CRRT treatment intelligent safety early warning system based on multi-mode fusion

Abstract

The invention relates to the technical field of medical information, and discloses a CRRT treatment intelligent safety early warning system based on multi-modal fusion, and the system comprises a data collection and fusion module which is used for collecting multi-modal data from a plurality of heterogeneous data sources in real time, carrying out the standardization and fusion processing of the multi-modal data, and generating fusion data with a unified time sequence; the knowledge model library is used for storing medical knowledge, clinical guidelines and historical data related to CRRT treatment and trained early warning and decision models; according to the system, through multi-modal data fusion and machine learning, the crossing from passive alarm to active early warning is realized, and the safety of a patient is remarkably improved. The intelligent decision support can output an individualized treatment scheme, promote precise medical treatment and optimize resource utilization, the system integration and grading alarm mechanism effectively reduces the cognitive load of medical care, and the self-learning ability of the system ensures that an early warning model can be continuously evolved and promotes the continuous improvement of clinical treatment quality.

Description

TECHNICAL FIELD

[0001] The application relates to the field of medical information technology, in particular to a CRRT treatment intelligent safety warning system based on multi-modal fusion. BACKGROUND

[0002] Continuous renal replacement therapy (CRRT) is a key life support technology for rescuing critical patients, especially patients with acute kidney injury (AKI) or multiple organ failure, in an intensive care unit (ICU). The CRRT treatment process is long (usually more than 24 hours), involves multiple links such as machine connection, steady-state operation, machine disconnection and patient condition management during the period, and is complex in operation, with numerous parameters, and the patient's condition changes rapidly, so that the safety risk in the treatment process is extremely high.

[0003] The current CRRT equipment itself has basic parameter alarm functions (such as pressure overrun, air detection, etc.), but such alarms are isolated and passive responses. The entire CRRT treatment process, from machine preparation, parameter setting, anticoagulation scheme selection, monitoring during treatment to filter replacement decision, is in a scattered management state, lacks an integrated intelligent system, and leads to increased patient risk, overworked medical staff, and treatment quality remaining at the level of experience-based medicine rather than modern precision medicine. Therefore, we propose a CRRT treatment intelligent safety warning system based on multi-modal fusion to solve the above problems. SUMMARY

[0004] In view of the deficiencies of the prior art, the application provides a CRRT treatment intelligent safety warning system based on multi-modal fusion, which solves the problems caused by data isolation, passive alarm and experience-dependent decision-making in current CRRT treatment.

[0005] To achieve the above purpose, the application is implemented by the following technical scheme: a CRRT treatment intelligent safety warning system based on multi-modal fusion, comprising:

[0006] A data acquisition and fusion module is used to acquire multi-modal data from multiple heterogeneous data sources in real time, and to standardize and fuse the data to generate unified time-series fusion data.

[0007] A knowledge model library is used to store medical knowledge, clinical guidelines, historical data related to CRRT treatment, and trained warning and decision-making models.

[0008] The early warning and decision support engine, as the core of the system, is in communication connection with the data acquisition fusion module and the knowledge model library, and is used for performing comprehensive risk assessment and generating comprehensive decision information including at least treatment parameter setting suggestions, anticoagulation scheme recommendations, filter life prediction and replacement early warning, and complication risk early warning based on the fusion data and calling knowledge of the knowledge model library;

[0009] The interactive alarm module is used for displaying the comprehensive decision information and performing differential alarm prompts according to risk levels.

[0010] Preferably, the multi-modal data includes:

[0011] The CRRT device real-time running parameters include arterial pressure, venous pressure, transmembrane pressure, filter front and back pressure, and flow parameters;

[0012] The patient physiological monitoring data include vital sign data, invasive hemodynamic data, laboratory examination results, and intake and output data;

[0013] The patient electronic medical record data include diagnosis information, medication records, and nursing records;

[0014] The environment and operation data include treatment start / end time, treated time length, and medical staff operation records.

