Real-time monitoring system and method for bladder filling degree based on near-infrared sensor

Abstract

The application discloses a bladder filling degree real-time monitoring system and method based on a near-infrared sensor, and relates to the field of bladder filling degree monitoring. The system simultaneously detects an abdominal region by using a group of measuring light paths and a group of adjacent reference light paths. The measuring light paths pass through a bladder, and signal changes of the measuring light paths contain bladder filling information and physiological noise of shallow tissues; and the reference light paths only pass through the shallow tissues, and signals of the reference light paths mainly reflect the physiological noise. Through optical density conversion of the double-wavelength light signals collected by the two light paths and real-time differential processing in space, the differential operation is equivalent to dynamically calibrating the measuring channel by using the signals of the reference channel, so that global physiological fluctuations irrelevant to the bladder filling are separated from the total signals. Finally, a bladder water content index model is established based on the pure signals after the differential operation, and state estimation is carried out, so that accurate and reliable monitoring of the bladder filling degree is realized, and the false alarm rate is significantly reduced.

Description

Technical Field

[0001] This application relates to the field of bladder fullness monitoring, and more specifically, to a real-time bladder fullness monitoring system and method based on a near-infrared sensor. Background Technology

[0002] Real-time monitoring of bladder fullness has significant clinical and lifestyle value for specific populations, such as patients with spinal cord injuries, patients with neurogenic bladder dysfunction, and children and the elderly requiring bladder training. Traditional bladder capacity monitoring methods, such as ultrasound scanning or catheter measurement, while accurate, suffer from problems such as complex operation, high cost, invasiveness, or the need for professional personnel, making continuous, non-invasive, home-based real-time monitoring difficult. Therefore, developing a portable, non-invasive, wearable real-time bladder fullness monitoring solution is urgently needed to improve patients' quality of life and prevent complications such as urinary incontinence or urinary retention. Near-infrared spectroscopy (NIRS) technology, due to its good penetration depth into biological tissues and sensitivity to water molecules, offers a highly promising technological approach to achieving this goal.

[0003] In existing technological explorations, the basic idea behind using near-infrared sensors to monitor bladder fullness is to infer changes in bladder volume (mainly water) by measuring the attenuation of near-infrared light passing through abdominal tissues (including the bladder). However, a core technical challenge arises when extending this technology from the laboratory to everyday applications: how to effectively distinguish between the true signal changes caused by bladder fullness and baseline drift caused by other physiological factors. In real-world scenarios, a user's drinking, sweating, posture changes, and even emotional fluctuations can cause changes in body fluids and blood flow, both systemically and locally, which also lead to alterations in near-infrared light absorption. For example, after drinking water, the water content of skin and fat tissues increases slightly, resulting in a slow increase in light absorption. This interference signal is highly similar in amplitude and rate of change to the target signal of slow bladder fullness, making traditional single-point, dual-wavelength measurement methods prone to misjudgment and difficult to accurately separate the target signal from complex physiological background noise. The fundamental reason is that near-infrared light passes through a complex multi-layered tissue structure, and the measured light attenuation is the sum of the entire optical path. There is a lack of an effective reference benchmark to measure and subtract this non-bladder-related global physiological drift in real time.

[0004] To address the problem of physiological baseline drift and the difficulty in separating the target signal, a new solution is needed. Summary of the Invention

[0005] To address the aforementioned problems, according to one aspect of this application, a method for real-time monitoring of bladder fullness based on a near-infrared sensor is provided, comprising: S1: acquiring a first wavelength light intensity signal of a first measurement channel and a second wavelength light intensity signal of a second measurement channel collected by a measurement optical path, and a first wavelength light intensity signal of a first reference channel and a second wavelength light intensity signal of a second reference channel collected by a reference optical path, wherein the reference optical path is adjacent to the measurement optical path.

[0006] S2: Perform light intensity-to-light density conversion on the first wavelength light intensity signal of the first measurement channel, the second wavelength light intensity signal of the second measurement channel, the first wavelength light intensity signal of the first reference channel, and the second wavelength light intensity signal of the second reference channel to obtain light density change data of the first wavelength light intensity signal of the first measurement channel, light density change data of the second wavelength light intensity signal of the second measurement channel, light density change data of the first wavelength light intensity signal of the first reference channel, and light density change data of the second wavelength light intensity signal of the second reference channel.

[0007] S3: Spatial difference is performed on the first wavelength optical density change data of the first measurement channel, the first wavelength optical density change data of the first reference channel, the second wavelength optical density change data of the second measurement channel, and the second wavelength optical density change data of the second reference channel to obtain the first wavelength optical density spatial difference data and the second wavelength optical density spatial difference data.

[0008] S4: Model the bladder water content index based on dual-wavelength information using the spatial difference data of the first wavelength optical density and the spatial difference data of the second wavelength optical density to obtain the bladder water content index modeling vector.

[0009] S5: Model the bladder water content index vector to estimate the bladder fullness state and estimate the bladder capacity.

[0010] According to another aspect of this application, a real-time bladder fullness monitoring system based on a near-infrared sensor is provided, comprising: a light intensity signal acquisition module, used to acquire a first wavelength light intensity signal of a first measurement channel and a second wavelength light intensity signal of a second measurement channel collected by a measurement optical path, and a first wavelength light intensity signal of a first reference channel and a second wavelength light intensity signal of a second reference channel collected by a reference optical path, wherein the reference optical path is adjacent to the measurement optical path.

[0011] The light intensity-to-light density conversion module is used to perform light intensity-to-light density conversion on the first wavelength light intensity signal of the first measurement channel, the second wavelength light intensity signal of the second measurement channel, the first wavelength light intensity signal of the first reference channel, and the second wavelength light intensity signal of the second reference channel to obtain light density change data of the first wavelength of the first measurement channel, light density change data of the second wavelength of the second measurement channel, light density change data of the first wavelength of the first reference channel, and light density change data of the second wavelength of the second reference channel.

[0012] The optical density spatial difference calculation module is used to perform spatial difference calculation on the optical density change data of the first wavelength of the first measurement channel, the optical density change data of the first wavelength of the first reference channel, the optical density change data of the second wavelength of the second measurement channel, and the optical density change data of the second wavelength of the second reference channel to obtain the optical density spatial difference data of the first wavelength and the optical density spatial difference data of the second wavelength.

[0013] The bladder water content index generation module is used to model the bladder water content index based on dual-wavelength information from the first wavelength optical density spatial difference data and the second wavelength optical density spatial difference data to obtain the bladder water content index modeling vector.

[0014] The bladder filling state estimation module is used to estimate the filling state of the bladder by modeling the bladder water content index vector to obtain the bladder filling state and estimate the capacity.

