warping distance algorithm can be an algorithm for measuring the similarity between two time series, even if there is a non-linear alignment between these two series on the time axis (for example, the rhythm speeds are inconsistent or the peak occurrence time points are different). DTW calculates the minimum alignment cost by elastically "stretching" or "compressing" the series on the time axis to obtain the best match in the overall trend.

[0085] Among them, the similarity metric matrix can be a symmetric two-dimensional matrix, which is used to represent the similarity degree between individual rhythm time series vectors. Each element in the matrix represents the matching degree of any two department staff in the rhythm pattern, which is usually a value calculated by DTW or other similarity algorithms.

[0086] Specifically, all individual rhythm time series vectors are unified into a comparable time series format to ensure that their time granularity and value range are consistent. The dynamic time warping (DTW) algorithm is used to compare the rhythm time series vectors of any two department staff pairwise. Among them, DTW can match different behavior cycles and rhythm fluctuations through elastic alignment, so as to accurately capture the rhythm similarity of individuals in the time dimension, especially applicable to the situation where the occurrence times of rhythm peaks are inconsistent. During the processing, a minimum matching distance is calculated for each pair of time series vectors through DTW, which represents the difference degree in their rhythm patterns. Finally, the DTW distance results between all personnel are summarized to construct a complete similarity metric matrix, and each element in the matrix corresponds to the rhythm similarity score of a pair of personnel.

[0087] Step 504, construct a rhythm conflict penalty matrix according to the similarity metric matrix and the rhythm conflict rules.

[0088] Among them, the rhythm conflict rules can be predefined judgment criteria, which are used to identify and label the behavior patterns that may cause resource conflicts or efficiency decline during the rhythm operation between individuals. For example, if two personnel are both in a high-load rhythm state within the same time period and their task execution or equipment dependence highly overlaps, the conflict rules will be triggered.

[0089] Among them, the rhythm conflict penalty matrix can be another type of matrix constructed based on the similarity metric according to the rhythm conflict rules, which is used to record the potential conflict degree between pairs of individuals. Each element in the matrix represents the penalty weight corresponding to when there is a high-intensity synchronization peak in the rhythm time series vectors of two personnel.

[0090] Specifically, for each pair of individual rhythm time series vectors, analyze the time position and duration of their rhythm peaks, and identify whether there is a high-intensity synchronization peak phenomenon on the time axis between the two. If the rhythm curves of two personnel are both in a high-load state within a similar time period, according to the conflict rules (such as exceeding the set time overlap threshold or the activity intensity superposition limit), mark this pair of individuals as having a rhythm conflict, and assign a certain penalty weight to these pairs of conflicting personnel. The higher the value, the more serious the conflict. Based on this, construct a rhythm conflict penalty matrix with the same structure as the similarity metric matrix, where each element represents the penalty value of the corresponding pair of personnel in terms of the rhythm overlap risk.

[0091] Step 506: Based on the similarity metric matrix and the rhythm conflict penalty matrix, construct a conflict penalty function in the original spectral clustering algorithm to obtain the rhythm clustering algorithm.

[0092] Among them, the conflict penalty function can be an optimization function introduced into the original clustering model, which is used to balance the goals between "maximizing rhythm similarity" and "minimizing conflicts". It takes the similarity metric matrix as the forward drive and the conflict penalty matrix as the reverse constraint, and modifies the objective function in spectral clustering or other clustering models, so that the model preferentially selects combinations with both rhythm similarity and low conflicts during the population division process, thereby optimizing the overall clustering quality and actual scheduling effect.

[0093] Among them, the rhythm clustering algorithm can be a clustering algorithm specifically designed for the rhythm data of personnel in the medical imaging department, which integrates the similarity metric and conflict penalty mechanisms. Its core is to introduce a conflict penalty function on the basis of the traditional spectral clustering algorithm, construct a weighted similarity graph and a modified Laplacian matrix, and perform eigenvalue decomposition and clustering operations.

[0094] Specifically, in the process of spectral clustering, a weighted similarity graph needs to be constructed. The edge weights of this graph originally come from the similarity metric matrix between individuals; at this time, by introducing the rhythm conflict penalty matrix, the original edge weights are corrected, so that the edge weights of individual pairs with high similarity but rhythm conflicts are weakened or a penalty term is added. Then, according to the above processing results, a conflict penalty term is added to the original spectral clustering algorithm to suppress potential peak conflicts during the population division process; during the definition process, by reconstructing the graph Laplacian matrix and comprehensively considering similarity and conflict in the eigenvalue decomposition step, the rhythm clustering algorithm is finally obtained.

[0095] Step 508: Input the individual rhythm time series vectors into the rhythm clustering algorithm to obtain each rhythm synchronous sub-group.

