A cloud platform-based hemodialysis big data management system
Patent Information
- Application Number
- CN202511953570.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-23
- Publication Date
- 2026-10-13
AI Technical Summary
[0003]但在医疗系统中，对于透析时间安排缺乏数据支撑，难以精准捕捉患者的惯性透析规律，易出现透析间隔不合理的情况，同时对于诊疗流程中的检验节点异常情况未量化分析，无法提前锁定异常情况，增加了临床风险，最后，就诊资源分配多为事后调整，医护人力、耗材等资源与患者实际需求的匹配度低，既易造成资源闲置，也可能因资源不足影响透析安全性与效率，难以适配血液透析个体化、精准化的管理需求
本发明通过采集血液透析患者信息，透析时间分析模块提取透析时间，计算日期离散值和间隔离散值并比较，确定目标患者的惯性透析时间，进而摆脱人工经验依赖，精准识别患者透析时间规律，保障透析安排的稳定性，降低透析间隔不合理的风险；
Smart Images
Figure CN122907243A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of medical data management technology, and in particular to a cloud-based big data management system for hemodialysis. Background Technology
[0002] Hemodialysis is one of the renal replacement therapies for patients with acute and chronic renal failure. It involves draining blood from the body and passing it through a dialyzer composed of numerous hollow fibers. The blood and an electrolyte solution (dialysis fluid) of similar concentration to the body exchange substances through diffusion / convection inside and outside the hollow fibers, removing metabolic waste, maintaining electrolyte and acid-base balance, and removing excess water. The entire process of returning the purified blood is called hemodialysis.
[0003] However, the medical system lacks data support for dialysis scheduling, making it difficult to accurately capture patients' habitual dialysis patterns and leading to unreasonable dialysis intervals. Furthermore, the lack of quantitative analysis of abnormalities at testing points in the treatment process makes it impossible to identify abnormalities in advance, increasing clinical risks. Finally, the allocation of medical resources is mostly adjusted after the fact, resulting in a low degree of matching between medical staff, consumables, and other resources and patients' actual needs. This can easily lead to idle resources or, due to insufficient resources, affect the safety and efficiency of dialysis, making it difficult to meet the individualized and precise management needs of hemodialysis. Summary of the Invention
[0004] The purpose of this invention is to solve the problems in the background art by proposing a cloud-based big data management system for hemodialysis.
[0005] To achieve the above objectives, the present invention adopts the following technical solution: A cloud-based hemodialysis big data management system includes: The user information collection module is used to identify hemodialysis patients in the medical system and collect their patient information. The dialysis time analysis module is used to extract dialysis time from patient information and analyze the dialysis time of dialysis patients. Based on the dialysis cycle time and interval time, it calculates the date discrete value and interval discrete value of the target patient respectively, compares the date discrete value and interval discrete value, and determines the inertial dialysis time of the target patient. The user information analysis module is used to extract the diagnosis and treatment process from the patient information, divide the diagnosis and treatment process into several test nodes, calculate the abnormal frequency of each test node, and determine the key diagnostic nodes for dialysis patients based on the abnormal frequency. The appointment prediction module is used to receive the inertial dialysis time of dialysis patients and, based on the inertial dialysis time, to determine the target patients to be seen within the target date. The dynamic resource adjustment module is used to receive the target patients within the target date and obtain the key diagnostic nodes for each target patient. Then, medical staff adjust the medical resources in advance based on the number of target patients and the corresponding key diagnostic nodes.
[0006] As a further aspect of the present invention, the method for obtaining the dialysis cycle time includes: Dialysis patients are sequentially marked as target patients. The medical orders and historical dialysis times of the target patients are obtained. Based on the medical order information, the dialysis cycle time of the target patients is identified. The medical order information refers to the dialysis time arrangement of patients led and formulated by the doctor.
[0007] As a further aspect of the present invention, the method for calculating the discrete value of the date includes: Set the cycle time based on the dialysis cycle duration; Obtain the target patient's historical dialysis time and mark the historical dialysis time in all cycles to obtain the labeled dates. Then, align all cycles in chronological order, starting from the first date of the cycle and label this date as date 1. Identify the number of labeled dates Bb within date 1, and then obtain the total number of cycles Bx. Then, use the calculation formula F1=Bb÷Bx to obtain the visit frequency F1 of date 1. Then, calculate the visit frequency Fi of each date in the cycle according to the above method, where i represents the date in the cycle. The frequency Fi of all dates in the cycle is arranged in descending order to form a frequency sequence. Then, based on the cycle frequency in the medical order information, the first position in the frequency sequence is set as the starting position, the final position corresponding to the cycle frequency is identified, the frequency of visits between the starting position and the ending position is identified, and the date corresponding to this frequency of visits is marked as the key date. Here, the cycle frequency refers to the number of times the target patient needs dialysis in the dialysis time cycle in the medical order information. Next, obtain all the cycle periods, match the tag date with the key date. If the key date and the tag date coincide, generate the information value ME=0; otherwise, if the tag date and the key date do not coincide, generate the information value ME=1. Extract all information values MEj, where j represents different information values, and then use the formula The discrete date values UR are obtained, where J represents the total number of information values and MEa is the mean of the information values MEj.
