Risk early warning system based on enteral nutrition management of patient

Abstract

The invention discloses a risk early warning system based on enteral nutrition management of a patient, and belongs to the technical field of medical technology and biomedicine. A risk early warning system based on enteral nutrition management of a patient comprises a patient monitoring module, a score calculation module, a nutrition control module, an early warning notification module, a nutrition scheme adjustment module and a continuous observation and feedback module. The problems that in the prior art, symptoms of critical patients with the acute respiratory distress syndrome are not monitored in time, the regulation efficiency of enteral nutrition supply is low, errors are prone to occurring, and an early warning mechanism is incomplete are solved. By introducing a real-time monitoring and intelligent quantitative evaluation tool, an automatic regulation and control mechanism and multi-level early warning notification and detailed guidance, main problems existing in enteral nutrition management of the severe patient with the acute respiratory distress syndrome at present are effectively solved, and the rehabilitation process and the treatment effect of the patient are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of medical technology and biomedical technology, and in particular to a risk early warning system based on enteral nutrition management of patients. Background Art

[0002] Enteral nutrition management is essential for critically ill patients with acute respiratory distress syndrome (ARDS). However, during the implementation of enteral nutrition, patients may experience a variety of intolerance symptoms, such as abdominal distension, abdominal pain, nausea, vomiting, constipation or diarrhea, which may lead to serious complications and affect the patient's recovery process. The current nutrition management methods and technologies still have the following shortcomings:

[0003] Untimely symptom monitoring: Traditionally, the evaluation of patients’ nutritional status and related complications relies on regular ward rounds by medical staff, and relies on the experience and judgment of medical staff. There is a lack of real-time monitoring and quantitative assessment tools for patients’ physiological status. This approach is not only inefficient, but also prone to missing early symptoms, resulting in missing the best time for intervention.

[0004] Manual operation is inefficient and error-prone: Traditionally, medical staff need to frequently manually adjust the speed and amount of enteral nutrition supply, which is not only time-consuming and labor-intensive, but also prone to risks due to human errors. Especially in intensive care environments, the heavy workload of medical staff increases the possibility of operational errors.

[0005] Imperfect early warning mechanism: The existing early warning system usually only issues an alarm when serious problems occur, but fails to identify potential risk factors in advance, making preventive management impossible. In addition, the early warning information is not detailed enough to guide medical staff to take targeted measures. Summary of the invention

[0006] The purpose of the present invention is to provide a risk warning system based on enteral nutrition management of patients. By introducing real-time monitoring and intelligent quantitative evaluation tools, automated control mechanisms, and multi-level warning notifications and detailed guidance, the main problems existing in the current enteral nutrition management of ARDS patients are effectively solved, the patient's recovery process and treatment effect are significantly improved, and the problems raised in the above-mentioned background technology are solved.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] The risk warning system based on patient enteral nutrition management includes:

[0009] A patient monitoring module, comprising at least one intelligent single-chip computer, for acquiring and recording data of the patient's abdominal distension, abdominal pain, nausea, vomiting, diarrhea or constipation symptoms in real time;

[0010] The scoring calculation module, which is built with an AI chip, automatically calculates the total tolerance score according to the patient symptom data provided by the data acquisition module and in combination with the NUTRIC scoring results of ARDS prone ventilation patients, in accordance with the preset enteral nutrition tolerance assessment criteria;

[0011] The nutrition control module determines whether it is necessary to adjust the enteral nutrition supply for the patient according to the total tolerance score obtained by the scoring calculation module;

[0012] The early warning notification module, when the total tolerance score reaches a specific threshold, sends an early warning message to medical staff and conducts tests on the patient's blood, liver and kidneys, gastrointestinal tract and excreta;

[0013] The nutrition plan adjustment module proposes targeted nutrition plan adjustment suggestions based on the test results and the patient's clinical symptoms;

[0014] The continuous observation and feedback module tracks the patient's reaction to the adjusted nutrition plan to ensure that the nutrient intake meets the standard.

[0015] Preferably, an intelligent single-chip microcomputer is set in the patient monitoring module for connecting the data interfaces of various types of monitoring sensors. The monitoring sensors include:

[0016] An abdominal pressure sensor for real-time monitoring of changes in intra-abdominal pressure and transmitting the data to the intelligent single-chip microcomputer for processing;

[0017] A physiological parameter wearable monitor is worn on the patient's wrist for monitoring the skin electrical response and heart rate variability of the patient, judging the occurrence of nausea and / or vomiting, and transmitting data to the intelligent single-chip microcomputer;

[0018] A humidity sensor, whose data receiving end is fixed in a feces container under the hospital bed or inside the mattress, is used to detect the water content of feces to judge the degree of constipation or diarrhea, and analyze the humidity fluctuation record to judge the number of excretions in a day, and transmit data to the intelligent single-chip microcomputer.

[0019] Preferably, the scoring calculation module is further used for:

[0020] Receiving the readings of the abdominal pressure sensor, the signals of the physiological parameter wearable monitor, and the feedback of the humidity sensor, and using the algorithm implanted in the AI chip to clean and standardize the original data, removing noise and outliers to ensure the accuracy of subsequent analysis;

[0021] Based on the deep learning model, quantifying the degree of abdominal distension and / or abdominal pain, nausea and / or vomiting, diarrhea or constipation of the patient, and converting the received continuous data into discrete scoring data through the AI chip, assigning values in sequence according to the degree of symptoms from mild to severe, and corresponding different scores to different severity levels to achieve precise assessment of the patient's tolerance;

[0022] Sum up the scores of each individual symptom to obtain the patient's preliminary total tolerance score;

[0023] Receive the NUTRIC score results of ARDS patients in prone position ventilation as additional parameters, and the risk-weighted algorithm run by the AI chip adjusts the weights and thresholds of enteral nutrition tolerance scores according to different NUTRIC score intervals;

[0024] Generate the patient's final total tolerance score through the AI chip according to the integrated symptom scores and adjusted weights.

