Medical monitoring bracelet for receiving infusion information of patient

Abstract

The invention relates to the field of medical information monitoring, and discloses a doctor monitoring bracelet for receiving infusion information of a patient, comprising: a multispectral sensing module for collecting an original physiological light intensity signal of the wrist of the patient; the motion sensing module is used for collecting motion signals; the micro-electro-mechanical motion compensation module is configured on the multispectral sensing module; the processor is electrically connected with the multispectral sensing module, the motion sensing module and the micro-electro-mechanical motion compensation module; and the communication module is electrically connected with the processor. By integrating a high-sensitivity multispectral sensing module and a micro-electro-mechanical active motion compensation system and combining an advanced signal processing algorithm, the accuracy of real-time monitoring of key vital signs of a patient in a complex infusion scene is remarkably improved, more reliable patient infusion information is provided for a subsequent multi-head infusion intelligent switching control system, and the infusion safety of the patient is improved. High fidelity of a data source is ensured, and tedious work of medical staff is reduced.

Description

Technical Field

[0001] The present invention relates to the field of medical information monitoring, and in particular to a doctor monitoring wristband for receiving patient infusion information. Background Art

[0002] In current clinical practice, especially during complex multi-infusion treatments, monitoring of a patient's physiological status primarily relies on traditional bedside monitoring devices and intermittent manual observation by medical staff. While bedside monitoring devices can provide relatively accurate vital signs, their wired nature limits the patient's freedom of movement and typically lacks the ability to conduct in-depth data analysis and proactive early warning for specific infusion scenarios (such as intelligently switching between insulin and glucose during blood sugar fluctuations). Manual observation, on the other hand, is subjective and discontinuous, making it difficult to capture all key physiological changes in real time. This presents a challenge, especially at night or when medical resources are limited, in obtaining timely patient infusion information.

[0003] Currently, management of the infusion process relies primarily on medical staff manually setting and adjusting infusion pump parameters based on doctor's orders and clinical experience. While multi-channel infusion pumps or workstations are available that can manage multiple fluids simultaneously, these devices typically lack the ability to intelligently integrate with patients' real-time physiological status (particularly dynamic data continuously monitored by wearable devices). They often rely on executing preset programs rather than autonomously optimizing or recommending adjustments to multi-infusion intelligent switching strategies based on immediate patient feedback.

[0004] Current infusion management still relies heavily on isolated monitoring methods and manual judgment, lacking an integrated, patient-centric, closed-loop or semi-closed-loop system capable of providing real-time intelligent decision support. This situation limits the in-depth mining and efficient utilization of patient infusion information, and also poses numerous challenges to achieving truly intelligent, personalized, multi-head infusion management. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention provides a doctor monitoring bracelet for receiving patient infusion information, which solves the problem that current infusion management still relies heavily on separate monitoring methods and manual judgment.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a doctor monitoring wristband for receiving patient infusion information, comprising: Multispectral sensing module, used to collect the original physiological light intensity signal from the patient's wrist; Motion sensing module, used to collect motion signals; a micro-electromechanical motion compensation module, configured in the multi-spectral sensing module, for actively adjusting the posture of the multi-spectral sensing module according to a motion compensation control instruction; a processor electrically connected to the multispectral sensing module, the motion sensing module, and the micro-electromechanical motion compensation module, the processor being configured to analyze the motion signal and generate the motion compensation control instruction; Using a dynamic filtering algorithm to process the physiological light intensity signal after compensation by the micro-electromechanical motion compensation module to filter out residual noise; Based on a preset signal reconstruction algorithm and light-tissue interaction model, a pure physiological spectral signal is extracted from the filtered physiological light intensity signal and at least one vital sign value is calculated, the vital sign value including but not limited to blood sugar, respiration, blood lipids, and body temperature; generating linkage control instructions for the infusion device based on the comparison result of the vital sign value and the preset threshold value; The communication module is electrically connected to the processor and is used to wirelessly transmit the vital sign values ​​to the medical monitoring terminal and send the linkage control instructions.

