A remote care guidance system

Abstract

The present application relates to a kind of remote nursing instruction system, including remote control module, data acquisition module, data mining analysis module, display module, data storage module, information feedback module, terminal module, cloud module, remote medical staff is examined and palpation to the body of patient by respectively medical machine hand, provides remote consultation service, improves diagnosis and treatment efficiency, reduces the cost of manpower and material resources of medical monitoring, improves the quality of medical monitoring.

Description

Technical Field

[0001] This invention belongs to the technical field of remote control, and specifically relates to a remote nursing guidance system. Background Technology

[0002] With the continuous development of internet technology and its widespread application in the field of medical and health monitoring, opportunities have been created for the development of telemedicine. In traditional medical treatment, doctors need to be on-site to test patients' physiological indicators and analyze the results to diagnose and treat their conditions. Telemedicine breaks down the limitations of time and space, narrowing the gap in medical and health resources between different regions. Its development is now gradually influencing traditional medical models. Bedridden patients can have their physiological data collected at home or in a local hospital through telemedicine and IoT medical devices. Remote medical staff can then examine the data and inform them of their physiological status, improving the efficiency of physical examinations and greatly simplifying the examination process. The development of telemedicine is of great significance to areas with scarce medical resources, allowing people in these areas to receive medical treatment similar to that in cities. Meanwhile, my country's aging population is evident, and the number of people suffering from chronic diseases is gradually increasing. Patients with aging and chronic diseases require long-term observation and treatment, but due to limitations and other factors, they typically only visit hospitals for scheduled checkups, leading to poor treatment outcomes. This not only wastes human and financial resources but also significantly impacts the patients' physical and mental well-being. For patients with chronic diseases, their vital signs should be monitored in real time; otherwise, it will seriously affect the accuracy of doctors' diagnoses. Some patients are unable to care for themselves, requiring real-time monitoring for better care. With the rapid development of wireless communication technology, remote health monitoring systems based on wireless sensor networks have gained widespread attention and application in rural and community healthcare services. Non-medical health monitoring systems can not only focus on caring for terminally ill patients but also provide caregivers with proactive guidance and routine monitoring, preventing them from being helpless in various situations. Furthermore, they can handle emergencies through emergency rescue channels. In conclusion, there is an urgent need for a practical and effective solution to address the social issue of remote nursing. Summary of the Invention

[0003] This invention proposes a remote nursing guidance system that can perform palpation on patients, provide remote consultation services, improve diagnostic and treatment efficiency, reduce the human and material costs of medical monitoring, and improve the quality of medical monitoring.

[0004] A remote nursing guidance system includes a remote control module, a data acquisition module, a data mining and analysis module, a display module, a data storage module, an information feedback module, a terminal module, and a cloud module. The remote control module includes a master / slave system, where the master system includes a master medical robotic hand corresponding to a remote control system, and the slave system is a slave medical robotic hand corresponding to a local control system. Remote medical personnel can control the position and pressure of the contact points of the slave medical robotic hand through the master medical robotic hand, and sense the temperature, heat flow, and hardness feedback from the slave medical robotic hand. The system controls palm heat conduction by setting a virtual thermal conductivity coefficient between two fingers of the master / slave medical robotic hand. The master medical robotic hand system alters the heat conduction of the slave medical robotic hand through virtual heat conduction. The slave medical robotic hand is equipped with two Peltier devices, one of which is connected to… One component is attached to the thumb, and the other to the index finger. Real-time thermal sensing can be shared between the master and slave systems, allowing control of heat conduction on the palm and sharing of temperature distribution in a remote system. Based on a thermal network approach, the thermal behavior of the thermal interface is modeled, representing the thermal phenomenon as a circuit composed of thermal resistance and thermal capacity. The data acquisition module monitors the temperature of the fingers and the contacting object through thermocouples installed on both sides of the medical robotic hand. Using the monitored temperature information, the data mining and analysis module calculates the virtual heat flow between the fingers and the contacting object from the medical robotic hand. The connectivity of each heat source is defined as "thermal conductivity." By changing the magnitude of the thermal conductivity, the thermal conductivity of the palm can be freely changed, enabling bidirectional thermal conductivity control of the heat sources attached to the fingers—that is, it can receive heat transfer and also output heat.

[0005] Furthermore, when the temperature of the thumb changes, the temperature of the index finger also changes accordingly. The rate of change is controlled based on the magnitude of the thermal conductivity. When a heat source is applied to the thumb, temperature information and heat flow are exchanged between the master and slave systems. The control objective is expressed as...