[0025] Preferably, the enteral nutrition tolerance evaluation criteria are divided into four indicators, namely abdominal distension and / or abdominal pain indicators, nausea and / or vomiting indicators, excrement humidity indicators, and excrement frequency indicators. Each indicator is specifically assigned values as follows according to the severity:

[0026] When there are no detectable symptoms for the indicator, the indicator is assigned 0 points;

[0027] When there are mild symptoms for the indicator, the indicator is assigned 1 point;

[0028] When there are obvious symptoms for the indicator, the indicator is assigned 2 points;

[0029] When there are severe symptoms for the indicator, the indicator is assigned 3 points.

[0030] Preferably, the nutrition control module is further used for:

[0031] Obtain the final total tolerance score from the score calculation module;

[0032] Perform the following operations according to different ranges of the final total tolerance score:

[0033] When the final total tolerance score is 0 - 3 points, continue enteral nutrition supplementation at the current speed or increase the nutrition supplementation speed;

[0034] When the final total tolerance score is 4 - 7 points, slow down the enteral nutrition supplementation speed and re-evaluate after 2 hours to observe whether the symptoms improve;

[0035] When the final total tolerance score is 8 points or above, suspend enteral nutrition supplementation and perform drug treatment or emergency rescue measures on the patient's condition.

[0036] Preferably, the nutrition control module supports two control modes: manual control and automatic control;

[0037] Among them, manual control is used for medical staff to set the initial nutrition supply plan according to the doctor's order and directly adjust the operation parameters of the nutrition pump through the operation panel;

[0038] Automatic regulation is used to automatically send control instructions to the nutrition pump according to the final total tolerance score, adjust the infusion rate and pause the infusion operation.

[0039] Preferably, the warning notification module includes a discrimination unit and a signal indication unit. The discrimination unit is used to receive and judge the patient's final total tolerance score; the signal indication units are respectively arranged at the nurse station and the doctor's studio. Each signal indication unit is provided with indicator light colors representing different warning levels, namely, three colors of light: green, yellow, and red. According to different ranges of the final total tolerance score, the signal indication unit performs the following operations:

[0040] When the final total tolerance score is 0 - 3 points, there is no warning information and the green light is always on;

[0041] When the final total tolerance score is 4 - 7 points, the yellow light flashes slowly, sending a low-level warning information, indicating that the patient's nutritional supply is poor, prompting the doctor to conduct blood glucose, blood routine, electrolyte, liver and kidney function, and stool routine tests, and selectively conduct gastroscopy and gastrointestinal B-ultrasound examinations;

[0042] When the final total tolerance score is 8 points or above, the red light flashes rapidly, sending a high-level warning information to the medical staff, prompting the medical staff to conduct a comprehensive physical examination on the patient and carry out medical rescue.

[0043] Preferably, the continuous observation and feedback module includes:

[0044] An observation and feedback unit, which is used to monitor the nutritional intake;

[0045] Among them, when the nutritional intake reaches 80% - 100% of the target amount, the original plan is maintained. When the nutritional intake does not reach the target amount, a warning is sent to the medical staff again, and a new round of enteral nutrition tolerance scoring is carried out;

[0046] A medical staff feedback interface, which is used to view the patient's observation data, scoring history, warning information and its response measures.

[0047] Preferably, the nutrition plan adjustment module proposes targeted nutrition plan adjustment suggestions based on the detection results and the patient's clinical symptoms, including:

[0048] According to the examination results, when the patient's nutritional supply is poor or extremely poor, analyze whether the patient has a past medical history. When the patient has a past medical history, retrieve the past medical history for the patient;

[0049] Analyze the impact of the patient's past medical history on nutrient digestion and absorption in sequence, determine the impact of the past medical history on the patient's nutrient digestion and absorption, and obtain the nutrient digestion and absorption influence of the past medical history;

[0050] According to the nutrient digestion and absorption influence of the past medical history, screen the effective data for the patient's past medical history, and screen out the past medical history that affects the patient's nutrient digestion and absorption to obtain the target past medical history;

[0051] Analyze the characteristics of the impact of the target past medical history on the patient's nutrient digestion and absorption to obtain impact characteristic information;

[0052] Obtain the patient's clinical symptoms, and match the nutritional adjustment plan according to the patient's clinical symptoms respectively, determine the nutritional adjustment plan corresponding to the clinical symptoms, and obtain multiple sets of first initial nutritional adjustment plans; among them, each clinical symptom corresponds to a nutritional adjustment plan;

[0053] Revise the first initial nutritional adjustment plan according to the impact characteristic information to obtain the second initial nutritional adjustment plan;

[0054] Integrate the second initial nutritional adjustment plan, integrate the second initial nutritional adjustment plan into a set of nutritional adjustment plans to obtain the target nutritional adjustment plan;

[0055] Generate nutritional plan adjustment suggestions based on the target nutritional adjustment plan.