[0015] Compared with existing technologies, this application provides a real-time bladder fullness monitoring system and method based on near-infrared sensors. It addresses the interference of physiological baseline drift on bladder fullness monitoring by introducing a spatial differential near-infrared spectroscopy measurement method. Specifically, it utilizes a set of measurement optical paths and a set of adjacent reference optical paths to simultaneously detect the abdominal region. The measurement optical path passes through the bladder, and its signal changes include bladder fullness information and physiological noise from superficial tissues; while the reference optical path only passes through superficial tissues, and its signal mainly reflects physiological noise. By performing optical density conversion on the dual-wavelength optical signals acquired by the two optical paths and performing real-time spatial differential processing, common-mode interference signals caused by factors such as systemic hydration status and local blood flow, i.e., physiological baseline drift mentioned in the background art, can be effectively eliminated or suppressed. This differential operation is equivalent to dynamically calibrating the measurement channel with the signal from the reference channel, thereby separating global physiological fluctuations unrelated to bladder fullness from the total signal. Finally, a bladder water content index model is established based on the differentially purified signal, and state estimation is performed, thus achieving accurate and reliable monitoring of bladder fullness and significantly reducing the false alarm rate. Attached Figure Description

[0016] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0017] Figure 1 This is a flowchart of a method for real-time monitoring of bladder fullness based on a near-infrared sensor according to an embodiment of this application.

[0018] Figure 2This is a schematic diagram of the data flow of a real-time bladder filling monitoring method based on a near-infrared sensor according to an embodiment of this application.

[0019] Figure 3 This is a flowchart of step S2 in the real-time bladder filling monitoring method based on a near-infrared sensor according to an embodiment of this application.

[0020] Figure 4 This is a flowchart of step S5 in the real-time bladder filling monitoring method based on a near-infrared sensor according to an embodiment of this application.

[0021] Figure 5 This is a block diagram of a real-time bladder fullness monitoring system based on a near-infrared sensor according to an embodiment of this application. Detailed Implementation

[0022] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0023] In response to the problems mentioned in the background description, this application proposes a method for real-time monitoring of bladder fullness based on a near-infrared sensor. Figure 1 This is a flowchart of a method for real-time monitoring of bladder fullness based on a near-infrared sensor according to an embodiment of this application. Figure 2 This is a schematic diagram of the data flow in a real-time bladder fullness monitoring method based on a near-infrared sensor according to an embodiment of this application. Figure 1 and Figure 2As shown, the real-time bladder fullness monitoring method based on a near-infrared sensor according to an embodiment of this application includes: S1: acquiring a first wavelength light intensity signal of a first measurement channel and a second wavelength light intensity signal of a second measurement channel collected by a measurement optical path, and a first wavelength light intensity signal of a first reference channel and a second wavelength light intensity signal of a second reference channel collected by a reference optical path, wherein the reference optical path is adjacent to the measurement optical path; S2: performing light intensity-to-light density conversion on the first wavelength light intensity signal of the first measurement channel, the second wavelength light intensity signal of the second measurement channel, the first wavelength light intensity signal of the first reference channel, and the second wavelength light intensity signal of the second reference channel to obtain light density change data of the first wavelength of the first measurement channel, light density change data of the second wavelength of the second measurement channel, and light density change data of the first reference channel. S3: Spatial difference is performed on the optical density change data of the first wavelength of the first measurement channel, the first wavelength of the first reference channel, the second wavelength of the second measurement channel, and the second wavelength of the second reference channel to obtain spatial difference data of the first wavelength optical density and spatial difference data of the second wavelength optical density; S4: Bladder water content index modeling is performed on the spatial difference data of the first wavelength optical density and the spatial difference data of the second wavelength optical density based on dual-wavelength information to obtain bladder water content index modeling vector; S5: Bladder water content index modeling vector is used to estimate the bladder filling state to obtain bladder filling state and estimated capacity.

[0024] In step S1, the first wavelength light intensity signal of the first measurement channel and the second wavelength light intensity signal of the second measurement channel are acquired by the measurement optical path, and the first wavelength light intensity signal of the first reference channel and the second wavelength light intensity signal of the second reference channel are acquired by the reference optical path, wherein the reference optical path is adjacent to the measurement optical path. It should be understood that a key challenge in non-invasive bladder filling monitoring using near-infrared technology stems from the complexity of biological signals. The light intensity attenuation detected by the sensor not only includes the target signal caused by the increase in urine in the bladder, but also a large number of interference signals generated by the physiological activities of superficial tissues (such as skin and fat). For example, a user's drinking behavior can cause minor changes in overall hydration status, while changes in posture or emotion can lead to fluctuations in local blood flow. The baseline drift caused by these physiological activities is highly similar to the bladder filling signal in the time domain, making it difficult for traditional single-point measurement methods to accurately distinguish between the two, resulting in frequent misjudgments. To overcome this technical bottleneck, this application sets up an adjacent reference optical path next to the measurement optical path, the purpose of which is to capture a reference signal that is independent of the target signal but can reflect the physiological noise of superficial tissues in real time. In this way, by performing differential processing on the two optical signals, common-mode interference can be effectively eliminated, thereby accurately separating the pure signal caused by changes in bladder capacity.

[0025] In an exemplary embodiment of this application, the operation flow of step S1 is as follows: The light source module integrates two light-emitting diodes with different wavelengths. In an exemplary embodiment of this application, the first wavelength is 970nm and the second wavelength is 850nm. That is, the 970nm wavelength is located near the strong absorption peak of water, making it highly sensitive to changes in the content of urine in the bladder (whose main component is water), and is a key wavelength for detecting target signals. The 850nm wavelength is more sensitive to hemoglobin and deoxyhemoglobin, and is close to their isoabsorption point, which can effectively reflect changes in light absorption caused by background factors such as tissue blood flow. At the same time, its absorption coefficient for water is significantly different from that of 970nm, which provides a basis for the subsequent separation of the influence of different physiological components using a dual-wavelength algorithm.

[0026] In terms of sensor layout, two photodetector modules, such as photodiodes, are asymmetrically placed on one side of the light source module. The distance between the first photodetector (detector A) and the light source module is set relatively far, for example, 3 to 4 centimeters. This distance ensures that photons emitted from the light source can form a deeper optical path. This optical path, after penetrating superficial tissues such as skin, fat, and muscle, can effectively reach and partially pass through the deep bladder, and is ultimately received by detector A. This deep optical path is defined as the measurement optical path. In contrast, the distance between the second photodetector (detector B) and the light source module is set relatively close, for example, 1 to 1.5 centimeters. Its detection depth is correspondingly shallower, and the formed optical path is mainly limited to the skin and subcutaneous fat layer, basically not involving the deep bladder tissue. This shallow optical path is defined as the reference optical path. Since the reference optical path and the measurement optical path are spatially adjacent on the skin surface, it can be reasonably assumed that the superficial tissues they both pass through have highly similar physiological characteristics and background noise.

[0027] During continuous monitoring, the control circuit precisely drives the 970nm and 850nm LEDs alternately using high-frequency time-division multiplexing, and simultaneously acquires data to form a continuous data stream. Specifically, when the 970nm LED is driven to emit light, detectors A and B synchronously sample the light intensity. Over time, detector A continuously outputs a series of light intensity values, forming a time series of the first wavelength light intensity signal for the first measurement channel. This series dynamically reflects the total attenuation change of the 970nm light across the entire measurement optical path, including the bladder. Simultaneously, detector B also generates a time series of the first wavelength light intensity signal for the first reference channel, which mainly depicts the attenuation change of the 970nm light by superficial tissue. Then, the control circuit switches to drive the 850nm LED, and detectors A and B again synchronously acquire data, generating continuous time series of the second wavelength light intensity signal for the second measurement channel and the second wavelength light intensity signal for the second reference channel, respectively.