[0006] here, in, The main medical robotic arm shares the sum of the temperatures from the medical robotic arm and the doctor's thumb. This is the sum of the heat source temperature of the thumb of the medical robotic hand and the temperature of the object it is in contact with. The temperature of the doctor's thumb in the main medical machine, The temperature of the object being contacted is shared with the main medical robotic arm. To measure the temperature of the thumb heat source in the medical robotic hand. The temperature of the object being touched, received from the medical robotic arm. This refers to the virtual heat flow between the finger and the object it is touching. To create a virtual heat flow between the object and the finger, To achieve these control objectives, the reference value for a heat source in the master / slave system is calculated as follows:

[0007]

[0008] and K represents the heat flux reference values ​​for the thumb and index finger systems, respectively. pt K dt s, K ph K dh s are all reference coefficients. Each temperature and heat flow is controlled by a proportional-derivative controller according to the control target value in the above formula, thereby controlling the temperature and heat flow of the master / slave system to achieve thermal sensation from one side to the other.

[0009] Furthermore, one or more Peltier devices are connected to the interface. Thermal conductivity represents the connectivity of each Peltier device and is related to the virtual heat conduction characteristics between heat sources on the palm. By changing the thermal conductivity, different temperature distributions can be presented on the same interface.

[0010] The input temperature is the temperature recorded by the main medical machine operator. The thermal conductivity can be changed from 0 to an integer to the input temperature of the heat source of the medical robotic hand finger, and the heat flow between the thumb and index finger can be adjusted as needed. When there is no thermal interference, the thermal conductivity is 0. When the thermal conductivity increases, the heat flow is considered. The connectivity of the heat source is changed by controlling the thermal conductivity between the heat sources connected to the index finger. The heat conduction is reproduced by utilizing the connectivity of the heat source without having to share information from all heat sources. The temperature distribution can be freely changed by utilizing the connectivity of the heat source.

[0011] Furthermore, multiple pressure sensors are installed on the fingers of the master / slave medical robotic hand. The pressure distribution and the position of the center of gravity are simultaneously displayed on the display module along with the palpation image. This allows doctors to understand the abdominal position and pressure changes. The response of the piezocapacitive sensor originates from the capacitance change caused by the change in the distance between the upper and lower electrodes of the parallel plate capacitor. An equivalent circuit is used to simulate the electrical signal response of the pressure sensor when there is or is no finger contact. R2 and CPE are connected in parallel and then connected in series with R1 and L. Here, R1 is the contact resistance of the sensor circuit; R2 is the charge transfer resistance in the sensing element; and L is the inductance of the sensor. A constant phase angle element CPE is used to simulate the electrical characteristics of a non-uniform system. The sensing element is composed of porous fibers, and CPE can be represented by the following equation:

[0012] in, ω = 2πf, where ω is the angular frequency in rads. -1 f is the frequency, measured in Hz. CPE consists of q and n, where q is the capacitance value and n (0 ≤ n ≤ 1) is the fractional exponent of the capacitance. Before and after contact with the sensor, the contact resistance R1 of the circuit remained almost unchanged, while the charge transfer resistance R2 of the sensor increased significantly, which is consistent with the sensor's signal measurement results. When the finger was not in contact, the average capacitance of the pressure sensor was 21.633 pF / s. p-1 It increases to 56.9 pF s when touched by a finger. p-1 This represents a 2.26-fold improvement, with the sensor only exhibiting inductive characteristics when touched by a finger.

[0013] Furthermore, before and after contacting the sensor, the contact resistance R1 of the circuit remained almost unchanged, while the charge transfer resistance R2 of the sensor increased significantly. When the finger was not in contact, the average capacitance of the pressure sensor was 21.633 pF. p-1 It increases to 56.9 pF s when touched by a finger. p-1 This represents a 2.26-fold improvement, as the pressure sensor only exhibits inductive characteristics when touched by a finger.

[0014] Furthermore, the pressure sensor is a flexible pressure sensor based on carbon nanofibers / polybutadiene-styrene-butadiene, and its preparation method is as follows: (1) Preparation of carbon nanofiber / polybutadiene-styrene-butadiene membrane Different grades of commercial sandpaper (#400, #600, #1000, #2000) were cleaned with ethanol and then dried for later use. Polybutadiene-styrene-butadiene particles were completely dissolved in ethyl acetate solution under magnetic stirring at 500 rpm. Carbon nanofibers were dispersed in ethyl acetate solution by ultrasonic treatment. The polybutadiene-styrene-butadiene solution was then mixed with the carbon nanofiber dispersion solution in different proportions to prepare carbon nanofiber / polybutadiene-styrene-butadiene solution. Finally, the prepared carbon nanofiber / polybutadiene-styrene-butadiene solution was poured onto pre-cleaned sandpaper and placed in a vacuum desiccator at room temperature to remove air bubbles. After the ethyl acetate was completely evaporated, the carbon nanofiber / polybutadiene-styrene-butadiene was carefully peeled off the sandpaper and cut into 1.5 cm × 1 cm sizes for use. (2) Two 1cm×1cm carbon nanofiber / polybutadiene-styrene-butadiene membranes are bonded to one end with commercial soft silver wires using conductive adhesive and sealed with transparent tape using face-to-face stacking. The silver wires at both ends are connected to a resistance meter to form a resistive flexible pressure sensor.

