Nephropathy medication interaction risk screening system

Abstract

The invention relates to the technical field of medical care informatics, in particular to a nephropathy medication interaction risk screening system which comprises a data acquisition unit, an SOC chip processing and analysis unit, a risk early warning unit and a man-machine interaction and communication unit. Through multi-source data access, dynamic weight scheduling and Kalman filtering fusion correction, accurate and on-demand acquisition of nephropathy core indexes is realized, and real-time intelligent discrimination of personalized medication risks is realized by relying on knowledge base support, adaptive path adjustment and a safe atomization updating mechanism, so that the risk of the nephropathy is accurately determined. Therefore, the problems that traditional system data acquisition lacks specialized pertinence, risk assessment personalization is insufficient, and model updating is lagged are effectively solved, and the accuracy and timeliness of nephropathy medication interaction risk screening are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of healthcare informatics technology, specifically to a kidney disease medication interaction risk screening system. Background Technology

[0002] The kidneys, as the core excretory and endocrine organs of the human body, perform key physiological functions such as filtering metabolic waste from the blood, maintaining water, electrolyte, and acid-base balance, and secreting erythropoietin (to regulate hematopoiesis) and renin (to regulate blood pressure). Kidney disease, also known as nephropathy, refers to a group of diseases caused by various factors such as infection, immune abnormalities, metabolic disorders (e.g., diabetes, hypertension), drug toxicity, and genetics, leading to damage to the structure or function of the kidneys and resulting in the aforementioned physiological dysfunctions. It encompasses various types, including acute kidney injury, chronic kidney disease, nephrotic syndrome, and glomerulonephritis.

[0003] Because patients with kidney disease generally have pathological characteristics such as decreased glomerular filtration rate and weakened renal tubular excretion capacity, their absorption, distribution, metabolism and excretion of drugs differ significantly from those of the general population. At the same time, patients with kidney disease often need to use multiple drugs in combination, such as immunosuppressants, antihypertensive drugs, and diuretics. The risk of drug-drug and drug-disease interactions is much higher than that of ordinary patients. If medication is used inappropriately, it may not only reduce the treatment effect, but also accelerate the deterioration of kidney function and even induce serious complications.

[0004] Therefore, medication risk screening for kidney disease patients needs to rely on "specialized data support, personalized assessment logic, and dynamic model iteration" to achieve precise prevention and control. Current technologies suffer from several problems: traditional medication risk screening systems lack nephrology-specific data collection, rely on general rules for risk assessment with insufficient personalization, and suffer from lagging model and rule updates and poor safety.

[0005] Based on this, the present invention provides a kidney disease medication interaction risk screening system to solve the above-mentioned technical problems. Summary of the Invention

[0006] The purpose of this invention is to provide a kidney disease medication interaction risk screening system. This invention utilizes multi-source data access, dynamic weight scheduling, and Kalman filter fusion correction to achieve accurate and on-demand collection of core kidney disease indicators. Furthermore, relying on knowledge base support, adaptive path adjustment, and a secure atomic update mechanism, it enables real-time intelligent identification of personalized medication risks. This effectively solves the problems of traditional systems, such as lack of specialized data collection, insufficient personalization of risk assessment, and lagging model updates, significantly improving the accuracy and timeliness of kidney disease medication interaction risk screening.

[0007] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a kidney disease medication interaction risk screening system, comprising a data acquisition unit, a SOC chip processing and analysis unit, a risk warning unit, and a human-computer interaction and communication unit, wherein: The data acquisition unit: based on a dynamic weight acquisition mechanism, it collects physiological parameters and medication information of kidney disease patients through multi-source data fusion and correction; The SOC chip processing and analysis unit: based on the dynamic characteristics of the patient's physiological state and medication data, performs adaptive risk assessment model analysis to intelligently identify and update the risks of personalized medication interaction in real time; The risk warning unit is used to identify potential medication risks for kidney disease patients based on the analysis results and generate graded warning prompts. The human-computer interaction and communication unit is used for information exchange between doctors, patients and the system, and supports data uploading and remote monitoring through a secure communication protocol.

