A hemodialysis auxiliary care system and method

Abstract

The application relates to the field of intelligent auxiliary nursing, and specifically discloses a hemodialysis auxiliary nursing system and method. First, a temperature and humidity information acquisition module uses a sensing device to obtain humidity and temperature data in a dialysis area. A blood pressure influence parameter module analyzes the influence of the data on the blood pressure of a patient to obtain a blood pressure influence parameter, and then a dialysis influence parameter calculation module obtains a dialysis influence parameter. Meanwhile, a comfort parameter calculation module evaluates the influence of the environmental factors on the comfort of the patient to obtain a comfort perception influence parameter. A proportion calculation module determines the environmental comfort influence proportion by analyzing a patient's perception image array. Finally, a scheme generation module synthesizes all the parameters to formulate an optimal auxiliary nursing scheme. The scheme realizes personalized and accurate nursing scheme optimization, and improves the stability of the dialysis effect and the comfort of the patient.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intelligent assisted nursing, and more specifically, to a hemodialysis assisted nursing system and method. BACKGROUND

[0002] As a common renal replacement therapy, hemodialysis mainly targets patients with renal failure, simulates kidney function with the help of dialysis equipment, and removes metabolic waste such as creatinine, urea nitrogen, and excess water from the patient's body through diffusion, convection, etc., to maintain the stability of the body's internal environment. However, traditional nursing often ignores the influence of environmental factors such as temperature and humidity on dialysis effect and patient comfort, resulting in unstable nursing quality.

[0003] To this end, the existing patent CN118762796A proposes an intelligent hemodialysis assisted nursing method and system based on dynamic environment perception, which collects humidity and temperature information arrays of multiple sensing points in the dialysis area with the help of sensing devices. Through these information analysis, the influence on the blood pressure of dialysis patients is obtained, and the blood pressure influence parameters are classified as dialysis influence parameters; at the same time, comfort influence analysis is performed to obtain sensing comfort influence parameters. Then, the patient sensing image array is collected, and the actual comfort influence parameters are analyzed to calculate the environmental comfort influence proportion. Finally, the above-mentioned parameters are integrated to optimize the assisted nursing scheme, and the optimal temperature and humidity adjustment scheme of the dialysis environment is determined, thereby solving the problem of traditional nursing ignoring environmental factors and individual differences, and realizing personalized nursing optimization.

[0004] In the existing patent, the calculation of blood pressure influence parameters is based on static input humidity and temperature information arrays, without modeling the continuous changes in the time dimension. This method ignores the cumulative effect and lagging influence of environmental parameter fluctuation trends (such as sudden temperature rise or gradual drop) on blood pressure, resulting in the inability to accurately capture the influence of dynamic changes on patient blood pressure. In addition, relying only on a simple regression model to learn feature associations makes it difficult to represent complex cross-modal spatiotemporal interaction mechanisms, such as the synergistic effect of local high-temperature regions and adjacent high-humidity regions on blood pressure. This limitation makes the prediction results less robust and accurate in the face of changing real-world environments.

[0005] Therefore, an optimized hemodialysis assisted nursing scheme is desired. SUMMARY

[0006] To solve the above technical problems, the present application is proposed.

[0007] According to one aspect of the present application, a hemodialysis assisted nursing system is provided, comprising:

[0008] a temperature and humidity information acquisition module, configured to acquire a humidity information array and a temperature information array in a hemodialysis area through a sensing device;

[0009] a blood pressure influence parameter module configured to perform blood pressure influence analysis on the humidity information array and the temperature information array based on the dialysis patient to obtain a blood pressure influence parameter, wherein the blood pressure influence parameter module is configured to perform temperature-humidity spatiotemporal interaction chain reasoning analysis on the humidity information array and the temperature information array based on time domain features to obtain the blood pressure influence parameter;

[0010] a dialysis influence parameter calculation module configured to obtain a dialysis influence parameter based on the blood pressure influence parameter;

[0011] a comfort parameter calculation module configured to perform comfort influence analysis on the humidity information array and the temperature information array based on the dialysis patient to obtain a comfort perception influence parameter;

[0012] a proportion calculation module configured to obtain a perception image array of the dialysis patient, and perform environmental comfort analysis on the perception image array to obtain an environmental comfort influence proportion;

[0013] a scheme generation module configured to obtain an optimal auxiliary nursing scheme based on the dialysis influence parameter, the comfort perception influence parameter, and the environmental comfort influence proportion.

[0014] According to another aspect of the present application, a blood dialysis auxiliary nursing method is provided, comprising:

[0015] obtaining a humidity information array and a temperature information array in a blood dialysis area through a perception device;

[0016] performing blood pressure influence analysis on the humidity information array and the temperature information array based on the dialysis patient to obtain a blood pressure influence parameter, comprising: performing temperature-humidity spatiotemporal interaction chain reasoning analysis on the humidity information array and the temperature information array based on time domain features to obtain the blood pressure influence parameter;

[0017] obtaining a dialysis influence parameter based on the blood pressure influence parameter;

[0018] performing comfort influence analysis on the humidity information array and the temperature information array based on the dialysis patient to obtain a comfort perception influence parameter;

[0019] obtaining a perception image array of the dialysis patient, and performing environmental comfort analysis on the perception image array to obtain an environmental comfort influence proportion;

[0020] obtaining an optimal auxiliary nursing scheme based on the dialysis influence parameter, the comfort perception influence parameter, and the environmental comfort influence proportion.

[0021] Compared with the prior art, the blood dialysis auxiliary nursing system and method provided by the application first acquires the humidity and temperature data in the dialysis area by using a sensing device in a temperature and humidity information acquisition module. A blood pressure influence parameter module analyzes the influence of these data on the blood pressure of the patient to obtain blood pressure influence parameters, and then a dialysis influence parameter calculation module obtains dialysis influence parameters. At the same time, a comfort parameter calculation module evaluates the influence of these environmental factors on the comfort of the patient to obtain comfort perception influence parameters. A proportion calculation module determines the environmental comfort influence proportion by analyzing the patient's perception image array. Finally, a scheme generation module synthesizes all the parameters to develop an optimal auxiliary nursing scheme. This scheme realizes personalized and accurate nursing scheme optimization, improves the stability of the dialysis effect and the comfort of the patient. BRIEF DESCRIPTION OF DRAWINGS

[0022] The above and other objects, features and advantages of the present application will become more apparent from the following detailed description of embodiments of the present application taken in conjunction with the accompanying drawings. The drawings provided in the Drawings serve to provide a further understanding of the embodiments of the present application, constitute a part of the specification and are included to explain the present application, and do not limit the present application. In the drawings, the same reference numerals generally refer to the same components or steps throughout the drawings.

[0023] Figure 1 A block diagram of a blood dialysis auxiliary nursing system according to an embodiment of the present application.

[0024] Figure 2 A block diagram of a blood pressure influence parameter module in a blood dialysis auxiliary nursing system according to an embodiment of the present application.

[0025] Figure 3 A block diagram of a temperature-humidity space-time interaction unit in a blood dialysis auxiliary nursing system according to an embodiment of the present application.

[0026] Figure 4 A flowchart of a blood dialysis auxiliary nursing method according to an embodiment of the present application. DETAILED DESCRIPTION

[0027] Embodiments of the present disclosure will be described in greater detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms, and should not be interpreted as being limited to the embodiments set forth herein, but rather, these embodiments are provided to more thoroughly and completely understand the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes, and are not intended to limit the scope of protection of the present disclosure.

[0028] It is worth noting that in this application, all actions of obtaining signals, information or data are carried out in accordance with the corresponding data protection regulations and policies of the country where the device is located, and with the authorization given by the owner of the corresponding device.