[0007] Preferably, the multispectral sensing module includes a quantum dot enhanced optical sensor array, and the array is configured with light sources of at least three different wavelengths and corresponding photoelectric detection units.

[0008] Preferably, the micro-electromechanical motion compensation module is an active stabilization platform based on a six-degree-of-freedom parallel mechanism, and the motion compensation control instruction drives the active stabilization platform to perform sub-micron displacement compensation.

[0009] Preferably, the dynamic filtering algorithm adopted by the processor is a fractional-order filtering algorithm, wherein the filtering order of the fractional-order filtering algorithm is adaptively adjusted according to the real-time signal-to-noise ratio of the physiological light intensity signal after motion compensation.

[0010] Preferably, the filter order of the fractional order filter algorithm is The adaptive adjustment method is: When the real-time signal-to-noise ratio is higher than a first preset signal-to-noise ratio threshold, the filter order is adjusted to a first preset order range, where the first preset order range corresponds to or is adjacent to an integer order, so as to facilitate retaining low-frequency physiological signal characteristics; When the real-time signal-to-noise ratio is lower than a second preset signal-to-noise ratio threshold, the filter order is adjusted to a second preset order range, which corresponds to a preset lower fractional order value interval to enhance the suppression of high-frequency noise.

[0011] Preferably, the signal reconstruction algorithm adopted by the processor is a compressed sensing reconstruction algorithm, which adopts a joint dictionary containing physiological signal feature basis functions and motion artifact feature basis functions to reconstruct the pure physiological spectral signal by solving a sparse optimization problem.

[0012] Preferably, the light-tissue interaction model adopted by the processor is a fractional-order light transmission model, which expresses the absorption and scattering characteristics of light by tissue in the form of a fractional-order differential equation related to the concentration of the vital sign.

[0013] Preferably, the communication module uses the low-power Bluetooth protocol for periodic data transmission when the vital sign value is within the normal range, and switches to the long-distance low-energy wide area network protocol for immediate alarm data transmission when the vital sign value triggers the warning condition.

[0014] Preferably, the linkage control instructions generated by the processor include: when the vital sign value continues to be higher than the first preset vital sign value threshold for a preset time period, an instruction for reducing the infusion rate is generated; when the vital sign value is lower than the second preset vital sign value threshold, an instruction for pausing the infusion and triggering an alarm is generated.

[0015] A monitoring method for receiving patient infusion information comprises the following steps: Collecting original physiological light intensity signals from the patient's wrist through the multispectral sensing module and collecting motion signals through the motion sensing module; The processor analyzes the motion signal and generates a motion compensation control instruction; The micro-electromechanical motion compensation module actively adjusts the posture of the multi-spectral sensing module according to the motion compensation control instruction to obtain a physiological light intensity signal after motion compensation; The processor processes the motion-compensated physiological light intensity signal using the dynamic filtering algorithm to filter out residual noise and obtain a filtered physiological light intensity signal; The processor extracts a pure physiological spectrum signal from the filtered physiological light intensity signal based on the preset signal reconstruction algorithm and the light-tissue interaction model and calculates at least one vital sign value, the vital sign value including but not limited to blood sugar, respiration, blood lipids, and body temperature; The communication module wirelessly transmits the vital sign values ​​to the medical monitoring terminal; The processor generates a linkage control instruction for the infusion device based on the comparison result of the vital sign value and the preset threshold value, and sends it to the infusion device or the medical monitoring terminal through the communication module.

[0016] The present invention provides a doctor monitoring wristband for receiving patient infusion information. It has the following beneficial effects: 1. This invention significantly improves the accuracy of real-time monitoring of key patient vital signs in complex infusion scenarios by integrating a highly sensitive multispectral sensing module with a micro-electromechanical active motion compensation system, combined with advanced signal processing algorithms. This enables the doctor monitoring wristband to overcome the vulnerability of traditional monitoring methods to motion artifacts, providing more reliable patient infusion information for subsequent multi-infusion intelligent switching control systems. The synergistic effect of physical compensation and algorithmic correction ensures high fidelity of the data source.