[0008] The data acquisition unit includes a multi-source data access module, a dynamic weight scheduling module, and a data fusion and correction module, wherein: The multi-source data access module is used to connect to electronic medical records, prescription systems, and portable testing devices to obtain patients' physiological parameters and medication data. The dynamic weight scheduling module is used to dynamically allocate the collection priority and frequency of core renal function indicators according to the needs of renal disease risk analysis. The data fusion and correction module is used to perform spatiotemporal alignment and error correction on the acquired multi-source data.

[0009] The physiological parameters include glomerular filtration rate, serum creatinine, blood urea nitrogen, urinary protein quantification, blood pressure, heart rate, blood glucose, serum potassium, serum sodium, serum phosphorus, serum uric acid, and hemoglobin; the medication data includes the generic name of the drug, dosage form, specification, single dose, frequency of administration, route of administration, start and end time of administration, history of previous drug allergies, history of adverse reactions to similar drugs, and list of combined medications; the core indicators of renal function include glomerular filtration rate, serum creatinine, urinary protein quantification, and blood urea nitrogen.

[0010] The dynamic weight scheduling module dynamically allocates the collection priority and frequency of core renal function indicators according to the needs of renal disease risk analysis. The specific operation is as follows: A1: Risk scenario identification: Obtain the current patient's kidney disease stage, whether nephrotoxic drugs are being used concurrently, and whether there are signs of acute kidney injury risk, and determine the corresponding risk analysis needs level of high / medium / low; A2: Weighting Factor Assignment: For core indicators of kidney function, a weighting factor matrix is ​​preset to match the risk requirement level. ① Under high-risk conditions, the weighting factors for glomerular filtration rate and serum creatinine are assigned to 0.4-0.5, and the weighting factors for urinary protein quantification and blood urea nitrogen are assigned to 0.25-0.3; ② Under medium / low risk conditions, the weight of core indicators will be reduced proportionally, while the auxiliary weight of basic physiological parameters will be increased; A3: Data Acquisition Parameter Scheduling: Data acquisition priority is determined based on weighting factors, and the data acquisition frequency is dynamically adjusted based on risk level. ① Under high-risk requirements, the collection frequency of the top 2 priority core indicators is set to 1-2 hours / time; ② Medium-risk requirements are set at 4-6 hours / time; ③ Low-risk needs are set at 12-24 hours / time.

[0011] The SOC chip processing and analysis unit includes a knowledge base management module, an adaptive computing engine module, and a real-time update agent module, wherein: The knowledge base management module is used to store and manage drug interaction rules, pharmaceutical knowledge graphs, and patient-specific parameter models. The adaptive computing engine module is used to load and run the risk assessment algorithm, dynamically adjust the model calculation path according to the input physiological parameters, and perform personalized risk identification. The real-time update agent module is used to receive and verify incremental update packages of models or rules from the cloud in the background.

[0012] The adaptive computing engine module loads and runs a risk assessment algorithm, dynamically adjusting the model's calculation path based on the input physiological parameters to perform personalized risk identification. The specific operations are as follows: B1: Receives patient physiological parameters input from the data acquisition unit, prioritizes the parameters based on the risk correlation of kidney disease medication, and selects the top 3 core influencing parameters; B2: Pre-set multiple risk assessment sub-pathways matched with core physiological parameters: ① If the core parameter is "eGFR < 30 ml / min + combined with diabetes", the "medication pathway for renal failure with metabolic abnormalities" will be triggered; ② If the core parameter is "urine protein quantification > 3.5g / 24h + combined use of immunosuppressants", the "nephrotic syndrome immunosuppression pathway" is triggered; B3: For the triggered calculation path, call the patient's historical medication data and make personalized corrections to the risk assessment algorithm parameters within the path; B4: If multiple core parameters trigger cross paths, the calculation results of each path are integrated through the "weighted voting method" to finally output a judgment result that includes "risk level, key influencing factors, and personalized avoidance suggestions".

[0013] In B4, the calculation results of each path are combined using a "weighted voting method," and the specific expression is as follows: ; in, For the final comprehensive risk score, Let i be the weight of the i-th triggering path. For the risk assessment sub-model of the i-th dedicated route, n is the number of paths activated, representing a personalized parameter adjusted based on the patient's historical medication data.

