Gastrointestinal function risk warning system for sap patients after colloidal fluid resuscitation
Patent Information
- Application Number
- CN202610977024.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-02
- Publication Date
- 2026-09-15
Smart Images
Figure CN122762286A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of medical data processing technology, specifically to a gastrointestinal function risk warning system for SAP patients after colloid resuscitation. Background Technology
[0002] During severe acute pancreatitis (SAP), the systemic inflammatory response leads to abnormally increased capillary permeability, causing fluid leakage into the interstitial space. Clinically, a resuscitation strategy typically employs early colloid fluid resuscitation combined with diuretic dehydration to maintain effective circulating blood volume and reduce tissue edema.
[0003] Currently, existing medical information systems generally use simple arithmetic addition and subtraction to calculate macroscopic fluid intake and output when managing fluid resuscitation, i.e., input minus output. However, in scenarios involving combined diuretic intervention, this arithmetic logic has serious masking flaws. For example, a large amount of colloid fluid penetrates capillaries and seeps into the periphery of the gastrointestinal tract, and is subsequently drawn out again through high-frequency diuresis, appearing as a stable, albeit non-significantly increased, result in a flat overall load on the arithmetic daily report. This repeated infiltration and extraction of free fluid causes the delicate gastrointestinal smooth muscle to undergo frequent volume expansion and contraction. The cumulative load on the tissues caused by this repeated volume fluctuation is masked by the arithmetic cancellation of positive and negative values.
[0004] Furthermore, inflammatory and immunological test data used to reflect tissue vulnerability, such as C-reactive protein, are low-frequency indicators, while intravenous infusion and urine output data are high-frequency indicators. There is a significant asynchronous gap between these two data points in time, preventing existing monitoring systems from accurately correlating discrete inflammatory markers with continuous fluid fluctuations. Clinicians are forced to rely on subjective experience for vague warnings, lacking data-driven objective thresholds, resulting in low accuracy in predicting gastrointestinal risks in patients. Summary of the Invention
[0005] To address the technical problem of low accuracy in predicting gastrointestinal function risks in patients, the present invention aims to provide a gastrointestinal function risk prediction system for SAP patients after colloid resuscitation. The specific technical solution adopted is as follows: In a first aspect, embodiments of the present invention provide a gastrointestinal function risk warning system for SAP patients after colloid resuscitation, the system comprising: The interval division module is used to obtain the test time interval corresponding to the patient's hematocrit test report; The leakage separation module is used to calculate the net retained fluid volume within a test time interval by combining the total intravenous infusion volume and the total urine outflow volume; to determine the intravascular expansion volume using the measured hematocrit values at the start and end points of the test time interval; and to determine the interstitial leakage fluid volume based on the net retained fluid volume and the intravascular expansion volume. The feature recombination module is used to extract the volume change cycle, which includes the volume change amplitude and the basic liquid phase retention, based on the time series sequence of interstitial leakage fluid volume. The volume change amplitude is the maximum amplitude difference of interstitial leakage fluid volume in the volume change cycle, and the basic liquid phase retention is the minimum non-zero interstitial leakage fluid volume in the volume change cycle. The weighted fusion module is used to match the volume change cycle with the corresponding test time interval; based on the measured C-reactive protein value and the initial baseline CT score in the matched test time interval, the inflammation severity weight of the test time interval is determined; combined with the inflammation severity weight, volume change amplitude and baseline liquid retention, the cumulative damage index of gastrointestinal smooth muscle is obtained. The early warning module generates a prompt signal to assist in medical decision-making when the cumulative damage index exceeds a preset damage threshold.
[0006] Secondly, a method for early warning of gastrointestinal function risk in SAP patients after colloid resuscitation is provided, the method comprising the following steps: Obtain the laboratory time interval corresponding to the patient's hematocrit test report; Within the testing time interval, the net retained fluid volume is calculated by the total intravenous infusion volume and the total urine outflow volume; the intravascular expansion volume is determined by the measured hematocrit values at the beginning and end points of the testing time interval; and the interstitial leakage fluid volume is determined based on the net retained fluid volume and the intravascular expansion volume. Based on the time series of interstitial leakage fluid volume, the volume change period, which includes the volume change amplitude and the basic liquid phase retention, is extracted; where the volume change amplitude is the maximum amplitude difference of interstitial leakage fluid volume in the volume change period, and the basic liquid phase retention is the minimum non-zero interstitial leakage fluid volume in the volume change period. The corresponding test time intervals were matched with the volume change cycle; the inflammation severity weight of the test time interval was determined based on the measured C-reactive protein value and the initial baseline CT score in the matched test time interval; and the cumulative damage index of gastrointestinal smooth muscle was obtained by combining the inflammation severity weight, volume change amplitude and baseline liquid retention. When the cumulative damage index exceeds the preset damage threshold, a prompt signal is generated to assist in medical decision-making.