[0015] Furthermore, the information feedback module employs a time series forecasting method. Time series are generated and arranged chronologically, change over time, and are interconnected. It forecasts future changes based on the inherent patterns of variable variation, using a neural network for time series forecasting. Given a time series {x(1), x(2), ..., x(N)}, time series forecasting is to predict the value x(t+k) at time t+k based on the historical data {x(t), x(tl), ..., x(t-m+1)} of the time series, i.e., to find the relationship between x(t+k) and the historical data {x(t), x(t-1), ..., x(t-m+1)}. When k=1, it is a one-step forecast; when k>1, it is a multi-step forecast. The m parameter is the embedding dimension. The signal is decomposed into 6 levels using the dbN wavelet, and the decomposition relationship is: S=d1+d2+d3+d4+d5+d6+a6. The following formula is used to extract the feature values ​​of each high-frequency part of the wavelet decomposition, and the frequency characteristics of each level are expressed numerically:

[0016] Among them, a m d represents the average value of the signal at layer m. mi Let represent the signal vector of the m-th layer, where n represents the signal dimension. Based on wavelet decomposition, the feature vector of the signal is defined as: [dal, da2, da3, da4, da5, da6, aa6]. The calculated feature vector of 7 elements is used as the input of the neural network.

[0017] Furthermore, the data mining and analysis module also performs noise processing on the physiological signals: Line filtering is used to average the adjacent line data of the two-dimensional image data to average the impact of random noise. After acquiring the data for an entire image, the data of each adjacent group of lines is then horizontally averaged. This is achieved by averaging the first and last two lines of a frame with their adjacent line, and averaging the middle lines with their left and right adjacent lines. The calculation process uses the following formula:

[0018] In the formula, g(x, y) refers to the gray value of the data point after line smoothing, f(x, y) refers to the gray value of the data point before denoising, and a, b, and c are the average coefficients of line smoothing, which can be adjusted. To highlight the pixel value of the center point, the value of a can be increased. Line enhancement on 2D image data aims to highlight the contours of structures within the image, particularly for pixels with significant variations in grayscale between adjacent pixels. This makes the highlighted parts of the image easier to identify. Line enhancement increases the pixel value differences at edges, and is calculated using the following formula:

[0019] In the formula, Enh is the enhancement index. The implementation steps are as follows: First, calculate the values ​​of each point on the line: keep the first row of data unchanged, start from the second row and take the difference between the pixel values ​​of the adjacent points, multiply the difference by the enhancement index Enh, and finally add the difference obtained by multiplying the coefficient to this pixel.

[0020] Furthermore, the main medical robotic arm is wearable.

[0021] Furthermore, the cloud module includes a signal receiving module, and the data mining and analysis module and the data storage module are adapted to receive and process information from the data acquisition module. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of a remote nursing guidance system.

[0023] Figure 2 This is the equivalent circuit of a pressure sensor. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] Human sensory signals originate from the complex life activities of the human body, manifesting as time-series signals based on sensor sampling. Human body temperature is not constant; it fluctuates with factors such as gender, age, day / night cycle, exercise, health, and emotions. Furthermore, the heat transfer through the skin can represent the abundance and functional state of the body's Yang energy. Low heat transfer efficiency indicates slow blood circulation, suggesting coldness, insufficient Yang energy, or inactive Yang energy. Since blood is the sole carrier bringing nutrients and removing waste, this invention innovatively applies thermal sensation to remote medical diagnosis and treatment. It utilizes control technology to develop a human support system, considering the interaction between the human body and a medical robotic hand. The focus is on a thermal sensation rendering technology. By using thermal technology, the medical robotic hand can provide doctors with accurate judgments, such as hot / cold sensations and the intensity of touch. Tactile sensation can be achieved by combining thermal technology with other systems. Humans perceive heat through thermal receptors within their skin. This method can transmit real temperature and heat between the medical robotic hand and the object it contacts, and can share online thermal sensing between the master and slave systems. Specifically, the heat flow and online thermal sensing between the medical robotic hand and the object being detected can be shared between the master and slave systems. The master system is the wearable master medical robotic hand corresponding to the remote control system. Remote medical personnel can control the position and pressure of the slave medical robotic hand's contact points through the master medical robotic hand and sense the temperature, heat flow, and hardness feedback from the slave medical robotic hand. The slave system is the slave medical robotic hand corresponding to the local control system.