[0014] The real-time update agent module receives and verifies incremental update packages for models or rules from the cloud in the background. The specific operations are as follows: C1: Update package proactive detection and request: Sends version verification requests to the cloud update server at preset intervals to obtain the metadata of the package to be updated; C2: Incremental packet encrypted transmission: A secure channel with the cloud is established using an asymmetric encryption algorithm to receive encrypted incremental update packets; C3: Multi-level validation: ① Integrity verification: Calculate the actual hash value of the received packet and compare it with the hash value in the metadata. If they do not match, the update is rejected. ② Compatibility verification: Verify whether the system version adapted by the update package matches the firmware version of the local SOC chip processing and analysis unit. If they do not match, trigger downgrade compatibility processing. ③Signature verification: Verify whether the digital signature of the update package comes from the authorization server to prevent malicious tampering and injection of the package; C4: Atomic Update and Rollback: If the verification passes, update the local model / rule base using atomic replacement; if the verification fails, log the error and report the exception information to the cloud.

[0015] The risk warning unit includes a risk decision-making module, a prompt generation module, and a hierarchical output driving module, wherein: The risk decision-making module is used to receive analysis results and automatically determine the severity level of the current risk based on a preset risk level threshold matrix. The prompt generation module is used to convert the risk level and interaction mechanism determined by the machine into natural language descriptions and specific intervention suggestions for doctors or patients. The graded output driver module is used to map different levels of warning signals into different forms of sound, light, and electrical prompts, and drive the UI interface and alarm to provide differentiated alarms.

[0016] The human-computer interaction and communication unit includes an interactive interface module, a secure communication gateway module, and a permission management module, wherein: The interactive interface module is used to provide a graphical user interface, receive user input, and clearly present warning information, medication lists, and patient data. The secure communication gateway module is used to encrypt and encapsulate all data to be transmitted, and to reliably exchange data with the cloud server through an encrypted communication protocol. The permission management module is used to verify user identity and assign different data access and operation permissions according to roles.

[0017] Compared with the prior art, the beneficial effects of the present invention are: This invention achieves accurate and on-demand collection of core indicators of kidney disease through multi-source data access, dynamic weight scheduling, and Kalman filter fusion correction. Relying on knowledge base support, adaptive path adjustment, and secure atomic update mechanism, it realizes real-time intelligent identification of personalized medication risks. This effectively solves the problems of traditional systems, such as lack of specialized targeting in data collection, insufficient personalization of risk assessment, and lagging model updates, and significantly improves the accuracy and timeliness of interactive risk screening for kidney disease medication. Attached Figure Description

[0018] Figure 1 This is a system diagram of an interactive risk screening system for kidney disease medication according to the present invention.

[0019] Figure 2 This is a flowchart of dynamic weight scheduling in a kidney disease medication interaction risk screening system of the present invention.

[0020] Figure 3 This is a flowchart of an adaptive risk assessment system for interactive risk screening of kidney disease medication according to the present invention.

[0021] Explanation of icon numbers: 1. Data Acquisition Unit; 11. Multi-Source Data Access Module; 12. Dynamic Weight Scheduling Module; 13. Data Fusion and Correction Module; 2. SOC Chip Processing and Analysis Unit; 21. Knowledge Base Management Module; 22. Adaptive Computing Engine Module; 23. Real-time Update Agent Module; 3. Risk Warning Unit; 31. Risk Decision Module; 32. Prompt Generation Module; 33. Hierarchical Output Driver Module; 4. Human-Computer Interaction and Communication Unit; 41. Interactive Interface Module; 42. Secure Communication Gateway Module; 43. Access Control Module. Detailed Implementation

[0022] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0023] like Figure 1-3 As shown, this embodiment provides a kidney disease medication interaction risk screening system, including a data acquisition unit 1, a SOC chip processing and analysis unit 2, a risk warning unit 3, and a human-computer interaction and communication unit 4, wherein: the data acquisition unit 1: based on a dynamic weight acquisition mechanism, it collects physiological parameters and medication information of kidney disease patients through multi-source data fusion and correction; the SOC chip processing and analysis unit 2: based on the dynamic characteristics of the patient's physiological state and medication data, it performs adaptive risk assessment model analysis to intelligently identify and update personalized medication interaction risks in real time; the risk warning unit 3: it is used to identify potential medication risks of kidney disease patients based on the analysis results and generate graded warning prompts; the human-computer interaction and communication unit 4: it is used for information interaction between doctors, patients and the system, and supports data upload and remote monitoring through a secure communication protocol.