[0015] Preferably, the early warning and decision support engine includes:

[0016] The treatment parameter recommendation unit is used for recommending initial or adjusted blood flow rate, dehydration amount, replacement fluid mode, and dose based on preset rules or machine learning models according to patient's body weight, disease diagnosis, fluid balance target, and laboratory index before or during treatment;

[0017] The anticoagulation scheme recommendation unit is used for recommending individualized anticoagulation scheme and anticoagulant dose and predicting system blood coagulation risk according to patient's blood coagulation function index, platelet count, and bleeding risk;

[0018] The filter performance evaluation and early warning unit is used for dynamically predicting filter remaining effective life through time series analysis or regression model based on filter front and back pressure change trend, transmembrane pressure rise rate, and running time, and issuing replacement early warning before filter performance decreases or is about to be blocked;

[0019] The complication risk evaluation unit is used for identifying early risk features of hypotension, electrolyte disorder, bleeding, or thrombus event through machine learning models based on fusion data and issuing early warning.

[0020] Preferably, the filter performance evaluation and early warning unit is used for:

[0021] Construct an input vector characterized by filter running time, pressure parameter change rate, and concentration of formed elements in waste liquid;

[0022] The input vector is fed into a trained filter lifetime prediction model, which outputs the remaining effective time of the filter or the probability of coagulation risk.

[0023] When the remaining effective time is lower than the first threshold or the probability of coagulation risk is higher than the second threshold, a filter replacement warning is generated.

[0024] Preferably, the anticoagulation scheme recommendation unit is specifically used for:

[0025] Based on the patient's APTT, PT, INR, platelet count, and whether there is active bleeding, the decision tree model is used to classify the patient into high bleeding risk, high coagulation risk, or balanced risk categories.

[0026] Based on different risk categories, corresponding anticoagulation strategies are recommended, including no anticoagulation, local citrate anticoagulation, low molecular weight heparin or unfractionated heparin anticoagulation, and initial dose recommendations are given.

[0027] Preferably, the interactive alarm module includes:

[0028] Emergency alarm: In cases where life is immediately in danger, the highest level of audible and visual alarm will be activated.

[0029] Important alarms correspond to high-risk situations requiring timely intervention and employ audible and visual alerts that differ from those of Level 1 alarms.

[0030] General reminders, for auxiliary suggestions, are provided through visual prompts on the interface or message notifications.

[0031] Preferably, the system further includes:

[0032] The self-learning processing module records data on the operational feedback of medical staff and the subsequent reactions of patients, and uses this data to iteratively update the models in the knowledge model base to improve the accuracy of early warnings and recommendations.

[0033] The present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory, wherein when the processor executes the computer program, it implements the functions of the system.

[0034] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the functions of a system.

[0035] Beneficial effects

[0036] This invention provides an intelligent safety early warning system for CRRT treatment based on multimodal fusion. Compared with existing technologies, it has the following advantages:

[0037] This intelligent safety early warning system for CRRT treatment based on multimodal fusion, through multimodal data fusion and machine learning models, can identify complex and hidden early risk characteristics, enabling early warning and comprehensive risk assessment. This allows medical staff to shift from passively handling alarms to proactively intervening, taking measures before complications occur, effectively reducing the incidence of life-threatening events, and providing a higher level of safety for critically ill patients. It integrates comprehensive physiological, medical record, and real-time treatment data of patients, and outputs individualized initial settings and dynamic adjustment suggestions for treatment parameters, as well as anticoagulation regimens based on bleeding / coagulation risk stratification through intelligent models. This effectively reduces treatment deviations caused by experience differences, enabling CRRT treatment to move from experience-driven to a data- and knowledge-driven precision medicine model, which is expected to improve the adequacy of treatment and patient survival rates.

[0038] By predicting the remaining effective lifespan of the filter, an accurate replacement warning is issued before the efficiency declines or the filter is about to become clogged, and the operation procedure is pushed out. This can not only avoid the waste of consumables caused by premature filter replacement, but also prevent the risk of sudden blood clotting of the filter interrupting treatment, thus ensuring the continuity of treatment, while improving the planning and efficiency of nursing work.