[0029] To solve the problems in the background art, the present application provides a hemodialysis auxiliary nursing system. Figure 1 The block diagram of the hemodialysis auxiliary nursing system according to the embodiments of the present application is shown in FIG. 1. Specifically, as shown in FIG. 1, the hemodialysis auxiliary nursing system 100 according to the embodiments of the present application comprises a temperature and humidity information acquisition module 110, a blood pressure influence parameter module 120, a dialysis influence parameter calculation module 130, a comfort parameter calculation module 140, a proportion calculation module 150, and a scheme generation module 160. Figure 1 The temperature and humidity information acquisition module 110 is configured to acquire a humidity information array and a temperature information array in a hemodialysis area through a sensing device. The blood pressure influence parameter module 120 is configured to perform blood pressure influence analysis on the humidity information array and the temperature information array based on a dialysis patient to obtain a blood pressure influence parameter. The dialysis influence parameter calculation module 130 is configured to obtain a dialysis influence parameter based on the blood pressure influence parameter. The comfort parameter calculation module 140 is configured to perform comfort influence analysis on the humidity information array and the temperature information array based on the dialysis patient to obtain a comfort perception influence parameter. The proportion calculation module 150 is configured to acquire a perception image array of the dialysis patient and perform environmental comfort analysis on the perception image array to obtain an environmental comfort influence proportion. The scheme generation module 160 is configured to obtain an optimal auxiliary nursing scheme based on the dialysis influence parameter, the comfort perception influence parameter, and the environmental comfort influence proportion.

[0030] In the embodiments of the present application, the temperature and humidity information acquisition module 110 is configured to acquire a humidity information array and a temperature information array in a hemodialysis area through a sensing device. Specifically, the sensing device comprises a sensor array arranged at multiple sensing points around a dialysis position. It should be understood that the humidity information array is a set of humidity data at the multiple sensing points around the dialysis position, and the temperature information array is a set of temperature data at the multiple sensing points around the dialysis position. The temperature and humidity in the dialysis environment are in dynamic change, and the data of a single sensing point cannot accurately reflect the overall environmental conditions. In actual dialysis scenarios, different parts of the patient's body have different feelings and reactions to temperature and humidity, and the layout of multiple sensing points can collect data from multiple dimensions, covering the temperature and humidity of each area around the patient, so as to obtain more comprehensive and accurate environmental information.

[0031] The specific implementation process is as follows: In the hemodialysis area, multiple sensing points are carefully arranged around the dialysis position. The layout of these sensing points is scientifically planned and distributed at key positions such as the patient's head and limbs. At each sensing point, a sensor array composed of high-precision temperature sensors and humidity sensors is arranged. These sensors are like sensitive "environmental antennae" that can accurately capture the temperature and humidity changes at the location in real time.

[0032] When dialysis begins, the sensor array enters a working state and continuously collects temperature and humidity data at each sensing point. The temperature data collected by the numerous sensing points is aggregated to form a temperature information array, which details the temperature distribution in different areas of the dialysis environment, such as whether the temperature in the patient's head area is stable, whether there are temperature differences near the limbs, and the like. Similarly, the humidity data at each sensing point is aggregated to form a humidity information array, which provides detailed information about the changes in the humidity of the dialysis environment, such as whether the humidity fluctuates within an appropriate range, whether there are local areas of excessive or insufficient humidity, and the like.

[0033] These real-time collected humidity information array and temperature information array are rapidly transmitted to the subsequent data processing system. They provide indispensable basic data for subsequent key links such as blood pressure influence analysis and comfort influence analysis. Through this multi-point sensing layout and accurate data collection method, the system can sensitively capture the small changes in the dialysis environment, lay a solid data foundation for subsequent accurate adjustment of the temperature and humidity in the dialysis environment, and improve the dialysis effect and comfort of patients. The entire blood dialysis auxiliary nursing process is more scientific and effective.

[0034] In the embodiment of the present application, the blood pressure influence parameter module 120 is configured to perform blood pressure influence analysis on the humidity information array and the temperature information array based on the blood pressure of the dialysis patient to obtain blood pressure influence parameters. Specifically, in the embodiment of the present application, the blood pressure influence parameter module is configured to perform temperature-humidity space-time interaction chain reasoning analysis based on time domain features on the humidity information array and the temperature information array to obtain the blood pressure influence parameters. Accordingly, environmental factors such as temperature and humidity will affect the dialysis effect, and blood pressure is an important indicator reflecting the dialysis effect and the patient's physical condition. Changes in temperature and humidity can affect the physiological functions of the human body, and dialysis patients are more fragile and more sensitive to environmental changes. Humidity can affect water metabolism in the human body, and temperature can affect the contraction and relaxation of blood vessels, thereby affecting blood pressure. Different combinations of temperature and humidity have different effects on the blood pressure of patients, so it is necessary to analyze the influence of the humidity information array and the temperature information array on the blood pressure of dialysis patients.

[0035] Thus, in view of the technical problems in the background art, the technical concept of the present application is to first model each sensing point in the temperature and humidity information array dynamically in time sequence, capturing the trend change characteristics of each sensing point temperature and humidity in the time dimension (such as the cumulative effect of temperature slow decline or the lag response of humidity sudden rise). On this basis, a local interaction chain reasoning mechanism is constructed to model the cross-modal coupling relationship between temperature and humidity parameters in the spatio-temporal interwoven feature space (for example, the dynamic influence of the synergistic effect of high temperature area and adjacent high humidity area on blood vessel contraction), and the nonlinear action law between local areas is gradually excavated through the chain transmission of the interactive encoder. Finally, the spatio-temporal synergistic interaction vector is decoded into a comprehensive parameter reflecting the blood pressure fluctuation risk. This scheme effectively solves the problems of missing trend response and ignoring local interaction in traditional static modeling by capturing the spatio-temporal evolution law of temperature and humidity parameters and their nonlinear synergistic mechanism, enabling blood pressure influence prediction to reflect the lag effect in dynamic fluctuations of environmental parameters and regional synergistic effect, thereby improving the adaptability of auxiliary care programs to complex environmental changes.

[0036] Figure 2 A block diagram of a blood pressure influencing parameter module according to the blood dialysis auxiliary care system of the embodiments of the present application. Specifically, as shown in Figure 2 the blood pressure influencing parameter module 120 includes: a humidity information time sequence encoding unit 121 for performing time series analysis on each humidity row vector in the humidity information array to obtain a set of single-point humidity time sequence correlation feature encoding vectors; a temperature information time sequence encoding unit 122 for performing time series analysis on each temperature row vector in the temperature information array to obtain a set of single-point temperature time sequence correlation feature encoding vectors; a temperature-humidity spatio-temporal interaction unit 123 for performing temperature-humidity spatio-temporal interaction chain reasoning on the set of single-point humidity time sequence correlation feature encoding vectors and the set of single-point temperature time sequence correlation feature encoding vectors to obtain a temperature-humidity spatio-temporal synergistic interaction encoding vector; and a temperature-humidity spatio-temporal interaction feature decoding unit 124 for decoding the temperature-humidity spatio-temporal synergistic interaction encoding vector to obtain the blood pressure influencing parameter.