[0024] It should be noted that the data acquisition unit 1 transmits the accurately collected physiological and medication information to the SOC chip processing and analysis unit 2 for real-time intelligent risk assessment. The analysis results generated trigger the risk warning unit 3 to produce graded warning prompts, and the human-computer interaction and communication unit 4 completes internal and external information exchange and remote monitoring.

[0025] In this embodiment, it should also be noted that the data acquisition unit 1 includes a multi-source data access module 11, a dynamic weight scheduling module 12, and a data fusion correction module 13. Specifically: the multi-source data access module 11 is used to interface with electronic medical records, prescription systems, and portable testing devices to acquire patient physiological parameters and medication data; physiological parameters include glomerular filtration rate, serum creatinine, blood urea nitrogen, quantified urine protein, blood pressure, heart rate, blood glucose, serum potassium, serum sodium, serum phosphorus, serum uric acid, and hemoglobin; medication data includes the generic name of the drug, dosage form, specifications, single dose, frequency of administration, route of administration, start and end time of administration, history of previous drug allergies, history of adverse reactions to similar drugs, and a list of combined medications; the dynamic weight scheduling module 12 is used to dynamically allocate the collection priority and frequency of core renal function indicators according to the needs of renal disease risk analysis; core renal function indicators include glomerular filtration rate, serum creatinine, quantified urine protein, and blood urea nitrogen. The specific operations are as follows: A1: Risk Scenario Identification: Obtain the current patient's kidney disease stage, whether nephrotoxic drugs are being used concurrently, and whether there are signs of acute kidney injury risk, and determine the corresponding risk analysis requirement level as high / medium / low; A2: Weighting Factor Assignment: For core kidney function indicators, preset a weighting factor matrix matching the risk requirement level: ① Under high-risk requirements, the weighting factor for glomerular filtration rate and serum creatinine is assigned to 0.4-0.5, and the weighting factor for urinary protein quantification and blood urea nitrogen is assigned to 0.25-0.3; ② Under medium / low-risk requirements, the weight of core indicators is proportionally reduced, while the auxiliary weight of basic physiological parameters is increased; A3: Collection Parameter Scheduling: Determine the collection priority according to the weighting factor ranking, and dynamically adjust the collection frequency based on the risk level: ① Under high-risk requirements, the collection frequency of the top 2 priority core indicators is set to 1-2 hours / time; ② Under medium-risk requirements, it is set to 4-6 hours / time; ③ Under low-risk requirements, it is set to 12-24 hours / time. Data Fusion Correction Module 13: Used for spatiotemporal alignment and error correction of the acquired multi-source data.

[0026] It should be noted that the multi-source data access module 11 is responsible for obtaining raw physiological and medication data from heterogeneous data sources, the dynamic weight scheduling module 12 dynamically calculates the collection priority and frequency of core indicators based on real-time risk scenarios, and the data fusion correction module 13 performs spatiotemporal alignment and error correction on the multi-source data collected according to priority scheduling.

[0027] Furthermore, it should be noted that the definitions and criteria for kidney disease staging, nephrotoxic drugs, and acute kidney injury signals in A1 are as follows: Kidney disease staging: determined according to the KDIGO guidelines (e.g., CKD stage 1: eGFR ≥ 90 ml / min; CKD stage 5: eGFR < 15 ml / min); Nephrotoxic drugs: a pre-set list of nephrotoxic drugs (e.g., aminoglycoside antibiotics, nonsteroidal anti-inflammatory drugs, traditional Chinese medicines containing aristolochic acid, etc.) is used for identification through drug ATC code matching; Acute kidney injury signals: any of the following conditions are considered a risk signal: ① serum creatinine increases by ≥ 26.5 μmol / L within 48 hours; ② serum creatinine increases by ≥ 1.5 times the baseline value within 7 days; ③ urine output < 0.5 ml / (kg·h) lasts for more than 6 hours. The specific values ​​and dynamic adjustment logic of the weights for medium-risk A2: Weight factor matrix under medium-risk requirements: glomerular filtration rate (0.35), serum creatinine (0.3), urinary protein quantification (0.2), blood urea nitrogen (0.15), with the weights of basic physiological parameters (blood pressure, blood glucose) increased to 0.2; Weight factor matrix under low-risk requirements: glomerular filtration rate (0.25), serum creatinine (0.2), urinary protein quantification (0.15), blood urea nitrogen (0.1), with the weights of basic physiological parameters increased to 0.3; Weight adjustment trigger condition: If the patient experiences "eGFR decrease of ≥10ml / min within 72 hours" or "urinary protein quantification sudden increase of ≥1g / 24h", the risk requirement level will be automatically increased by 1 level (e.g., from medium risk to high risk), and the weight factors will be updated synchronously. Error correction algorithm: The Kalman filter algorithm is used to correct the random error of the portable device, and the formula is as follows: ① State prediction: ② Error covariance prediction: ③ Kalman gain: ④ Status Update: ⑤ Error covariance update: ;in: The data value after correction at time k. Let be the original detection value at time k, A be the state transition matrix, H be the observation matrix, Q be the process noise covariance, and R be the observation noise covariance. Let B be the corrected data value at time K-1, and let B be the control input matrix. This is the control input vector at time K-1. Let K be the error covariance matrix at time K-1. It is an identity matrix.