[0007] The embodiments of the present invention have at least the following beneficial effects: This invention first determines the interstitial leakage fluid volume based on the net retained fluid volume and intravascular expansion volume. To avoid masking repeated fluctuations in body fluid using conventional addition and subtraction methods, it extracts the volume change cycle, which includes the amplitude of volume change and the baseline fluid retention, by analyzing the amplitude characteristics of the closure cycle in the time series of interstitial leakage fluid volume. Addressing data gaps caused by asynchronous clinical sampling, this invention matches the volume change cycle with the laboratory time interval, achieving precise alignment of multimodal time series data. Finally, by combining inflammation severity weights, volume change amplitude, and baseline fluid retention, this invention obtains a cumulative damage index for gastrointestinal smooth muscle. This index objectively characterizes the true load evolution process of gastrointestinal smooth muscle under repeated expansion and contraction, providing a reliable and objective quantitative auxiliary decision-making basis for early warning of gastrointestinal functional risks after colloid fluid resuscitation. Attached Figure Description
[0008] Figure 1 This is a system block diagram of a gastrointestinal function risk warning system for SAP patients after colloid resuscitation, provided in one embodiment of the present invention. Figure 2 This is a flowchart of a method for early warning of gastrointestinal function risks in SAP patients after colloid resuscitation, provided by an embodiment of the present invention. Detailed Implementation
[0009] The following description, in conjunction with the accompanying drawings, details the specific scheme of the gastrointestinal function risk warning system for SAP patients after colloid resuscitation provided by the present invention.
[0010] Please see Figure 1 The diagram illustrates a system block diagram of a gastrointestinal function risk warning system for SAP patients after colloid resuscitation, according to an embodiment of the present invention. The system includes the following modules: Interval division module 10 is used to obtain the test time interval corresponding to the patient's hematocrit test report.
[0011] First, after the patient is diagnosed with severe acute pancreatitis, retrieve the patient's first hematocrit (HCT) test report. Extract the absolute timestamp of the first test report and use this absolute timestamp as the initial observation baseline. Simultaneously, the initial observation reference point is extracted. Corresponding initial value of hematocrit This invention establishes an initial observation reference point. This solved the problem of lacking historical comparison benchmarks in the early stages of data monitoring.
[0012] After establishing the initial observation baseline, the timestamps of subsequent hematocrit test reports were extracted in chronological order according to the actual calendar time. .in, It is a positive integer, representing the sequence number of the nth subsequent hematocrit test report.
[0013] Subsequently, an iterative approach was adopted to divide the testing time intervals. Specifically, the initial observation baseline was established. As the starting point, the timestamp of the second hematocrit test report obtained will be used. As the end point of the time interval, a closed time period is generated, which serves as the testing time interval. The system marks this as the first testing time interval, i.e. And so on, until the process reaches the... When performing the next interval segmentation, the timestamp of the previous hematocrit test report is extracted. As a starting point, the timestamp of the next adjacent hematocrit test report will be used. As the endpoint, the nth generation of testing time intervals is used. Through the above iterative operations, a series of consecutive testing time intervals are generated.
[0014] The leakage separation module 20 is used to calculate the net retained fluid volume within the test time interval by using the total intravenous infusion volume and the total urine outflow volume; to determine the intravascular expansion volume using the measured hematocrit values at the beginning and end points of the test time interval; and to determine the interstitial leakage fluid volume based on the net retained fluid volume and the intravascular expansion volume.
[0015] After receiving the defined test time intervals, the corresponding data are retrieved from the hospital's medical order system and nursing records. By analyzing the total macroscopic body fluid volume and the red blood cell dilution ratio, the actual volume of interstitial fluid leaking through capillaries into the surrounding tissues is determined.
[0016] The amount of fluid in the blood vessels can be inferred from the dilution of hematocrit (HCT). This calculation method is valid only if the patient's absolute total red blood cell count does not undergo drastic artificial increases or decreases, or pathological losses, within the corresponding time period. Therefore, certain constraints must be set for this calculation.
[0017] Taking the nth test time interval as an example, before performing numerical calculations, a query command is first sent to the hospital's medical order system. This retrieves the patient's information within that test time interval. Within the body, is there a record of red blood cell suspension infusion or a confirmed diagnosis of active gastrointestinal bleeding?