[0056] Preferably, use the algorithm implanted in the AI chip to clean and standardize the original data, including:

[0057] Determine the standards for the original data, and standardize the data standards of the abdominal pressure sensor readings, the signals of the physiological parameter wearable monitors, and the feedback of the humidity sensors; the original data includes the abdominal pressure sensor readings, the signals of the physiological parameter wearable monitors, and the feedback of the humidity sensors;

[0058] Conduct a preliminary analysis on the abdominal pressure sensor readings, the signals of the physiological parameter wearable monitors, and the feedback of the humidity sensors according to the data standards, and judge whether the abdominal pressure sensor readings, the signals of the physiological parameter wearable monitors, and the feedback of the humidity sensors meet the corresponding data standards to obtain the first analysis and judgment result;

[0059] When the first analysis and judgment result is that it does not meet the data standards, convert the original data that does not meet the data standards according to the corresponding data standards to obtain the standardized original data;

[0060] Preprocess the standardized original data, obtain the cleaned data, and synchronize the time stamps to obtain a data sequence;

[0061] Anomaly analysis is performed on the data sequence through the following formula:

[0062]

[0063] In the above formula, K i is the anomaly analysis value of the original data at time i, A i is the original data at time i, A0 is the original data at the initial time, α is the first analysis parameter, k1 is the first weight coefficient, k2 is the second weight coefficient, A i-1 is the original data at time i-1, K i-1 is the anomaly analysis value of the original data at time i-1, σ is the second analysis parameter, A i-2 is the original data at time i-2, A i-3 is the original data at time i-3, ΔT i is the time difference between time i and the initial time;

[0064] Determine whether the original data is abnormal according to the anomaly analysis value of the original data. When the anomaly analysis value of the original data is greater than 0, the corresponding original data is an abnormal value.

[0065] Compared with the prior art, the beneficial effects of the present invention are:

[0066] 1. The present invention realizes continuous monitoring of key physiological parameters of patients by adopting a variety of sensors, and automatically analyzes these data through a scoring calculation module with a built-in AI chip to provide a quantitative tolerance score, improves the recognition ability of early intolerance symptoms of patients, reduces the risk of complications caused by missing the best intervention opportunity, realizes continuous monitoring of the physiological state of patients, and enhances the scientificity and accuracy of medical decisions.

[0067] 2. The present invention sets up an automatic regulation mechanism. The system is equipped with a nutrition control module, which can automatically send control instructions to the nutrition pump according to the total tolerance score obtained by the scoring calculation module to adjust the infusion speed and pause the infusion operation. At the same time, it supports a manual regulation mode, allowing medical staff to make targeted adjustments according to specific situations, significantly improving the efficiency of nutrition management, reducing the work burden of medical staff, and the automatic regulation reduces the occurrence probability of human errors and improves the safety and reliability of treatment.

[0068] 3. The present invention realizes preventive management by setting multi-level warning notifications and detailed guidance, setting green, yellow, and red signal lights to represent different warning levels, and sending detailed warning information to medical staff when specific thresholds are reached, including specific measures such as suggesting relevant examinations such as blood glucose and blood routine or comprehensive examinations and carrying out medical rescue, and helps to reduce the incidence of serious complications. Brief Description of the Drawings

[0069] Figure 1 is the system module structure diagram of the present invention;

[0070] Figure 2 is the system operation flow chart of the present invention. Detailed implementation manners

[0071] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0072] In order to solve the problems of untimely symptom monitoring of critically ill patients with acute respiratory distress syndrome (ARDS), low efficiency and easy error in the regulation of enteral nutrition supply, and imperfect early warning mechanism in the prior art, please refer to Figure 1-2 This embodiment provides the following technical solutions:

[0073] A risk early warning system based on the enteral nutrition management of patients, including:

[0074] A patient monitoring module, including at least one intelligent single-chip microcomputer, for real-time obtaining and recording data of the patient's abdominal distension, abdominal pain, nausea, vomiting, diarrhea or constipation symptoms;

[0075] A scoring calculation module, built-in with an AI chip, automatically calculates the total tolerance score according to the patient symptom data provided by the data acquisition module and in combination with the NUTRIC score result of the ARDS prone position ventilation patient according to the preset enteral nutrition tolerance evaluation criteria;

[0076] A nutrition control module, determines whether it is necessary to adjust the enteral nutrition supply of the patient according to the total tolerance score obtained by the scoring calculation module;

[0077] An early warning notification module, when the total tolerance score reaches a specific threshold, sends an early warning message to medical staff and detects the patient's blood, liver and kidneys, stomach and excreta;

[0078] A nutrition plan adjustment module, based on the detection results and the patient's clinical indications, proposes targeted nutrition plan adjustment suggestions;

[0079] A continuous observation and feedback module, tracks the patient's reaction to the adjusted nutrition plan to ensure that the nutritional intake meets the standard.