[0008] As a further aspect of the present invention, the method for setting the cycle period includes: Identify the current real-time date and time, and then set the dialysis cycle time as the window length of the sliding window based on the dialysis cycle time. Obtain the time before the current date and time, and set the sliding window according to the natural cycle, that is, align the sliding window with the natural cycle. Then, set the movement step size according to the length of the dialysis cycle time, and push forward sequentially according to the set movement step size to generate continuous sliding windows. Mark each sliding window as a cycle period.
[0009] As a further aspect of the present invention, the method for calculating the interval discrete values includes: Obtain the target patient's historical dialysis times, calculate the time interval Tn between adjacent dialysis times, and then use the calculation formula... The discrete values UT are obtained, where n represents different interval times, N represents the total number of interval times, and Ta is the mean of the interval times.
[0010] As a further aspect of the present invention, the method for determining the inertial dialysis time includes: The date discrete value UR and the interval discrete value UT are compared. If UR≤UT, the key dates in the cycle are marked as the inertial dialysis time of the target patient according to the cycle period. If UR>UT, the most recent dialysis time of the target patient and the mean interval time Ta are obtained. The next dialysis time of the target patient is determined according to the mean interval time Ta and marked as the inertial dialysis time.
[0011] As a further aspect of the present invention, the method for determining key diagnostic nodes includes: Based on basic medical information, the routine testing process for dialysis patients is obtained and divided into several testing nodes; Arbitrarily select a test node, obtain all test reports of this test node, and identify abnormal reports among the test reports. Abnormal reports refer to test reports whose values are not within the normal range of the target patient. Count the number of abnormal reports and test reports respectively, divide the number of abnormal reports by the number of test reports, and mark the result as the abnormality frequency Rf. The abnormal frequency Rf is compared with the abnormal threshold Ry. If Rf < Ry, the test node is marked as a normal node. If Rf ≥ Ry, the test node is marked as a key diagnostic node.
[0012] As a further aspect of the present invention, the method for determining the target patient includes: Set a target date, which is the estimated date and time for dialysis patients to visit the clinic. Obtain the inertial dialysis time of all dialysis patients, identify dialysis patients whose inertial dialysis time is the same as the target date, and mark these patients as target patients.
[0013] As a further aspect of the present invention, after the target date is set, the day before the target date is obtained and marked as the preceding date. The target patients within the preceding date are obtained, and the target patients within the preceding date are compared with the actual patients. The target patients within the preceding date are identified, and the target patients who have not actually received treatment within the preceding date are remarked as the target patients for the target date.
[0014] As a further aspect of the present invention, a medical information collection module is also included, which is used to collect basic medical information, including medical care information and medical resource information. The medical information collection module then transmits the collected basic medical information to the user information analysis module.
[0015] Compared with existing technologies, the advantages of this invention are: This invention collects information from hemodialysis patients, extracts dialysis time using a dialysis time analysis module, calculates and compares date and interval discrete values, and determines the inertial dialysis time of the target patient. This eliminates reliance on manual experience, accurately identifies the pattern of the patient's dialysis time, ensures the stability of dialysis scheduling, and reduces the risk of unreasonable dialysis intervals. The user information analysis module breaks down the diagnosis and treatment process into several testing nodes, calculates the abnormal frequency of each node to identify key diagnostic nodes, and then quantifies the risk links in the diagnosis and treatment process, identifies high-risk diagnostic nodes for patients in advance, provides precise guidance for clinical intervention, and improves the safety of diagnosis and treatment. The appointment prediction module determines the target date for patients based on inertial dialysis time, while the dynamic resource adjustment module adjusts appointment resources based on the number of patients and key diagnostic nodes, achieving precise pre-allocation of resources, improving the utilization efficiency of medical staff and consumables, avoiding resource idleness or shortage, and adapting to the individualized management needs of hemodialysis. Attached Figure Description
[0016] Figure 1 This is a schematic diagram of the system structure of the present invention. Detailed Implementation
[0017] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.