[0096] Specifically, after constructing the rhythm clustering algorithm with conflict constraints, input the individual rhythm time series vectors into the improved spectral clustering algorithm, and construct a weighted similarity graph according to the graph structure that integrates the similarity metric matrix and the rhythm conflict penalty matrix; then use this graph to generate a modified graph Laplacian matrix, and extract key low-dimensional feature representations through eigenvalue decomposition to retain the main information of the rhythm structure between personnel. On this basis, use a clustering method (such as K-means) to divide the extracted spectral feature vectors to achieve population aggregation. The clustering process classifies personnel with highly synchronous rhythms but low resource conflict risks into the same class, thereby forming several rhythm synchronous sub-groups. The members in each sub-group are highly similar in terms of rhythm cycle, active period, load curve, etc., and avoid generating superimposed conflicts during the same peak period.

[0097] In this embodiment, by introducing the dynamic time warping distance algorithm to measure the similarity of individual rhythm time series vectors, the non-linear matching relationship of rhythm patterns in the time dimension is accurately captured. Then, a conflict penalty matrix is constructed by combining rhythm conflict rules, and similarity and conflict factors are jointly introduced into the spectral clustering algorithm to form an improved clustering model with the ability to suppress rhythm conflicts. During the clustering process, not only the similarity of rhythms is considered, but also potential peak period conflicts are actively avoided, effectively improving the scientificity and practicality of the division of rhythm synchronization sub-groups. Thus, a more accurate and rhythm-friendly data basis is provided for medical imaging departments in aspects such as personnel grouping, collaborative scheduling, and task allocation, significantly enhancing the rhythm coordination and load safety of system operation.

[0098] In an exemplary embodiment, as Figure 6 shown, according to the rhythm peak overlap degree and physiological load balance degree between each rhythm synchronization sub-group, department rhythm analysis data is calculated, including steps 602 to 606. Among them:

[0099] Step 602, generate a rhythm synergy feature vector according to the rhythm peak overlap degree and physiological load balance degree.

[0100] Among them, the rhythm synergy feature vector can be a data vector used to describe the overall rhythm operation state of the medical imaging department, which is composed of the rhythm peak overlap degree, physiological load balance degree between multiple sub-groups, and possible additional indicators (such as the number of groups, rhythm fluctuation range, etc.).

[0101] Specifically, extract the rhythm peaks of each rhythm synchronization sub-group, identify the high-load working periods (such as the daily task peak period) of each sub-group in the time dimension, and quantify the rhythm peak overlap degree by calculating the intersection ratio of the peak time periods between different groups. This index is used to evaluate the possibility of resource conflict and the severity of time overlap. At the same time, key data such as the number of tasks, average load level, load fluctuation range, and number of personnel of each rhythm synchronization sub-group are counted, and the standard deviation or coefficient of variation of its load distribution is calculated to measure the balance of load distribution between groups, and the physiological load balance degree is obtained. The rhythm peak overlap degree and physiological load balance degree are uniformly encoded, and combined with additional information such as time tags and the number of sub-groups to construct a structured rhythm synergy feature vector.

[0102] Step 604, input the rhythm synergy feature vector into the current rhythm state evaluation model to obtain the current rhythm coordination state.

[0103] Among them, the current rhythm state evaluation model can be a classification or scoring model constructed based on supervised learning, which is used to judge the rhythm coordination level of the medical imaging department at the current time point by analyzing the rhythm synergy feature vector.

[0104] Among them, the current rhythm coordination state can be the analysis result output by the current rhythm state evaluation model, reflecting whether the personnel and resources in the medical imaging department are rhythmically coordinated at the current time, and whether there are problems such as task backlog, equipment conflict or uneven load.

[0105] Specifically, the rhythm synergy feature vector is input into the current rhythm state evaluation model to obtain the rhythm coordination state of the medical imaging department at the current moment. The current rhythm state evaluation model is a trained evaluation model, which is constructed based on historical rhythm data and coordinated state labels marked by experts. Supervised learning algorithms such as random forest, support vector machine (SVM) or multi-layer perceptron (MLP) are often used, and it has the ability to comprehensively discriminate multi-dimensional input features. During the processing of the current rhythm state evaluation model, the current rhythm state evaluation model analyzes whether the current department is in a high coordination, medium coordination or low coordination state according to the internally learned feature weights and pattern recognition capabilities, and outputs corresponding analysis results, which may be probability distributions, score values or grade labels, etc., to obtain the current rhythm coordination state.

[0106] Step 606, input the rhythm synergy feature vector into the predicted rhythm state evaluation model to obtain the predicted rhythm coordination state.