[0026] This invention controls palm heat conduction by setting a virtual thermal conductivity between two fingers of a master / slave medical robotic hand. The master medical robotic hand system alters the heat conduction of the slave medical robotic hand through virtual heat conduction. The interface and control algorithm used in this invention can be used to share spatial thermal sensation between the master and slave systems. Experimental results demonstrate the effectiveness of this method, and this interface and its control system hold promise for connecting remote personnel. On one hand, this invention shares thermal sensation from the slave medical robotic hand to the master medical robotic hand; on the other hand, by controlling the heat source temperature of the slave medical robotic hand through the master medical robotic hand, and thereby monitoring the body's response under different heat conduction conditions, it provides doctors with more accurate diagnostic information.

[0027] The medical robotic hand features two Peltier devices, one attached to the thumb and the other to the index finger. Real-time thermal sensation can be shared between the master and slave systems. Furthermore, it allows for control of heat conduction on the palm and sharing of temperature distribution between remote locations.

[0028] The Peltier device is used in most research because it can heat and cool by applying an electric current and can function as a control system. The device has the property of converting the electric current into a temperature gradient, thereby generating heat. This is called the Peltier effect, and the reference heat flux is expressed as...

[0029] However, existing research has not linked the principle of the Peltier device to individual differences in the human body. As is well known, the spatial and temporal resolution of human thermal sensation varies with circumstances. Therefore, this invention detects small temperature differences by expanding the area that presents thermal sensation, and analyzes the relationship between the heating space and duration of the applied heat source, thereby determining the human body's sensation when temperature changes from multiple dynamically positioned heat sources.

[0030] Nonlinear components in the system, such as Joule heating, thermal interference near the device, and modeling errors, are defined as perturbed heat flow. Based on the thermal network method, the thermal behavior of the thermal interface is modeled, representing the thermal phenomenon as a circuit composed of thermal resistance and thermal capacity. Each thermal interface uses two Peltier devices; the device connected to the thumb of the interface is connected to the other side of the system via thermal conduction. The heat flow between the finger and the contact object is monitored by thermocouples mounted on both sides of the medical robotic hand, which monitor the temperatures of the thumb and the contact object, respectively. Using the monitored temperature information, the virtual heat flow from the finger in the medical robotic hand to the contact object is calculated.

[0031] in, This refers to the virtual heat flow between the finger and the object it is touching. The temperature of the fingers, The temperature of the object in contact. This represents the virtual thermal resistance between the finger and the object it is in contact with.

[0032] On the other hand, the total temperature between the finger and the object being touched Represented as

[0033] Temperature and heat flux can be represented as the sum and temperature difference between the finger and the object it contacts. This model is used to perform bidirectional thermal control of a device attached to the finger.

[0034] For heat flow between fingers: The thermal interface has two Peltier devices for each master / slave system. The heat flow between the thumb and forefinger is represented as follows:

[0035] in This indicates the heat flow between the thumb and index finger. Indicates the temperature of the thumb. Indicates the temperature of the index finger. This represents the virtual thermal resistance located between the thumb and index finger.

[0036] Thermal conductivity G is the reciprocal of thermal resistance, expressed as:

[0037] Therefore, thermal conductivity varies with the material of the object in contact. Heat conduction in the hand can be adjusted based on thermal conductivity. Furthermore, the temperature and heat flow between the index fingers in the master-slave system are controlled through bilateral thermal control when in contact with the same object.

[0038] This invention has two objectives: one is to achieve bidirectional thermal control via a single heat source between master and slave systems; the other is to control the thermal conductivity of two heat sources within the same medical robotic hand. For bidirectional thermal control, the temperatures between the master and slave systems can be synchronized. Simultaneously, the sum of heat fluxes becomes zero, and thermal energy between the two systems can be conserved. To apply thermal conductivity control to a remote control system, the connectivity of each heat source is defined as "thermal conductivity." By changing the magnitude of the thermal conductivity, the thermal conductivity of the hand can be freely altered.

[0039] A bidirectional thermal conductivity control is implemented for a heat source attached to a finger, allowing it to both receive and transmit heat. When the temperature of the thumb changes, the temperature of the index finger also changes, with the rate of change controlled based on the thermal conductivity. When a heat source is used at the thumb, temperature information and heat flow are exchanged between the master and slave systems. The control objective is expressed as... here, in, The main medical robotic arm shares the sum of the temperatures from the medical robotic arm and the doctor's thumb. This is the sum of the heat source temperature of the thumb of the medical robotic hand and the temperature of the object it is in contact with. The temperature of the doctor's thumb in the main medical machine, The temperature of the object being contacted is shared with the main medical robotic arm. To measure the temperature of the thumb heat source in the medical robotic hand. The temperature of the object being touched, received from the medical robotic arm. This refers to the virtual heat flow between the finger and the object it is touching. This refers to the virtual heat flow between the object in contact with the finger.