[0080] An intelligent single-chip microcomputer is set in the patient monitoring module, and the intelligent single-chip microcomputer is configured with a plurality of data interfaces for connecting different types of sensors. The sensors include:

[0081] An abdominal pressure sensor is fixed to the patient's abdomen by attachment or wearing, used to monitor the change of intra-abdominal pressure in real time, reflect the degree of intra-abdominal pressure, and transmit the data to an intelligent single-chip microcomputer for processing;

[0082] A physiological parameter wearable monitor is worn on the patient's wrist, used to monitor the skin electrical response and heart rate variability of the patient, judge the occurrence of nausea and / or vomiting, and transmit data to the intelligent single-chip microcomputer;

[0083] Specifically, the physiological parameter wearable monitor is provided with a skin electrical response sensor (GSR sensor) and a heart rate variability monitor (HRV monitor). The GSR sensor is used to measure the change in skin conductivity that occurs when the patient has nausea or vomiting. When the GSR signal is detected to be enhanced, it indicates that the patient is experiencing nausea or vomiting and its associated reaction symptoms; the HRV monitor evaluates the activity of the autonomic nervous system by analyzing the standard deviation of consecutive heartbeats, and is used to identify the discomfort symptoms during nausea and vomiting. When the patient has nausea or vomiting, the HRV observation value will decrease.

[0084] By combining the GSR sensor and the HRV monitor, and combining body temperature and the patient's subjective report, the state of the patient can be understood more comprehensively; due to individual differences, the GSR signal and the HRV signal have non-fixed thresholds, so medical staff need to adjust the initial setting parameters of the GSR sensor and the HRV monitor according to the patient's personal physical examination indicators before monitoring;

[0085] A humidity sensor, whose data receiving end is fixed in a feces container under the hospital bed or inside the mattress, is used to detect the water content of feces to judge the degree of constipation or diarrhea, analyze the humidity fluctuation record to judge the number of excretions in a day, and transmit data to the intelligent single-chip microcomputer.

[0086] By integrating a variety of sensors and monitoring devices, the physiological state changes of the patient can be comprehensively captured, improving the accuracy and real-time nature of data collection, and providing a reliable basis for subsequent score calculation.

[0087] The intelligent single-chip microcomputer further includes:

[0088] An integration unit, used to receive the data streams of the abdominal pressure sensor, the physiological parameter wearable monitor and the humidity sensor, and convert them into a unified format;

[0089] A preliminary analysis unit, which preliminarily analyzes the integrated data based on a preset algorithm to identify abnormal conditions that occur during the detection process;

[0090] A communication unit, which supports wireless communication with external devices to upload the original data and preliminary analysis results to the hospital management system;

[0091] A storage unit for storing data and set parameters, facilitating subsequent review analysis and trend prediction of the data; a self-optimization unit for continuously optimizing the symptom recognition accuracy and warning threshold setting according to the patient's historical data.

[0092] The scoring calculation module is further used for:

[0093] Receiving the readings of the abdominal pressure sensor, the signals of the physiological parameter wearable monitor, and the feedback of the humidity sensor, and using the algorithm implanted in the AI chip to clean and standardize the original data, removing noise and outliers to ensure the accuracy of subsequent analysis;

[0094] For abdominal distension and abdominal pain, judging its severity according to the intra-abdominal pressure measurement value and the patient's subjective statement in the waking state, and determining the score for nausea and vomiting symptoms based on the GSR signal and HRV signal combined with electrogastrogram information. The degree of constipation or diarrhea is comprehensively evaluated through the data of the humidity sensor for diarrhea.

[0095] Based on the deep learning model, quantifying the degree of the patient's abdominal distension and / or abdominal pain, nausea and / or vomiting, diarrhea or constipation, and converting the received continuous data into discrete scoring data through the AI chip, assigning values in sequence according to the degree of symptoms from mild to severe, corresponding to different scores according to different severity levels to achieve precise assessment of the patient's tolerance.

[0096] Using the AI chip to process the original data from the sensors and quantifying the symptom severity with the deep learning model improves the processing ability of complex data, making the scoring more accurate, helping to detect potential problems early and intervene in a timely manner.

[0097] The scoring is based on the pre-established enteral nutrition tolerance assessment criteria as a reference. The enteral nutrition tolerance assessment criteria specifically include four indicators, namely, the abdominal distension and / or abdominal pain indicator, the GSR and HRV indicator, the excrement humidity indicator, and the excrement frequency indicator. The specific assignments are as follows:

[0098] When there are no detectable symptoms for the indicator, the score for this indicator is 0;

[0099] When there are mild symptoms for the indicator, the score for this indicator is 1;

[0100] When there are obvious symptoms for the indicator, the score for this indicator is 2;

[0101] When there are severe symptoms for the indicator, the score for this indicator is 3;

[0102] The specific symptom manifestations corresponding to the assignments are shown in Table 1:

[0103]

[0104]

[0105] Table 1 Enteral nutrition tolerance assessment table

[0106] The scores of each individual symptom were summed up to obtain the patient's preliminary total tolerance score;

[0107] The NUTRIC score results of ARDS patients undergoing prone ventilation are received as an additional parameter. The risk-weighted algorithm run by the AI ​​chip adjusts the weight and threshold of the enteral nutrition tolerance score according to different NUTRIC score intervals. For example, in high-risk situations (NUTRIC score equal to or greater than 5), mild symptoms are given a higher weight, resulting in stricter recommendations for intervention measures.

[0108] Based on the integrated symptom score and adjusted weights, the patient's final total tolerance score is generated by the AI ​​chip.

[0109] Here, dedicated AI accelerators such as GPU and FPGA are selected to meet the requirements of real-time performance and high efficiency.