[0107] Among them, the predicted rhythm state evaluation model can be a machine learning model used to predict the rhythm coordination state within a certain future time period, often constructed based on time series modeling methods (such as LSTM, GRU, time series regression, etc.). It uses the current rhythm synergy feature vector and historical rhythm state trends as inputs to predict whether the rhythm state will improve, deteriorate or fluctuate in the future, so as to provide a basis for early intervention.

[0108] Among them, the predicted rhythm coordination state can be the judgment result of the predicted rhythm state evaluation model on the rhythm operation trend in a certain future time period, usually including the analysis sequence, trend direction (such as rising, falling) or risk analysis data of the future rhythm state, and is used to early warn potential rhythm conflicts or operation incoordination problems.

[0109] Specifically, similar to the previous calculation process, the rhythm synergy feature vector is input into the predicted rhythm state evaluation model to obtain the rhythm coordination trend in a future period of time. The predicted rhythm state evaluation model is an evaluation model based on time series prediction. This model is trained through historical rhythm state change data and corresponding feature evolution trajectories and often uses recurrent neural networks (such as LSTM, GRU) or time series regression models, and has the ability to capture rhythm evolution patterns and trend fluctuations. During the processing of the predicted rhythm state evaluation model, the rhythm synergy feature vector is used as the starting point for input, and combined with necessary historical feature sequences or sliding window mechanisms, time forward calculation is performed to predict the change in the rhythm coordination state within one or more future time steps. The output of the predicted rhythm state evaluation model can be an analysis sequence of the coordination state, trend direction (rising / falling), risk analysis, etc. for the future time period as the predicted rhythm coordination state, which is used to determine whether there are potential problems such as upcoming rhythm conflicts, resource load imbalances, or increased rhythm fluctuations.

[0110] Step 608: Perform policy function fusion on the current rhythm coordination state and the predicted rhythm coordination state to obtain the department rhythm analysis data.

[0111] Among them, policy function fusion can be a calculation method for comprehensively analyzing the current rhythm coordination state and the predicted rhythm coordination state. By setting fusion rules or weighting strategies, the state information in two time dimensions is integrated into a processing method that reflects the overall operation quality and risk trend.

[0112] Specifically, unified standardization processing is performed according to the current rhythm coordination state and the predicted rhythm coordination state in different states (such as scores, grade labels, probability distributions) to ensure their comparability on the same evaluation scale. Then, according to a preset set of fusion policy functions, including weighted average, time-sensitive weighting (such as assigning higher weights to predicted states to enhance forward-looking), confidence interval adjustment (dynamically adjusting weights according to model uncertainty), or rule-based state mapping (such as marking as "critical" when the current is highly coordinated but the prediction is declining), comprehensively considering the immediate rhythm stability represented by the current state and the future trend risk reflected by the predicted state, a set of multi-dimensional rhythm analysis indicators are output, and finally the department rhythm analysis data is generated, including the overall rhythm synergy score, future fluctuation trend, risk level identification, etc.

[0113] In this embodiment, by constructing a rhythm synergy feature vector based on the rhythm peak overlap degree and the physiological load balance degree, the operation synergy characteristics within the department are accurately characterized in the time dimension. Further, the feature vector is respectively input into the current rhythm state evaluation model and the prediction model to obtain the immediate rhythm coordination state and the future rhythm evolution trend, and the organic integration of timeliness and forward-looking information is realized through the fusion of the policy function. This not only improves the dynamic adaptability and prediction ability of rhythm analysis, but also provides a scientific basis for rhythm imbalance warning, resource allocation optimization, and operation rhythm intervention, thus significantly enhancing the rhythm sensitivity, risk perception ability, and decision-making intelligence level of the operation of the medical imaging department.

[0114] In an exemplary embodiment, as Figure 7 shown, the real-time state data of the department and the rhythm analysis data of the department are input into the resource scheduling algorithm for the medical imaging department to obtain each resource scheduling parameter group, including steps 702 to 706. Among them:

[0115] Step 702, identify the current resource availability parameters from the real-time state data of the department.

[0116] Among them, the current resource availability parameters can be the resource state information extracted from the real-time state data of the medical imaging department, reflecting the resources available for scheduling at the current moment, covering idle or soon-to-be-idle devices (such as CT, MRI, etc.), deployable personnel (such as technicians and doctors who are not overloaded in the schedule), and the current task queue situation (such as the number of waiting tasks, task urgency, etc.).