[0028] In this embodiment, it should also be noted that the SOC chip processing and analysis unit 2 includes a knowledge base management module 21, an adaptive computing engine module 22, and a real-time update agent module 23, wherein: the knowledge base management module 21 is used to store and manage drug interaction rules, pharmaceutical knowledge graphs, and patient-specific parameter models; the adaptive computing engine module 22 is used to load and run risk assessment algorithms, dynamically adjust the model calculation path according to the input physiological parameters, and perform personalized risk discrimination; the specific operation is as follows: B1: Receives patient physiological parameters input by the data acquisition unit 1, prioritizes the parameters based on the risk correlation of kidney disease medication, and selects the TOP3 core influencing parameters; B2 The system pre-defines multiple risk assessment sub-pathways matched with core physiological parameters: ① If the core parameter is "eGFR < 30 ml / min + combined with diabetes," the "Renal Failure Combined with Metabolic Abnormalities Medication Pathway" is triggered; ② If the core parameter is "Urine Protein Quantitative > 3.5 g / 24 h + Combined with Immunosuppressants," the "Nephropathy Syndrome Immunosuppression Pathway" is triggered; B3: For the triggered calculation path, the system calls the patient's historical medication data and performs personalized adjustments to the risk assessment algorithm parameters within the path; B4: If multiple core parameters trigger intersecting paths, the calculation results of each path are merged using a "weighted voting method," and the final output includes a judgment result containing "risk level, key influencing factors, and personalized avoidance suggestions." The specific expression for merging the calculation results of each path using the "weighted voting method" in B4 is as follows: ;

[0029] in, For the final comprehensive risk score, Let be the weight of the i-th trigger path. The value is determined based on the close correlation between the triggering path and the patient's current kidney disease status. It is dynamically assigned by the system based on a pharmaceutical knowledge graph and a personalized parameter model for the patient, with a value range of [range missing]. And satisfy , For the risk assessment sub-model of the i-th dedicated path, the risk assessment sub-model is... For machine learning-based regression models (such as logistic regression or random forest), the output range is [0, 1], representing the risk probability under a single path. This is achieved by associating the risk probability of each path with its corresponding weight. Multiply and then sum to achieve a weighted average, ultimately The value ranges from [0, 1]. A larger value indicates a higher overall risk. x is the vector of input patient physiological parameters. n represents the number of activated paths, representing personalized parameters corrected based on the patient's historical medication data. Real-time update agent module 23: used to receive and verify incremental update packages for models or rules from the cloud in the background. Specific operations are as follows: C1: Update package active detection and request: Sends a version verification request to the cloud update server at a preset cycle to obtain the metadata of the package to be updated, including version number, hash value, and compatible system version range; C2: Incremental package encrypted transmission: Establishes a secure channel with the cloud using an asymmetric encryption algorithm to receive encrypted incremental update packages; C3: Multi-level verification: ① Integrity verification: Calculates the actual hash value of the received package and compares it with the hash value in the metadata; if inconsistent, the update is rejected; ② Compatibility verification: Verifies whether the system version adapted by the update package matches the firmware version of the local SOC chip processing and analysis unit 2; if incompatible, triggers downgrade compatibility processing; ③ Signature verification: Verifies whether the digital signature of the update package comes from the authorized server to prevent malicious tampering of the package; C4: Atomic update and rollback: If verification passes, updates the local model / rule base using atomic replacement; if verification fails, records an error log and reports the abnormal information to the cloud.