[0110] The nutrition control module is further used to:

[0111] Obtain the final total tolerance score from the score calculation module;

[0112] Depending on the range of the final total tolerance score, the following actions are performed:

[0113] When the final total tolerance score is 0-3 points, continue to provide enteral nutrition at the current rate or increase the rate of nutrition supply. That is, if the patient's symptoms are mild or there is no obvious discomfort, it is recommended to maintain the current nutrition supply rate, and even consider increasing it appropriately to meet higher nutritional needs;

[0114] When the final total tolerance score is 4-7 points, slow down the enteral nutrition supply rate and re-evaluate after 2 hours. By reducing the nutrition supply rate, the stomach and intestines can have more time to digest and absorb. At the same time, re-evaluate after 2 hours to observe whether the symptoms have improved.

[0115] When the final total tolerance score is 8 points or above, enteral nutrition is suspended, and the patient's condition is treated with medication or emergency treatment measures.

[0116] The nutrition control module supports two control modes: manual control and automatic control;

[0117] Among them, manual control is used for medical staff to set the initial nutrition supply plan according to the doctor's advice, and directly adjust the operating parameters of the nutrition pump through the operation panel;

[0118] Automatic regulation is used to automatically send control instructions to the nutrition pump according to the final total tolerance score, adjust the infusion rate and pause the infusion operation.

[0119] Manual regulation in combination with automatic regulation improves flexibility on the one hand, allowing medical staff to make targeted adjustments according to specific situations, effectively supporting the professional judgment of medical staff. On the other hand, it can quickly respond to the scoring results, improve the response speed and accuracy, reduce human delay, improve the treatment efficiency of patients, and reduce the risk caused by human errors in the care process of patients.

[0120] The nutrition plan adjustment module further includes:

[0121] Recommend specific intervention measures based on the patient's condition, including adjusting the nutritional formula, changing the feeding method (such as raising the head of the bed by 30 to 45 degrees) and using drug treatment (such as metoclopramide injection, erythromycin), and optimizing the implementation of the plan through an intelligent single-chip microcomputer or an AI chip.

[0122] The early warning notification module includes a discrimination unit and a signal indication unit. The discrimination unit is used to receive and judge the final total tolerance score of the patient; the signal indication units are respectively set in the nurse station and the doctor's studio. Each signal indication unit is provided with indicator light colors representing different early warning levels, namely three colors of green, yellow, and red light. According to different ranges of the final total tolerance score, the signal indication unit performs the following operations:

[0123] When the final total tolerance score is 0 - 3 points, there is no early warning information and the green light is always on;

[0124] When the final total tolerance score is 4 - 7 points, the yellow light flashes slowly, sending a low-level early warning information to remind the patient's nutritional supply situation, prompting the doctor to conduct blood sugar, blood routine, electrolyte, liver and kidney function, and stool routine tests, and selectively conduct gastroscopy and gastrointestinal B-ultrasound examinations;

[0125] When the final total tolerance score is 8 points or above, the red light flashes rapidly, sending a high-level early warning information to medical staff, prompting medical staff to conduct a comprehensive physical examination on the patient and carry out medical rescue, and formulate a subsequent treatment plan.

[0126] By clarifying the early warning level, it helps medical staff better understand the severity of the patient's symptoms, so as to promptly switch the response plan, prevent the symptoms from worsening, and ensure the patient's life safety.

[0127] The continuous observation and feedback module includes:

[0128] An observation and feedback unit for monitoring the nutritional intake;

[0129] Among them, when the nutrient intake reaches 80%-100% of the target amount, the original plan is maintained. When the nutrient intake does not reach the target amount, a warning is sent to the medical staff again, and a new round of enteral nutrition tolerance scoring is carried out, so as to ensure that the nutrient supply is neither excessive nor insufficient, maintained within a reasonable range, and the effectiveness of nutritional management is improved;

[0130] The medical staff feedback interface is used to view the patient's observation data, scoring history, warning information and its response measures. This feedback interface can be integrated into the mobile device through the SOC chip, which is convenient to access anytime and anywhere, helps the medical staff to view the patient's physical signs and change trends at any time, and at the same time can enhance the communication efficiency among the medical teams and improve the overall nursing quality.

[0131] Through the design and implementation of the above-mentioned modules, the risk warning system not only realizes the accurate assessment of the patient's enteral nutrition tolerance, but also provides flexible and efficient management means. It combines advanced sensing technology, AI algorithms and user-friendly interface design to ensure the practicality and reliability of the system. This not only improves the patient's treatment experience and effect, but also provides a powerful support tool for the medical staff, significantly improving work efficiency and decision-making accuracy.

[0132] The nutrition plan adjustment module puts forward targeted nutrition plan adjustment suggestions based on the detection results and the patient's clinical symptoms, including:

[0133] According to the examination results, when the patient's nutritional supply is poor or extremely poor, analyze whether the patient has a past medical history, and when the patient has a past medical history, retrieve it according to the past medical history to obtain the patient's past medical history;

[0134] Analyze the influence of the past medical history on the digestion and absorption of nutrients in turn for the patient's past medical history, determine the influence of the past medical history on the patient's digestion and absorption of nutrients, and obtain the influence of the past medical history on the digestion and absorption of nutrients;

[0135] Effectively screen the patient's past medical history according to the influence of the past medical history on the digestion and absorption of nutrients, and screen out the past medical history that affects the patient's digestion and absorption of nutrients to obtain the target past medical history;

[0136] Analyze the characteristics of the influence of the target past medical history on the patient's digestion and absorption of nutrients to obtain the influence characteristic information;

[0137] Obtain the patient's clinical symptoms, and respectively match the nutrition adjustment plan according to the patient's clinical symptoms to determine the nutrition adjustment plan corresponding to the clinical symptoms, and obtain multiple sets of first initial nutrition adjustment plans; among them, each clinical symptom corresponds to a set of nutrition adjustment plans;

[0138] Revise the first initial nutrition adjustment plan according to the influencing characteristic information to obtain the second initial nutrition adjustment plan;

[0139] Integrate the second initial nutrition adjustment plan into a set of nutrition adjustment plans based on the second initial nutrition adjustment plan to obtain the target nutrition adjustment plan;

[0140] Generate nutrition plan adjustment suggestions based on the target nutrition adjustment plan.