[0117] Specifically, the real-time state data of the department is classified and extracted to obtain the core state information related to equipment, personnel, and tasks respectively. The equipment state includes the current operating state (such as in operation, idle, under maintenance) of each imaging device, the remaining working hours, the number of completed tasks, etc.; the personnel state includes the scheduling situation of medical technicians in each position, the current task progress, working hours, and remaining available working hours, etc.; the task state covers the current number of queued tasks, the estimated execution time of each task, priority, and the required resource type. Then, logical judgments and statistical analyses are performed on the data of the above different core states. For example, it is judged which devices are in a callable state, which personnel are idle or soon-to-be-idle, and which time periods have a scheduling space for inserting tasks. The above analysis results are aggregated to form the current resource availability parameter set.

[0118] Step 704, determine the scheduling constraint factor and the scheduling soft limit variable according to the rhythm analysis data of the department.

[0119] Among them, the scheduling constraint factors can be rules or restrictive conditions that must be strictly adhered to during the resource scheduling process, usually derived from the results of rhythm analysis and management regulations. For example, the continuous operation time of equipment shall not exceed the set upper limit, the continuous working time of personnel shall not exceed a certain threshold, and specific tasks must be completed within a specified time period, etc. These factors are used to ensure that no problems in terms of safety, efficiency, or human health are caused during the resource allocation process, belonging to the "hard constraints" in the scheduling model and must be forcibly satisfied by the algorithm.

[0120] Among them, the scheduling soft limit variables can be optimization preference items that are not enforced but are given priority consideration during the resource scheduling process, and are often used to improve the overall efficiency or operation experience of the system. They include, for example, "prioritize inspection tasks during off-peak periods", "suggest that doctors schedule shifts according to individual rhythm preferences", "try to shorten the waiting time of abnormal objects", etc. These variables participate in the optimization as part of the objective function in the scheduling model. Although they can be violated, they will affect the scoring and ranking of the scheduling plan and are used to achieve a more user-friendly and high-quality scheduling result.

[0121] Specifically, identify key rhythm patterns from the rhythm analysis data of the department, including the prediction of peak periods of each rhythm synchronization subgroup, the risk of resource conflicts, and the uneven load sections, etc. According to the identified patterns, set scheduling constraint factors as rules that must be strictly adhered to. For example, prohibit over-arranging equipment tasks during certain high-load periods, limit the continuous working time of medical technicians in high-risk fatigue intervals, or enforce the rotation mechanism of specific human resources. Subsequently, extract scheduling soft limit variables that can be used as optimization objectives. These variables include, for example, "suggest scheduling low-priority tasks during low-load periods", "try to balance the task volume of each group", "follow the individual rhythm preferences of personnel", etc. Although they are not mandatory constraints, they will be used as preference items in the algorithm to improve the scheduling quality.

[0122] Step 706, according to the current resource availability parameters, scheduling constraint factors, and scheduling soft limit variables, conduct scheduling allocation analysis on the hardware resources and personnel resources of the medical imaging department to obtain each resource scheduling parameter group.

[0123] Among them, the scheduling allocation analysis can be a series of calculations and optimization operations after the resource availability, constraint factors, and soft limit variables are jointly input into the scheduling algorithm, aiming to generate one or more specific resource configuration plans. The analysis process usually adopts a multi-objective optimization algorithm. By evaluating the performance of various scheduling combinations in terms of efficiency, stability, coordination, etc., gradually screen out the scheduling parameter group that performs optimally on the soft objectives on the premise of meeting the hard constraints.

[0124] Specifically, taking the current resource availability parameters, scheduling constraint factors, and scheduling soft limit variables as combined inputs, the scheduling algorithm is driven to comprehensively schedule and allocate the hardware resources (such as CT, MRI, DR, etc. devices) and personnel resources (such as radiologists, technicians, nurses) in the medical imaging department. In the specific calculation process, the scheduling algorithm filters out unavailable or unreachable devices and personnel according to the resource availability parameters to ensure the basic feasibility of the scheduling plan; subsequently, the scheduling constraint factors are embedded into the scheduling model as hard limit conditions, such as device operation time limits, upper limits of personnel work schedules, or non-crossing queuing of inspection tasks, etc., to strictly control the safety boundary of the scheduling plan. On this basis, the scheduling soft limit variables are introduced as optimization objectives, and a multi-objective optimization algorithm (such as genetic algorithm, particle swarm algorithm, or simulated annealing, etc.) is used for iterative calculation to dynamically balance factors such as task waiting time, device utilization rate, and personnel load balance, generating multiple feasible and relatively high-quality resource scheduling parameter groups. Each group of parameters includes detailed configurations such as device usage time period allocation, task priority sorting, and the optimal matching relationship between personnel and tasks.