[0017] 2. This invention significantly enhances the safety and responsiveness of multi-infusion intelligent switching processes through the intelligent infusion linkage control logic executed by the wristband's built-in processor, combined with its reliable wireless communication module. When monitoring vital signs (such as blood sugar) showing adverse trends or reaching warning thresholds, the wristband can rapidly generate and transmit intervention recommendations or alerts to the switching system and medical staff. This rapid feedback mechanism, based on real-time patient infusion information, is key to ensuring infusion safety and avoiding treatment delays. The key to this invention lies in the integration of local intelligent decision-making and rapid communication.

[0018] 3. By combining the precise capture of patient activity by motion sensors, the physical stabilization of optical sensors by microelectromechanical systems, and the deep suppression of residual motion artifacts by advanced algorithms such as dynamic fractional-order filtering and compressed sensing implemented by the processor, this invention enables the physician monitoring wristband to stably and continuously acquire high-quality physiological data even while the patient is engaged in daily activities. This robustness to motion interference ensures the continuity of the patient infusion information flow and provides a solid foundation for the reliable operation of the multi-head infusion intelligent switching system in real clinical environments. The key to this invention lies in the multi-level active and passive comprehensive motion artifact suppression strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Schematic diagram of the wristband module of the present invention. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the present specification. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0021] Please see the attached Figure 1An embodiment of the present invention provides a doctor monitoring wristband for receiving patient infusion information. The doctor monitoring wristband described in this embodiment is intended to provide medical staff with a wearable device that can continuously, non-invasively, and in real time monitor the key vital signs of the wearer (usually a hospitalized patient receiving infusion treatment), especially the vital sign values, and can be intelligently linked with the infusion system.

[0022] The bracelet body adopts a wristband design, with built-in multispectral sensing module, motion sensing module, micro-electromechanical system (MEMS) active motion compensation module, central processing unit (CPU) and memory, wireless communication module and power management module.

[0023] The multispectral sensing module is the core front-end for accurate and continuous monitoring of the patient's vital status. Its performance directly determines the accuracy and timeliness of the decision-making of the back-end "multi-head infusion intelligent switching control system" (hereinafter referred to as the "switching system"). The design goal of this module is to provide the switching system with high-quality, real-time vital signs, especially the vital sign values ​​( ), blood oxygen saturation (SpO2) and heart rate (HR), so that the switching system can intelligently select or adjust the type, concentration or rate of the currently infused liquid according to the patient's immediate physiological needs, or switch safely and effectively between multiple infusion pathways.

[0024] Assume that a patient is connected to a "multi-head infusion intelligent switching device" at the same time. The device may contain multiple infusion pump heads or pathways, each connected to a different type of fluid, for example: Pathway A: basic nutrient solution (such as compound amino acid injection); Pathway B: High-concentration glucose solution, such as 25% or 50% glucose injection; Pathway C: other solutions; Pathway D: Electrolyte balance solution or special therapeutic drugs.

[0025] The multispectral sensing module continuously monitors the patient's physiological status through the wristband and transmits the data in real time to the control unit of the switching system. The switching system makes decisions based on this data, combined with the preset treatment plan and safety thresholds: Fluid adjustment under physiological stress conditions In addition to blood glucose, the module also monitors SpO2 and HR. For example, if the module detects a sudden drop in SpO2 or abnormal fluctuations in HR while the patient is infusing a certain medication, this may indicate an adverse reaction or fluid overload. After receiving these signals, the switching system can: Suspend the current infusion of the suspected drug.

[0026] Switch to normal saline or basic support fluid access to maintain intravenous access patency.

[0027] At the same time, an alarm is issued to prompt medical staff to assess the patient's condition.