[0037] Specifically, in the embodiment of the present application, the humidity information time series encoding unit 121 is configured to perform bidirectional recurrent network-based time series analysis on each humidity row vector in the humidity information array to obtain a set of single-point humidity time series correlation feature encoding vectors. It should be understood that the influence of dynamic changes in environmental humidity on the physiological state of the patient has significant time series correlation. Dialysis treatment usually lasts for several hours, and humidity fluctuations in the area where the patient is located (such as sudden local humidity rise / slow drop caused by equipment heat dissipation, body fluid evaporation, or air conditioning operation) will have cumulative effects or lag effects on vascular tone through skin microcirculation and the respiratory system. For example, the sustained decline in the humidity of a certain sensing point in the first half of the dialysis may exacerbate vasoconstriction through skin dehydration, and this process needs to be analyzed bidirectionally in combination with historical trends and future potential changes. Traditional one-way time series modeling can only capture the correlation in a single direction (such as from the past to the present), and cannot represent the evolution law of the humidity parameter on the complete time axis (such as the predictability of current humidity changes on subsequent trends), making it difficult to predict the dynamic correlation between the humidity parameter and blood pressure fluctuations in real-time monitoring scenarios. Based on this, the present application performs bidirectional recurrent network-based time series analysis on each humidity row vector in the humidity information array to obtain a set of single-point humidity time series correlation feature encoding vectors. Specifically, through the bidirectional network structure, the context-dependent relationship of the humidity parameter is extracted from both the forward (past to present) and backward (future to present) dimensions. Specifically, the forward layer captures the driving mechanism of historical humidity changes on the current state (such as the cumulative effect of three consecutive hours of humidity decline on vasoconstriction), and the backward layer learns the potential influence of the current humidity value on future trends (such as the current humidity surge may indicate subsequent accelerated evaporation rate). Through the bidirectional gating mechanism, the features of the two directions are fused, and finally the single-point humidity time series correlation feature encoding vector of each monitoring point is generated, which fully represents the dynamic evolution law of the humidity parameter in the time dimension and its potential causal chain with blood pressure fluctuations.

[0038] Specifically, in the embodiments of the present application, the temperature information time series encoding unit 122 is configured to perform the time series analysis based on the bidirectional recurrent network on each temperature row vector in the temperature information array to obtain a set of single-point temperature time series correlation feature encoding vectors. Correspondingly, it is considered that the adjustment of the cardiovascular system of the patient has significant time series dependence in view of the dynamic change of the environmental temperature. The temperature fluctuation in the dialysis area (such as the local temperature sudden rise / slow drop caused by the equipment heat production, air conditioning regulation or patient body surface heat dissipation) will affect the vasomotor state of the peripheral blood vessels through the heat radiation and conduction mechanism, and then cause the delayed or cumulative abnormality of the blood pressure. For example, the continuous rise of the temperature of a certain sensing point in the middle of dialysis can cause the blood vessels to expand, but this physiological effect can be shown several minutes after the temperature reaches the peak, and the influence degree is closely related to the temperature change rate and historical trend. The traditional one-way time series modeling can only extract the one-way evolution feature of the temperature parameter (such as the linear correlation from the past to the present), and cannot capture the bidirectional correlation of the temperature parameter on the complete time axis (such as the predictability of the current temperature value to the future trend), which limits the analysis ability of the complex causal chain between the temperature dynamic change and the blood pressure response. Based on this, in the technical solutions of the present application, the time series analysis based on the bidirectional recurrent network is performed on each temperature row vector in the temperature information array to obtain the nonlinear evolution law of the temperature parameter in the time dimension, and a set of single-point temperature time series correlation feature encoding vectors are obtained.

[0039] Specifically, the temperature-humidity spatiotemporal interaction unit 123 is configured to perform temperature-humidity spatiotemporal interaction chain reasoning on the set of single-point humidity time sequence correlation feature encoding vectors and the set of single-point temperature time sequence correlation feature encoding vectors to obtain a temperature-humidity spatiotemporal synergistic interaction encoding vector. Furthermore, considering that the influence of environmental temperature and humidity parameters on the blood pressure of a patient is not a single time or space dimension effect, but a complex process of spatiotemporal dynamic coupling. The temperature and humidity distribution at different monitoring points in the dialysis area has spatial heterogeneity (for example, the temperature near the dialysis equipment is relatively high, and the humidity gradient changes on the surface of the patient), and the time sequence fluctuation of local temperature and humidity parameters (for example, the temperature in a certain area rises slowly, accompanied by a sharp drop in humidity in the adjacent area) will produce cross-modal synergistic effect through heat conduction and evaporation effect. For example, high-temperature areas accelerate the evaporation of water on the surface of the patient, resulting in a decrease in local humidity, while a high-humidity environment may inhibit the heat dissipation efficiency and thus exacerbate the stimulation of temperature on blood vessels. The existing technology uses static or single-mode analysis, which cannot capture such cross-temporal, cross-modal nonlinear coupling mechanism, resulting in significant deviation in the evaluation of the blood pressure fluctuation risk. The traditional global interaction model is difficult to analyze the causal correlation network implied in the temperature and humidity field because it ignores the chain influence transmission between local areas. Based on this, the temperature-humidity spatiotemporal synergistic interaction encoding vector is obtained by performing temperature-humidity spatiotemporal interaction chain reasoning on the set of single-point humidity time sequence correlation feature encoding vectors and the set of single-point temperature time sequence correlation feature encoding vectors.

[0040] Figure 3 A block diagram of a temperature-humidity spatiotemporal interaction unit in a hemodialysis auxiliary nursing system according to an embodiment of the present application is shown. Specifically, as shown in Figure 3 The temperature-humidity spatiotemporal interaction unit 123 includes a temperature-humidity single-point time sequence feature interaction subunit 1231 configured to perform single-point time sequence feature interaction on each corresponding single-point humidity time sequence correlation feature encoding vector and single-point temperature time sequence correlation feature encoding vector in the set of single-point humidity time sequence correlation feature encoding vectors and the set of single-point temperature time sequence correlation feature encoding vectors to obtain a set of temperature-humidity local time sequence feature interaction encoding vectors; and a temperature-humidity single-point time sequence feature cross-modal synergistic subunit 1232 configured to perform cross-modal chain attention spatiotemporal synergistic coding on the set of temperature-humidity local time sequence feature interaction encoding vectors to obtain the temperature-humidity spatiotemporal synergistic interaction encoding vector.

[0041] Specifically, the temperature-humidity single-point time sequence feature interaction subunit 1231 is configured to perform single-point time sequence feature interaction on each corresponding single-point humidity time sequence correlation feature encoding vector and single-point temperature time sequence correlation feature encoding vector in the set of single-point humidity time sequence correlation feature encoding vectors and the set of single-point temperature time sequence correlation feature encoding vectors to obtain a set of temperature-humidity local time sequence feature interaction encoding vectors, which can be expressed by a formula as follows:

[0042] X = {x1, x2,..., x i ,...,x n}

[0043] Y = {y1, y2,..., y i ,...,y n}

[0044]

[0045] wherein, X is a set of single-point humidity time-series correlation feature encoding vectors, Y is a set of single-point temperature time-series correlation feature encoding vectors, x1, x2, x i and x n are the 1st, 2nd, i-th and n-th single-point humidity time-series correlation feature encoding vectors in the set of single-point humidity time-series correlation feature encoding vectors, y1, y2, y i and y n are the 1st, 2nd, i-th and n-th single-point temperature time-series correlation feature encoding vectors in the set of single-point temperature time-series correlation feature encoding vectors, n is the number of vectors in X and Y, and X and Y have the same length, is a point-by-point multiplication by position, is a point-by-point addition by position, is a point-by-point subtraction by position, concat{·; ·; ·} is a concatenation operation, W i is the i-th local time-series feature interaction weight matrix in the set of local time-series feature interaction weight matrices, b i is the i-th local time-series feature interaction bias vector in the set of local time-series feature interaction bias vectors, v i is the i-th temperature-humidity local time-series feature interaction encoding vector in the set of temperature-humidity local time-series feature interaction encoding vectors.