[0141] The above-mentioned nutrition plan adjustment suggestions for patients are realized through the nutrition plan adjustment module, enabling patients to supply nutrition according to the nutrition plan adjustment suggestions, providing guarantee for the rehabilitation of patients. By analyzing whether the patient has a past medical history, when the patient has a past medical history, the impact of the past medical history on the patient is clarified, the correctness of the nutrition plan adjustment suggestions is improved, and the effect of the nutrition plan adjustment suggestions is prevented from being affected by the past medical history. Moreover, by successively analyzing the impact of the past medical history on nutrition digestion and absorption, effective data screening is carried out for the patient's past medical history based on the influence of the past medical history on nutrition digestion and absorption, reducing the redundancy of data information on irrelevant past medical histories, so that only the target past medical history is analyzed for characteristics during subsequent analysis, reducing the waste of unnecessary time and improving the efficiency of characteristic analysis. And by matching the nutrition adjustment plan according to the patient's clinical symptoms respectively, the nutrition adjustment plan can be matched for different clinical symptoms synchronously, improving the efficiency of nutrition adjustment plan matching, obtaining the first initial nutrition adjustment plan in a short time, so that the target nutrition adjustment plan can be determined and nutrition plan adjustment suggestions can be generated in time. At the same time, when determining the target nutrition adjustment plan, based on the second initial nutrition adjustment plan, the mutual relationship between the second initial nutrition adjustment plans is sorted out and unified, simplifying the summary of the second initial nutrition adjustment plan, and optimizing the target nutrition adjustment plan; revising the first initial nutrition adjustment plan according to the influencing characteristic information, making the second initial nutrition adjustment plan more suitable for the patient's own condition, and further ensuring that the nutrition plan adjustment suggestions can bring ideal implementation effects to the patient.

[0142] Use the algorithm implanted in the AI chip to clean and standardize the original data, including:

[0143] Determine the standard for the original data, and standardize the data standards of the abdominal pressure sensor readings, the signals of the physiological parameter wearable monitors, and the feedback of the humidity sensors; the original data includes the abdominal pressure sensor readings, the signals of the physiological parameter wearable monitors, and the feedback of the humidity sensors;

[0144] Receiving and initially analyzing the abdominal pressure sensor readings, physiological parameter wearable monitor signals, and humidity sensor feedback according to the data standard, determining whether the abdominal pressure sensor readings, physiological parameter wearable monitor signals, and humidity sensor feedback meet the corresponding data standards, and obtaining the first analysis and judgment result;

[0145] When the first analysis and judgment result is that it does not meet the data standard, converting the original data that does not meet the data standard according to the corresponding data standard to obtain the standardized original data;

[0146] Preprocessing the standardized original data, obtaining the cleaned data, and synchronizing the time stamps to obtain a data sequence;

[0147] Performing anomaly analysis on the data sequence through the following formula:

[0148]

[0149] In the above formula, K i is the anomaly analysis value of the original data at time i, A i is the original data at time i, A0 is the original data at the initial time, α is the first analysis parameter, k1 is the first weight coefficient, k2 is the second weight coefficient, A i-1 is the original data at time i - 1, K i-1 is the anomaly analysis value of the original data at time i - 1, σ is the second analysis parameter, A i-2 is the original data at time i - 2, A i-3 is the original data at time i - 3, ΔT i is the time difference between time i and the initial time;

[0150] Determining whether the original data is abnormal according to the anomaly analysis value of the original data. When the anomaly analysis value of the original data is greater than 0, the corresponding original data is an abnormal value.

[0151] Among them, both the first weight coefficient k1 and the second weight coefficient k2 are constants greater than 0 and less than 1, and their sum is 1. The first analysis parameter α is determined in advance according to the normal fluctuation range of the original data, and the second analysis parameter σ is a preset threshold.

[0152] By determining the standard for the original data as described above, the original data from the same source can be standardized, avoiding large errors in the analysis of the original data caused by different abdominal pressure sensors, physiological parameter wearable monitors, and humidity sensors, which lead to different measurement standards for the same type of original data, and providing convenience for the analysis of the original data. Moreover, by preprocessing the standardized original data, data cleaning is achieved, removing chaotic and unusable data, and leaving normal and usable data, thereby improving the data quality, further ensuring the abnormal analysis of the original data. And when performing abnormal analysis, the abnormal analysis value is determined according to the original data at times i-1, i-2, and i-3 in combination with the data sequence. Compared with performing abnormal analysis only by combining the original data at time i-1, it is more accurate, reducing the large error of the abnormal analysis value caused by accidental factors. At the same time, it also combines the abnormal analysis value of the original data at time i-1, enabling the direct analysis and calculation of the abnormal analysis value based on the original data of the previous time, reducing the time for repeated calculations, and improving the calculation efficiency of the abnormal analysis value of the original data at time i.