[0028] In order to ensure the reliability of the switching system decision in the above scenario, the multispectral sensing module must have high precision, high sensitivity and high anti-interference ability. Its specific composition is as follows: Precision light source array: VCSEL (Vertical Cavity Surface Emitting Laser) Selection and Layout: For the glucose-sensitive wavelength range, a high-power, narrow-linewidth VCSEL with a central wavelength of 1300nm ± 5nm is selected. This wavelength not only has a relatively characteristic absorption peak for glucose molecules but also has a moderate tissue penetration depth. High power (for example, peak optical power can reach 5-10mW, but average power is controlled within a safe range through pulse modulation) helps improve the signal-to-noise ratio, while a narrow linewidth (for example, <1nm) reduces spectral crosstalk and enhances measurement specificity.

[0029] Blood oxygen and reference wavelength bands: VCSELs with center wavelengths of 850nm ± 3nm (red light) and 940nm ± 3nm (near-infrared light) are used. These wavelengths are the classic SpO2 measurement pair. They also provide important tissue reference information, which is used to correct for the effects of blood volume fluctuations, changes in tissue scattering properties, and other factors on the 1300nm blood glucose signal.

[0030] Optional additional bands: Depending on the switching system's need for other vital signs (such as tissue hydration status, which is not directly measured but can be indirectly correlated), consideration may be given to adding VCSELs such as 1550nm (sensitive to water absorption) or 760nm (different penetration depths).

[0031] Driving and modulation strategies: Constant current drive: Each VCSEL is driven by an independent, high-precision (e.g., current ripple <0.1%) constant current source to ensure the stability of light output intensity.

[0032] Time-division multiplexing pulse modulation: All VCSELs alternately emit light in a strict TDM sequence. For example, within a 1ms cycle, the 1300nm VCSEL emits light for 200µs, followed by the 850nm light for 200µs, and then the 940nm light for 200µs. The remaining 400µs is a dark period or for use by other sensors. The pulse frequency (here, 1kHz) must be chosen to avoid major physiological noise bands and power frequency interference. The duty cycle (here, 20% for each light source) must balance signal strength and average power consumption. This modulation method not only avoids spectral crosstalk but also facilitates subsequent improvement of the signal-to-noise ratio through synchronous demodulation techniques.

[0033] Quantum dot enhanced photodetection array: Detector selection: For the 1300nm band, an InGaAsPIN photodiode with a high peak response wavelength in this area is selected, and its dark current must be extremely low and the response must be high.

[0034] For the 850nm and 940nm bands, high-performance silicon-based PIN photodiodes can be used.

[0035] Quantum dot integration and its switching system: A layer of surface-passivated PbS (lead sulfide) core-shell quantum dots is precisely deposited on the photosensitive surface of the InGaAs detector via spin coating or inkjet printing. The absorption and emission wavelengths of the PbS quantum dots can be precisely controlled by size to cover the 1300nm region. For example, a PbS core with a diameter of approximately 4.5-5.5nm is selected, along with a suitable shell material (such as CdS or ZnS) to improve stability and quantum yield.

[0036] The role of the quantum dot layer here is to act as a highly efficient light capture and enhancement layer. When a 1300nm photon is incident: Quantum dots have extremely high molar extinction coefficients for specific wavelengths and can capture incident photons more efficiently than bulk materials.

[0037] Near-field enhancement / fluorescence re-radiation: The photon energy absorbed by the quantum dots can directly enhance the generation of photogenerated carriers in the InGaAs detector below through non-radiative energy transfer, or the quantum dots can be excited and efficiently radiate secondary fluorescence in a similar band (or a slightly longer band after the Stokes shift, but still within the response range of InGaAs), which is captured by the InGaAs detector.

[0038] Value to switching systems: Quantum dot enhancement makes even subtle differences in light absorption caused by minute changes in vital sign values ​​(e.g., fluctuations of 0.1-0.2 mmol / L) more easily detectable. This means the module can provide vital sign data with higher sensitivity and resolution to the switching system, enabling earlier identification of trends in vital sign values ​​and more precise intervention, resulting in smoother vital sign value control. This is crucial for closed-loop or semi-closed-loop intelligent infusion.