[0040] Using the above formula, thermal energy can be virtually stored, and the sensation of heat can be shared between remote locations.

[0041] To achieve these control objectives, the reference value for a heat source in the master / slave system is calculated as follows:

[0042]

[0043] and K represents the heat flux reference values ​​for the thumb and index finger systems, respectively. pt K dt s, K ph K dh 's' represents a reference coefficient. Therefore, each temperature and heat flux is controlled by a proportional-derivative controller based on the target value in the above equation. This controls the temperature and heat flux of the master / slave system, achieving thermal sensing from one side to the other.

[0044] One or more Peltier devices are connected to an interface. Thermal conductivity represents the connectivity of each Peltier device and is related to the virtual heat conduction characteristics between heat sources on the palm. By changing the thermal conductivity (which depends on the material of the contacting object), different temperature distributions can be presented on the same interface.

[0045]

[0046] The input temperature is the temperature recorded by the main medical machine operator. Input temperature from the heat source of the medical robotic hand's fingers The thermal conductivity can range from 0 to an integer, and the heat flow between the thumb and forefinger can be adjusted as needed. The thermal conductivity is 0 when no thermal interference occurs; as the thermal conductivity increases, heat flow is considered. The connectivity of the heat sources is altered by controlling the thermal conductivity between the heat sources connected to the forefinger and index finger. Heat conduction is reproduced using the connectivity of the heat sources without having to share information from all heat sources. Furthermore, the temperature distribution can be freely altered using the connectivity of the heat sources. This interface and its control algorithm can be used to share spatial thermal sensation of the palm. Experimental results demonstrate the effectiveness of the proposed method. The developed control system interface can connect remote personnel and has potential applications in medical settings.

[0047] At the same time, during palpation, the position, pressure, and hardness feedback of the contact points are also important diagnostic information.

[0048] Deep palpation aims to explore deep muscle tissue or internal organs. From superficial to deep, based on tissue structure, it can be roughly divided into the skin layer, fascia layer, superficial muscle groups, deep muscle groups, bony landmarks, and internal organs. Tender points can appear at any of these layers, so different pressures are needed to reach different levels of tenderness. The fingertips are used to push along the muscle, feeling the tension. The direction of the push is perpendicular to the muscle's direction. By repeatedly pushing along the muscle, the most rigid points or points with nodules can be found. Because it is specific to a particular muscle, repeated pushing along the muscle's direction allows for precise location. Palpation can be used on the breast, abdomen, etc. As is well known, magnetic resonance imaging (MRI) is one of the essential tools for comprehensive breast imaging diagnosis. It is beneficial for the characterization of masses and improves the detection rate of early and multifocal breast cancer. However, MRI is not sensitive to small calcifications in the breast; patients with pacemakers or other metallic foreign bodies in their bodies are prohibited from undergoing MRI examinations; and the images are easily affected by respiratory and cardiac pulsation artifacts. Furthermore, the scanning time is long, the operation is complex, and the cost is high. These characteristics determine that it is not the first choice for early breast cancer screening during health checkups. For abdominal palpation, a superficial palpation is first used, mainly using the palmar surface of the proximal fingers to gently touch the abdominal wall without sliding. The pressure is approximately 1 cm down the abdominal wall to detect abdominal wall tension and resistance. Deep palpation follows the superficial palpation, with the examiner using the metacarpophalangeal joints and the palmar surface of the distal fingers to apply deep pressure to the abdominal wall, approximately 2 cm down. Bilateral palpation is used to assess the condition of the liver, spleen, and kidneys; superficial and deep palpation is used to examine deep organs or masses with large amounts of ascites; and hooked finger palpation is mostly used for liver and spleen palpation. In addition, palpation can also be applied to the fascia and muscle layers. Fascia is divided into superficial fascia, intermediate fascia, and deep fascia. Superficial fascia is located subcutaneously at a relatively superficial level. Intermediate fascia wraps around muscles, and deep fascia wraps around nerves and blood vessels, located between bones and muscle groups. Clinically, fascia is distributed between two muscles, including the areas surrounding the muscles and where nerves pass through. These areas require focused examination and often contain tender points. Radiating sensations can radiate outwards in points, lines, or areas. When a radiating sensation is felt during palpation, it is generally felt within the tissue structures of fascia, muscles, nerves, or blood vessels. The tender point and the radiating area are usually closely related. The general practice is to slowly apply pressure to the tender point; a radiating sensation is also a clue to tenderness.