[0018] If this occurs, it indicates that the patient's absolute red blood cell count has changed. At this point, the fluid volume inversion calculation for the current testing time interval is stopped, the data value of the previous testing time interval is automatically maintained and output, and a data distortion prompt is displayed on the terminal.
[0019] If not, further obtain the measured hematocrit value for the current testing time interval. Measured values of hematocrit Perform a non-zero check to verify the measured hematocrit value. Is it greater than 0? If the measured hematocrit value is... If the value is less than or equal to 0, it indicates an malfunction in the testing equipment or data loss. This abnormal data point is discarded, and the calculation for the current testing time interval is terminated. This pre-emptive screening prevents the calculation model from violating physiological principles and avoids the risk of program crashes caused by a denominator of 0 in subsequent division formulas.
[0020] When the measured value of hematocrit When the value is greater than 0, retrieve the data falling within the nth test time interval from the hospital information system. All intravenous infusion and urination records are included, specifically: the total volume of all colloid solutions, the total volume of crystalloid solutions, and the total volume of urine excreted due to diuretic intervention or natural excretion within the testing time interval. It should be noted that the total volume of urine excreted is the same as the total amount of urine flowing out.
[0021] The total volume of all colloid solutions and crystalloid solutions input within the testing time interval is arithmetically added together to obtain the total input amount for that testing time interval, which is recorded as the total intravenous infusion volume; that is, the sum of the total volume of colloid solutions and the total volume of crystalloid solutions is taken as the total intravenous infusion volume.
[0022] The difference between the total intravenous infusion volume and the total urine output volume is taken as the net fluid retention volume of the interval. More specifically: the net fluid retention volume of the interval is obtained by subtracting the total urine output volume from the total intravenous infusion volume. Net liquid volume retained in the interval It represents the amount of water that a patient's body gains or loses on a macroscopic level.
[0023] Read the patient's weight data recorded upon admission. Based on the patient's weight data and a preset blood volume conversion constant, calculate the patient's baseline theoretical blood volume. Specifically, multiply the weight data by the preset blood volume conversion constant to calculate the patient's baseline theoretical blood volume. The blood volume conversion constant is set according to conventional medical standards. In this embodiment of the invention, the system for adult male patients is based on... Calculations were performed on the adult female patient system. Perform the calculation.
[0024] Obtaining basic theoretical blood volume Then, the measured hematocrit value at the starting point of the testing time interval was obtained. Measured hematocrit value at the end of the interval The blood dilution coefficient is obtained by comparing the measured hematocrit values at the beginning and end of the test time interval.
[0025] The intravascular expansion volume is obtained by combining the patient's basic theoretical blood volume and blood dilution coefficient.
[0026] In this embodiment of the invention, the intravascular volume expansion during the nth testing time interval The calculation formula is: ;in, This represents the blood dilution factor.
[0027] Among them, blood dilution coefficient Its function is to quantify the degree to which blood is diluted. When infused free water enters the bloodstream, red blood cells are diluted, leading to changes in the subsequent measured hematocrit value. It will inevitably decrease, thus reducing the blood dilution factor. Greater than 1. Then, use the blood dilution factor. Subtracting the constant 1 aims to eliminate the original base volume of blood, i.e., 1 times the base volume, thereby simply separating the additional fluid proportion. The system will then process this purely increased fluid proportion. Multiply by the basic theoretical blood volume This allows for the accurate calculation of the absolute increase in fluid volume that actually remains inside the blood vessels during the testing time interval.
[0028] Assuming bleeding is excluded, the denominator is... The smaller the value, the more severe the dilution. The larger the relative size, the greater the final intravascular expansion volume. The larger it gets, the bigger it becomes.
[0029] In this embodiment of the invention, the volume of interstitial leakage fluid is determined based on the net retained fluid volume of the interstitial space and the intravascular expansion volume, specifically: After obtaining the net retained fluid volume representing the total macroscopic retention and the intravascular expansion volume representing the microscopic intravascular retention, a difference subtraction operation was performed to extract the water hidden in the interstitial spaces surrounding the gastrointestinal tract. (Interstitial space leakage volume) The calculation formula is: .
[0030] The formula for calculating the volume of interstitial fluid leakage is based on the net retained fluid volume of the interval that characterizes the total increase in total fluid volume on a macroscopic scale in the patient's body. Subtract the intravascular expansion volume, which represents the effective amount of fluid retained within the blood vessel. The remaining difference must be the volume of water that leaks through the damaged capillary wall into the interstitial space outside the blood vessel.