[0153] Working principle: Use multiple sensors to obtain the physiological data of the patient's abdominal distension, abdominal pain, nausea, vomiting, diarrhea, and constipation. The sensors are connected and transmit data through an intelligent single-chip microcomputer. The scoring calculation module with an embedded AI chip receives the original data and uses a deep learning model to quantify the severity of the symptoms. According to the "Enteral Nutrition Tolerance Assessment Form", each symptom is converted into a specific score (0-3 points), and the total tolerance score is obtained by summarization. In addition, the scoring weight is adjusted in combination with the NUTRIC score results of ARDS prone position ventilation patients to improve the accuracy of risk prediction.

[0154] When the total tolerance score is 0-3 points, the system recommends maintaining or increasing the enteral nutrition supply rate. When the score is 4-7 points, the system automatically slows down the supply rate and re-evaluates after 2 hours; at the same time, a low-level warning is sent to the medical staff, prompting them to conduct relevant examinations such as blood glucose and blood routine. When the score ≥ 8 points, the system immediately suspends the enteral nutrition supply and triggers a high-level warning, notifying the doctor to conduct a comprehensive examination and emergency treatment. Based on the detection results and clinical manifestations, the nutrition plan adjustment module provides personalized suggestions, such as adjusting the nutritional formula, changing the feeding method, and administering drug treatment. These suggestions can be achieved through manual control or automatic control. Continuously monitor the patient's nutritional intake. When it reaches 80%-100% of the target amount, maintain the original plan; if not, issue a warning again and re-evaluate. The medical staff feedback interface allows viewing all relevant data, historical records, and warning information, facilitating long-term tracking and retrospective research.

[0155] It should be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0156] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention.

Claims

1. A risk warning system based on patient enteral nutrition management, characterized in that: include: A patient monitoring module, comprising at least one intelligent single-chip computer, for acquiring and recording data of the patient's abdominal distension, abdominal pain, nausea, vomiting, diarrhea or constipation symptoms in real time; The scoring calculation module has a built-in AI chip. It automatically calculates the total tolerance score according to the preset enteral nutrition tolerance assessment criteria based on the patient symptom data provided by the data acquisition module and the NUTRIC score results of ARDS prone ventilation patients; The nutrition control module determines whether the enteral nutrition supply of the patient needs to be adjusted according to the total tolerance score obtained by the score calculation module; The early warning notification module sends an early warning message to medical staff when the total tolerance score reaches a certain threshold, and tests the patient's blood, liver, kidney, stomach and excrement; Nutritional plan adjustment module, which provides targeted nutritional plan adjustment suggestions based on test results and the patient's clinical symptoms; The continuous observation and feedback module tracks the patient's response to the adjusted nutritional plan to ensure that the nutritional intake meets the target.

2. The risk warning system based on patient enteral nutrition management according to claim 1, characterized in that: The patient monitoring module is provided with an intelligent single chip microcomputer for connecting to data interfaces of various types of monitoring sensors, the monitoring sensors including: Abdominal pressure sensor, used to monitor changes in intra-abdominal pressure in real time and transmit data to the intelligent single-chip microcomputer for processing; The physiological parameter wearable monitor is worn on the patient's wrist to monitor the patient's skin electrical response and heart rate variability, determine the occurrence of nausea and / or vomiting, and transmit data to the intelligent single-chip computer; Humidity sensor, used to detect the moisture content of feces to determine the degree of constipation or diarrhea, analyze humidity fluctuation records to determine the number of excretions in a day, and transmit data to the intelligent microcontroller.

3. The risk warning system based on patient enteral nutrition management according to claim 2, characterized in that: The scoring calculation module is further used for: Receive abdominal pressure sensor readings, wearable physiological parameter monitor signals, and humidity sensor feedback, and use the algorithm embedded in the AI ​​chip to clean and standardize the raw data to remove noise and outliers; Based on the deep learning model, the patient's abdominal distension and / or abdominal pain, nausea and / or vomiting, diarrhea or constipation degree is quantified, and the received continuous data is converted into discrete scoring data through the AI ​​chip, and the values ​​are assigned in order from mild to severe symptoms; The scores of each individual symptom were summed up to obtain the patient's preliminary total tolerance score; The NUTRIC score results of ARDS patients undergoing prone ventilation were received as an additional parameter, and the risk-weighted algorithm run by the AI ​​chip adjusted the weight and threshold of the enteral nutrition tolerance score according to different NUTRIC score intervals; Based on the integrated symptom score and adjusted weights, the patient's final total tolerance score is generated by the AI ​​chip.

4. The risk warning system based on enteral nutrition management of patients according to claim 3, characterized in that: The enteral nutrition tolerance assessment standard is divided into four indicators, namely, abdominal distension and (or) abdominal pain indicator, nausea and (or) vomiting indicator, excrement humidity indicator and excrement frequency indicator. Each item is assigned a value according to the severity as follows: When there are no perceptible symptoms, the indicator is assigned a score of 0; When the indicator shows mild symptoms, the indicator is assigned 1 point; When the indicator shows obvious symptoms, the indicator is assigned 2 points; When the indicator shows severe symptoms, the indicator is assigned 3 points.