[0046] It should be understood that the temperature and humidity parameters of the microenvironment in which the patient is located have an inseparable correlation in the time dimension. The temperature rise at a certain sensing point in the dialysis area may accelerate the evaporation of body surface water, leading to a decrease in local humidity, and the humidity change may in turn affect the heat exchange efficiency, forming a dynamically coupled temperature and humidity action chain. If the temperature and humidity parameters are analyzed independently or simply superimposed, the non-linear synergistic mechanism of both in the same space-time unit is ignored (such as the superimposed effect of high temperature and low humidity, which may accelerate the rate of vasoconstriction). Especially in the middle and late stages of dialysis, the change in the patient's body fluid balance state causes the time sequence interaction of local temperature and humidity parameters (such as a gradual decrease in humidity after a sudden temperature rise) to have a combined effect on blood pressure fluctuations. Therefore, the application will encode the corresponding single-point humidity time sequence correlation feature vector and the single-point temperature time sequence correlation feature vector into a set of temperature-humidity local time sequence feature interaction encoding vectors that reflect the dynamic coupling relationship of temperature and humidity in the local space-time unit.

[0047] Specifically, in the embodiment of the application, the temperature-humidity single-point time sequence feature cross-modal synergistic subunit includes: a chain attention weight calculation secondary subunit for determining cross-modal interaction chain attention weights of each temperature-humidity local time sequence feature interaction encoding vector in the set of temperature-humidity local time sequence feature interaction encoding vectors based on the feature distribution characteristics of the temperature-humidity local time sequence feature interaction encoding vector to obtain a set of temperature-humidity local time sequence feature cross-modal interaction chain attention weights; a temperature-humidity interaction feature modulation secondary subunit for weighting and modulating the set of temperature-humidity local time sequence feature interaction encoding vectors based on the set of temperature-humidity local time sequence feature cross-modal interaction chain attention weights to obtain a set of temperature-humidity local time sequence feature interaction modulation encoding vectors; and a temperature-humidity interaction feature time sequence reasoning secondary subunit for inputting the set of temperature-humidity local time sequence feature interaction modulation encoding vectors into a time sequence reasoner based on a forward LSTM model to obtain the temperature-humidity space-time synergistic interaction encoding vector.

[0048] Specifically, the chain attention weight calculation secondary subunit is configured to determine cross-modal interaction chain attention weights of each temperature-humidity local time sequence feature interaction encoding vector in the set of temperature-humidity local time sequence feature interaction encoding vectors based on the feature distribution characteristics of the temperature-humidity local time sequence feature interaction encoding vector to obtain a set of temperature-humidity local time sequence feature cross-modal interaction chain attention weights, and the process is represented by a formula as follows:

[0049]

[0050] wherein v iis the i-th temperature-humidity local time-series feature cross-modal interaction chain attention weight in the set of temperature-humidity local time-series feature cross-modal interaction chain attention weights, v i,j is the i-th temperature-humidity local time-series feature cross-modal interaction chain attention weight in the set of temperature-humidity local time-series feature cross-modal interaction chain attention weights, v i is the j-th feature value in v 2 is the square of the Euclidean norm of the calculation vector, L is the number of feature values in v i is the number of feature values in v i is the i-th temperature-humidity local time-series feature cross-modal interaction chain attention weight in the set of temperature-humidity local time-series feature cross-modal interaction chain attention weights, v

[0051] Correspondingly, due to the significant difference in the influence of temperature and humidity interaction at different times and spaces on blood pressure fluctuations. For example, the coordinated change of temperature and humidity in the trunk of the patient may directly act on the cardiovascular system through the core body temperature regulation mechanism, while the temperature and humidity interaction in the foot area may have a lagging effect due to the difference in microcirculation. If equal weight is given to all local interactions, the dominant role of the interaction features of the key physiological sensitive area (such as the synergistic effect of the temperature of the carotid artery region and the sudden change of the adjacent humidity) in the overall blood pressure fluctuation is ignored. Based on this, in the technical solution of the present application, based on the feature distribution characteristics of each temperature-humidity local time-series feature interaction encoding vector, the cross-modal interaction chain attention weight of each temperature-humidity local time-series feature interaction encoding vector is determined to obtain a set of temperature-humidity local time-series feature cross-modal interaction chain attention weights, which can strengthen the attention to the interaction unit with strong physiological correlation (such as the dangerous signal of continuous high variance in the temperature and humidity interaction of the patient's body surface), and suppress non-critical areas or noise interference (such as stable fluctuations in the equipment heat dissipation area away from the patient).

[0052] More specifically, in the embodiment of the present application, the temperature-humidity interaction feature modulation secondary subunit is used for:

[0053] The set of temperature-humidity local time-series feature cross-modal interaction chain attention weights is subjected to weight correction based on interaction rule decomposition to obtain a set of temperature-humidity local time-series feature cross-modal interaction chain attention correction weights, and the process is represented by the formula:

[0054]

[0055] a′ i = (ω a × ξ i + ω b × ζ i )a i

[0056] where x iis the ith single-point humidity time-series correlation feature encoding vector in the set of single-point humidity time-series correlation feature encoding vectors, y i is the ith single-point temperature time-series correlation feature encoding vector in the set of single-point temperature time-series correlation feature encoding vectors, a i is the ith temperature-humidity local time-series feature cross-modal interaction chain attention weight in the set of temperature-humidity local time-series feature cross-modal interaction chain attention weights, a i is the ith temperature-humidity local time-series feature linear displacement strength in the set of temperature-humidity local time-series feature linear displacement strengths, b i is the ith temperature-humidity local time-series feature displacement difference strength in the set of temperature-humidity local time-series feature displacement difference strengths, g i is the ith temperature-humidity local time-series feature oscillation coupling strength in the set of temperature-humidity local time-series feature oscillation coupling strengths, ln is the logarithmic function value with the natural constant e as the base, x i is the ith temperature-humidity local time-series feature period dimension dynamic correction factor in the set of temperature-humidity local time-series feature period dimension dynamic correction factors, z i is the ith temperature-humidity local time-series feature period phase auxiliary modulation term in the set of temperature-humidity local time-series feature period phase auxiliary modulation terms, w a and w b are the corresponding weight coefficients, a′ i and z′ i are the corresponding weight coefficients, b′ i is the ith temperature-humidity local time-series feature cross-modal interaction chain attention correction weight in the set of temperature-humidity local time-series feature cross-modal interaction chain attention correction weights;

[0057] Based on the set of temperature-humidity local time-series feature cross-modal interaction chain attention correction weights, the set of temperature-humidity local time-series feature interaction encoding vectors is weighted and modulated to obtain the set of temperature-humidity local time-series feature interaction modulation encoding vectors, and this process is represented by the formula:

[0058] I i = a′ i · v i

[0059] I = {I1, I2,..., I i ,..., I n}

[0060] where v i is the ith temperature-humidity local time-series feature interaction encoding vector in the set of temperature-humidity local time-series feature interaction encoding vectors, a′ iis the ith temperature-humidity local time-series feature cross-modal interaction chain attention revision weight in the set of temperature-humidity local time-series feature cross-modal interaction chain attention revision weights, I1, I2, I i and I n are the 1st, 2nd, ith and nth temperature-humidity local time-series feature interaction modulation encoding vectors in the set of temperature-humidity local time-series feature interaction modulation encoding vectors, respectively, I is the set of temperature-humidity local time-series feature interaction modulation encoding vectors.

[0061] In particular, when calculating each temperature-humidity local time-series feature interaction encoding vector, the interaction feature between the corresponding single-point humidity time-series correlation feature encoding vector and the single-point temperature time-series correlation feature encoding vector needs to be introduced, such as x i ⊙y i , and the like. These operations essentially correspond to different interaction rules, thereby forming differentiated regional constraint coupling in the interaction space.