[0028] In step S2, the light intensity signals of the first wavelength of the first measurement channel, the second wavelength of the second measurement channel, the first wavelength of the first reference channel, and the second wavelength of the second reference channel are converted from light intensity to light density to obtain light density change data of the first wavelength of the first measurement channel, the second wavelength of the second measurement channel, the first wavelength of the first reference channel, and the second wavelength of the second reference channel. Correspondingly, the raw signal directly acquired by the photodetector is a light intensity signal, the value of which is related to various factors such as light source intensity, detector sensitivity, skin coupling, and the optical properties of the tissue, and its relationship with the concentration of the absorbed substance is non-linear. For quantitative or semi-quantitative physiological analysis, this physical quantity needs to be converted into a parameter that can more directly and linearly reflect changes in tissue absorption. According to the Beer-Lambert law, changes in light density are proportional to changes in the concentration of the absorbed substance. Therefore, converting the raw light intensity signal into light density change data is to linearize the non-linear light intensity signal, making it directly correlated with the relative changes in the concentration of light-absorbing substances (such as water and hemoglobin) within the tissue. This transformation is a necessary prerequisite for subsequent signal differentiation, feature extraction, and physiological state modeling, and it provides an effective mathematical basis for accurately analyzing the absorption changes caused by bladder fullness.

[0029] In one exemplary embodiment of this application, Figure 3 This is a flowchart of step S2 in the real-time bladder fullness monitoring method based on a near-infrared sensor according to an embodiment of this application. Figure 3 As shown, step S2 includes: S21, extracting the baseline light intensity of the first wavelength of the first measurement channel; S22, based on the baseline light intensity of the first wavelength of the first measurement channel, performing light intensity-to-density conversion on the light intensity at each moment in the light intensity signal of the first wavelength of the first measurement channel to obtain the light density change data of the first wavelength of the first measurement channel.

[0030] In the exemplary embodiment described above, the operation flow of step S2 is as follows: First, a stable baseline light intensity needs to be determined for each signal. It should be understood that in near-infrared spectroscopy-based physiological monitoring, the measured optical density values ​​are usually relative changes rather than absolute values. This is because absolute light absorption values ​​are significantly affected by various static or slowly varying factors, such as individual differences (e.g., skin color, fat thickness), the coupling pressure between the sensor and the skin, and the device's own parameters (e.g., light source intensity, detector gain). Directly using instantaneous light intensity for calculation cannot distinguish whether the difference in signals is due to target physiological changes or these individual differences. To eliminate the interference of these static differences and focus on the changes in light absorption caused by dynamic physiological processes such as bladder filling, a stable and reliable reference benchmark needs to be established for each measurement channel. Therefore, extracting a baseline light intensity in an initial state can provide a normalized starting point for subsequent optical density change calculations, ensuring that the calculated optical density change values ​​can realistically and comparably reflect changes in tissue optical properties caused by physiological activities, rather than individual or device differences.

[0031] Specifically, this explanation uses the baseline light intensity of the first wavelength of the first measurement channel as an example. First, step S21 is executed. After the user first puts on the monitoring device and starts the monitoring program, the program enters a preset baseline acquisition phase. The duration of this phase can be preset, for example, to 3 minutes. This time length is chosen to ensure that the user is in a relatively still and physiologically stable state, while changes in bladder capacity during this period are negligible. During these 3 minutes, the device continuously acquires the light intensity signal of the first wavelength of the first measurement channel at a preset sampling frequency, for example, 10 Hz, thereby obtaining a time series containing 1800 light intensity data points (3 minutes × 60 seconds / minute × 10 Hz). Next, this time series data needs to be processed to extract the baseline light intensity. To enhance the robustness of the baseline and avoid the influence of single-point noise or transient disturbances, single data points are not used directly. A preferred method is to calculate the statistical characteristic value of the time series. For example, the arithmetic mean of these 1800 light intensity data points can be calculated. Alternatively, to better resist occasional spike interference, a more robust statistic, such as the median, can be used. In another optional implementation, the data can be filtered first, for example, using a low-pass filter to remove high-frequency noise, and then the average value of the filtered signal can be calculated. Through calculation, the average value of the first wavelength light intensity signal of the first measurement channel within this initial window is obtained as a specific value, such as 1.25 volts. This value of 1.25 volts is then determined as the baseline light intensity of the first wavelength of the first measurement channel. It should be emphasized that this explanation uses the first wavelength of the first measurement channel as an example; however, in practical applications, this baseline extraction process is applied in parallel and independently to all four light intensity signal channels, that is, extracting the baseline light intensity values ​​for the first wavelength of the first measurement channel, the second wavelength of the second measurement channel, the first wavelength of the first reference channel, and the second wavelength of the second reference channel, respectively.

[0032] After obtaining the baseline light intensity for each channel, the subsequently acquired real-time light intensity signals can be converted from light intensity to optical density. Accordingly, each measurement channel establishes a baseline light intensity representing the initial physiological state. However, the sensor acquires the raw light intensity signal stream during continuous monitoring, the value of which has a non-linear relationship with the concentration of light-absorbing substances in the tissue, and is not standardized. For meaningful physiological analysis, this dynamic light intensity signal stream needs to be converted into a physical quantity that can directly and linearly reflect changes in the tissue's optical properties. According to Beer-Lambert's law, the change in optical density (OD) is proportional to the change in the concentration of light-absorbing substances. Therefore, this application uses the established baseline light intensity as a reference, converting the instantaneous light intensity value at each subsequent moment into an optical density change value relative to that baseline.

[0033] Taking the first wavelength of the first measurement channel as an example again, step S22 is then executed. In an exemplary embodiment of this application, step S22, based on the baseline light intensity of the first wavelength of the first measurement channel, performs light intensity-to-density conversion on the light intensity of the first wavelength light intensity signal of the first measurement channel at various times to obtain the light density change data of the first wavelength of the first measurement channel, including:

[0034] Based on the baseline light intensity of the first wavelength of the first measurement channel, the light intensity at each moment in the first wavelength light intensity signal of the first measurement channel is converted from light intensity to light density using the following formula: ;in, The light intensity at each moment in the first wavelength light intensity signal of the first measurement channel. The baseline light intensity at the first wavelength of the first measurement channel. This refers to the optical density at various moments within the optical density change data of the first wavelength in the first measurement channel. Specifically, the calculation process first calculates the instantaneous light intensity. The ratio of the light intensity I1base to the baseline light intensity. This ratio is a normalized relative light transmittance that quantifies how the ability of light to penetrate tissue at the current moment changes relative to the initial baseline state. Because... This is a constant measured under initial steady-state conditions. This ratio effectively eliminates the influence of static factors such as inherent device gain and initial coupling state. As the bladder begins to fill, the volume of urine (mainly water) increases, leading to an increase in the overall water content of the tissue traversed by the measuring optical path. Since water is a strong absorber in the near-infrared band (especially around 970 nm), the increased water content inevitably leads to enhanced light absorption. The direct consequence of enhanced light absorption is a reduction in the number of photons that can successfully penetrate the tissue and reach the detector, thus affecting the instantaneous light intensity measured by the detector. Accordingly, it decreases. Therefore, during bladder filling, the ratio will be a value less than 1 that gradually decreases over time. Next, the formula performs a base-10 logarithmic operation on this ratio less than 1. According to the mathematical properties of logarithms, the logarithm of a positive number less than 1 must be negative. The magnitude of this negative value is related to the degree to which the ratio deviates from 1; that is, the stronger the light absorption, the smaller the ratio, and the larger the absolute value of its logarithm. Finally, by adding a negative sign before the logarithmic result, this negative value is converted to a positive value, resulting in the value of . Optical density. This calculation process performs a conversion on each data point in the first wavelength light intensity signal of the first measurement channel, thereby generating an optical density change data stream synchronized with the original light intensity signal stream. Each output optical density value represents the change in light absorption relative to the initial baseline state at the current moment. Similarly, in the actual processing flow, this light intensity-to-optical density conversion operation is applied in parallel to all four channels. That is, the second wavelength light intensity signal of the second measurement channel, the first wavelength light intensity signal of the first reference channel, and the second wavelength light intensity signal of the second reference channel are also converted using their respective baseline light intensities determined in step S21, through the same formula.

[0035] In step S3, spatial difference is performed on the first wavelength optical density change data of the first measurement channel, the first wavelength optical density change data of the first reference channel, the second wavelength optical density change data of the second measurement channel, and the second wavelength optical density change data of the second reference channel to obtain spatial difference data of the first wavelength optical density and spatial difference data of the second wavelength optical density. It is understood that although step S2 produces optical density data that linearly reflects changes in tissue light absorption, these data are still mixed signals. The optical density changes in the measurement channel contain both target information from bladder fullness and superficial tissue physiological noise caused by factors such as changes in systemic hydration and local blood flow regulation. Similarly, the optical density changes in the reference channel are also mainly composed of these physiological noises. Since these interference sources typically exhibit spatial homogeneity or high correlation within a local area, they simultaneously affect the measurement optical path and the adjacent reference optical path, forming so-called common-mode interference. To accurately extract signals related only to deep bladder activity from the mixed measurement signals, a method needs to be found to eliminate or suppress this common-mode interference. To this end, this application introduces spatial differential processing to use the physiological noise signal captured by the reference channel to correct the signal of the measurement channel in real time, thereby eliminating common-mode, non-target physiological baseline drift and purifying the differential signal that reflects the bladder filling state.

[0036] In an exemplary embodiment of this application, step S3 includes: S31, calculating the positional difference between the first wavelength optical density change data of the first measurement channel and the first wavelength optical density change data of the first reference channel to obtain the first wavelength optical density spatial difference data; S32, calculating the positional difference between the second wavelength optical density change data of the second measurement channel and the second wavelength optical density change data of the second reference channel to obtain the second wavelength optical density spatial difference data.

[0037] In the above exemplary embodiment, the operation flow of step S3 is as follows: In a specific embodiment, the spatial difference processing is performed one by one by wavelength and time point. First, step S31 is executed to process the signal with a first wavelength, such as 970nm. The processor synchronously acquires the first wavelength optical density change data of the first measurement channel and the first wavelength optical density change data of the first reference channel at the same time i. Then, the difference between the two is calculated. This calculation is position-differentiated, meaning that the same subtraction operation is performed for each time point in the time series. Since the superficial tissue physiological changes (such as skin hydration) shared by the measurement optical path and the reference optical path are approximately equal in these two data, the subtraction operation can effectively cancel out this part of the common-mode signal. And since the reference optical path does not pass through the bladder, its signal does not contain bladder filling information, so the difference mainly retains the optical density change caused by bladder filling that exists only in the measurement channel. By performing point-by-point calculation on the entire time series, a completely new time series is finally generated, namely, the spatial difference data of the first wavelength optical density.

[0038] Simultaneously, step S32 is executed, applying the same logic to a signal with a second wavelength, such as 850nm. The processor acquires the second wavelength optical density change data of the second measurement channel and the second wavelength optical density change data of the second reference channel at the same time i. Then, the positional difference between them is calculated. This difference also eliminates common-mode physiological noise at the second wavelength, primarily reflecting the optical characteristic changes of the bladder region at the second wavelength. Calculations are performed on the entire time series to obtain the spatial difference data of the second wavelength optical density.

[0039] In step S4, the bladder water content index is modeled based on dual-wavelength information using the spatial difference data of the first and second wavelength optical densities to obtain the bladder water content index modeling vector. It should be understood that after spatial difference processing, although most of the common-mode noise caused by superficial tissues is eliminated, the two differential signals are still a mixture of information. This is because the bladder region itself is a complex biological tissue, and its optical changes not only stem from the increase in urine (water) but may also be affected by factors such as changes in bladder wall hemodynamics. For example, bladder filling may be accompanied by bladder wall stretching and changes in local blood volume. Since water and hemoglobin have different absorption characteristics at the two wavelengths (970nm and 850nm), these two differential signals are actually a linear superposition of multiple factors, including changes in water content and hemoglobin concentration. To separate the water content change caused by increased urine from this superposition effect, decoupling using dual-wavelength information is necessary. Therefore, by constructing a mathematical model based on dual-wavelength information, the two differential signals are effectively combined to maximize the contribution of water content, while suppressing or eliminating residual interference caused by changes in other components (such as hemoglobin), thereby generating an index that can more purely and accurately characterize changes in bladder water content.

[0040] In an exemplary embodiment of this application, step S4 includes: calculating the weighted difference between the first wavelength optical density spatial difference data and the second wavelength optical density spatial difference data based on the first wavelength weight and the second wavelength weight to obtain the bladder water content index modeling vector.