[0031] The formula calculates the output. It is a numerical characteristic with positive and negative polarity. When the volume of fluid leaking from the interstitial space... A positive value indicates a significant outflow of fluid from the blood vessels during the current testing time interval, placing an outward volume expansion load on the gastrointestinal smooth muscle; when fluid is extracted due to the use of large doses of diuretics, the volume of fluid leaking from this interstitial space is [value missing]. A negative value indicates a decrease in water content within the interstitial space, with the smooth muscle bearing the inward volumetric load.
[0032] The feature recombination module 30 is used to extract the volume change cycle, which includes the volume change amplitude and the basic liquid phase retention, based on the time series sequence of interstitial leakage fluid volume; wherein, the volume change amplitude is the maximum amplitude difference of interstitial leakage fluid volume in the volume change cycle, and the basic liquid phase retention is the minimum non-zero interstitial leakage fluid volume in the volume change cycle.
[0033] Since the absolute free water volume in a patient's body is a gradually accumulating process, and under normal physiological conditions, the lower limit of free water volume is the base state of having no edema, i.e., a volume of 0, a negative volume is impossible. When a negative value appears, it indicates an abnormal situation where water is extracted due to the use of large doses of diuretics. To prevent negative values from exceeding the limit and violating objective natural law when diuresis is excessive, a non-negative forced truncation constraint is introduced when splicing sequences.
[0034] The time series of interstitial leakage fluid volume is truncated to obtain a non-negative time series of interstitial leakage fluid volume.
[0035] Specifically: It receives the total volume of interstitial leakage fluid from the first testing time interval to the current testing time interval, i.e. to Following a chronological order, these data are progressively summed to generate a time-series sequence representing the fluctuation of absolute free water volume in the interstitial space. That is, the time-series sequence of interstitial space leakage fluid volume contains... .
[0036] During each step-by-step accumulation operation, the final result value of the current accumulation node is determined. If the calculation result of the current accumulation node is greater than or equal to 0, the system retains this calculated value as the sequence value of that node; if the calculation result of the current accumulation node is less than 0, the sequence value of that node is forcibly truncated and modified to 0. By introducing this truncation logic, it is ensured that the generated continuous time series always remains non-negative, making it fully consistent with the real logic of biological volume. This achieves non-negative truncation processing of the time series of interstitial leakage fluid volume, resulting in a non-negative time series of interstitial leakage fluid volume. It should be noted that, to avoid the problem of insufficient data on interstitial leakage fluid volume, interpolation compensation processing is performed on the time series of interstitial leakage fluid volume. In this embodiment of the invention, linear interpolation can be used to perform interpolation compensation on the time series of interstitial leakage fluid volume.
[0037] After obtaining the non-negative time-series sequence of the interstitial fluid leakage volume, the hidden fluctuation amplitude is extracted. Traditional statistical algorithms, such as variance or standard deviation calculations, can only assess the overall dispersion of the dataset and cannot extract closed deformation drops with temporal causal relationships. Therefore, this invention employs a rainflow counting algorithm capable of extracting temporal peak-valley pairing features.
[0038] Using the rainflow counting algorithm, multiple closed volume change cycles are extracted from a non-negative time series.
[0039] To accommodate the continuously increasing flow of medical monitoring data, an online sliding window mode is adopted to run the rainflow counting logic. The specific operation is as follows: The constructed non-negative time series is sent to the online rainflow feature extraction module. First, the non-negative time series is scanned within the sliding window to identify the local maxima where the data changes from rising to falling, which is defined as a peak; and the local minima where the data changes from falling to rising are identified, which are defined as troughs.
[0040] After identifying peaks and troughs, the algorithm uses the rain path interception principle of the rainflow counting algorithm to determine whether the current extreme point can form a complete closed drop. Instead of waiting for the entire course of the disease to end, a strategy of caching half-cycle data is adopted: when the algorithm detects that an upward seepage segment is completely enveloped and intercepted by a downward pumping segment of equal or greater depth, it determines that a complete expansion-to-retraction segment has been identified. The upward seepage segment is from trough to peak, and the downward pumping segment is from peak to trough. This complete expansion-to-retraction segment is then truncated from the chaotic non-negative time series and reassembled into an independent, closed-loop volume change cycle.