[0062] To enhance the influence of rule dynamic gain on temperature-humidity local time-series feature cross-modal interaction chain attention weights, the weights need to be revised based on the spatial effectiveness of the interaction rules. Specifically, x and can be regarded as linear displacement effects (i.e., the gradient direction is consistent with the spatial interaction direction), while x i ⊙y i corresponds to an oscillatory coupling response (the gradient direction is orthogonal to the spatial interaction direction). By analyzing the statistics of different action vectors (such as γ i = ||x i ⊙y i || 2 ), it can be found that the oscillatory coupling response in the direction of linear displacement effect will induce regional periodic dynamic characteristics, and the corresponding temperature-humidity local time-series feature periodic dimension dynamic revision factor can be represented as:

[0063]

[0064] where α i + β i as a linear local representation, indicates that the strength of the oscillatory coupling response increases with the expansion of the linear displacement effect on the logarithmic scale.

[0065] At the same time, the oscillatory coupling response will also cause a phase shift of the linear displacement effect, thereby generating a temperature-humidity local time-series feature periodic phase auxiliary modulation term:

[0066] Finally, the original temperature-humidity local time sequence feature cross-modal interaction chain attention weight a is corrected by the weighted sum of the above two i : a′ i = (ω a × ξ i + ω b × ζ i )a i

[0067] The method significantly improves the modeling accuracy of the region constraint coupling correlation in the interaction space by distinguishing the action mechanism of different interaction rules in the spatial constraint rule framework, thereby optimizing the calculation robustness of the temperature-humidity local time sequence feature cross-modal interaction chain attention weight.

[0068] Then, based on the set of temperature-humidity local time sequence feature cross-modal interaction chain attention correction weights, the set of temperature-humidity local time sequence feature interaction encoding vectors is weighted and modulated to obtain the set of temperature-humidity local time sequence feature interaction modulation encoding vectors. For each spatio-temporal interaction encoding vector, the feature amplitude is scaled according to its attention correction weight value: high weight vectors (such as the interaction mode of sustained high temperature and slow humidity decrease in the trunk region) amplify their contribution through feature channels, and low weight vectors (such as stable temperature and humidity interaction in the device remote area) reduce their interference through feature suppression. During the modulation process, the attention correction weight and the channel dimension of the interaction encoding vector are multiplied element by element, retaining the spatial distribution pattern of the key interaction features (such as the abnormal waveform of sudden temperature rise accompanied by sudden humidity drop in the patient's back), while filtering out noise patterns with weak blood pressure fluctuation correlation (such as periodic temperature and humidity oscillation around the dialysis chair fixed support). The modulated feature set not only maintains the spatio-temporal topology, but also highlights the interaction hotspots with high physiological correlation.

[0069] Specifically, the temperature-humidity interaction feature time sequence reasoning secondary subunit is used to input the set of temperature-humidity local time sequence feature interaction modulation encoding vectors into a time sequence reasoner based on a forward LSTM model to obtain the temperature-humidity spatio-temporal collaborative interaction encoding vector, which is represented by the formula as follows:

[0070]

[0071] where I is the set of temperature-humidity local time sequence feature interaction modulation encoding vectors, is the forward LSTM encoding, v f is the temperature-humidity spatio-temporal collaborative interaction encoding vector.

[0072] Finally, the set of temperature-humidity local temporal feature interaction modulation encoding vectors is input into a forward LSTM model-based temporal reasoner to obtain the temperature-humidity spatiotemporal synergistic interaction encoding vector. That is, the memory unit of the LSTM filters the newly added features (such as the abnormal pulse signal of the temperature-humidity interaction in the back of the hand region) through the input gate, eliminates historical irrelevant information (such as the stabilized early interaction features in the torso) through the forget gate, and controls the contribution of the current state to the global reasoning through the output gate when processing the modulation encoding vectors one by one. As the sequence progresses, the model gradually accumulates spatiotemporal memories of cross-region interactions (such as the conduction path of foot temperature rise → lower leg humidity response → thigh blood vessel contraction), and finally fuses to form a high-order encoding vector in the hidden state that represents the synergistic effect of temperature and humidity during the entire treatment period.

[0073] Specifically, the temperature-humidity spatiotemporal interaction feature decoding unit 124 is configured to decode the temperature-humidity spatiotemporal synergistic interaction encoding vector to obtain the blood pressure influence parameter. Specifically, in the embodiments of the present application, the temperature-humidity spatiotemporal interaction feature decoding unit is configured to: use a decoder-based blood pressure influence analyzer to decode the temperature-humidity spatiotemporal synergistic interaction encoding vector to obtain the blood pressure influence parameter. That is, the temperature-humidity spatiotemporal synergistic interaction encoding vector is obtained through complex spatiotemporal interaction chain reasoning analysis. Its form and features are designed to facilitate the model to capture and process spatiotemporal information of temperature and humidity, but it is not intuitive to directly represent the blood pressure influence parameter. It is necessary to convert it into a parameter form that can directly reflect the blood pressure influence through the decoder, so as to realize the mapping from the complex encoding space to the blood pressure influence parameter space with practical significance. Specifically, the decoder-based blood pressure influence analyzer can deeply mine the hidden blood pressure-related information in the temperature-humidity spatiotemporal synergistic interaction encoding vector. The decoder can learn the potential relationship between different features in the encoding vector and the blood pressure influence, and present these potential relationships in the form of blood pressure influence parameters through the decoding process, thereby revealing the specific influence mechanism of the temperature and humidity in the dialysis environment on the blood pressure of the patient.

[0074] The decoder-based blood pressure influence analyzer is used to decode the encoding vector. This blood pressure influence analyzer is specially trained, and its training data comes from a large amount of historical hemodialysis data. During the training process, the analyzer learns the internal relationship between the temperature-humidity spatiotemporal synergistic interaction encoding vector and the blood pressure influence parameter. When the current temperature-humidity spatiotemporal synergistic interaction encoding vector is input, the analyzer will decode the vector according to the patterns and rules learned before. It will extract the feature information related to the blood pressure influence from the vector, and through the internal calculation logic, it will convert these features into specific blood pressure influence parameters.

[0075] For example, the analyzer may analyze the trend of temperature change at different times and locations in the encoding vector, as well as the interaction relationship between humidity and it, combined with the experience in the historical data, to judge the influence degree of the current environment on the patient's blood pressure, and finally output a quantified blood pressure influence parameter. This parameter can intuitively reflect the potential influence of temperature and humidity factors in the current dialysis environment on the patient's blood pressure, providing an important basis for subsequent evaluation of dialysis impact and development of nursing plan.

[0076] In the embodiment of the present application, the dialysis impact parameter calculation module 130 is configured to obtain a dialysis impact parameter based on the blood pressure influence parameter. It should be understood that although the blood pressure influence parameter can reflect the effect of temperature and humidity environment on the blood pressure of dialysis patients, the dialysis process is a complex physiological process affected by multiple factors, and blood pressure is only one of the key aspects. Focusing only on blood pressure is not enough to fully measure the effectiveness and safety of the entire dialysis process. By further obtaining a dialysis impact parameter based on the blood pressure influence parameter, the role of blood pressure factors in the entire dialysis process can be considered comprehensively, and other potential influencing factors can be combined to more comprehensively and accurately evaluate the overall impact of environmental factors on the dialysis effect.

[0077] Specifically, first, a large amount of historical hemodialysis data is screened and collected to obtain a sample humidity information array set and a sample temperature information array set. These data cover environmental temperature and humidity conditions in many different dialysis scenarios, providing a rich variety of samples for subsequent analysis. At the same time, combined with the corresponding blood pressure test data, the sample blood pressure influence parameter set is carefully labeled according to the blood pressure change parameters therein. This set records detailed quantitative data on the impact of blood pressure on dialysis patients in different temperature and humidity environments.

[0078] Next, using these collected and labeled data, machine learning algorithms such as neural networks or random forests are used to train the environmental blood pressure influence analysis channel. During the training process, the algorithm continuously learns the complex internal relationship between humidity, temperature and blood pressure change. After repeated iteration and optimization of a large amount of data, the environmental blood pressure influence analysis channel has the ability to accurately predict the potential impact of environmental factors on blood pressure. Then, the humidity information array and temperature information array collected in real time in the current dialysis environment are input into the trained environmental blood pressure influence analysis channel, and the blood pressure influence parameters under the current environmental conditions are obtained through the operation and analysis of the channel. These parameters quantitatively describe the potential influence of the current environment on the patient's blood pressure.