[0041] In the above exemplary embodiment, the operation flow of step S4 is as follows: Specifically, for any time i in the first wavelength optical density spatial difference data and the second wavelength optical density spatial difference data, the calculation formula for the bladder water content index (BWI) can be expressed as: BWI(i) = W1*h1(i) - W2*h2(i). Where h1(i) and h2(i) are the values ​​at the i-th time in the first wavelength optical density spatial difference data and the second wavelength optical density spatial difference data, respectively, and W1 and W2 are the first wavelength weight and the second wavelength weight, respectively. These are key parameters for achieving target signal purification. The determination of these weights is based on precise physical principles, namely, the modified Beer-Lambert law. In an exemplary embodiment of this application, the first wavelength weight and the second wavelength weight are determined based on the extinction coefficients of the first wavelength and the second wavelength for water and proteins, that is, based on the different absorption characteristics of each wavelength for the target substance (water) and the main interfering substances (such as proteins or hemoglobin in blood). The extinction coefficient is a physical quantity describing the absorption capacity of a substance per unit concentration and unit path length for a specific wavelength of light, and is an inherent property of the substance. The weighting is designed to make the weighted index BWI most sensitive to changes in water concentration, while remaining insensitive to changes in other coexisting, potentially interfering substances (such as hemoglobin), ideally resulting in a response of zero or near-zero. To achieve this, a constraint equation can be constructed using the extinction coefficients of the interfering substances when calculating the weight ratio. Taking water as the target substance and hemoglobin as the main interfering substance as an example, the extinction coefficients of hemoglobin at wavelengths λ1 and λ2 are b1 and b2, respectively. To eliminate the influence of hemoglobin concentration changes on BWI, the weights W1 and W2 must satisfy a constraint condition: W1*b1 - W2*b2 ≈ 0. By solving this system of equations, the ratio of W1 to W2 can be obtained. For example, W2 can be set to 1, then W1 = b2 / b1. These extinction coefficient values ​​can be obtained from publicly available scientific literature or databases. For example, regarding 970nm and 850nm, literature shows that the extinction coefficient of hemoglobin at 850nm is approximately 0.7, and at 970nm it is approximately 1.0. Based on this derivation, if W2=1, then W1=0.7 / 1.0=0.7. Therefore, a specific weight pair could be W1=0.7 and W2=1. Applying the calculated weights W1 and W2 to the aforementioned weighted difference formula, and performing point-by-point calculations on the time series of the two input difference signals, a single time series can be obtained, namely the bladder water content index modeling vector. Each value in this vector has undergone dual purification through spatial difference and spectral decoupling, enabling it to reflect the net change in water content caused by the increase in urine in the bladder to the greatest extent and linearly.

[0042] In step S5, the bladder fluid content index modeling vector is used to estimate the bladder fullness state and estimated capacity. That is, the bladder fluid content index modeling vector itself is still just a continuously changing numerical sequence. Although it is positively correlated with bladder fullness, it has not yet been transformed into actionable information that provides direct guidance to the user. Users, especially patients or caregivers, need to know specific fullness states such as empty, urgency, or needing to urinate, as well as an approximate volume value (e.g., 200 ml). Furthermore, the original index vector may still contain some short-term fluctuations caused by minor body movements or electronic noise. To transform this internally calculated index into a final, practical, and stable monitoring result, and to provide users with clear and explicit physiological state feedback and quantitative estimates, the index vector needs final interpretation and processing.

[0043] In one exemplary embodiment of this application, Figure 4 This is a flowchart of step S5 in the real-time bladder filling monitoring method based on a near-infrared sensor according to an embodiment of this application. Figure 4 As shown, step S5 includes: S51, filtering and smoothing the bladder water content index modeling vector to obtain a filtered bladder water content index vector; S52, extracting the current bladder water content index from the filtered bladder water content index vector; S53, generating the bladder fullness state based on the comparison between the current bladder water content index and a preset threshold; S54, inputting the current bladder water content index into a calibration model to obtain the estimated capacity.

[0044] In the above exemplary embodiment, the operation flow of step S5 is as follows: First, step S51 is executed. It should be understood that while the bladder water content index modeling vector can effectively track the macroscopic trend of bladder filling, in actual wearing scenarios, even the slightest changes in body posture, abdominal muscle contraction, respiratory depth, or even minute changes in the pressure of the sensor against the skin can introduce non-physiological, transient high-frequency fluctuations or spikes into the signal. Although these noises are small in amplitude, if used directly for subsequent state judgment and capacity estimation without processing, they may cause frequent jumps in threshold judgment or severe fluctuations in capacity readings, thereby reducing the stability and reliability of the monitoring results and causing confusion for users. To extract the core signal that truly reflects the slow, continuous filling process of the bladder and to provide a stable and reliable value for subsequent threshold comparison and model input, the index vector needs further purification processing.

[0045] Specifically, in one embodiment, this filtering and smoothing process can be achieved by applying a digital low-pass filter. Bladder filling is a slow physiological process measured in minutes or even hours, corresponding to extremely low signal frequencies. Therefore, a filter with a very low cutoff frequency can be chosen to filter out high-frequency noise unrelated to the filling process. One such filter is a moving average filter. This filter is implemented by setting a fixed-length time window and then sliding the window across a vector modeling bladder water content indicators. For each data point in the vector, its filtered new value is replaced by the arithmetic mean of that point and all points around it (within the window). Setting the window size is a critical parameter, requiring a trade-off between smoothing effect and signal response delay. An overly large window will over-smooth the flow, potentially masking the true changes in the filling rate; an overly small window will result in poor filtering. For example, a window of 30 seconds can be set. If the data acquisition frequency is 1 Hz, the window size is 30 data points. The filtered index value BWI_filtered(i) at time i is calculated as BWI_filtered(i) = (1 / 30) * Σ[BWI(j)], where j ranges from i-14 to i+15 (for a central moving average). Alternatively, a Kalman filter can be used. A Kalman filter is a recursive filter that predicts the current state based on the state of the previous time step and corrects the prediction using the current observation. Its model architecture includes state prediction equations and observation update equations, where parameters such as process noise covariance and measurement noise covariance need to be preset or adaptively adjusted based on the statistical characteristics of the signal. Compared to a moving average, the Kalman filter typically preserves the dynamic characteristics of the signal better while providing good smoothing, and has a smaller response delay. Regardless of the filtering method used, the input bladder water content index modeling vector is processed point by point, ultimately outputting a filtered bladder water content index vector of the same length but with a smoother curve and significantly reduced noise.

[0046] Next, step S52 is executed. The bladder fluid content index vector comprehensively records the filling trend from the start of monitoring to the present. However, for a real-time monitoring application, its core value lies in providing immediate feedback reflecting the current physiological state. Whether it's issuing a urination reminder to the user or estimating the current bladder capacity, a specific value representing the current point in time is needed as the basis for decision-making, rather than the entire historical data stream. In order to transform the continuous, historical data stream into a discrete value that can be used for immediate judgment and calculation, the specific index representing the current state needs to be accurately located and extracted from this smooth vector.

[0047] Specifically, in one embodiment, the extraction process is performed in real time. The processing unit continuously maintains and updates the filtered bladder fluid content index vector. Since this is a time series, the data points in the vector are arranged in chronological order. Therefore, extracting the current index value is operationally equivalent to obtaining the last element of the filtered bladder fluid content index vector, or the data point with the latest timestamp, as the current bladder fluid content index.

[0048] Then, proceed to step S53. It should be understood that for end users, a continuously changing abstract numerical value is not intuitive and difficult to directly guide their behavior. Users need a clear, easy-to-understand qualitative description to inform them of their current physiological state, such as whether their bladder is empty, whether they need to urinate, or whether the urge to urinate is very urgent. To transform this quantified, continuous internal indicator into discrete, actionable status information that is meaningful to the user, thereby achieving the ultimate goal of monitoring—instant reminders and status notifications—a set of rules needs to be established to map this indicator value to predefined physiological state categories.