[0049] This invention incorporates multiple pressure sensors on the fingers of a master / slave medical robotic hand. The pressure distribution and the position of the center of gravity are simultaneously displayed on a monitor along with the palpation image. This allows doctors to understand the abdominal position and pressure changes. The response of the piezocapacitive sensor originates from the capacitance change caused by the change in the distance between the upper and lower electrodes of a parallel-plate capacitor, thus offering the advantage of low power consumption. [The last sentence appears to be incomplete and possibly refers to a different invention.] Figure 2The equivalent circuit shown simulates the electrical signal response of a pressure sensor in the presence or absence of finger contact. R1 is the contact resistance of the sensor circuit; R2 is the charge transfer resistance in the sensing element (PEDOT / PSS-PVA fiber); and L is the inductance of the sensor, using a constant phase angle element (CPE) to simulate the electrical characteristics of the non-uniform system (PEDOT / PSS-PVA fiber). The sensing element is composed of porous fibers with a non-uniform capacitance density distribution, which therefore changes with frequency. The CPE can be represented by the following equation:

[0050] in, ω = 2πf, where ω is the angular frequency in rads. -1 f is the frequency, measured in Hz. CPE consists of q and n, where q is the capacitance value and n (0 ≤ n ≤ 1) is the fractional exponent of the capacitance.

[0051] Before and after contact with the sensor, the contact resistance (R1) of the circuit remained almost unchanged, while the charge transfer resistance (R2) of the sensor increased significantly, consistent with the sensor's signal measurement results. When the finger was not in contact, the average capacitance of the pressure sensor was 21.633 pF / s. p-1 It increases to 56.9 pF s when touched by a finger. p-1 This represents a 2.26-fold improvement. Furthermore, the sensor only exhibits inductive characteristics upon finger contact. This indicates that human contact causes a redistribution of charges within the sensor, ultimately leading to an increase in the sensor's high-frequency impedance.

[0052] Generally, to detect minute changes, the dielectric layer or flexible electrode of the sensor must be able to undergo significant deformation under relatively small pressure, thereby improving the pressure response sensitivity of the flexible pressure sensor. This application introduces micro / nano structures on the sensor electrodes, enabling the sensitive layer to undergo significant deformation under pressure. Typically, templates for micro / nano structures are prepared using laser etching or chemical etching methods; however, these methods are generally complex and time-consuming, unsuitable for large-scale production. This invention employs a simple sandpaper template transfer method to prepare randomly distributed micro / nano structure electrodes. Due to the high aspect ratio of CNF, it can effectively reduce the percolation threshold of composite conductive materials. During the experiment, by adjusting the size of the micro / nano structure and the conductivity of the sensing layer, we obtained a resistive flexible pressure sensor with high sensitivity and high stability.

[0053] The fabrication method of the flexible pressure sensor based on carbon nanofibers / polybutadiene-styrene-butadiene is as follows: (1) Preparation of CNF / SBS membrane Different grades of commercial sandpaper (#400, #600, #1000, #2000) were cleaned with ethanol and then dried for later use. SBS particles were completely dissolved in an ethyl acetate solution under magnetic stirring at 500 rpm. CNF was dispersed in the ethyl acetate solution by ultrasonication. Then, the SBS solution (20%, wt) was mixed with the CNF dispersion in different proportions to prepare CNF / SBS solutions (mass fractions of 2%, 3%, 4%, 5%, 7%, 10%, 15%, and 20%). Finally, the prepared CNF / SBS solutions were poured onto pre-cleaned sandpaper and placed in a vacuum desiccator at room temperature to remove air bubbles. After complete evaporation of the ethyl acetate, the CNF / SBS was carefully peeled off the sandpaper and cut into 1.5 cm × 1 cm pieces for use.

[0054] (2) One end of each of the two CNF / SBS membranes (1cm×1cm) is bonded to a commercial soft silver wire with conductive adhesive and sealed with transparent tape by face-to-face stacking. The silver wires at both ends are connected to a resistance meter to form a resistive flexible pressure sensor.

[0055] Since sensitivity is affected not only by the microstructure's deformation capability but also by the conductivity of the sensing layer, this invention significantly improves the sensor's sensitivity by fabricating a flexible electrode with randomly distributed characteristics and adjusting the sensor's conductivity. Thirty cycles of tensile and release tests were conducted on the thin film under different strains. The experimental results show that the strain sensor exhibits a sensitivity of 769.2 kPa. -1 It features high sensitivity, a low detection limit of 5 Pa, and high reliability in 1000 cycles.

[0056] This invention displays the detected temperature, heat transfer, and pressure-hardness distribution in a remote doctor's intelligent monitoring and management system through feature extraction.

[0057] The algorithms used for different target shapes, detection environments, and tactile signals vary greatly. The intelligent remote monitoring and management system described in this invention adopts the time series forecasting method. The time series is a sequence of observation data that is generated and arranged in chronological order and changes over time and is interconnected. Time series analysis mainly establishes a forecasting model for the time series and forecasts future changes based on the changing patterns of the variables themselves. Neural networks are used for time series forecasting.