[0079] To further obtain the dialysis influence parameters, it is also necessary to collect sample dialysis influence parameter sets again from historical hemodialysis data. These parameter sets are obtained by in-depth analysis of historical dialysis data and contain key indicators such as dialysis clearance rate, electrolyte balance, and other indicators reflecting the specific influence of different dialysis environmental conditions on dialysis effect. Then, the sample blood pressure influence parameter set obtained before is combined with the collected sample dialysis influence parameter set to construct a dialysis influence classification channel. This channel is like an intelligent classifier that can accurately classify different blood pressure influence parameters into corresponding dialysis effect categories, helping medical staff clearly identify which dialysis influence parameters are significantly affected by environmental factors under different environmental conditions.

[0080] Finally, the current blood pressure influence parameters obtained through the environmental blood pressure influence analysis channel are input into the dialysis influence classification channel. The dialysis influence classification channel classifies the input blood pressure influence parameters based on the previously learned rules and models, and finally outputs the dialysis influence parameters. These dialysis influence parameters accurately evaluate the specific influence of the current environmental conditions on the dialysis effect, providing a scientific and reliable basis for subsequent development and optimization of auxiliary nursing plans, helping medical staff adjust nursing strategies according to actual conditions and improve the treatment effect of hemodialysis and the comfort of patients.

[0081] In the embodiments of the present application, the comfort parameter calculation module 140 is configured to perform comfort influence analysis on the humidity information array and the temperature information array based on the comfort of the dialysis patient to obtain comfort perception influence parameters. Accordingly, since traditional hemodialysis nursing often ignores the influence of environmental factors on patient comfort, patient comfort during dialysis is crucial. Humidity and temperature are important environmental factors that affect comfort, and different patients have different sensitivities to changes in temperature and humidity, and changes in temperature and humidity will bring different comfort experiences to patients. To make up for the shortcomings of traditional nursing and fully consider individual differences, it is necessary to analyze the influence of temperature and humidity on comfort to obtain comfort perception influence parameters, thereby deeply understanding the relationship between environmental factors and comfort.

[0082] To implement this process, the first step is to collect a large number of patient data records in the hemodialysis environment, which covers detailed environmental data of different dialysis periods and different patients. From these records, a sample humidity information array set and a sample temperature information array set are selected to comprehensively reflect the diverse environmental temperature and humidity conditions during the dialysis process. At the same time, comfort inquiry data of patients in the same period are collected, which are the intuitive and subjective feedback of patients on the comfort of the current dialysis environment. The environmental data and comfort inquiry data are matched one by one, and a sample comfort perception influence parameter set is labeled according to specific labeling rules, so that the data set contains both environmental feature information and corresponding comfort influence information, providing solid data support for subsequent analysis.

[0083] After the data collection and labeling are completed, a machine learning algorithm such as a support vector machine, decision tree, or deep learning model is used to train the perception comfort influence analysis channel. During the training phase, the algorithm deeply mines and learns the sample humidity information array set, sample temperature information array set, and sample comfort perception influence parameter set, and analyzes the complex relationship between humidity, temperature, and patient comfort. Through repeated training and verification of a large amount of sample data, the model parameters are continuously optimized, the prediction accuracy and generalization ability of the perception comfort influence analysis channel are improved, and it is ensured that the channel can accurately identify the potential influence of temperature and humidity on patient comfort under different environmental conditions and adapt to various dialysis environments and individual differences of patients.

[0084] When the perception comfort influence analysis channel is trained and reaches the expected performance, it enters the actual application stage. The real-time collected humidity information array and temperature information array are input into the trained perception comfort influence analysis channel. The channel automatically operates and analyzes the input data according to the previously learned relationship model between environmental temperature and humidity and comfort. Through a series of complex calculation processes, comfort perception influence parameters corresponding to the current environmental temperature and humidity are output. These parameters accurately describe the specific influence of the current dialysis environment on patient comfort in a quantitative form, for example, the higher the value, the greater the positive influence of the environment on patient comfort, and vice versa.

[0085] In the embodiments of the present application, the proportion calculation module 150 is configured to obtain the perception image array of the dialysis patient and perform environmental comfort analysis to obtain the environmental comfort influence proportion. It should be understood that the comfort analysis based on only humidity and temperature information has limitations and cannot fully reflect the real comfort status of the patient. The comfort of the patient is influenced by multiple factors, including not only environmental temperature and humidity but also the patient's psychological state, health status, etc. The perception image can capture the patient's facial expression, body posture, skin state, and other external features, which can intuitively reflect the patient's comfort. Combining the comfort perception influence parameters predicted based on environmental temperature and humidity with the actual comfort influence parameters obtained based on image analysis can more comprehensively and accurately evaluate the proportion of environmental factors in the overall comfort of the patient, making up for the shortcomings of a single analysis method.

[0086] The specific implementation process is as follows: In the blood dialysis area, multiple perception points will be deployed with cameras or image sensors and other devices to collect the perception image array of the dialysis patient. These perception points are carefully planned and distributed in positions that can fully capture the patient's state, ensuring that the perception image array obtained can cover the patient's facial expression, body posture, skin state, blood vessel color, and other visual features during the dialysis process. These real-time image data provide intuitive external visual clues for subsequent analysis of the patient's comfort.

[0087] Next, the patient's comfort level analysis data is extracted from historical data, and based on these data, a set of sample perception images is collected. This set covers the image state of the patient at different comfort levels, with rich diversity. For each sample perception image, detailed labeling is performed in combination with the actual patient comfort situation corresponding to it, so as to obtain a set of sample actual comfort influence parameters. This step closely links the patient's subjective comfort feeling with the objective performance in the image, providing a key data basis for subsequent model training.

[0088] After that, the sample perception image set and the sample actual comfort influence parameter set are used to train the actual comfort influence analysis channel with the help of a convolutional neural network (CNN). As a powerful deep learning model, CNN has unique advantages in image recognition and analysis. Through repeated training on a large number of sample data, CNN can automatically extract and learn the key features in the image, gradually mastering the complex mapping relationship between image features and actual comfort influence factors. After sufficient training and optimization, the actual comfort influence analysis channel has the ability to accurately identify the actual comfort influence factors of the patient.

[0089] After the model training is completed, the trained actual comfort influence analysis channel is used to recognize and analyze the current perception image array collected. The channel processes each of the multiple perception images in the perception image array to generate multiple actual comfort influence analysis results. Based on these analysis results, various factors are considered through specific calculation methods, and finally the actual comfort influence parameter of the current dialysis patient is calculated. This parameter is an objective evaluation of the patient's real comfort level in the current dialysis environment, reflecting the actual level of the patient's physiological and psychological comfort.

[0090] At the same time, the comfort perception influence parameter has been obtained by previously performing comfort influence analysis on the humidity information array and the temperature information array. At this time, by calculating the ratio of the comfort perception influence parameter to the actual comfort influence parameter, the environmental comfort influence ratio can be obtained. This ratio is a key indicator for measuring the degree of influence of the dialysis environment on the patient's comfort. When the ratio is small, it means that the patient's comfort is relatively less affected by environmental factors, and more likely to be determined by psychological state or health condition and other factors; on the contrary, if the ratio is large, it indicates that environmental factors are the main reason affecting the patient's comfort, and in the subsequent optimization of the nursing plan, it is necessary to focus on adjusting the dialysis environment, such as adjusting the temperature and humidity, to improve the patient's dialysis comfort and optimize the entire nursing process.