[0125] In this embodiment, by dynamically identifying the current resource availability parameters from the real-time status data of the department, accurately grasping the actual schedulable status of devices and personnel, and combining with the department rhythm analysis data, scheduling constraint factors and soft limit variables that highly match the operation rhythm are set, comprehensively considering the requirements of operation safety and rhythm coordination; on this basis, comprehensive scheduling and allocation analysis is carried out to generate multiple groups of resource scheduling parameter schemes. It not only realizes the refined, dynamic, and rhythm-sensitive scheduling of resources, but also significantly improves the resource utilization efficiency, task response speed, and system operation stability, providing scientific, efficient, and executable resource scheduling decision support for the medical imaging department.

[0126] In an exemplary embodiment, as Figure 8 shown, according to the current resource availability parameters, scheduling constraint factors, and scheduling soft limit variables, scheduling and allocation analysis is carried out on the hardware resources and personnel resources in the medical imaging department to obtain each resource scheduling parameter group, including steps 802 to 806. Among them:

[0127] Step 802, identify the device resources and personnel resources that meet the operation conditions from the current resource availability parameters to obtain resource feasibility set information.

[0128] Among them, the resource feasibility set information can refer to a set of resources that meet the scheduling operation conditions obtained by screening all device and personnel resources in the medical imaging department at the current time point. It includes information such as the status, idle time period, technical ability, and task matching of all available devices and personnel, excluding resource units that do not have scheduling qualifications (such as faulty devices, overloaded personnel, etc.).

[0129] Specifically, extract the basic information of all devices and personnel from the current resource availability parameters, including the current status of the devices (such as idle, running, maintaining), the remaining operating capacity (such as the remaining available duration per day), the supported inspection types and technical requirements, as well as the scheduling status, task load, remaining working hours, and skill matching degree of the personnel. Conduct conditional screening on the data obtained above one by one: For example, eliminate faulty devices, devices that are under maintenance or have reached the maximum operating threshold, and exclude personnel who are fully scheduled, have exceeded the working hour limit, or are unable to perform specific tasks. Further mark the serviceable time period, acceptable task types, and idle windows for the remaining resources, construct a structured resource entity, and uniformly organize all eligible devices and personnel into resource feasibility set information, which defines the legal resource units that the scheduling system can currently call and their schedulable boundaries.

[0130] Step 804, perform coordinated scheduling according to the scheduling constraint factors and scheduling soft limit variables to obtain scheduling limit information.

[0131] Among them, coordinated scheduling can be to comprehensively consider the scheduling constraint factors and soft limit variables during the resource scheduling process to ensure that the scheduling plan not only complies with mandatory rules (such as the maximum working hours of personnel, equipment operation restrictions, etc.), but also fits the optimization objectives as much as possible (such as task rhythm matching, load balancing, etc.), emphasizing the dynamic balance between "constraint satisfaction" and "scheduling quality".

[0132] Among them, the scheduling limit information can be a set of scheduling boundaries and priority rules generated after analyzing the resource feasibility set in combination with the scheduling constraint factors (hard constraints) and soft limit variables (optimization preferences). It not only includes which resource combinations cannot be scheduled (violating the rules), but also identifies the recommended priority options (preference matching), providing an accurate scheduling operation space and scoring basis for the multi-objective optimization algorithm.

[0133] Specifically, for each resource unit (device or personnel), conduct rule verification according to the scheduling constraint factors. For example, judge whether the maximum continuous operating duration of the device is exceeded, whether the personnel are approaching the upper limit of continuous working hours, or whether there are tasks that must be completed within a specified time period, etc., and eliminate resource combinations that do not meet the hard conditions; these verification results constitute a hard constraint filter to ensure the security and legality of the system operation; then, for the remaining resource options, apply the scheduling soft limit variables to perform priority scoring, such as whether the device is at the low peak of the rhythm load, whether the personnel are in the rhythm preference interval, and whether the task arrangement is conducive to load balancing and rhythm synchronization. Finally, output the resource options filtered by the hard constraints and the scoring results based on the soft limits together to form structured scheduling limit information.

[0134] Step 806: Using the minimization of task waiting time, the maximization of resource utilization, and the optimization of rhythm adaptability as optimization conditions, perform multi-objective optimization on the resource feasibility set information and the scheduling constraint information to obtain each resource scheduling parameter group.

[0135] Among them, the optimization of rhythm adaptability can be that when the scheduling system arranges equipment tasks and personnel schedules, it tries to match the physiological rhythm states (such as high-efficiency working periods and low-load periods) of individuals or groups as much as possible to reduce fatigue, improve work efficiency, and system stability.