[0058] Given a time series {x(1), x(2), ..., x(N)}, time series forecasting is to predict the value x(t+k) at time t+k based on the historical data {x(t), x(tl), ..., x(t-m+1)} of the time series, that is, to find the relationship between x(t+k) and the historical data {x(t), x(t-1), ..., x(t-m+1)}. When k=1, it is a one-step forecast; when k>1, it is a multi-step forecast. The m parameter is the embedding dimension.

[0059] The determination of eigenvalues ​​plays a crucial role in neural network recognition, as the selection of input eigenvalues ​​directly affects the accuracy of the network recognition method and results. The signal is decomposed into 6 levels using the dbN wavelet, with the decomposition relationship: S = d1 + d2 + d3 + d4 + d5 + d6 + a6. The following formula is used to extract eigenvalues ​​from the high-frequency components of each level of the wavelet decomposition, expressing the frequency characteristics of each level numerically.

[0060] Among them, a m d represents the average value of the signal at layer m. mi Let represent the signal vector of the m-th layer, where n represents the signal dimension. Based on wavelet decomposition, the feature vector of the signal is defined as: [dal, da2, da3, da4, da5, da6, aa6]. The calculated feature vector of 7 elements is used as the input of the neural network.

[0061] Neural network technology enables people to predict the behavior of some nonlinear systems. Compared with linear models, using neural network technology for time series forecasting, and analyzing time series data through intelligent learning mechanisms, no longer requires assuming randomness as a basic characteristic of time series data and systems, nor does it require linear assumptions as a premise for time series analysis. This, to some extent, makes up for the shortcomings of stochastic time series analysis technology. However, neural networks also have defects such as being prone to getting trapped in local minima and insufficient generalization ability. This is because monitoring signals often generate varying degrees of noise during signal acquisition or transmission due to the influence of sensors or transmission paths. The presence of noise seriously affects the analysis and diagnosis of physiological signals. Therefore, physiological signal noise processing is a very important part of physiological parameter analysis and diagnosis.

[0062] This invention employs line filtering to average adjacent line data in a two-dimensional image, thus averaging the impact of random noise. After acquiring the entire image data, the data of each adjacent group of lines is then horizontally averaged. The process involves averaging the first and last lines of a frame with their adjacent lines, and averaging the middle lines with their left and right adjacent lines. The calculation uses the following formula:

[0063] In the formula, g(x, y) refers to the gray value of the data point after line smoothing, f(x, y) refers to the gray value of the data point without noise reduction, and a, b, c are the average coefficients of line smoothing, which can be adjusted. If you want to highlight the pixel value of the center point, you can increase the value of a.

[0064] Line enhancement on 2D image data aims to highlight the contours of elements within the image, particularly pixels with significant variations in grayscale, making the desired parts of the image easier to identify. Line enhancement increases the difference in pixel values ​​at edges, calculated using the following formula:

[0065] In the formula, Enh is the enhancement index. The implementation steps are as follows: First, calculate for each point on the line: keep the first row of data unchanged, starting from the second row, take the difference between the pixel values ​​of adjacent points, multiply the difference by the enhancement index Enh, and finally add the difference obtained by multiplying the coefficient to this pixel.

[0066] In summary, the remote nursing platform is a system based on remote management. Remote medical staff can use separate medical robotic arms to examine and measure the patient's body. It can also be used for multidisciplinary professional teams to collaborate, perform palpation on patients, provide remote consultation services, improve diagnostic and treatment efficiency, reduce the human and material costs of medical monitoring, and improve the quality of medical monitoring.

[0067] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention shall fall within the scope of the technical solution of the present invention.