[0030] It should be noted that the knowledge base management module 21 provides the system with core drug rules and model support, the adaptive computing engine module 22 dynamically calls and personally corrects the model paths in the knowledge base based on real-time data to make risk judgments, and the real-time update agent module 23 ensures that the knowledge base can safely and reliably receive incremental updates from the cloud.

[0031] Furthermore, it should be noted that the calculation method for the risk association of kidney disease medication uses the Pearson correlation coefficient to calculate the association between physiological parameters and medication risk, as shown in the following formula: Where: r is the correlation (range [-1, 1], the larger the absolute value, the stronger the correlation), x is the physiological parameter value (e.g., eGFR), y is the marker of drug risk occurrence (1 = risk occurred, 0 = no risk occurred), and n is the sample size; Screening rules: retain parameters with an absolute value of r ≥ 0.6, sort them from largest to smallest absolute value of r, and select the TOP3 as core influencing parameters (e.g., eGFR: r = -0.75; serum creatinine: r = 0.72; urinary protein quantification: r = 0.68). Atomic update process in C4: ① Backup: Fully back up the existing local model / rule base to a separate partition (backup file naming format: "backup_date_version number", such as "backup_20240520_V2.1"); ② Replace: Unzip the verified incremental update package and overwrite the old version files (using a "write first, delete later" mechanism to avoid data loss due to interruption of file replacement); ③ Verification: After the update is completed, calculate the hash value of the new model / rule base and compare it with the verification value provided by the cloud to confirm successful replacement.

[0032] Rollback trigger conditions: If step ② is interrupted (e.g., power failure) or step ③ verification fails, the backup file will be automatically loaded to restore the state before the update, and "update failed + reason for failure" will be sent to the cloud.

[0033] In this embodiment, it should also be noted that the risk warning unit 3 includes a risk decision module 31, a prompt generation module 32, and a graded output driving module 33, wherein: the risk decision module 31 is used to receive the analysis results and automatically determine the severity level of the current risk based on a preset risk level threshold matrix; the prompt generation module 32 is used to convert the risk level determined by the machine and the interaction mechanism into a natural language description and specific intervention suggestions for doctors or patients; the graded output driving module 33 is used to map different levels of warning signals into different forms of sound, light, and electricity prompts, and drive the UI interface and alarm to provide differentiated alarms.

[0034] It should be noted that the risk decision module 31 determines the risk level based on the analysis results, the prompt generation module 32 converts the level and mechanism into natural language suggestions, and the hierarchical output driving module 33 maps the warning results into a multimodal alarm form.

[0035] Furthermore, it should be noted that the risk level threshold matrix is ​​as follows:

[0036] Decision logic: If If the score is at the threshold of 50, the risk level is automatically adjusted based on whether the patient has diabetes / hypertension comorbidities (if comorbidities are present, the risk level is raised to medium risk; otherwise, the risk level remains low risk).

[0037] In this embodiment, it should also be noted that the human-computer interaction and communication unit 4 includes an interactive interface module 41, a secure communication gateway module 42, and a permission management module 43, wherein: the interactive interface module 41 is used to provide a graphical user interface, receive user input, and clearly present warning information, medication lists, and patient data; the secure communication gateway module 42 is used to encrypt and encapsulate all data to be transmitted, and to reliably exchange data with the cloud server through an encrypted communication protocol; the permission management module 43 is used to verify the user's identity and assign different data access and operation permissions according to the role.

[0038] It should be noted that the permission management module 43 verifies the user's identity and grants permissions, the interactive interface module 41 provides data interaction and presentation functions for the corresponding user roles based on permissions, and the secure communication gateway module 42 encrypts all internal and external communication data.