[0041] Taking the k-th volume change cycle as an example, the volume change amplitude of the k-th volume change cycle is extracted. The extraction method is as follows: the difference between the maximum and minimum interstitial fluid leakage volume within the k-th volume change period is calculated as the volume change amplitude. The difference between the maximum and minimum interstitial interstitial fluid leakage volume is also the maximum amplitude difference of interstitial interstitial fluid leakage volume during the volume change period. This value... The larger the value, the more intense the unidirectional deformation of the gastrointestinal tract periphery during this specific closure period, characterized by the expansion of free fluid followed by diuresis and subsequent shrinkage, and the heavier the corresponding tissue mechanical tensile load.
[0042] Then, extract the basic liquid phase retention volume for the kth volume change cycle. The extraction method is as follows: obtain the minimum non-zero interstitial leakage fluid volume within the k-th volume change cycle, i.e., the amount of water at the bottom when the volume fluctuation has not dropped to 0. This value... The larger the value, the greater the volume of static stagnant water that remains in the interstitial spaces for a long time and does not participate in flow exchange during the occurrence of the fluctuation.
[0043] The weighted fusion module 40 is used to match the volume change cycle with the corresponding test time interval; based on the measured C-reactive protein value and the initial baseline CT score in the matched test time interval, the inflammation severity weight of the test time interval is determined; combined with the inflammation severity weight, volume change amplitude and baseline liquid retention, the cumulative damage index of gastrointestinal smooth muscle is obtained.
[0044] Within the k-th volume change cycle, the maximum interstitial fluid leakage volume is obtained, and the absolute calendar time coordinate corresponding to the maximum interstitial fluid leakage volume is obtained as the absolute peak timestamp of the k-th volume change cycle. The testing time interval corresponding to the largest interstitial leakage volume is then used as the testing time interval to match the volume change cycle. More specifically: the absolute peak timestamp of the kth volume change cycle is retrieved. Using this unique time anchor point, compare the start and end times of all defined test time intervals to determine the absolute peak timestamp. Specifically, it falls into a specific testing time interval, thus obtaining the testing time interval matched for each volume change cycle.
[0045] Since the actual degree of damage to the gastrointestinal wall depends not only on physical volumetric stretching but also on the non-linear decrease in tolerance caused by inflammatory damage, this invention uses absolute peak timestamps as indexes to find biochemical and imaging inflammatory data within the corresponding laboratory time intervals. Subsequently, through joint calculation, the volumetric change amplitude without time attributes is fused with discrete inflammatory weights to generate a final unidirectionally increasing cumulative damage index.
[0046] Since the frequency of C-reactive protein (CRP) testing, which characterizes systemic inflammatory response, is much lower than the monitoring frequency of urine output during intravenous infusion, a historical proximity query strategy is adopted to ensure that each independent test time interval corresponds to an accurate tissue vulnerability assessment.
[0047] The modified CT severity index recorded at the time of admission and diagnosis was retrieved from the medical record database and used as the initial baseline CT score. Subsequently, for the currently processing test time interval, the system extracts the end timestamp of that test time interval. Using this endpoint timestamp as a constraint, the system initiates a query to the hospital's laboratory database, searching for tests conducted earlier than or equal to [the specified time]. All C-reactive protein history records, and extract the timestamps from the endpoint. The most recent test result will be used as the most recent historical C-reactive protein measurement value for the current testing time period. This constraint effectively prevents logical errors in the algorithm from accessing future data.
[0048] After obtaining the baseline CT score and measured C-reactive protein (CRP) value, since both indicators may be zero when the patient is healthy, directly performing arithmetic operations could easily lead to a product of zero. To avoid algorithm crashes, the system introduces a natural base constant and bias addition, using the following logarithmic fusion formula to calculate the local tissue inflammation severity weight for the current interval. More specifically: the measured CRP value is compared with the upper reference CRP value to determine the protein ratio coefficient; the initial baseline CT score is compared with the CT score full-score standard value to determine the image scoring coefficient; combining the protein ratio coefficient and the image scoring coefficient, the inflammation severity weight for the testing time interval is obtained. More specifically: the protein ratio coefficient and the image scoring coefficient are weighted and fused to obtain the inflammation severity weight for the testing time interval.
[0049] In this embodiment of the invention, the inflammation severity weighting of the testing time interval is... The calculation formula is: ;in, This indicates the most recent historical C-reactive protein measurement value retrieved. This refers to the upper reference limit for C-reactive protein set in the system's embedded medical dictionary. In this embodiment of the invention, it is used as a system calibration parameter and can be set as follows: ; This represents the natural constant, approximately equal to 2.718; This indicates the patient's initial baseline CT score at the time of diagnosis; The standard value of the CT score corresponding to the improved CT severity index is set to 10 points in this embodiment of the invention. This is the protein ratio coefficient; This is the image rating coefficient.