[0049] Specifically, in one embodiment, the state generation process is achieved by comparing the current bladder fluid content index with a set of preset thresholds. These thresholds define the boundaries between different physiological states. The setting of the thresholds directly determines the timeliness and accuracy of the state alert. These thresholds can be determined in two main ways: one is a universal threshold derived from statistical analysis of a large amount of clinical trial data; the other is a more preferred individualized calibration method, which records the index value at the time of initial device use, combined with the user's actual feelings (such as the first time feeling the urge to urinate, or when the urge is strong), thereby setting a specific threshold for that particular user. For example, two thresholds, Th1 and Th2, can be preset to divide the bladder filling process into three states. For example, through calibration, it can be determined that for a certain user, Th1=0.45 and Th2=0.80. The processing logic is as follows: upon receiving the current bladder fluid content index, a conditional judgment logic is executed. If the current bladder fluid content index is less than or equal to Th1, for example, an index value of 0.3, then the bladder is determined to be in a low-fill or empty state. If the current bladder fluid content index is greater than Th1 but less than or equal to Th2, for example, an index value of 0.6, the bladder is determined to be moderately full or in a state of urgency to urinate. If the current bladder fluid content index is greater than Th2, for example, an index value of 0.9, the bladder is determined to be highly full or in a state of urgency to urinate. This judgment process is performed in real time; whenever a new current bladder fluid content index is extracted, this logic is executed once. The final bladder fullness status is then generated.

[0050] Ideally, bladder fluid content should rise smoothly as the bladder fills. However, because the optical paths of the measurement and reference channels do not perfectly overlap, spatial misalignment prevents the complete cancellation of common-mode noise, such as local blood flow changes caused by user movement or muscle tension. This results in a dynamic signal-to-noise ratio (SNR) for the final bladder fluid content: a smooth signal and high SNR when the user is stationary, but greater uncertainty and a lower SNR when the user is active. Using a fixed threshold window for state determination leads to sluggish response when signal quality is good (too wide a window), and frequent noise crossing the threshold causes state jitter and false alarms when signal quality is poor (too narrow a window). Therefore, to ensure both response sensitivity and stability in complex environments, a mechanism is needed to adaptively adjust the threshold window width based on real-time signal quality. In other words, the difference between the hysteresis window width for state transitions—the difference between the upgrade and downgrade thresholds of the state machine—should be positively correlated with the real-time uncertainty of the signal. The more unstable the signal, the wider the hysteresis window, and the more conservative the state machine; while the more stable the signal, the narrower the hysteresis window, and the more sensitive the state machine.

[0051] Based on this, in a preferred exemplary embodiment of this application, determining the preset threshold includes: first, obtaining a set of baseline state transition thresholds containing multiple state boundary points. It should be understood that, firstly, it is necessary to obtain a set of baseline state transition thresholds containing multiple state boundary points to provide a static, universal reference framework for subsequent dynamic adjustments. This set of baseline thresholds is preset through statistical analysis of a large amount of clinical data or individualized calibration, defining the theoretical boundary points between different filling states (such as low, medium, and high). For example, a set of standardized quantile values, such as [2;3;7;8], can be selected and defined as follows: 2 as the degradation threshold benchmark from medium to low filling, 3 as the upgrade threshold benchmark from low to medium filling, 7 as the degradation threshold benchmark from high to medium filling, and 8 as the upgrade threshold benchmark from medium to high filling. Alternatively, [2;4;6;8] can be used as state transition benchmarks from medium to low, low to medium, high to medium, and medium to high, respectively. This establishes an anchor point around which all dynamic adjustments revolve, ensuring the orderliness and consistency of the adjustment.

[0052] Then, based on the local time window of the filtered bladder water content index vector, a responsive index value is constructed, namely: ;in, It is the mean of the numerical sequence of the current bladder fluid content index within a local time window of the filtered bladder fluid content index vector. A preset fixed length is determined experimentally, and the latest data sequence is always captured in a sliding manner. For example, the length of the local time window can be preset to 10 seconds. It is the variance of the numerical sequence. This is a responsiveness index value. Accordingly, to achieve adaptive adjustment, the quality of the current signal must first be quantitatively evaluated. By analyzing the bladder fluid content index sequence within a sliding local time window, its dynamic characteristics can be captured in real time. Specifically, the responsiveness index value is constructed using a formula, where... The value within the window, such as the mean, represents the numerical sequence and reflects the signal strength. This is the variance of the sequence, reflecting the noise level. The normalized response value obtained in this way can reflect the determinism of the signal in real time and quantitatively: when the signal is smooth and stable, Approaching 1; when the signal fluctuates drastically, Approaching 0.

[0053] Next, based on the responsiveness index value, the state threshold confidence score is obtained, i.e.: ;in, This is the state threshold confidence score. It should be understandable that by transforming the response using this exponential decay function, the response can be... As a confidence count at the state level, the confidence score triggered by the state change threshold is obtained. In practice, meaningful signals cause the responsiveness index value to... Within the interval [0.5, 1], this makes the calculated confidence score... Accordingly, it falls within the effective working range of [0.293, 0.5], a range that basically matches the reasonable distribution range of the state threshold. When the signal quality is good, the confidence score... A confidence score approaching the upper limit of 0.5 indicates a high probability of a state transition; conversely, a lower confidence score indicates poor signal quality. When the value approaches the lower limit of the interval, it indicates that it is not advisable to switch states at this point.

[0054] Based on the confidence score of the state threshold, a window modulation is applied to the state transition threshold group to obtain the modulated state transition threshold as a preset threshold. Finally, the confidence score calculated in the previous steps is used to actually and dynamically adjust the threshold window. Specifically, The value is used to adjust the difference between the upgrade and downgrade thresholds in the baseline state transition threshold group, i.e., the width of the hysteresis window. When the confidence score... When the confidence score is high, the hysteresis window is narrowed, making the state machine more sensitive and able to respond quickly to real physiological changes; when the confidence score is high... At low hysteresis, the hysteresis window is widened, making the state machine more conservative and effectively suppressing misjudgments caused by noise, thus achieving stable locking of non-critical states. The final set of modulated state transition thresholds generated will serve as the final preset thresholds for state determination at the current moment.

[0055] In a specific numerical embodiment, the modulation is state-dependent. If the system was previously in a moderately full state, modulation is only applied to the thresholds leaving this state (threshold 2 for downgrading to low fullness and threshold 8 for upgrading to high fullness), while thresholds 3 and 7 entering this state remain unchanged. A modulation offset Δ = k * (C_max - cᵢ) is defined, where k is a reasonable scaling factor determined empirically or experimentally (e.g., k = 1), and C_max is the maximum confidence level of 0.5. In scenarios with high signal quality, the calculated cᵢ is close to 0.5, for example, 0.495, resulting in a very small modulation offset Δ of 1 * (0.5 - 0.495) = 0.005. To make the system more sensitive (narrowing the window), the thresholds are shifted inward: the downgrading threshold becomes 2 + Δ = 2.005, and the upgrading threshold becomes 8 - Δ = 7.995. The complete threshold set is then [2.005; 3; 7; 7.995], resulting in a more sensitive system response. Conversely, in scenarios with poor signal quality, the calculated cᵢ is lower, for example, 0.3, so the modulation offset Δ increases to 1 * (0.5 - 0.3) = 0.2. To make the system more conservative (widen the window), the thresholds are shifted outward: the downgrade threshold becomes 2 - Δ = 1.8, and the upgrade threshold becomes 8 + Δ = 8.2. At this point, the complete threshold set is [1.8; 3; 7; 8.2], the system becomes stable, and false alarms caused by noise fluctuations are effectively avoided.