[0018] Reference Figure 1 A cloud-based big data management system for hemodialysis includes a medical information collection module, a user information collection module, a dialysis time analysis module, a user information analysis module, a visit prediction module, and a dynamic resource adjustment module. The medical information collection module is used to collect basic medical information and transmit the collected basic medical information to the user information analysis module. The basic medical information includes medical care information and medical resource information. In this embodiment, the medical information collection module is specifically used to collect basic medical information for those who need to undergo hemodialysis. Furthermore, medical care information refers to the specific dialysis treatment process during hemodialysis, and medical resource information refers to the medical staff in the dialysis room and the corresponding medical consumables. The user information collection module is used to collect patient information already recorded in the medical system. In this embodiment, all patients are hemodialysis patients. The patient information includes dialysis time, treatment process, and medical orders. Then, the user information collection module establishes a one-way communication connection with the dialysis time analysis module and the user information analysis module respectively, and transmits the patient information to the user information analysis module and the dialysis time analysis module respectively. It should be further explained that the collection of basic medical information and patient information is all carried out on the cloud platform; The dialysis time analysis module is used to acquire patient information, extract dialysis time from the patient information, analyze the dialysis time, and determine the patient's inertial dialysis time. Specific methods for determining the inertial dialysis time include: S1: Randomly select a dialysis patient and mark this dialysis patient as the target patient. Obtain the target patient's medical orders and historical dialysis times. Based on the medical order information, identify the target patient's dialysis cycle time. The medical order information refers to the dialysis time determined by the doctor, such as twice a week, three times a week, or once every three days. At this time, the dialysis cycle time refers to the weekly time or interval. For example, when the medical order is twice a week or three times a week, the dialysis cycle time is set to one week. If the medical order is once every three days, the dialysis cycle time is set to one month. S2: Based on the dialysis cycle time and the target patient's historical dialysis time, calculate the date discrete value and interval discrete value for the target patient, and then compare the date discrete value and interval discrete value to determine the target patient's inertial dialysis time. Furthermore, methods for calculating discrete date values include: The system identifies the current real-time date and time, then sets the dialysis cycle time as the window length of the sliding window based on the dialysis cycle time. It obtains the time prior to the current real-time date and time and sets the sliding window according to the natural cycle, aligning the sliding window with the natural week cycle. Then, it sets the movement step size according to the length of the dialysis cycle time and moves forward sequentially according to the set movement step size to generate continuous sliding windows. Each sliding window is marked as a cycle. For example, if the target patient's dialysis cycle time is one week, the system obtains the current real-time date and time, sets the window length of the sliding window to 7 days, and sets Monday to Sunday as a sliding window according to the natural cycle. Then, it determines all the sliding windows according to the movement step size and marks each sliding window as a sliding cycle. Obtain the target patient's historical dialysis time and mark the historical dialysis time in all cycles to obtain the labeled dates. Then, align all cycles in chronological order, starting from the first date of the cycle and label this date as date 1. Identify the number of labeled dates Bb within date 1, and then obtain the total number of cycles Bx. Then, use the calculation formula F1=Bb÷Bx to obtain the visit frequency F1 of date 1. Then, calculate the visit frequency Fi of each date in the cycle according to the above method, where i represents the date in the cycle. The frequency Fi of all dates in the cycle is arranged in descending order to form a frequency sequence. Then, based on the cycle frequency in the medical order information, the first position in the frequency sequence is set as the starting position. The final position corresponding to the cycle frequency is identified, and the frequency of visits between the starting position and the ending position is identified. The date corresponding to this frequency of visits is marked as a key date. Here, the cycle frequency refers to the number of times the target patient needs dialysis in the dialysis time cycle in the medical order information. For example, when the medical order information is three times a week, three times is the cycle frequency. Next, obtain all the cycle periods, match the tag date with the key date. If the key date and the tag date coincide, generate the information value ME=0; otherwise, if the tag date and the key date do not coincide, generate the information value ME=1. Extract all information values MEj, where j represents different information values, and then use the formula The date discrete values UR are obtained, where J represents the total number of information values and MEa is the mean of the information values MEj. Methods for calculating discrete interval values include: Obtain the target patient's historical dialysis times, calculate the time interval Tn between adjacent dialysis times, and then use the calculation formula... The discrete values UT are obtained, where n represents different interval times, N represents the total number of interval times, and Ta is the mean of the interval times. Methods for determining inertial dialysis time include: The date discrete