[0050] Among them, the arithmetic terms Its function is to calculate the fold increase in C-reactive protein levels and combine it with the natural constant. Establish a baseline to prevent a return to zero. When a patient's measured C-reactive protein level is completely normal, the excess factor is 0, plus the natural constant. Later, due to This ensures that the minimum value of the preceding logarithmic term is 1. (Explanation terms) The purpose of this is to smooth out the influence of the image score. The numerator is increased by 1 to ensure that even if the CT score is 0, the subsequent product term will not be 0.
[0051] Natural logarithm function The introduction of this method suppressed the extremely high CRP levels during severe infection, preventing an explosive increase in the product of these levels. The final calculated inflammation severity weights... It is always a coefficient greater than zero. The larger the value, the more severe the inflammatory damage to the gastrointestinal wall tissue at that moment, and the lower its physical tolerance to water expansion and stretching.
[0052] This allows us to obtain the inflammation severity weights for the corresponding testing time intervals for each volume change cycle.
[0053] After parameter alignment, the cumulative damage index of gastrointestinal smooth muscle is obtained by combining the inflammation severity weight, volume change amplitude, and baseline fluid retention. Specifically, the three characteristic parameters—inflammation severity weight, volume change amplitude, and baseline fluid retention—are fed into a joint summator. More specifically, for each testing time interval corresponding to a volume change cycle, the interval damage index for each testing time interval is obtained by combining the inflammation severity weight, volume change amplitude, and baseline fluid retention within that interval. The interval damage indices of all testing time intervals are then summed to obtain the cumulative damage index of gastrointestinal smooth muscle.
[0054] In this embodiment of the invention, the formula for calculating the cumulative damage index of gastrointestinal smooth muscle is as follows: ; in, This represents the total number of volume change cycles identified up to the current processing node. The traversal index variable represents the summation and accumulation, and it increments gradually from the first volume change cycle. This represents the weight of the local tissue inflammation severity in the laboratory time interval corresponding to the kth volume change cycle; This represents the basic liquid phase retention during the k-th volume change cycle; This represents the volume change magnitude during the k-th volume change cycle; The basic theoretical blood volume is denoted as ; norm is the normalization function. In this embodiment of the invention, the maximum and minimum value normalization method is adopted, wherein the maximum and minimum values are the maximum and minimum values in the historical data within the past month. When the normalized value exceeds 1, it is truncated to 1; when the normalized value is less than 0, it is truncated to 0.
[0055] Among them, for the terms of the expression The mandatory square operation on the volume change amplitude is due to the mathematical properties of the square operation, which unifies positive and negative seepage expansion values into non-negative positive square values. This square operation successfully transforms the macroscopically offsetting positive and negative water flow into an equivalent volume fluctuation damage that only increases and never decreases. Multiplying this square equivalent by the static water pressure at the bottom layer, i.e., multiplying by the formula term... Then multiply by the inflammation severity weight. The product of these three factors represents the interval damage index caused by a single waveband. The summation sign on the outer layer of the formula ensures that the interval damage index calculated for each new complete volume change cycle is continuously added in. The final output is the cumulative damage index of the gastrointestinal smooth muscle. It shows an absolute unidirectional increasing trend. This index eliminates the illusory stability caused by simple arithmetic addition and subtraction. The larger the value, the more objectively and quantitatively it represents the heavier adverse load accumulated by repeated stretching of the patient's gastrointestinal smooth muscle.
[0056] The early warning module 50 is used to generate a prompt signal to assist medical decision-making when the cumulative damage index exceeds a preset damage threshold.
[0057] Due to the cumulative damage index of gastrointestinal smooth muscle It is a unidirectionally increasing absolute value, and if it deviates from the preset reference benchmark, the computer program's automatic judgment mechanism will not be triggered. Therefore, this invention introduces an objective static comparison threshold to screen the cumulative damage index output by the above calculation for exceeding the limit. When the limit is confirmed to be exceeded, a prompt signal to assist clinical medical staff in decision-making is output, thereby completing the data flow closed loop of risk warning.
[0058] To ensure readily available diagnostic benchmarks during initial deployment, default calibration logic was configured during the initialization phase. Specifically, the system connects to and accesses a database of past clinical records, filtering for historical data samples of patients with severe acute pancreatitis who had experienced severe gastrointestinal dysfunction. The same algorithm described above is used to extract the cumulative impairment index of these samples at the onset of diagnosed symptoms. Subsequently, the median of the cumulative impairment index from these historical data is calculated and automatically set as the default threshold for smooth muscle damage. The extraction of this preset damage threshold provides objective historical data to support the judgment.