[0091] In the embodiment of the present application, the scheme generation module 160 is configured to obtain an optimal auxiliary nursing scheme based on the dialysis impact parameter, the comfort perception impact parameter and the environmental comfort impact proportion. Accordingly, in the hemodialysis nursing, a single parameter cannot comprehensively reflect the patient's demand and the dialysis condition. The dialysis impact parameter reflects the effect of the environment on the dialysis effect, the comfort perception impact parameter reflects the influence of temperature and humidity on the subjective feeling of the patient, and the environmental comfort impact proportion measures the proportion of environmental factors in the overall comfort. By comprehensively considering these parameters, comprehensive information can be obtained from the dialysis effect, patient comfort and environmental impact, which provides a basis for formulating a more accurate and effective nursing scheme and avoids the one-sidedness of decision-making based on a single factor.

[0092] The specific implementation process is as follows: first, an auxiliary nursing function is constructed based on the above three key parameters. The function expression is as follows:

[0093]

[0094] wherein ANF represents the nursing adaptability, which is used to measure the pros and cons of the nursing scheme; ω1 and ω2 are the dialysis weight and the comfort weight respectively, and the sum of the two is 1. Their role is to balance the importance of dialysis effect and patient comfort in the nursing scheme; AFF n is the dialysis impact parameter of the environmental characteristic information array after nursing according to the auxiliary nursing scheme, AFF b is the current dialysis impact parameter; K s is the environmental comfort impact proportion; CFF n is the comfort perception impact parameter of the environmental characteristic information array after nursing according to the auxiliary nursing scheme, CFF b is the current comfort perception impact parameter. This function combines the three parameters organically, takes the maximization of the nursing adaptability as the goal, and provides a quantitative basis for subsequent optimization of the nursing scheme. In particular, after constructing the auxiliary nursing scheme space and generating a specific scheme, the blood pressure impact analysis and classification of the dialysis patients are carried out based on the temperature and humidity adjustment settings in the scheme, and finally the parameter is obtained. Taking the first auxiliary nursing scheme generated as an example, the scheme includes a first temperature adjustment scheme and a first humidity adjustment scheme. Based on this, the relevant analysis of the dialysis patients is carried out, and the parameter used to measure the dialysis effect is "the dialysis impact parameter of the environmental characteristic information array after nursing according to the auxiliary nursing scheme" (which is the first dialysis impact parameter at this time). It is a quantitative index, which can intuitively show the influence of the environment on the dialysis under the current nursing scheme.

[0095] Next, the temperature adjustment range and the humidity adjustment range of the hemodialysis environment are obtained, and a secondary care scheme space is constructed based on the two ranges. The space contains all possible temperature and humidity adjustment combinations, which constitute the set of care schemes available for selection in the system optimization process. Within the secondary care scheme space, a first secondary care scheme is randomly generated. This scheme includes a first temperature adjustment scheme and a first humidity adjustment scheme, according to which a blood pressure impact analysis and classification of the dialysis patient is performed, obtaining a first dialysis impact parameter, while a comfort impact analysis is also performed, obtaining a first comfort perception impact parameter. Then, the first dialysis impact parameter, the first comfort perception impact parameter, and the known environmental comfort impact proportion are substituted into the secondary care function to calculate the first care fitness, which reflects the comprehensive performance of this scheme in balancing dialysis effectiveness and patient comfort.

[0096] Subsequently, the first temperature adjustment scheme and the first humidity adjustment scheme are adjusted appropriately to generate a second secondary care scheme. The above analysis and calculation process is repeated, i.e., a blood pressure impact analysis and a comfort impact analysis are performed on the second secondary care scheme to obtain the corresponding parameters, which are then substituted into the secondary care function to calculate the second care fitness. Then, the first care fitness and the second care fitness are compared. If the second care fitness is greater than the first care fitness, the second secondary care scheme is directly retained; if the second care fitness is not greater than the first care fitness, the ratio of the two is calculated, and combined with the environmental comfort impact proportion, the retention probability is calculated. For example, the ratio is multiplied by the environmental comfort impact proportion to obtain the retention probability, and then a random number between 0 and 1 is generated and compared with the retention probability to determine whether to retain the first dialysis impact parameter and the corresponding first secondary care scheme, or the second dialysis impact parameter and the second secondary care scheme. This approach not only considers the fitness difference between schemes, but also takes into account the importance of environmental comfort to patient experience, improving the scientificity and globality of the optimization process.

[0097] Finally, based on the retained secondary care scheme, within the secondary care scheme space, the secondary care function is used to continue adjustment and iterative optimization. The temperature adjustment scheme and the humidity adjustment scheme are constantly changed, and the care fitness is recalculated. This process continues until the care fitness no longer improves, reaching a state of convergence. At this time, the scheme with the highest care fitness output is the optimal secondary care scheme, which contains the optimal temperature adjustment scheme and the optimal humidity adjustment scheme in the dialysis environment, and can provide the best secondary care service for patients, effectively improving the dialysis effectiveness and patient comfort. In particular, since this step is not the focus of the present application, it is not specifically described, please refer to the existing patent CN118762796A for details.

[0098] In summary, the blood dialysis auxiliary nursing system 100 based on the embodiments of the present application is illustrated as follows: firstly, the temperature and humidity information acquisition module uses the sensing device to obtain the temperature and humidity data in the dialysis area. The blood pressure influence parameter module analyzes the influence of these data on the patient's blood pressure to obtain the blood pressure influence parameter, and then the dialysis influence parameter calculation module obtains the dialysis influence parameter. At the same time, the comfort parameter calculation module evaluates the influence of these environmental factors on the patient's comfort to obtain the comfort perception influence parameter. The proportion calculation module determines the environmental comfort influence proportion by analyzing the patient's perception image array. Finally, the scheme generation module synthesizes all the parameters to develop the best auxiliary nursing scheme. This scheme realizes the individualized and accurate nursing scheme optimization, and improves the stability of the dialysis effect and the patient's comfort.

[0099] As described above, the blood dialysis auxiliary nursing system 100 according to the embodiments of the present application can be implemented in various wireless terminals, such as a server with a blood dialysis auxiliary nursing algorithm, etc. In one possible implementation, the blood dialysis auxiliary nursing system 100 according to the embodiments of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the blood dialysis auxiliary nursing system 100 can be a software module in the operating system of the wireless terminal, or can be an application program developed for the wireless terminal; of course, the blood dialysis auxiliary nursing system 100 can also be one of the many hardware modules of the wireless terminal.

[0100] Alternatively, in another example, the blood dialysis auxiliary nursing system 100 and the wireless terminal can also be separate devices, and the blood dialysis auxiliary nursing system 100 can be connected to the wireless terminal through a wired and / or wireless network, and transmit interactive information in a conventional data format.

[0101] Figure 4 The flowchart of the blood dialysis auxiliary nursing method according to the embodiments of the present application is shown in FIG. 6. As shown in FIG. 6, the blood dialysis auxiliary nursing method according to the embodiments of the present application includes the following steps: Figure 4As shown, the hemodialysis auxiliary nursing method according to the embodiment of the present application comprises: S110, collecting a humidity information array and a temperature information array in a hemodialysis area by a sensing device; S120, performing blood pressure influence analysis based on the humidity information array and the temperature information array to obtain a blood pressure influence parameter, comprising: performing temperature-humidity space-time interaction chain reasoning analysis based on time domain features on the humidity information array and the temperature information array to obtain the blood pressure influence parameter; S130, obtaining a dialysis influence parameter based on the blood pressure influence parameter; S140, performing comfort influence analysis based on a dialysis patient on the humidity information array and the temperature information array to obtain a comfort perception influence parameter; S150, obtaining a sensing image array of the dialysis patient, and performing environmental comfort analysis to obtain an environmental comfort influence proportion; S160, obtaining an optimal auxiliary nursing scheme based on the dialysis influence parameter, the comfort perception influence parameter and the environmental comfort influence proportion.