value UR and the interval discrete value UT are compared. If UR≤UT, the key dates in the cycle are marked as the inertial dialysis time of the target patient according to the cycle period. If UR>UT, the most recent dialysis time of the target patient and the mean interval time Ta are obtained. The next dialysis time of the target patient is determined according to the mean interval time Ta and marked as the inertial dialysis time. All dialysis patients were sequentially marked as target patients and processed according to the above method to determine the inertial dialysis time for each dialysis patient; Then, a one-way communication connection is established between the dialysis time analysis module and the visit prediction module, and the inertial dialysis time of dialysis patients is transmitted to the visit prediction module. The user information analysis module receives patient information from dialysis patients, extracts the treatment process from the patient information, analyzes the treatment process, and identifies key diagnostic points for dialysis patients. Specific methods for identifying key diagnostic points include: Based on basic medical information, the routine testing process for dialysis patients is obtained and divided into several testing nodes, including weight measurement, blood pressure, heart rate and blood glucose. Arbitrarily select a test node, obtain all test reports of this test node, and identify abnormal reports among the test reports. Abnormal reports refer to test reports whose values are not within the normal range of the target patient. Count the number of abnormal reports and test reports respectively, divide the number of abnormal reports by the number of test reports, and mark the result as the abnormality frequency Rf. The abnormal frequency Rf is compared with the abnormal threshold Ry. If Rf < Ry, the test node is marked as a normal node. If Rf ≥ Ry, the test node is marked as a key diagnostic node. The specific value of the abnormal threshold Ry is obtained by those skilled in the art based on big data calculations. All testing points were processed using the methods described above to identify key diagnostic points for each dialysis patient. Then, a one-way communication connection is established between the user information analysis module and the dynamic resource adjustment module, and the key diagnostic nodes of dialysis patients are transmitted to the dynamic resource adjustment module. The appointment prediction module receives the inertial dialysis time of dialysis patients and, based on this time, determines the target patients for appointment. Specific methods for this determination include: Set a target date, where the target date refers to the estimated date and time for dialysis patients to visit the clinic. Obtain the inertial dialysis time of all dialysis patients, identify dialysis patients whose inertial dialysis time is the same as the target date, and mark these patients as target patients. It should be further explained that after the target date is set, the day before the target date is obtained and marked as the previous date. The target patients within the previous date are obtained, and the target patients within the previous date are compared with the actual patients. The target patients who did not actually receive treatment within the previous date are identified, and the target patients who did not actually receive treatment are remarked as the target patients for the target date. Then, a one-way communication connection is established between the medical visit prediction module and the dynamic resource adjustment module, and the target medical visit patient is transmitted to the dynamic resource adjustment module; The dynamic resource adjustment module is used to receive the target patients within the target date and obtain the key diagnostic nodes of each target patient. Then, based on the number of target patients and the corresponding key diagnostic nodes, medical staff adjust the medical resources in advance within the target date. Further, resource adjustment includes precise matching of medical staff, targeted preparation of medical consumables, and adjustment of equipment and venues.
[0019] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. A cloud-based big data management system for hemodialysis, characterized in that, include: The user information collection module is used to identify hemodialysis patients in the medical system and collect their patient information. The dialysis time analysis module is used to extract dialysis time from patient information and analyze the dialysis time of dialysis patients. Based on the dialysis cycle time and interval time, it calculates the date discrete value and interval discrete value of the target patient respectively, compares the date discrete value and interval discrete value, and determines the inertial dialysis time of the target patient. The user information analysis module is used to extract the diagnosis and treatment process from the patient information, divide the diagnosis and treatment process into several test nodes, calculate the abnormal frequency of each test node, and determine the key diagnostic nodes for dialysis patients based on the abnormal frequency. The appointment prediction module is used to receive the inertial dialysis time of dialysis patients and, based on the inertial dialysis time, to determine the target patients to be seen within the target date. The dynamic resource adjustment module is used to receive the target patients within the target date and obtain the key diagnostic nodes for each target patient. Then, medical staff adjust the medical resources in advance based on the number of target patients and the corresponding key diagnostic nodes.
2. The hemodialysis big data management system based on a cloud platform according to claim 1, characterized in that, Methods for obtaining dialysis cycle time include: Dialysis patients are sequentially marked as target patients. The medical orders and historical dialysis times of the target patients are obtained. Based on the medical order information, the dialysis cycle time of the target patients is identified. The medical order information refers to the dialysis time arrangement of patients led and formulated by the doctor.