[0039] By integrating scattered device alarms, laboratory results, and medical records into a unified intelligent system with decision support capabilities, it automatically completes data integration, risk calculation, and generates intuitive decision suggestions. Through a hierarchical alarm mechanism, it pushes the most critical information to medical staff, which greatly reduces the mental stress of medical staff in identifying and judging complex information, allowing them to focus more on core clinical decisions and patient care, and alleviating their overworked state.

[0040] By continuously collecting clinical feedback and patient outcome data and iteratively optimizing the model, the system can continuously adapt to the characteristics of patients and clinical practice preferences of the institution. Its early warning and decision-making will become more and more accurate over time, ultimately driving the continuous improvement of CRRT treatment quality throughout the department and even the field. Attached Figure Description

[0041] Figure 1 This is a connection block diagram of an intelligent safety early warning system for CRRT treatment based on multimodal fusion according to the present invention;

[0042] Figure 2 This is a schematic diagram of an electronic device according to the present invention. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] like Figure 1 As shown:

[0045] A multimodal fusion-based intelligent safety early warning system for CRRT treatment includes:

[0046] The data acquisition and fusion module is used to collect multimodal data from multiple heterogeneous data sources in real time, and to standardize and fuse the data to generate unified time-series fused data.

[0047] Multimodal data specifically includes:

[0048] Real-time operating parameters of CRRT equipment, including core operating data such as arterial pressure, venous pressure, transmembrane pressure, pre- and post-filter pressure, blood flow, replacement fluid flow, and waste fluid flow, are sampled at a frequency of 1 time per second.

[0049] Patient physiological monitoring data includes vital signs data, such as heart rate, blood pressure, blood oxygen saturation, and body temperature; invasive hemodynamic data, such as central venous pressure and cardiac output; laboratory test results, such as complete blood count, coagulation function, electrolytes, renal function, and lactate; and fluid intake and output data, such as urine output, replacement fluid volume, waste fluid volume, and infusion volume. Vital signs data are sampled once per minute, and laboratory test results are updated in real time through the laboratory information system.

[0050] Patient electronic medical record data is synchronized from the hospital information system / electronic medical record system, including basic patient information such as age, weight, and gender; diagnostic information such as AKI stage and underlying diseases; medication records such as anticoagulants, vasoactive drugs, and diuretics; and nursing records such as changes in body position and operation records.

[0051] Environmental and operational data, including treatment start / end time, treatment duration, medical staff operation records such as parameter adjustment time / amplitude, filter replacement records, alarm handling records; treatment environment temperature / humidity;

[0052] By standardizing the format of the collected heterogeneous data, cleaning outliers, and filling in missing values, and aligning multimodal data based on timestamps, a feature-level fusion method is adopted to map device parameters, physiological data, and medical record information into time-series feature vectors of a unified dimension, generating a three-dimensional fusion dataset of patients, devices, and time.

[0053] A knowledge model library is used to store medical knowledge, clinical guidelines, historical data, and trained early warning and decision-making models related to CRRT treatment.

[0054] The early warning and decision support engine, as the core of the system, communicates with the data acquisition and fusion module and the knowledge model library. It is used to call the knowledge of the knowledge model library based on the fused data, conduct comprehensive risk assessment, and generate comprehensive decision information including at least treatment parameter setting suggestions, anticoagulation regimen recommendations, filter life prediction and replacement early warning, and complication risk early warning. Specifically, it includes 4 functional units.

[0055] The treatment parameter recommendation unit sets initial parameters before treatment and adjusts dynamic parameters during treatment. It accepts input of patient weight, AKI stage, fluid balance target, and laboratory data. Before treatment, based on preset rules and combined with machine learning models such as random forests, it outputs initial recommended values ​​for blood flow rate and replacement fluid dosage. During treatment, it monitors the patient's fluid balance in real time, such as the deviation of cumulative dehydration from the target and electrolyte levels, such as serum potassium and sodium. When indicators exceed safe ranges, the model automatically calculates the parameter adjustment range and provides adjustment suggestions.

[0056] The anticoagulation regimen recommendation unit is used to recommend individualized anticoagulation regimens and anticoagulant dosages based on the patient's coagulation function indicators, platelet count, and bleeding risk, and to predict systemic coagulation risk.