[0102] Here, those skilled in the art can understand that the specific operations of each step in the above hemodialysis auxiliary nursing method have been described in detail above with reference to the description of the hemodialysis auxiliary nursing system of the above Figures 1 to 3 , and therefore, the repeated description thereof will be omitted.

[0103] The above detailed description further describes the purpose, technical solutions and beneficial effects of the present application, and it should be understood that the above is only a specific embodiment of the present application and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A hemodialysis ancillary care system, characterized by, The method comprises the following steps: A temperature and humidity information acquisition module is configured to acquire a humidity information array and a temperature information array in a hemodialysis area through a sensing device; A blood pressure influence parameter module is configured to perform blood pressure influence analysis on the humidity information array and the temperature information array based on a dialysis patient to obtain a blood pressure influence parameter, wherein the blood pressure influence parameter module comprises: a humidity information time series encoding unit configured to perform time series analysis on each humidity row vector in the humidity information array to obtain a set of single-point humidity time series correlation feature encoding vectors; a temperature information time series encoding unit configured to perform time series analysis on each temperature row vector in the temperature information array to obtain a set of single-point temperature time series correlation feature encoding vectors; a temperature-humidity spatio-temporal interaction unit configured to perform temperature-humidity spatio-temporal interaction chain reasoning on the set of single-point humidity time series correlation feature encoding vectors and the set of single-point temperature time series correlation feature encoding vectors to obtain a temperature-humidity spatio-temporal collaborative interaction encoding vector; and a temperature-humidity spatio-temporal interaction feature decoding unit configured to perform feature decoding on the temperature-humidity spatio-temporal collaborative interaction encoding vector to obtain the blood pressure influence parameter; A dialysis influence parameter calculation module is configured to obtain a dialysis influence parameter based on the blood pressure influence parameter; A comfort parameter calculation module is configured to perform comfort influence analysis on the humidity information array and the temperature information array based on a dialysis patient to obtain a comfort perception influence parameter; A proportion calculation module is configured to obtain a perception image array of the dialysis patient and perform environmental comfort analysis to obtain an environmental comfort influence proportion; A scheme generation module is configured to obtain an optimal auxiliary nursing scheme based on the dialysis influence parameter, the comfort perception influence parameter, and the environmental comfort influence proportion. The temperature-humidity spatio-temporal interaction unit comprises: a temperature and humidity single-point time series feature interaction subunit configured to perform single-point time series feature interaction on each corresponding single-point humidity time series correlation feature encoding vector and single-point temperature time series correlation feature encoding vector in the set of single-point humidity time series correlation feature encoding vectors and the set of single-point temperature time series correlation feature encoding vectors to obtain a set of temperature-humidity local time series feature interaction encoding vectors; and a temperature and humidity single-point time series feature cross-modal collaborative subunit configured to perform cross-modal chain attention spatio-temporal collaborative coding on the set of temperature-humidity local time series feature interaction encoding vectors to obtain the temperature-humidity spatio-temporal collaborative interaction encoding vector.

2. The hemodialysis secondary nursing system according to claim 1, characterized by The humidity information time series encoding unit is configured to perform bidirectional recurrent network-based time series analysis on each humidity row vector in the humidity information array to obtain the set of single-point humidity time series correlation feature encoding vectors.

3. The hemodialysis ancillary care system according to claim 2, characterized in that, The temperature information time series encoding unit is configured to perform the bidirectional recurrent network-based time series analysis on each temperature row vector in the temperature information array to obtain the set of single-point temperature time series correlation feature encoding vectors.

4. The hemodialysis ancillary nursing system according to claim 3, characterized by The temperature and humidity single-point time series feature cross-modal collaborative subunit comprises: a chain attention weight calculation secondary subunit configured to determine, based on feature distribution characteristics of each temperature-humidity local time sequence feature interaction encoding vector in the set of temperature-humidity local time sequence feature interaction encoding vectors, a cross-modal interaction chain attention weight of the temperature-humidity local time sequence feature interaction encoding vector to obtain a set of temperature-humidity local time sequence feature cross-modal interaction chain attention weights; a temperature-humidity interaction feature modulation secondary subunit configured to perform weighted modulation on the set of temperature-humidity local time sequence feature interaction encoding vectors based on the set of temperature-humidity local time sequence feature cross-modal interaction chain attention weights to obtain a set of temperature-humidity local time sequence feature interaction modulation encoding vectors; a temperature-humidity interaction feature time sequence inference secondary subunit configured to input the set of temperature-humidity local time sequence feature interaction modulation encoding vectors into a time sequence inference device based on a forward LSTM model to obtain the temperature-humidity spatio-temporal collaborative interaction encoding vector.

5. The hemodialysis ancillary care system according to claim 4, wherein The temperature-humidity interaction feature modulation secondary subunit is configured to: perform weight correction on the set of temperature-humidity local time sequence feature cross-modal interaction chain attention weights based on an interaction rule to obtain a set of temperature-humidity local time sequence feature cross-modal interaction chain attention correction weights; perform weighted modulation on the set of temperature-humidity local time sequence feature interaction encoding vectors based on the set of temperature-humidity local time sequence feature cross-modal interaction chain attention correction weights to obtain the set of temperature-humidity local time sequence feature interaction modulation encoding vectors.

6. The hemodialysis ancillary care system according to claim 5, wherein, The temperature-humidity spatio-temporal interaction feature decoding unit is configured to: perform feature decoding on the temperature-humidity spatio-temporal collaborative interaction encoding vector using a decoder-based blood pressure influence analyzer to obtain the blood pressure influence parameter.

7. A method of hemodialysis ancillary care, characterized by, The method comprises: collecting, by a perception device, a humidity information array and a temperature information array in a hemodialysis area; performing blood pressure influence analysis on the humidity information array and the temperature information array based on a dialysis patient to obtain a blood pressure influence parameter, comprising: performing time sequence analysis on each humidity row vector in the humidity information array to obtain a set of single-point humidity time sequence correlation feature encoding vectors; performing time sequence analysis on each temperature row vector in the temperature information array to obtain a set of single-point temperature time sequence correlation feature encoding vectors; performing temperature-humidity spatio-temporal interaction chain inference on the set of single-point humidity time sequence correlation feature encoding vectors and the set of single-point temperature time sequence correlation feature encoding vectors to obtain a temperature-humidity spatio-temporal collaborative interaction encoding vector; and performing feature decoding on the temperature-humidity spatio-temporal collaborative interaction encoding vector to obtain the blood pressure influence parameter; obtaining a dialysis influence parameter based on the blood pressure influence parameter; performing comfort influence analysis on the humidity information array and the temperature information array based on a dialysis patient to obtain a comfort perception influence parameter; obtaining a perception image array of the dialysis patient and performing environmental comfort analysis thereon to obtain an environmental comfort influence proportion; Based on the dialysis influence parameter, the comfort perception influence parameter and the environmental comfort influence proportion, an optimal auxiliary nursing scheme is obtained; The temperature-humidity spatio-temporal interaction chain reasoning on the set of single-point humidity time series correlation feature encoding vectors and the set of single-point temperature time series correlation feature encoding vectors to obtain a temperature-humidity spatio-temporal cooperative interaction encoding vector includes: performing single-point time series feature interaction on each corresponding single-point humidity time series correlation feature encoding vector and single-point temperature time series correlation feature encoding vector in the set of single-point humidity time series correlation feature encoding vectors and the set of single-point temperature time series correlation feature encoding vectors, respectively, to obtain a set of temperature-humidity local time series feature interaction encoding vectors; and performing cross-modal chain attention spatio-temporal cooperative coding on the set of temperature-humidity local time series feature interaction encoding vectors to obtain the temperature-humidity spatio-temporal cooperative interaction encoding vector.

Images