3. The hemodialysis big data management system based on a cloud platform according to claim 1, characterized in that, Methods for calculating discrete date values include: Set the cycle time based on the dialysis cycle duration; Obtain the target patient's historical dialysis time and mark the historical dialysis time in all cycles to obtain the labeled dates. Then, align all cycles in chronological order, starting from the first date of the cycle and label this date as date 1. Identify the number of labeled dates Bb within date 1, and then obtain the total number of cycles Bx. Then, use the calculation formula F1=Bb÷Bx to obtain the visit frequency F1 of date 1. Then, calculate the visit frequency Fi of each date in the cycle according to the above method, where i represents the date in the cycle. The frequency Fi of all dates in the cycle is arranged in descending order to form a frequency sequence. Then, based on the cycle frequency in the medical order information, the first position in the frequency sequence is set as the starting position, the final position corresponding to the cycle frequency is identified, the frequency of visits between the starting position and the ending position is identified, and the date corresponding to this frequency of visits is marked as the key date. Here, the cycle frequency refers to the number of times the target patient needs dialysis in the dialysis time cycle in the medical order information. Next, obtain all the cycle periods, match the tag date with the key date. If the key date and the tag date coincide, generate the information value ME=0; otherwise, if the tag date and the key date do not coincide, generate the information value ME=1. Extract all information values MEj, where j represents different information values, and then use the formula The date discrete values UR are obtained, where J represents the total number of information values and MEa is the mean of the information values MEj.
4. The hemodialysis big data management system based on a cloud platform according to claim 3, characterized in that, Methods for setting the cycle period include: Identify the current real-time date and time, and then set the dialysis cycle time as the window length of the sliding window based on the dialysis cycle time. Obtain the time before the current date and time, and set the sliding window according to the natural cycle, that is, align the sliding window with the natural cycle. Then, set the movement step size according to the length of the dialysis cycle time, and push forward sequentially according to the set movement step size to generate continuous sliding windows. Mark each sliding window as a cycle period.
5. A cloud-based hemodialysis big data management system according to claim 1, characterized in that, Methods for calculating discrete interval values include: Obtain the target patient's historical dialysis times, calculate the time interval Tn between adjacent dialysis times, and then use the calculation formula... The discrete values UT are obtained, where n represents different interval times, N represents the total number of interval times, and Ta is the mean of the interval times.
6. The hemodialysis big data management system based on a cloud platform according to claim 1, characterized in that, Methods for determining inertial dialysis time include: The date discrete value UR and the interval discrete value UT are compared. If UR≤UT, the key dates in the cycle are marked as the inertial dialysis time of the target patient according to the cycle period. If UR>UT, the most recent dialysis time of the target patient and the mean interval time Ta are obtained. The next dialysis time of the target patient is determined according to the mean interval time Ta and marked as the inertial dialysis time.
7. A cloud-based hemodialysis big data management system according to claim 1, characterized in that, The methods for determining key diagnostic nodes include: Based on basic medical information, the routine testing process for dialysis patients is obtained and divided into several testing nodes; Arbitrarily select a test node, obtain all test reports of this test node, and identify abnormal reports among the test reports. Abnormal reports refer to test reports whose values are not within the normal range of the target patient. Count the number of abnormal reports and test reports respectively, divide the number of abnormal reports by the number of test reports, and mark the result as the abnormality frequency Rf. The abnormal frequency Rf is compared with the abnormal threshold Ry. If Rf < Ry, the test node is marked as a normal node. If Rf ≥ Ry, the test node is marked as a key diagnostic node.
8. A cloud-based hemodialysis big data management system according to claim 1, characterized in that, Methods for identifying target patients include: Set a target date, which is the estimated date and time for dialysis patients to visit the clinic. Obtain the inertial dialysis time of all dialysis patients, identify dialysis patients whose inertial dialysis time is the same as the target date, and mark these patients as target patients.
9. A cloud-based hemodialysis big data management system according to claim 8, characterized in that, Once the target date is set, retrieve the day before the target date and mark it as the preceding date. Retrieve the target patients within the preceding date, compare the target patients within the preceding date with the actual patients, identify the target patients within the preceding date who did not actually receive treatment, and remark the target patients who did not actually receive treatment as target patients for the target date.
10. A cloud-based hemodialysis big data management system according to claim 1, characterized in that, It also includes a medical information collection module, which is used to collect basic medical information, including medical care information and medical resource information. The medical information collection module then transmits the collected basic medical information to the user information analysis module.