[0136] Specifically, it is necessary to use "minimization of task waiting time", "maximization of resource utilization", and "optimization of rhythm adaptability" as the core optimization objectives, and perform multi-objective optimization on the resource feasibility set information and the scheduling constraint information. First, regard each resource (equipment and personnel) as a scheduling variable, use the feasibility set obtained in the previous two steps as the candidate resource pool, and at the same time use the scheduling constraint factor as a hard limit condition to ensure that the scheduling plan is within the safety and operation boundaries. Then embed the soft limit variable into the objective function and assign different optimization weights. For example, set a part of the objective function to compress the task waiting time, another part to measure the utilization rate of equipment and personnel, and another part to specifically evaluate the rhythm matching degree (such as whether it is scheduled in the rhythm efficient interval). On this basis, introduce multi-objective optimization algorithms such as genetic algorithms, particle swarm algorithms, or NSGA-II. By initializing the scheduling population, iteratively generating candidate solutions, evaluating fitness, performing crossover and mutation operations, etc., continuously converge to a set of scheduling results that perform excellently in each objective dimension. Finally, output the resource scheduling parameter group, including the task execution order, the optimal matching of personnel and equipment, the scheduling time window, and the priority ranking.

[0137] In this embodiment, by screening out the equipment and personnel that meet the operating conditions from the current resource availability parameters, a resource feasibility set is constructed to ensure the accuracy and real-time nature of the scheduling basis; at the same time, combined with the scheduling constraint factor and the soft limit variable, a coordinated scheduling model is established, fully considering safety, rhythm matching, and operation preferences, and forming refined scheduling constraint information; finally, with the minimization of task waiting time, the maximization of resource utilization, and the optimization of rhythm adaptability as the goals, perform multi-objective optimization operations to generate multiple high-quality resource scheduling parameter groups. Achieve the dynamic balance among efficiency, rhythm, and resource health of the scheduling results, and effectively improve the resource regulation ability and operation resilience of the medical imaging department in a high-load environment.

[0138] In an exemplary embodiment, as Figure 9 shown, according to the equipment health data of the medical imaging department, optimize each resource scheduling parameter group to obtain the target department resource scheduling parameter group, including steps 902 to 906. Among them:

[0139] Step 902: Predict the normal usage information of each hardware resource in the medical imaging department based on the department's equipment health data.

[0140] Among them, the normal usage information of the hardware can be the expected results of the operating status of each device in a specific future time period obtained through a prediction model based on the imaging equipment health data, mainly including indicators such as the normal operation probability of the device, remaining useful life (RUL), fault warning time point, and continuous working duration.

[0141] Specifically, standardize the department's equipment health data (including historical operation duration, power-on and -off frequency, fault records, maintenance cycle, mean time between failures (MTBF), real-time performance monitoring data (such as temperature, voltage, load, etc.), and life assessment indicators provided by the manufacturer), and apply the prediction model for time series analysis or life modeling. For example, use the LSTM network to predict the operating status of the device in a specific future time window, or evaluate the remaining useful life (RUL) based on degradation models (such as Weibull distribution, Cox proportional hazards model). The model output results include the normal operation probability, expected fault data, maintenance warning data, etc. of each device per day or per hour in the future, and determine the normal usage information of the hardware according to the data output by the model.

[0142] Step 904: Analyze the operation risks of each hardware resource based on the normal usage information of the hardware and each resource scheduling parameter to obtain hardware risk analysis data.

[0143] Among them, the operation risk can be the possibility of the device having a fault, performance degradation, or task interruption due to insufficient health status during the actual scheduling and execution of tasks.

[0144] Among them, the hardware risk analysis data can be the structured results formed after evaluating the operating security of each group of resource scheduling parameter schemes at the device level, including indicators such as the risk score, expected fault probability, potential task failure points, and load overrun statistics of each device under this scheduling scheme.

[0145] Specifically, analyze each set of resource scheduling parameters, clarify the expected usage plan of each device, including its startup frequency, continuous working duration, task density, and cooperation relationship with other devices in each time period. Align the above clarified data with the normal usage information of the hardware in the time dimension, and calculate whether the scheduling intensity exceeds the healthy prediction safety range of the device, such as whether high-intensity tasks are arranged during the high-risk window period, whether the continuous operation time exceeds the predicted remaining life value, etc. Match and calculate the scheduling pressure of the device with its health status to obtain the comprehensive risk analysis data of each device under each scheduling plan, and at the same time mark the status such as "low risk", "medium risk" or "high risk", and output key indicators such as potential failure probability, expected early repair time, load overrun ratio, etc., to obtain the hardware risk analysis data.

[0146] Step 906, according to the hardware risk analysis data, optimize each set of resource scheduling parameters until the hardware risk analysis data meets the hardware risk control threshold, and obtain the target department resource scheduling parameter set.