[0056] The bladder fullness state is then determined by comparing this preset value with the current bladder fluid content index. If the recorded state is low fullness (i.e., the previous state), it will only focus on whether the conditions for upgrading to medium fullness are met, i.e., whether the latest current bladder fluid content index is greater than the threshold 3; otherwise, the state will remain unchanged. When in the most common medium fullness state, it will use a wide stable zone consisting of a downgrade threshold of 1.8 and an upgrade threshold of 8.2 for judgment. Only when the index value clearly falls below 1.8 or exceeds 8.2 will the state be updated to low fullness or high fullness, respectively; otherwise, it will stably remain in the medium fullness state. Similarly, if it is already in the high fullness state, it will only monitor whether the index value drops below the downgrade threshold of 7 to decide whether to return to the medium fullness state.

[0057] Finally, step S54 is executed. That is, while the current bladder fluid content indicator has significant indicative value for the user, it remains qualitative. In many clinical and nursing scenarios, simply knowing a vague state is insufficient; healthcare professionals or the user themselves often require more precise quantitative information to aid decision-making, such as choosing the timing for intermittent catheterization or assessing the effectiveness of renal function and fluid management. The current bladder fluid content indicator itself is only a dimensionless relative value, and its absolute magnitude varies due to individual physiological differences (such as abdominal wall thickness and bladder size). To transform this internal, relative indicator into a universally medically significant, standardized physical quantity—bladder capacity in milliliters—this application utilizes an individually calibrated model to precisely map the abstract indicator value to a specific estimated capacity.

[0058] Specifically, in one particular embodiment, this conversion process relies on a calibration model. This calibration model is essentially a mathematical function that describes the mapping between the current bladder fluid content index and the actual bladder capacity. Considering the differences in physiological responses among individuals, the specific form of this model can be flexibly chosen; for example, a linear function can be used, or a more complex polynomial function can be used to pursue higher fitting accuracy. A simple and effective calibration model is a linear regression model, whose structure can be expressed as: Estimated capacity = a * (current bladder fluid content index) + b. The parameters slope 'a' and intercept 'b' in the model are user-specific and need to be determined through an individualized calibration process. This calibration process is crucial for obtaining the model parameters. For example, when using the device for the first time, the user can be guided to perform the following operations: First, after completely emptying the bladder (at which point the actual capacity is 0 ml), wear the device and record the stable current bladder fluid content index, denoted as Index_empty. After a period of time, when the user feels a strong urge to urinate, they go to a medical institution to have their accurate bladder capacity measured using standard equipment (such as ultrasound), for example, Volume_full of 300 ml. Simultaneously, the current bladder fluid content index is recorded as Index_full. With these two sets of paired data points ((Index_empty, 0) and (Index_full, 300)), parameters a and b can be determined by solving a system of two linear equations. The value of a can be calculated using (Volume_full - 0) / (Index_full - Index_empty), while b is the intercept that ensures the line passes through the calibration point, which can be calculated using b = -a * I_empty. For example, if Index_empty is 0.1 and Index_full is 0.7, then a = 300 / (0.7 - 0.1) = 500, and b = -50. Therefore, the user's calibration model is: Estimated volume = 500 * (current bladder fluid content index) - 50. Once this individualized calibration model is established and stored, it can be directly applied in subsequent real-time monitoring. Whenever a new current bladder fluid content index is reached, such as 0.6, this value is substituted into the above formula for calculation: Estimated volume = 500 * 0.6 - 50 = 250. Ultimately, this calculated result of 250 ml is output as the estimated volume, providing the user with intuitive and quantitative information on bladder fullness.

[0059] In summary, the real-time bladder fullness monitoring method based on near-infrared sensors, as described in this application, addresses the interference of physiological baseline drift on bladder fullness monitoring by introducing a spatial differential near-infrared spectroscopy measurement method. Specifically, it utilizes a set of measurement optical paths and a set of adjacent reference optical paths to simultaneously detect the abdominal region. The measurement optical path passes through the bladder, and its signal changes include bladder fullness information and physiological noise from superficial tissues; while the reference optical path only passes through superficial tissues, and its signal mainly reflects physiological noise. By performing optical density conversion on the dual-wavelength optical signals acquired by the two optical paths and performing real-time spatial differential processing, common-mode interference signals caused by factors such as systemic hydration status and local blood flow can be effectively eliminated or suppressed, i.e., physiological baseline drift mentioned in the background art. This differential operation is equivalent to dynamically calibrating the measurement channel with the signal from the reference channel, thereby separating global physiological fluctuations unrelated to bladder fullness from the total signal. Finally, a bladder water content index model is established based on the differentially purified signal, and state estimation is performed, thereby achieving accurate and reliable monitoring of bladder fullness and significantly reducing the false alarm rate.

[0060] Figure 5 This is a block diagram of a real-time bladder fullness monitoring system based on a near-infrared sensor according to an embodiment of this application. Figure 5As shown, the bladder fullness real-time monitoring system 100 based on a near-infrared sensor according to an embodiment of this application includes: a light intensity signal acquisition module 110, used to acquire a first wavelength light intensity signal of a first measurement channel and a second wavelength light intensity signal of a second measurement channel collected by a measurement optical path, and a first wavelength light intensity signal of a first reference channel and a second wavelength light intensity signal of a second reference channel collected by a reference optical path, wherein the reference optical path is adjacent to the measurement optical path; and a light intensity-to-light density conversion module 120, used to perform light intensity-to-light density conversion on the first wavelength light intensity signal of the first measurement channel, the second wavelength light intensity signal of the second measurement channel, the first wavelength light intensity signal of the first reference channel, and the second wavelength light intensity signal of the second reference channel to obtain light density change data of the first wavelength of the first measurement channel, light density change data of the second wavelength of the second measurement channel, and light density change data of the first wavelength of the first reference channel. The system includes: optical density change data and second wavelength optical density change data of the second reference channel; an optical density spatial difference calculation module 130, used to perform spatial difference calculation on the first wavelength optical density change data of the first measurement channel, the first wavelength optical density change data of the first reference channel, the second wavelength optical density change data of the second measurement channel, and the second wavelength optical density change data of the second reference channel to obtain first wavelength optical density spatial difference data and second wavelength optical density spatial difference data; a bladder water content index generation module 140, used to perform bladder water content index modeling based on dual-wavelength information on the first wavelength optical density spatial difference data and the second wavelength optical density spatial difference data to obtain a bladder water content index modeling vector; and a bladder filling state estimation module 150, used to estimate the filling state of the bladder water content index modeling vector to obtain the bladder filling state and estimate the capacity.

[0061] As described above, the near-infrared sensor-based real-time bladder fullness monitoring system 100 according to embodiments of this application can be implemented in various wireless terminals, such as servers with a near-infrared sensor-based real-time bladder fullness monitoring algorithm. In one possible implementation, the near-infrared sensor-based real-time bladder fullness monitoring system 100 according to embodiments of this application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the near-infrared sensor-based real-time bladder fullness monitoring system 100 can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the near-infrared sensor-based real-time bladder fullness monitoring system 100 can also be one of many hardware modules of the wireless terminal.