[0039] Arrayed Layout: Detectors are arranged in a 1x2 or 2x2 array, combined with a specific light source geometry, to simultaneously acquire signals from different optical path lengths. For example, close-range SDS signals reflect more epidermal information, while long-range SDS signals carry more information from deeper tissues (including blood vessels). Analyzing the differences between different SDS signals can help algorithms separate surface interference and extract purer blood-related physiological signals, which is crucial for improving blood glucose measurement accuracy and providing more reliable decision-making for switching systems.

[0040] High-performance analog front end (AFE) and signal conditioning: Ultra-low noise transimpedance amplifier (TIA): Each detector channel is equipped with an independent TIA, such as an operational amplifier with fA-level input bias current and nV / Hz-level input reference noise, feedback resistor Select based on expected maximum photocurrent and dynamic range, for example, between 10 MΩ and 1 GΩ.

[0041] Programmable Gain Amplifier (PGA): A PGA is cascaded after the TIA, with a gain range of, for example, 1x to 256x. The CPU dynamically adjusts the gain based on the signal strength to ensure that the ADC input signal is in the optimal range and avoid saturation or excessive quantization noise.

[0042] High-precision ADC: Use a 24-bit or higher resolution Σ-Δ ADC with a sampling rate strictly synchronized with the light source modulation frequency (for example, multiple samples are taken during the plateau period of each light source pulse and digitally integrated) to maximize dynamic range and suppress noise.

[0043] Synchronous demodulation: The CPU performs digital synchronous demodulation (lock-in amplification) on the digital signal stream output by the ADC according to the TDM sequence of the light source, extracting the signal amplitude synchronized with each specific wavelength light source, effectively suppressing ambient light interference and asynchronous noise.

[0044] Output: After the above processing, the module output is consistent with each wavelength The corresponding, preliminarily purified light absorption intensity sequence ,This data stream will serve as a direct input for the subsequent calculation of vital sign values ​​and other parameters, and ultimately serve the decision logic of the dry switching system.

[0045] In the medical monitoring bracelet, the motion sensing module does not directly measure vital signs, but the motion data it provides is key to ensuring the accuracy of the bracelet's core physiological monitoring functions (especially the optical monitoring performed by the multispectral sensing module).

[0046] Even when bedridden, patients receiving infusion therapy will inevitably experience various physical activities, such as turning over, moving their arms, coughing, and even slight tremors. These movements will directly affect the optical coupling between the multispectral sensor module worn on the wrist and the skin, resulting in: Changes in the contact pressure between the sensor and the skin.

[0047] Optically detects minute displacements of points.

[0048] Temporary change in the distribution of blood in the detection area.

[0049] These factors will introduce strong interference, namely motion artifacts, into the optical signals collected by the multispectral sensing module. The amplitude and shape of these artifacts may be comparable to or even larger than real physiological signals (such as the weak light absorption changes caused by changes in blood sugar or blood oxygen saturation), thus seriously contaminating the measurement results.

[0050] If the switching system receives physiological data contaminated by motion artifacts and makes decisions based on it, it may lead to disastrous consequences: Scenario 1: Misdiagnosis of hypoglycemia and incorrect intervention: Assume that a rapid movement of the patient's arm causes temporary distortion of the multispectral signal, which is mistakenly interpreted as a sharp drop in blood sugar. Without correction of motion information, the switching system may: Incorrectly pausing an ongoing insulin infusion.

[0051] Even more serious is the mistaken initiation of a high-concentration glucose infusion, which would cause the patient's actual blood sugar level to rise unnecessarily and even lead to the risk of hyperglycemia.

[0052] Scenario 2: Ignoring actual physiological deterioration On the contrary, if the patient actually experiences physiological deterioration (for example, a decrease in SpO2 due to a drug reaction), but the resulting physiological signal change is masked or offset by a motion artifact in the opposite direction, a switching system that lacks motion artifact recognition and compensation may: The actual SpO2 drop was not detected in time.