Claims

1. A remote nursing guidance system, characterized in that, It includes a remote control module, a data acquisition module, a data mining and analysis module, a display module, a data storage module, an information feedback module, a terminal module, and a cloud module; The remote control module includes a master / slave system, wherein the master system includes a master medical robot corresponding to the remote control system, and the slave system is a slave medical robot corresponding to the local control system. Remote medical personnel can control the position and pressure of the contact points of the slave medical robot through the master medical robot. Multiple pressure sensors are placed on the fingers of the master / slave medical robotic hand. The pressure distribution and the position of the center of gravity are simultaneously displayed on the display module along with the palpation image. This allows doctors to understand the abdominal position and pressure changes. The response of the piezocapacitive sensor originates from the capacitance change caused by the change in the distance between the upper and lower electrodes of the parallel plate capacitor. An equivalent circuit is used to simulate the electrical signal response of the pressure sensor with and without finger contact. R2 is connected in parallel with a constant phase angle element CPE, and then connected in series with R1 and L. Here, R1 is the contact resistance of the sensor circuit; R2 is the charge transfer resistance in the sensing element; and L is the inductance of the sensor. A constant phase angle element CPE is used to simulate the electrical characteristics of a non-uniform system. The sensing element is composed of porous fibers. The impedance of the constant phase angle element CPE is expressed by the following equation: in, ω = 2πf, where ω is the angular frequency in rads. -1 f is the frequency, measured in Hz. CPE consists of q and n, where q is the capacitance value and n (0 ≤ n ≤ 1) is the fractional exponent of the capacitance. Before and after contacting the sensor, the contact resistance R1 of the circuit remains almost unchanged, while the charge transfer resistance R2 of the sensor increases significantly. The information feedback module employs a time series forecasting method. Time series are generated and arranged chronologically, changing over time and interconnected. It forecasts future changes based on the inherent patterns of variable variation, using a neural network for time series forecasting. Given a time series {x(1), x(2), ..., x(N)}, time series forecasting is to predict the value x(t+k) at time t+k based on the historical data {x(t), x(tl), ..., x(t-m+1)} of the time series, i.e., to find the relationship between x(t+k) and the historical data {x(t), x(t-1), ..., x(t-m+1)}. When k=1, it is a one-step forecast; when k>1, it is a multi-step forecast. The m parameter is the embedding dimension. The signal is decomposed into 6 levels using the dbN wavelet, and the decomposition relationship is: S=d1+d2+d3+d4+d5+d6+a6. The following formula is used to extract the feature values ​​of each high-frequency part of the wavelet decomposition, and the frequency characteristics of each level are expressed numerically: Among them, a m d represents the average value of the signal at layer m. mi Let n represent the signal vector of the m-th layer, and n represent the signal dimension. According to wavelet decomposition, the feature vector of the signal is defined as: [dal, da2, da3, da4, da5, da6, aa6]. The feature vector of the calculated 7 elements is used as the input of the neural network. The data mining and analysis module also performs noise processing on the physiological signals: Line filtering is used to average the adjacent line data of the two-dimensional image data, averaging the impact of random noise. After acquiring the entire image data, the data of each adjacent group of lines are then horizontally averaged. This is achieved by averaging the first and last two lines of a frame with their adjacent line, and averaging the middle lines with their left and right adjacent lines. The calculation process uses the following formula: In the formula, g(x, y) refers to the gray value of the data point after line smoothing, f(x, y) refers to the gray value of the data point before denoising, and a, b, and c are the average coefficients of line smoothing, which can be adjusted. To highlight the pixel value of the center point, the value of a can be increased. Line enhancement on 2D image data aims to highlight the contours of structures within the image. Its effect is more pronounced on pixels with significant variations in adjacent gray levels, making the highlighted portions easier to identify. Line enhancement increases the pixel value differences at edges, and is calculated using the following formula: In the formula, Enh is the enhancement index. The implementation steps are as follows: First, calculate the values ​​of each point on the line: keep the first row of data unchanged, start from the second row and take the difference between the pixel values ​​of the adjacent points, multiply the difference by the enhancement index Enh, and finally add the difference obtained by multiplying the coefficient to this pixel.

2. The remote nursing guidance system as described in claim 1, characterized in that, The pressure sensor is a flexible pressure sensor based on carbon nanofibers / polybutadiene-styrene-butadiene, and its preparation method is as follows: (1) Preparation of carbon nanofiber / polybutadiene-styrene-butadiene membrane Different grades of commercial sandpaper (#400, #600, #1000, #2000) were cleaned with ethanol and then dried for later use. Polybutadiene-styrene-butadiene particles were completely dissolved in ethyl acetate solution under magnetic stirring at 500 rpm. Carbon nanofibers were dispersed in ethyl acetate solution by ultrasonic treatment. The polybutadiene-styrene-butadiene solution was then mixed with the carbon nanofiber dispersion solution in different proportions to prepare carbon nanofiber / polybutadiene-styrene-butadiene solution. Finally, the prepared carbon nanofiber / polybutadiene-styrene-butadiene solution was poured onto pre-cleaned sandpaper and placed in a vacuum desiccator at room temperature to remove air bubbles. After the ethyl acetate was completely evaporated, the carbon nanofiber / polybutadiene-styrene-butadiene was carefully peeled off the sandpaper and cut into 1.5 cm × 1 cm sizes for use. (2) Two 1cm×1cm carbon nanofiber / polybutadiene-styrene-butadiene membranes are bonded to one end with commercial soft silver wires using conductive adhesive and sealed with transparent tape using face-to-face stacking. The silver wires at both ends are connected to a resistance meter to form a resistive flexible pressure sensor.

3. The remote nursing guidance system as described in claim 1, characterized in that, The main medical robotic arm is wearable.

4. The remote nursing guidance system as described in claim 1, characterized in that, The cloud module includes a signal receiving module, and the data mining and analysis module and the data storage module are adapted to receive and process information from the data acquisition module.

Images