[0062] Alternatively, in another example, the near-infrared sensor-based real-time bladder fullness monitoring system 100 and the wireless terminal can also be separate devices, and the near-infrared sensor-based real-time bladder fullness monitoring system 100 can be connected to the wireless terminal via wired and / or wireless networks, and transmit interactive information in accordance with an agreed data format.

[0063] Here, those skilled in the art will understand that the specific operations of each step in the above-described near-infrared sensor-based real-time bladder fullness monitoring system have been referenced above. Figures 1 to 4 The method for real-time monitoring of bladder fullness based on near-infrared sensors has been described in detail, and therefore, its repeated description will be omitted.

Claims

1. A method for real-time monitoring of bladder fullness based on a near-infrared sensor, characterized in that, include: S1: Acquire the first wavelength light intensity signal of the first measurement channel and the second wavelength light intensity signal of the second measurement channel collected by the measurement optical path, and the first wavelength light intensity signal of the first reference channel and the second wavelength light intensity signal of the second reference channel collected by the reference optical path. The reference optical path is adjacent to the measurement optical path, and the first wavelength is 970nm and the second wavelength is 850nm. S2: Perform light intensity-to-light density conversion on the first wavelength light intensity signal of the first measurement channel, the second wavelength light intensity signal of the second measurement channel, the first wavelength light intensity signal of the first reference channel, and the second wavelength light intensity signal of the second reference channel to obtain light density change data of the first wavelength light intensity of the first measurement channel, light density change data of the second wavelength light intensity of the second measurement channel, light density change data of the first wavelength light intensity of the first reference channel, and light density change data of the second wavelength light intensity of the second reference channel. S3: Spatial difference is performed on the first wavelength optical density change data of the first measurement channel, the first wavelength optical density change data of the first reference channel, the second wavelength optical density change data of the second measurement channel, and the second wavelength optical density change data of the second reference channel to obtain the first wavelength optical density spatial difference data and the second wavelength optical density spatial difference data. S4: Model the bladder water content index based on dual-wavelength information using the first wavelength optical density spatial difference data and the second wavelength optical density spatial difference data to obtain the bladder water content index modeling vector. S5: Model the bladder water content index vector to estimate the bladder filling state and estimate the bladder capacity; Step S2 includes: Extract the baseline light intensity of the first wavelength in the first measurement channel; Based on the baseline light intensity of the first wavelength of the first measurement channel, the light intensity at each moment in the light intensity signal of the first wavelength of the first measurement channel is converted from light intensity to light density to obtain the light density change data of the first wavelength of the first measurement channel, including: Based on the baseline light intensity of the first wavelength of the first measurement channel, the light intensity at each moment in the first wavelength light intensity signal of the first measurement channel is converted from light intensity to light density using the following formula: ;in, The light intensity at each moment in the first wavelength light intensity signal of the first measurement channel. The baseline light intensity at the first wavelength of the first measurement channel. It is the optical density at each moment in the optical density change data of the first wavelength of the first measurement channel.

2. The method for real-time monitoring of bladder fullness based on a near-infrared sensor according to claim 1, characterized in that, Step S3 includes: The spatial difference data of the first wavelength optical density is obtained by calculating the positional difference between the first wavelength optical density change data of the first measurement channel and the first wavelength optical density change data of the first reference channel; The positional difference between the second wavelength optical density change data of the second measurement channel and the second wavelength optical density change data of the second reference channel is calculated to obtain the spatial difference data of the second wavelength optical density.

3. The method for real-time monitoring of bladder fullness based on a near-infrared sensor according to claim 1, characterized in that, Step S4 includes: calculating the weighted difference between the first wavelength optical density spatial difference data and the second wavelength optical density spatial difference data based on the first wavelength weight and the second wavelength weight to obtain the bladder water content index modeling vector.

4. The method for real-time monitoring of bladder fullness based on a near-infrared sensor according to claim 3, characterized in that, The first wavelength weight and the second wavelength weight are determined based on the extinction coefficients of the first wavelength and the second wavelength for water and protein.

5. The method for real-time monitoring of bladder fullness based on a near-infrared sensor according to claim 1, characterized in that, Step S5 includes: The bladder water content index modeling vector is filtered and smoothed to obtain the filtered bladder water content index vector. Extract the current bladder fluid content index from the filtered bladder fluid content index vector; The bladder fullness state is generated based on the comparison between the current bladder water content index and the preset threshold. The current bladder fluid content index is input into the calibration model to obtain the estimated capacity.

6. A real-time bladder fullness monitoring system based on a near-infrared sensor, characterized in that, include: The light intensity signal acquisition module is used to acquire the first wavelength light intensity signal of the first measurement channel and the second wavelength light intensity signal of the second measurement channel collected by the measurement optical path, as well as the first wavelength light intensity signal of the first reference channel and the second wavelength light intensity signal of the second reference channel collected by the reference optical path. The reference optical path is adjacent to the measurement optical path, and the first wavelength is 970nm and the second wavelength is 850nm. The light intensity-to-light density conversion module is used to perform light intensity-to-light density conversion on the first wavelength light intensity signal of the first measurement channel, the second wavelength light intensity signal of the second measurement channel, the first wavelength light intensity signal of the first reference channel, and the second wavelength light intensity signal of the second reference channel to obtain light density change data of the first wavelength light intensity of the first measurement channel, light density change data of the second wavelength light intensity of the second measurement channel, light density change data of the first wavelength light intensity of the first reference channel, and light density change data of the second wavelength light intensity of the second reference channel. The optical density spatial difference calculation module is used to perform spatial difference calculation on the optical density change data of the first wavelength of the first measurement channel, the optical density change data of the first wavelength of the first reference channel, the optical density change data of the second wavelength of the second measurement channel, and the optical density change data of the second wavelength of the second reference channel to obtain the optical density spatial difference data of the first wavelength and the optical density spatial difference data of the second wavelength. The bladder water content index generation module is used to model the bladder water content index based on dual-wavelength information from the first wavelength optical density spatial difference data and the second wavelength optical density spatial difference data to obtain the bladder water content index modeling vector. The bladder filling state estimation module is used to estimate the bladder filling state and estimate the capacity by modeling the bladder water content index vector. The light intensity-to-light density conversion module includes: Extract the baseline light intensity of the first wavelength in the first measurement channel; Based on the baseline light intensity of the first wavelength of the first measurement channel, the light intensity at each moment in the light intensity signal of the first wavelength of the first measurement channel is converted from light intensity to light density to obtain the light density change data of the first wavelength of the first measurement channel, including: Based on the baseline light intensity of the first wavelength of the first measurement channel, the light intensity at each moment in the first wavelength light intensity signal of the first measurement channel is converted from light intensity to light density using the following formula: ;in, The light intensity at each moment in the first wavelength light intensity signal of the first measurement channel. The baseline light intensity at the first wavelength of the first measurement channel. It is the optical density at each moment in the optical density change data of the first wavelength of the first measurement channel.

Images