[0057] Based on the patient's APTT, PT, INR, platelet count, and whether there is active bleeding, the decision tree model is used to classify the patient into high bleeding risk, high coagulation risk, or balanced risk categories.

[0058] High bleeding risk: No anticoagulation regimen or local citrate anticoagulation is recommended; High coagulation risk: Low molecular weight heparin or unfractionated heparin is recommended; Balanced risk: Regular dose of low molecular weight heparin or regular dose of unfractionated heparin is recommended.

[0059] Real-time monitoring of changes in coagulation function indicators, prediction of system coagulation risk, and triggering an early warning for adjusting anticoagulant dosage when the risk probability is >80%.

[0060] The filter performance assessment and early warning unit dynamically predicts the remaining effective life of the filter based on the pressure change trend before and after the filter, the rate of increase of transmembrane pressure, and the operating time, using time series analysis or regression models. It issues a replacement warning before the filter performance declines or is about to become clogged. The unit constructs an input vector characterized by filter operating time, pressure parameter change rate, and formed element concentration in waste liquid. The input vector is fed into a trained filter life prediction model, which outputs the remaining effective time of the filter or the probability of coagulation risk. When the remaining effective time is lower than a first threshold or the probability of coagulation risk is higher than a second threshold, a filter replacement warning is generated, and the filter replacement operation procedure is pushed out.

[0061] The complication risk assessment unit is used to identify early risk characteristics of hypotension, electrolyte imbalance, bleeding, or thrombosis based on fused data using a machine learning model, and to issue early warnings. The fused data includes trends in blood pressure, heart rate, electrolyte levels, coagulation parameters, urine output, and equipment operating parameters, which are then used to identify early risk characteristics of complications through an XGBoost classification model. For example:

[0062] Hypotension warning: When the model detects the combined characteristics of "systolic blood pressure <90 mmHg, heart rate >100 bpm, and blood flow rate >250 mL / min", it predicts the risk of hypotension and issues a warning 5-10 minutes in advance. Electrolyte disturbance warning: When serum potassium concentration continues to rise, such as an increase of 0.3 mmol / L per hour or serum sodium <130 mmol / L, an electrolyte disturbance warning is triggered, and adjustment of the replacement fluid electrolyte formula is recommended. Bleeding / thrombosis warning: When the platelet count decreases by >10 × 10⁻⁶ per hour... 9 When D-dimer is >10 mg / L and APTT >80s, a warning of bleeding risk is issued; when D-dimer is >10 mg / L and transmembrane pressure rises sharply, a warning of thrombosis risk is issued.

[0063] The interactive alarm module, serving as the interface between the system and medical staff, is responsible for displaying comprehensive decision-making information and providing differentiated alarm prompts based on risk levels, ensuring that medical staff can obtain key information and intervene in a timely manner. By using a touchscreen, it displays basic patient information, real-time fused data, current decision suggestions, and a list of warning information. It also supports connection with medical staff's mobile APP or PDA to push warning information and decision suggestions in real time, ensuring that medical staff can obtain key prompts even when they are not near the device.

[0064] The interactive alarm module includes a tiered alarm mechanism, specifically divided into emergency alarms, important alarms, and general alerts;

[0065] Emergency Alarm (Level 1): In cases of immediate life-threatening situations, the highest level of audible and visual alarm will be activated, such as flashing red lights + high-frequency buzzer with a volume of ≥85dB. At the same time, a mandatory confirmation window will pop up on the interface, and the alarm will only stop after medical staff click "Confirm Processing".

[0066] Important alarm (level 2), corresponding to high-risk situations requiring timely intervention, uses a flashing yellow light + medium frequency buzzer (volume ≥ 70dB, a pop-up prompt window on the interface, and the alarm stops after confirmation by medical staff);

[0067] General Reminder (Level 3): For auxiliary suggestions, only visual prompts (blinking green text) or APP message notifications are provided, without sound alarms, to avoid interfering with the normal work of medical staff.