[0059] Comparison of cumulative damage index and preset damage threshold : When the cumulative damage index is less than or equal to the preset damage threshold, it is determined that the current condition is still within the range of tolerable fluctuation load, and the background silent monitoring state is maintained, waiting for the data inflow of the next test time interval.
[0060] When the cumulative damage index exceeds a preset damage threshold, a warning signal is generated to assist in medical decision-making. This is because when the cumulative damage index exceeds the preset threshold, it indicates that the computer system has assessed that the physical fluctuation load accumulated by the patient due to repeated infiltration and removal of free fluids into and from tissues has exceeded the safety limits of historical severe cases. Based on the comparison results, the current patient data is determined to have entered the risk warning zone for excessive expansion and contraction of gastrointestinal smooth muscle.
[0061] Because directly cutting off intravenous infusion or forcibly administering medication via computer programs carries uncertain medical safety risks, this invention does not directly intervene in the automated physical control of any infusion pump or drug delivery device. When the cumulative damage index exceeds a preset damage threshold, auxiliary prompt information is immediately generated and distributed.
[0062] First, a warning electrical signal is sent to the display terminal located at the doctor's workstation or ward nursing station. This signal triggers the screen display module, forcibly popping up a prominent chart and text box on the user interface. The text box contains the patient's bed number, the specific value of the current cumulative damage index, and a warning notification that the volumetric fluctuation load has exceeded the limit.
[0063] Secondly, guiding text prompts for assisting clinical medical intervention decisions are generated simultaneously. Text suggestions recommending the discontinuation of intravenous infusion of macromolecular colloid solutions and the increase in the frequency of administration of diuretics such as furosemide are sent to the aforementioned display terminal. By outputting the above-mentioned text and chart prompts containing specific risk indicators and intervention recommendations, the risk of repeated microscopic fluid leakage, which is originally imperceptible to the naked eye, is numerically displayed, providing clinicians with objective and reliable data-driven decision-making references for further medical diagnosis and intervention. It should be noted that the prompts are for physician reference only.
[0064] Please see Figure 2 , Figure 2 This invention provides a flowchart of a method for early warning of gastrointestinal function risks in SAP patients after colloid resuscitation. The method includes: Obtain the laboratory time interval corresponding to the patient's hematocrit test report; Within the testing time interval, the net retained fluid volume is calculated by the total intravenous infusion volume and the total urine outflow volume; the intravascular expansion volume is determined by the measured hematocrit values at the beginning and end points of the testing time interval; and the interstitial leakage fluid volume is determined based on the net retained fluid volume and the intravascular expansion volume. Based on the time series of interstitial leakage fluid volume, the volume change period, which includes the volume change amplitude and the basic liquid phase retention, is extracted; where the volume change amplitude is the maximum amplitude difference of interstitial leakage fluid volume in the volume change period, and the basic liquid phase retention is the minimum non-zero interstitial leakage fluid volume in the volume change period. The corresponding test time intervals were matched with the volume change cycle; the inflammation severity weight of the test time interval was determined based on the measured C-reactive protein value and the initial baseline CT score in the matched test time interval; and the cumulative damage index of gastrointestinal smooth muscle was obtained by combining the inflammation severity weight, volume change amplitude and baseline liquid retention. When the cumulative damage index exceeds the preset damage threshold, a prompt signal is generated to assist in medical decision-making.
Claims
1. A system for predicting the risk of gastrointestinal function in a SAP patient after colloidal fluid resuscitation, characterized in that, The system includes the following modules: The interval division module is used to obtain the test time interval corresponding to the patient's hematocrit test report; The leakage separation module is used to calculate the net retained fluid volume within a test time interval by combining the total intravenous infusion volume and the total urine outflow volume; to determine the intravascular expansion volume using the measured hematocrit values at the start and end points of the test time interval; and to determine the interstitial leakage fluid volume based on the net retained fluid volume and the intravascular expansion volume. The feature recombination module is used to extract the volume change cycle, which includes the volume change amplitude and the basic liquid phase retention, based on the time series sequence of interstitial leakage fluid volume. The volume change amplitude is the maximum amplitude difference of interstitial leakage fluid volume in the volume change cycle, and the basic liquid phase retention is the minimum non-zero interstitial leakage fluid volume in the volume change cycle. The weighted fusion module is used to match the volume change cycle with the corresponding test time interval; based on the measured C-reactive protein value and the initial baseline CT score in the matched test time interval, the inflammation severity weight of the test time interval is determined; combined with the inflammation severity weight, volume change amplitude and baseline liquid retention, the cumulative damage index of gastrointestinal smooth muscle is obtained. The early warning module generates a prompt signal to assist in medical decision-making when the cumulative damage index exceeds a preset damage threshold.