[0147] Among them, the hardware risk control threshold can be the risk upper limit standard set to ensure the safe operation of the device, representing the maximum acceptable risk range that each device can bear during the scheduling process. For example, the maximum failure probability of a single device can be set not to exceed 10%, the daily operation duration does not exceed 90% of the recommended load upper limit of the device, and the continuous operation time shall not exceed the predicted life boundary, etc.

[0148] Specifically, set reasonable hardware risk control thresholds, such as the maximum allowable device failure probability (such as not exceeding 10%), the upper limit of the average load overrun times, the minimum guaranteed duration of the remaining available life, etc., as the judgment benchmark for the safety of the scheduling plan. Then conduct a risk comparison and analysis of each set of resource scheduling parameters one by one, and give priority to eliminating those high-risk plans that exceed the threshold. For the scheduling combinations close to the threshold edge, further perform parameter fine-tuning operations, such as adjusting the task assignment order, reducing the task density of high-risk devices, transferring tasks to devices with better health status, etc., and iteratively optimize the scheduling content to reduce the overall device load. At the same time, heuristic search or local optimization algorithms can be used to automatically search for a scheduling plan that meets the hardware risk control conditions and has better operating efficiency in the feasible solution space. Finally, when a certain set of scheduling parameters meets the preset threshold standards in terms of task coverage rate, operation rhythm matching degree, and device operation risk, it is determined as the target department resource scheduling parameter set.

[0149] In this embodiment, by modeling and analyzing the health data of department equipment, the normal use status of each hardware resource in the future is predicted, and potential failure risks and availability boundaries are identified in advance; then the prediction results are combined with each resource scheduling parameter group to carry out multi-dimensional operation risk assessment, generate hardware risk analysis data, and accurately reveal the equipment operation pressure and failure possibility under different scheduling schemes; finally, according to the set hardware risk control threshold, the scheduling scheme is iteratively optimized to ensure that the selected scheme takes into account equipment safety and continuous availability while achieving the best efficiency. The operation robustness and equipment protection capabilities of the scheduling scheme are significantly improved, and the deep integration of resource scheduling and equipment health status is achieved, which effectively extends the equipment life, reduces the sudden failure rate, and ensures the stability and reliability of the operation of the medical imaging department.

[0150] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0151] Based on the same inventive concept, the embodiment of the present application also provides a medical imaging department resource scheduling device for implementing the above-mentioned medical imaging department resource scheduling method. Figure 10 As shown, it includes: a data acquisition module 1002, a state virtual module 1004, a rhythm analysis module 1006, a scheduling analysis module 1008 and a scheduling optimization module 1010. The implementation solution for solving the problem provided by the device is similar to the implementation solution recorded in the above method. Therefore, the specific limitations in the embodiments of one or more medical imaging department resource scheduling devices provided below can refer to the limitations on a medical imaging department resource scheduling method in the above text, and will not be repeated here.

[0152] Each module in the above-mentioned medical imaging department resource scheduling device can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0153] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structural diagram may be as shown in Figure 11 . The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Those skilled in the art can understand that the structure shown in Figure 11 is only a block diagram of a part of the structure related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0154] In an embodiment, a computer device is further provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0155] In an embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0156] In an embodiment, a computer program product or a computer program is provided. The computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the steps in the above method embodiments.

[0157] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0158] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it may include the processes of the above method embodiments.

[0159] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope described in this specification.

[0160] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A method for scheduling medical imaging department resources, characterized in that: The method comprises: Obtain historical equipment operation data and historical physiological data of the medical imaging department; According to the historical department equipment operation data and the historical department physiological data, real-time virtual state modeling is performed on the medical imaging department to obtain real-time state data of the department; Performing physiological rhythm analysis on the medical imaging department according to the real-time status data of the department and the current physiological data of the medical imaging department to obtain department rhythm analysis data; Inputting the department real-time status data and the department rhythm analysis data into the resource scheduling algorithm for the medical imaging department to obtain each resource scheduling parameter group; According to the department equipment health data of the medical imaging department, each resource scheduling parameter group is optimized to obtain the target department resource scheduling parameter group.

2. The method according to claim 1, characterized in that The step of performing physiological rhythm analysis on the medical imaging department according to the real-time status data of the department and the current physiological data of the medical imaging department to obtain the department rhythm analysis data includes: Performing physiological decomposition on the current department's physiological data to obtain physiological data of personnel in each department; Performing periodic time series signal processing on the physiological data of the personnel in each department to obtain rhythm characteristic information of each body; The physiological rhythm aggregation processing is performed on each of the individual rhythm characteristic information to obtain the department rhythm analysis data.