[0068] The self-learning processing module enables the system to "continuously evolve." By recording the operational feedback of medical staff and the subsequent reactions of patients, iteratively optimizes the early warning and decision-making models in the knowledge model base, thereby improving the system's accuracy. Specific operational feedback data includes the adoption of system decision-making suggestions by medical staff and the processing results of early warning information. Subsequent patient reaction data includes changes in patients' physiological indicators and whether complications occur after medical staff implement intervention measures.

[0069] The above feedback data is automatically summarized weekly. After the data is anonymized, it is used to update the training dataset in the knowledge model base. Incremental training is used to retrain the original model based on the new dataset, retaining the core structure of the model and only adjusting the parameter weights. The new model must be clinically validated before it can replace the original model and be officially put into use.

[0070] like Figure 2 As shown:

[0071] The present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory. When the processor executes the computer program, it realizes the functions of the system.

[0072] The present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the functions of the system.

[0073] This solution uses a data acquisition and fusion module to collect multimodal data in real time from CRRT equipment, patient physiological monitoring systems, electronic medical record systems, and the operating environment. After standardization and feature-level fusion, a unified time-series fusion dataset is generated. Next, the early warning and decision support engine calls upon medical knowledge and trained models from the knowledge model library to comprehensively analyze the fusion data and generate four major decision support functions: a treatment parameter recommendation unit dynamically adjusts treatment parameters based on the patient's condition; an anticoagulation protocol recommendation unit recommends individualized anticoagulation protocols based on coagulation indicators; a filter efficiency assessment unit predicts filter lifespan and provides early warnings based on pressure trends; and a complication risk assessment unit uses machine learning models to identify early risks such as hypotension and electrolyte imbalances. Then, an interactive alarm module executes differentiated alarms based on risk levels, and in emergencies, a combined audio-visual forced confirmation method is used. Finally, a self-learning processing module records medical staff feedback and patient reaction data, periodically performing incremental training and optimization updates to continuously improve system performance.

[0074] This invention, through multimodal data fusion and machine learning models, can identify complex and hidden early risk characteristics, enabling early warning and comprehensive risk assessment. This allows medical staff to shift from passively handling alarms to proactively intervening, taking measures before complications occur, effectively reducing the incidence of life-threatening events, and providing a higher level of safety for critically ill patients. It integrates comprehensive physiological, medical record, and real-time treatment data of patients, and outputs personalized initial settings and dynamic adjustment suggestions for treatment parameters, as well as anticoagulation regimens based on bleeding / coagulation risk stratification through intelligent models. This effectively reduces treatment deviations caused by experience differences, enabling CRRT treatment to move from experience-driven to a data- and knowledge-driven precision medicine model, which is expected to improve the adequacy of treatment and patient survival rates.

[0075] By predicting the remaining effective lifespan of the filter, an accurate replacement warning is issued before the efficiency declines or the filter is about to become clogged, and the operation procedure is pushed out. This can not only avoid the waste of consumables caused by premature filter replacement, but also prevent the risk of sudden blood clotting of the filter interrupting treatment, thus ensuring the continuity of treatment, while improving the planning and efficiency of nursing work.

[0076] By integrating scattered equipment alarms, laboratory results, and medical records into a unified intelligent system with decision support capabilities, it automatically completes data integration, risk calculation, and generates intuitive decision suggestions. Through a hierarchical alarm mechanism, it pushes the most critical information to medical staff, which greatly reduces the mental stress of medical staff in identifying and judging complex information, allowing them to focus more on core clinical decisions and patient care, and alleviating their overworked state.

[0077] The invention’s unique self-learning module continuously collects clinical feedback and patient outcome data and iteratively optimizes the model accordingly. This allows the system to continuously adapt to the characteristics of patients and clinical practice preferences of the institution. Its warnings and decisions become more and more accurate over time, ultimately driving the continuous improvement of CRRT treatment quality throughout the department and even the field.