5. The risk warning system based on enteral nutrition management of patients according to claim 4, characterized in that: The nutrition control module is further used for: Obtain the final total tolerance score from the score calculation module; Depending on the range of the final total tolerance score, the following actions are performed: When the final total tolerance score is 0-3 points, continue to provide enteral nutrition at the current rate or increase the nutrition supply rate; When the final total tolerance score is 4-7 points, slow down the enteral nutrition supply and reassess after 2 hours; When the final total tolerance score is 8 points or above, enteral nutrition is suspended, and the patient's condition is treated with medication or emergency treatment measures.

6. The risk warning system based on enteral nutrition management of patients according to claim 5, characterized in that: The nutrition control module supports two control modes: manual control and automatic control; Among them, manual control is used for medical staff to set the initial nutrition supply plan according to the doctor's advice, and directly adjust the operating parameters of the nutrition pump through the operation panel; Automatic control is used to automatically send control instructions to the nutrition pump according to the final total tolerance score, adjust the infusion rate and pause the infusion operation.

7. The risk warning system based on patient enteral nutrition management according to claim 6, characterized in that: The early warning notification module includes a discrimination unit and a signal indication unit. The discrimination unit is used to receive and judge the patient's final total tolerance score; the signal indication units are respectively arranged in the nurse station and the doctor's studio. Each signal indication unit is provided with indicator lights indicating different early warning levels, which are green, yellow and red. According to different ranges of the final total tolerance score, the signal indication unit performs the following operations: When the final total tolerance score is 0-3 points, there is no warning information and the green light is always on; When the final total tolerance score is 4-7 points, the yellow light flashes slowly, sending a low-level warning message, indicating that the patient's nutritional supply is poor; When the final total tolerance score is 8 points or above, the red light flashes at a high speed, sending a high-level warning message to medical staff, indicating that the patient's nutritional supply is poor.

8. The risk warning system based on patient enteral nutrition management according to claim 7, characterized in that: The continuous observation and feedback module includes: Observation and feedback unit for monitoring nutritional intake; When the nutritional intake reaches 80%-100% of the target amount, the original plan is maintained. When the nutritional intake does not reach the target amount, an early warning is issued to the medical staff again, and a new round of enteral nutrition tolerance scoring is conducted. The medical feedback interface is used to view the patient's observation data, scoring history, warning information and response measures.

9. The risk warning system based on enteral nutrition management of patients according to claim 8, characterized in that: The nutrition plan adjustment module proposes targeted nutrition plan adjustment suggestions based on the test results and the patient's clinical symptoms, including: When the patient's nutritional supply is poor or the patient's nutritional supply is severe according to the examination results, analyzing whether the patient has a previous medical history, and if the patient has a previous medical history, retrieving the previous medical history to obtain the patient's previous medical history; Analyze the influence of the patient's medical history on nutrient digestion and absorption, determine the influence of the medical history on the patient's nutrient digestion and absorption, and obtain the influence of the medical history on nutrient digestion and absorption; According to the influence of the past medical history on the digestion and absorption of nutrition, effective data screening is performed on the patient's past medical history, and the past medical history that affects the patient's digestion and absorption of nutrition is screened out to obtain the target past medical history; Analyze the characteristics of the impact of the target's past medical history on the patient's nutrient digestion and absorption, and obtain the impact characteristic information; Obtaining the clinical symptoms of the patient, and matching the nutrition adjustment plans according to the clinical symptoms of the patient, determining the nutrition adjustment plans corresponding to the clinical symptoms, and obtaining multiple sets of first initial nutrition adjustment plans; wherein each clinical symptom corresponds to a set of nutrition adjustment plans; The first initial nutrition adjustment plan is revised according to the influencing characteristic information to obtain a second initial nutrition adjustment plan; Comprehensively integrating the second initial nutrition adjustment plan into a set of nutrition adjustment plans to obtain a target nutrition adjustment plan; Generate nutrition plan adjustment recommendations based on the target nutrition adjustment plan.

10. The risk warning system based on enteral nutrition management of patients according to claim 9, characterized in that: The raw data is cleaned and standardized using the algorithm embedded in the AI ​​chip, including: Determine the standard for the raw data, and standardize the data standards for the abdominal pressure sensor reading, the physiological parameter wearable monitor signal, and the humidity sensor feedback; the raw data includes the abdominal pressure sensor reading, the physiological parameter wearable monitor signal, and the humidity sensor feedback; Performing preliminary analysis on the abdominal pressure sensor reading, the physiological parameter wearable monitor signal, and the humidity sensor feedback according to the data standard, determining whether the abdominal pressure sensor reading, the physiological parameter wearable monitor signal, and the humidity sensor feedback meet the corresponding data standard, and obtaining a first analysis and determination result; When the first analysis result is that the data standard is not met, the original data that does not meet the data standard is converted according to the corresponding data standard to obtain standardized original data; Preprocess the standardized raw data, obtain the cleaned data, and synchronize the timestamp to obtain the data sequence; The following formula is used to perform abnormal analysis on the data series: In the above formula, K i is the abnormal analysis value of the original data at time i, A i is the original data at time i, A0 is the original data at the initial time, α is the first analysis parameter, k1 is the first weight coefficient, k2 is the second weight coefficient, A i-1 is the original data at time i-1, K i-1 is the abnormal analysis value of the original data at time i-1, σ is the second analysis parameter, A i-2 is the original data of time i-2, A i-3 is the original data of time i-3, ΔT i is the time difference between time i and the initial time; Whether the original data is abnormal is determined according to the abnormal analysis value of the original data. When the abnormal analysis value of the original data is greater than 0, the corresponding original data is an abnormal value.

Images