3. The method according to claim 2, characterized in that The physiological rhythm aggregation processing is performed on each of the individual rhythm characteristic information to obtain the department rhythm analysis data, including: Mapping each of the individual rhythm characteristic information into each body rhythm timing vector; Performing rhythm synchronization clustering processing according to the similarity information between the individual rhythm timing vectors to obtain each rhythm synchronization subgroup; The rhythm analysis data of the department is calculated based on the rhythm peak overlap and physiological load balance between each rhythm synchronization subgroup.

4. The method according to claim 3, characterized in that: The rhythm synchronization clustering process is performed according to the similarity information between the individual rhythm timing vectors to obtain each rhythm synchronization subgroup, including: Using a dynamic time warping distance algorithm, similarity processing is performed between the individual rhythm timing vectors to obtain a similarity measurement matrix; According to the similarity measurement matrix and the rhythm conflict rule, a rhythm conflict penalty matrix is ​​constructed; According to the similarity measurement matrix and the rhythm conflict penalty matrix, a conflict penalty function is constructed in the original spectral clustering algorithm to obtain a rhythm clustering algorithm; Each of the individual rhythm timing vectors is input into the rhythm clustering algorithm to obtain each of the rhythm synchronization sub-populations.

5. The method according to claim 3, characterized in that: The calculating of the department rhythm analysis data according to the rhythm peak overlap and physiological load balance between each rhythm synchronization subgroup includes: generating a rhythm synergy feature vector according to the rhythm peak overlap and the physiological load balance; Inputting the rhythm coordination feature vector into a current rhythm state evaluation model to obtain a current rhythm coordination state; and, inputting the rhythm coordination feature vector into a predicted rhythm state evaluation model to obtain a predicted rhythm coordination state; The current rhythm coordination state and the predicted rhythm coordination state are subjected to strategy function fusion to obtain the department rhythm analysis data.

6. The method according to claim 1, characterized in that The real-time status data of the department and the rhythm analysis data of the department are input into the resource scheduling algorithm of the medical imaging department to obtain each resource scheduling parameter group, including: identifying current resource availability parameters from the department real-time status data; Determine scheduling constraint factors and scheduling soft limit variables according to the department rhythm analysis data; According to the current resource availability parameter, the scheduling constraint factor and the scheduling soft limit variable, a scheduling allocation analysis is performed on the hardware resources and personnel resources of the medical imaging department to obtain each resource scheduling parameter group.

7. The method according to claim 6, characterized in that The scheduling and allocation analysis of the hardware resources and personnel resources of the medical imaging department is performed according to the current resource availability parameter, the scheduling constraint factor and the scheduling soft limit variable to obtain each resource scheduling parameter group, including: Identify equipment resources and personnel resources that meet the operating conditions from the current resource availability parameters to obtain resource feasibility set information; Perform coordinated scheduling according to the scheduling constraint factor and the scheduling soft limit variable to obtain scheduling restriction information; Taking minimization of task waiting time, maximization of resource utilization and optimization of rhythm adaptability as optimization conditions, multi-objective optimization is performed on the resource feasibility set information and the scheduling restriction information to obtain each resource scheduling parameter group.

8. The method according to claim 1, characterized in that: The optimizing of each resource scheduling parameter group according to the department equipment health data of the medical imaging department to obtain the target department resource scheduling parameter group includes: Predicting normal hardware usage information of each hardware resource of the medical imaging department based on the department equipment health data; Analyze the operation risk of each of the hardware resources according to the normal use information of the hardware and each of the resource scheduling parameters to obtain hardware risk analysis data; According to the hardware risk analysis data, each resource scheduling parameter group is optimized until the hardware risk analysis data meets the hardware risk control threshold, thereby obtaining the target department resource scheduling parameter group.

9. A medical imaging department resource scheduling device, characterized in that: The device comprises: A data acquisition module is used to acquire historical equipment operation data and historical physiological data of the medical imaging department; A state virtualization module is used to perform real-time virtual state modeling of the medical imaging department according to the historical department equipment operation data and the historical department physiological data to obtain real-time state data of the department; A rhythm analysis module, used to perform physiological rhythm analysis on the medical imaging department according to the real-time status data of the department and the current physiological data of the medical imaging department, so as to obtain department rhythm analysis data; A scheduling analysis module, used for inputting the real-time status data of the department and the rhythm analysis data of the department into a resource scheduling algorithm for the medical imaging department to obtain various resource scheduling parameter groups; The scheduling optimization module is used to optimize each resource scheduling parameter group according to the department equipment health data of the medical imaging department to obtain the target department resource scheduling parameter group.

10. 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, the steps of the method according to any one of claims 1 to 8 are implemented.

Images