[0039] Furthermore, it should be noted that the data encryption uses the AES-256 symmetric encryption algorithm (256-bit key length, CBC encryption mode, 16-byte initialization vector IV) to encrypt transmitted data. The communication protocol is as follows: Interaction with the hospital's HIS system uses HTTPS 1.2 protocol with the TLS encryption suite TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384; Interaction with the cloud server uses the MQTT 3.1.1 protocol, communicating via TCP port 1883, and enabling username / password authentication + SSL / TLS encryption. The authentication mechanism supports the following: Doctors / Pharmacists: "Account password (complexity requirement: at least 8 characters including uppercase and lowercase letters + numbers + special symbols) + hospital ID NFC dual verification"; Patients: "Mobile phone verification code + fingerprint verification"; Administrators: "Hardware key (USB-Key) + administrator password" triple verification.

[0040] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0041] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A kidney disease medication interaction risk screening system, characterized in that, It includes a data acquisition unit (1), a SOC chip processing and analysis unit (2), a risk warning unit (3), and a human-computer interaction and communication unit (4), wherein: The data acquisition unit (1) collects physiological parameters and medication information of kidney disease patients based on a dynamic weight acquisition mechanism and through multi-source data fusion correction. The SOC chip processing and analysis unit (2) performs adaptive risk assessment model analysis based on the dynamic characteristics of the patient's physiological state and medication data, and performs real-time intelligent judgment and update of personalized medication interaction risks. The risk warning unit (3) is used to identify potential medication risks for kidney disease patients based on the analysis results and generate graded warning prompts. The human-computer interaction and communication unit (4) is used for information interaction between doctors, patients and the system, and supports data uploading and remote monitoring through a secure communication protocol.

2. The kidney disease medication interaction risk screening system according to claim 1, characterized in that, The data acquisition unit (1) includes a multi-source data access module (11), a dynamic weight scheduling module (12), and a data fusion correction module (13), wherein: The multi-source data access module (11) is used to connect to electronic medical records, prescription systems, and portable testing devices to obtain patient physiological parameters and medication data. The dynamic weight scheduling module (12) is used to dynamically allocate the collection priority and frequency of core indicators of renal function according to the needs of renal disease risk analysis. The data fusion correction module (13) is used to perform spatiotemporal alignment and error correction on the acquired multi-source data.

3. The kidney disease medication interaction risk screening system according to claim 2, characterized in that, The physiological parameters include glomerular filtration rate, serum creatinine, blood urea nitrogen, urinary protein quantification, blood pressure, heart rate, blood glucose, serum potassium, serum sodium, serum phosphorus, serum uric acid, and hemoglobin; the medication data includes the generic name of the drug, dosage form, specification, single dose, frequency of administration, route of administration, start and end time of administration, history of previous drug allergies, history of adverse reactions to similar drugs, and list of combined medications; the core indicators of renal function include glomerular filtration rate, serum creatinine, urinary protein quantification, and blood urea nitrogen.

4. The kidney disease medication interaction risk screening system according to claim 2, characterized in that, The dynamic weight scheduling module (12) dynamically allocates the collection priority and frequency of core renal function indicators according to the needs of renal disease risk analysis. The specific operation is as follows: A1: Risk scenario identification: Obtain the current patient's kidney disease stage, whether nephrotoxic drugs are being used concurrently, and whether there are signs of acute kidney injury risk, and determine the corresponding risk analysis needs level of high / medium / low; A2: Weighting Factor Assignment: For core indicators of kidney function, a weighting factor matrix is ​​preset to match the risk requirement level. ① Under high-risk conditions, the weighting factors for glomerular filtration rate and serum creatinine are assigned to 0.4-0.5, and the weighting factors for urinary protein quantification and blood urea nitrogen are assigned to 0.25-0.3; ② Under medium / low risk conditions, the weight of core indicators will be reduced proportionally, while the auxiliary weight of basic physiological parameters will be increased; A3: Data Acquisition Parameter Scheduling: Data acquisition priority is determined based on weighting factors, and the data acquisition frequency is dynamically adjusted based on risk level. ① Under high-risk requirements, the collection frequency of the top 2 priority core indicators is set to 1-2 hours / time; ② Medium-risk requirements are set at 4-6 hours / time; ③ Low-risk needs are set at 12-24 hours / time.