2. The system for predicting the risk of gastrointestinal function in a SAP patient after colloidal fluid resuscitation according to claim 1, characterized in that, The calculation of the net retained fluid volume within the testing time interval, based on the total intravenous infusion volume and the total urine outflow, includes: For each testing time interval, obtain the total volume of all colloidal solutions, the total volume of crystalloid solutions, and the total volume of urine excreted within the testing time interval; The sum of the total volume of colloidal solution and the total volume of crystalloid solution is taken as the total intravenous infusion volume; The difference between the total intravenous infusion volume and the total urine output volume is taken as the net fluid retention volume of the interval.
3. The system for predicting the risk of gastrointestinal function in a SAP patient after colloidal fluid resuscitation according to claim 1, wherein, The determination of intravascular volume expansion using the measured hematocrit values at the start and end points of the testing time interval includes: Based on the patient's weight data and the preset blood volume conversion constant, the patient's baseline theoretical blood volume is obtained; The blood dilution factor is obtained by comparing the measured hematocrit values at the beginning and end of the test time interval. The intravascular expansion volume is obtained by combining the patient's basic theoretical blood volume and blood dilution coefficient.
4. The system for predicting the risk of gastrointestinal function in a SAP patient after colloidal fluid resuscitation according to claim 1, wherein, The time series sequence of interstitial leakage fluid volume extracts the volume change period, which includes the volume change amplitude and the basic liquid retention volume, including: The time series sequence of interstitial leakage fluid volume is non-negatively truncated to obtain the non-negative time series sequence of interstitial leakage fluid volume. Using the rainflow counting algorithm, multiple closed volume change cycles are extracted from a non-negative time series.
5. The system for predicting the risk of gastrointestinal function in a SAP patient after colloidal fluid resuscitation according to claim 1, wherein, The test time intervals corresponding to the volume change cycle include: Determine the maximum interstitial fluid leakage volume during the volume change cycle, and use the testing time interval containing the maximum interstitial fluid leakage volume as the testing time interval that matches the volume change cycle.
6. The system for predicting the risk of gastrointestinal function in a SAP patient after colloidal fluid resuscitation according to claim 1, wherein, The determination of the inflammation severity weight for each testing time interval, based on the measured C-reactive protein values and the initial baseline CT score within the matched testing time interval, includes: Compare the measured C-reactive protein value with the upper reference limit of C-reactive protein to determine the protein ratio coefficient; Compare the initial baseline CT score with the standard value of the full CT score to determine the image scoring coefficient; By combining the protein ratio coefficient and the image scoring coefficient, the inflammation severity weight for each testing time interval is obtained.
7. The gastrointestinal function risk warning system for SAP patients after colloid resuscitation according to claim 6, characterized in that, The method of combining the protein ratio coefficient and the image scoring coefficient to obtain the inflammation severity weight for the testing time interval includes: The protein ratio coefficient and the image scoring coefficient are weighted and fused to obtain the inflammation severity weight for the test time interval.
8. The gastrointestinal function risk warning system for SAP patients after colloid resuscitation according to claim 1, characterized in that, The cumulative damage index of gastrointestinal smooth muscle is obtained by combining the severity of inflammation, the magnitude of volume change, and the baseline fluid retention, including: For each test time interval corresponding to each volume change cycle, the interval impairment index for each test time interval is obtained by combining the inflammation severity weight, volume change amplitude and baseline liquid phase retention within the test time interval. The cumulative damage index of gastrointestinal smooth muscle is obtained by summing the interval damage indices of all test time intervals.
9. The gastrointestinal function risk warning system for SAP patients after colloid resuscitation according to claim 1, characterized in that, After obtaining the cumulative damage index of gastrointestinal smooth muscle, the method further includes: When the cumulative damage index is less than or equal to the preset damage threshold, the monitoring status is maintained, and the system waits for the data to flow in during the next testing time interval.
10. The gastrointestinal function risk warning system for SAP patients after colloid resuscitation according to claim 1, characterized in that, The method for determining the laboratory time interval corresponding to the patient's hematocrit test report is as follows: The timestamp of the patient's first hematocrit test report was used as the initial observation baseline. Based on the initial observation benchmark, discrete test time intervals with consecutive hematocrit test reports are divided according to the timestamps of consecutive hematocrit test reports.