[0078] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An intelligent safety early warning system for CRRT treatment based on multimodal fusion, characterized in that, include: The data acquisition and fusion module is used to collect multimodal data from multiple heterogeneous data sources in real time, and to standardize and fuse the data to generate unified time-series fused data. A knowledge model library is used to store medical knowledge, clinical guidelines, historical data, and trained early warning and decision-making models related to CRRT treatment. The early warning and decision support engine, as the core of the system, communicates with the data acquisition and fusion module and the knowledge model library. It is used to perform comprehensive risk assessment based on the fused data and call the knowledge of the knowledge model library, and generate comprehensive decision information including at least treatment parameter setting suggestions, anticoagulation regimen recommendations, filter life prediction and replacement early warning, and complication risk early warning. The interactive alarm module is used to display the comprehensive decision-making information and execute differentiated alarm prompts according to the risk level.

2. The intelligent safety early warning system for CRRT treatment based on multimodal fusion according to claim 1, characterized in that: The multimodal data includes: Real-time operating parameters of CRRT equipment, including arterial pressure, venous pressure, transmembrane pressure, pressure before and after the filter, and flow parameters; Patient physiological monitoring data, including vital signs data, invasive hemodynamic data, laboratory test results, and fluid intake and output data; Patient electronic medical record data, including diagnostic information, medication records, and nursing records; Environmental and operational data, including treatment start / end time, treatment duration, and medical staff operation records.

3. The intelligent safety early warning system for CRRT treatment based on multimodal fusion according to claim 2, characterized in that: The early warning and decision support engine includes: The treatment parameter recommendation unit is used to recommend initial or adjusted blood flow rate, dehydration volume, replacement fluid mode and dosage based on the patient's weight, disease diagnosis, fluid balance target and laboratory indicators before or during treatment, using preset rules or machine learning models. The anticoagulation regimen recommendation unit is used to recommend individualized anticoagulation regimens and anticoagulant dosages based on the patient's coagulation function indicators, platelet count, and bleeding risk, and to predict systemic coagulation risk. The filter performance assessment and early warning unit is used to dynamically predict the remaining effective life of the filter based on the pressure change trend before and after the filter, the rate of increase of transmembrane pressure, and the operating time, through time series analysis or regression model, and to issue a replacement warning before the filter performance declines or is about to become clogged. The complication risk assessment unit is used to identify early risk characteristics of hypotension, electrolyte imbalance, bleeding or thrombosis events based on fused data and through machine learning models, and to issue early warnings.

4. The intelligent safety early warning system for CRRT treatment based on multimodal fusion according to claim 2, characterized in that: The filter performance evaluation and early warning unit is used for: Construct an input vector characterized by filter running time, pressure parameter change rate, and concentration of formed elements in waste liquid; The input vector is fed into a trained filter lifetime prediction model, which outputs the remaining effective time of the filter or the probability of coagulation risk. When the remaining effective time is lower than the first threshold or the probability of coagulation risk is higher than the second threshold, a filter replacement warning is generated.

5. The intelligent safety early warning system for CRRT treatment based on multimodal fusion according to claim 1, characterized in that: The anticoagulation scheme recommendation unit is specifically used for: Based on the patient's APTT, PT, INR, platelet count, and whether there is active bleeding, the decision tree model is used to classify the patient into high bleeding risk, high coagulation risk, or balanced risk categories. Based on different risk categories, corresponding anticoagulation strategies are recommended, including no anticoagulation, local citrate anticoagulation, low molecular weight heparin or unfractionated heparin anticoagulation, and initial dose recommendations are given.

6. The intelligent safety early warning system for CRRT treatment based on multimodal fusion according to claim 1, characterized in that: The interactive alarm module includes: Emergency alarm: In cases where life is immediately in danger, the highest level of audible and visual alarm will be activated. Important alarms correspond to high-risk situations requiring timely intervention and employ audible and visual alerts that differ from those of Level 1 alarms. General reminders, for auxiliary suggestions, are provided through visual prompts on the interface or message notifications.

7. The intelligent safety early warning system for CRRT treatment based on multimodal fusion according to claim 1, characterized in that: The system also includes: The self-learning processing module records data on the operational feedback of medical staff and the subsequent reactions of patients, and uses this data to iteratively update the models in the knowledge model base to improve the accuracy of early warnings and recommendations.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory, characterized in that, When the processor executes the computer program, it implements the functions of the system as described in any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the functions of the system as described in any one of claims 1 to 7.

Images