5. The kidney disease medication interaction risk screening system according to claim 1, characterized in that, The SOC chip processing and analysis unit (2) includes a knowledge base management module (21), an adaptive computing engine module (22), and a real-time update agent module (23), wherein: The knowledge base management module (21) is used to store and manage drug interaction rules, pharmaceutical knowledge graphs and patient-personalized parameter models. The adaptive computing engine module (22) is used to load and run the risk assessment algorithm, dynamically adjust the model calculation path according to the input physiological parameters, and perform personalized risk identification. The real-time update agent module (23) is used to receive and verify incremental update packages of models or rules from the cloud in the background.

6. The kidney disease medication interaction risk screening system according to claim 5, characterized in that, The adaptive computing engine module (22) loads and runs a risk assessment algorithm, dynamically adjusts the model calculation path based on the input physiological parameters, and performs personalized risk identification. The specific operation is as follows: B1: Receive the patient's physiological parameters input by the data acquisition unit (1), prioritize the parameters based on the risk correlation of kidney disease medication, and select the TOP3 core influencing parameters; B2: Pre-set multiple risk assessment sub-pathways matched with core physiological parameters: ① If the core parameter is "eGFR < 30 ml / min + combined with diabetes", the "medication pathway for renal failure with metabolic abnormalities" will be triggered; ② If the core parameter is "urine protein quantification > 3.5g / 24h + combined use of immunosuppressants", the "nephrotic syndrome immunosuppression pathway" is triggered; B3: For the triggered calculation path, call the patient's historical medication data and make personalized corrections to the risk assessment algorithm parameters within the path; B4: If multiple core parameters trigger cross paths, the calculation results of each path are integrated through the "weighted voting method" to finally output a judgment result that includes "risk level, key influencing factors, and personalized avoidance suggestions".

7. A kidney disease medication interaction risk screening system according to claim 6, characterized in that, In B4, the calculation results of each path are combined using a "weighted voting method", and the specific expression is as follows: ； in, For the final comprehensive risk score, Let i be the weight of the i-th triggering path. For the risk assessment sub-model of the i-th dedicated route, n is the number of paths activated, representing a personalized parameter adjusted based on the patient's historical medication data.

8. A kidney disease medication interaction risk screening system according to claim 5, characterized in that, The real-time update agent module (23) receives and verifies incremental update packages of models or rules from the cloud in the background. The specific operation is as follows: C1: Update package proactive detection and request: Sends version verification requests to the cloud update server at preset intervals to obtain the metadata of the package to be updated; C2: Incremental packet encrypted transmission: A secure channel with the cloud is established using an asymmetric encryption algorithm to receive encrypted incremental update packets; C3: Multi-level validation: ① Integrity verification: Calculate the actual hash value of the received packet and compare it with the hash value in the metadata. If they do not match, the update is rejected. ② Compatibility verification: Verify whether the system version adapted by the update package matches the firmware version of the local SOC chip processing and analysis unit (2). If they do not match, a downgrade compatibility process will be triggered. ③Signature verification: Verify whether the digital signature of the update package comes from the authorization server to prevent malicious tampering and injection of the package; C4: Atomic Update and Rollback: If the verification passes, update the local model / rule base using atomic replacement; if the verification fails, log the error and report the exception information to the cloud.

9. A kidney disease medication interaction risk screening system according to claim 1, characterized in that, The risk warning unit (3) includes a risk decision module (31), a prompt generation module (32), and a hierarchical output driving module (33), wherein: The risk decision module (31) is used to receive the analysis results and automatically determine the severity level of the current risk based on the preset risk level threshold matrix. The prompt generation module (32) is used to convert the risk level and interaction mechanism determined by the machine into natural language descriptions and specific intervention suggestions for doctors or patients. The graded output driving module (33) is used to map different levels of warning signals into different forms of sound, light and electricity prompts, and drive the UI interface and alarm to perform differentiated alarms.

10. A kidney disease medication interaction risk screening system according to claim 1, characterized in that, The human-computer interaction and communication unit (4) includes an interactive interface module (41), a secure communication gateway module (42), and a permission management module (43), wherein: The interactive interface module (41) is used to provide a graphical user interface, receive user input, and clearly present warning information, medication list, and patient data; The secure communication gateway module (42) is used to encrypt and encapsulate all data to be transmitted, and to reliably exchange data with the cloud server through an encrypted communication protocol. The permission management module (43) is used to verify user identity and assign different data access and operation permissions according to roles.

Images