[0053] Failure to promptly sound the alarm or switch to necessary supportive infusions (such as adjusting oxygen concentration or replacing fluids) will delay necessary medical intervention.

[0054] Scenario 3: Unnecessarily frequent switching and alarms If the system is sensitive to motion artifacts, any slight disturbance may be misinterpreted as physiological fluctuations, resulting in frequent but unnecessary switching of infusion routes or a large number of false alarms, increasing the burden on medical staff and possibly reducing their trust in the system.

[0055] The core functions of the motion sensing module are: Capture the patient's movement status of the part where the bracelet is worn in real time and accurately.

[0056] It provides key motion reference data for subsequent signal processing algorithms to identify, quantify, and compensate or eliminate motion artifacts in multispectral physiological signals.

[0057] Provide the switching system with a confidence assessment of the quality of the current physiological data. For example, when intense exercise is detected, the system can mark the physiological data during this period as "low confidence" and adopt a more conservative decision-making strategy.

[0058] In some advanced applications, motion data itself can also be used as an indicator to assist in the assessment of physiological status, such as analyzing the patient's activity level and sleep quality (through body motion analysis). This information can indirectly assist the switching system in adjusting treatment plans over a longer period of time.

[0059] The motion sensing module in this embodiment uses a high-performance, low-power nine-axis inertial measurement unit: Core sensor components: 3-axis Accelerometer: used to measure the linear acceleration of the bracelet along its three orthogonal axes (X, Y, Z), including static gravity acceleration and dynamic motion acceleration.

[0060] Typical parameters for selection: Range: Typically, ±8g or ±16g is selected (g is the standard acceleration due to gravity). The ±8g range is sufficient to cover daily human activities and most sudden movements.

[0061] Resolution: Typically 14 or 16 bits, ensuring the ability to detect small changes in motion. For example, 16 bits corresponds to ±8g, resulting in a resolution of approximately 0.244mg / LSB.

[0062] Output Data Rate (ODR): Set to 200 Hz to 500 Hz. This rate needs to be higher than the signal sampling rate of the multispectral sensing module, or at least be able to effectively capture the motion frequency components that produce significant optical artifacts (usually <15 Hz, but its higher harmonics may affect higher frequencies).

[0063] Noise Density: For example, <150µg / √Hz. Low noise helps to more accurately identify subtle movements.

[0064] 3-axis gyroscope: used to measure the angular velocity of the bracelet around its three orthogonal axes.

[0065] Typical parameters for selection: Range: Usually choose ±1000dps (degrees per second) or ±2000dps, which is sufficient to cover the rapid rotation of the wrist.

[0066] Resolution: Typically 16 bits. For example, when 16 bits corresponds to ±1000 dps, the resolution is approximately 0.03 dps / LSB.

[0067] Output Data Rate (ODR): Synchronous with the accelerometer, for example also 200Hz to 500Hz.

[0068] Noise Density: For example, <0.01° / s / √Hz.

[0069] Optional: 3-axis magnetometer: Used to measure local magnetic field strength and assist in absolute attitude estimation (heading angle). It has a lower priority than the accelerometer and gyroscope in wrist motion artifact compensation, but is useful in scenarios where long-term, accurate tracking of the wristband's absolute spatial orientation is required.

[0070] Data output and synchronization: The IMU module is connected via SPI or I 2 The C interface outputs digital three-axis acceleration data to the wristband's central processing unit (CPU) and three-axis angular velocity data .

[0071] Sampling clock for motion sensing module The sampling clock of the multispectral sensing module must be strictly synchronized or have a precisely known timestamp correspondence. This is usually achieved by sharing a master clock source or using a hardware synchronization pulse. Precise time alignment is the cornerstone of the subsequent successful motion artifact reduction algorithm. Any significant time deviation will lead to inaccurate artifact estimation and poor compensation effect.

[0072] The application process of motion data in the multi-head infusion intelligent switching scenario: Synchronous acquisition: CPU simultaneously acquires the original light signal from the multispectral sensing module and motion data from the motion sensing module .

[0073] Motion artifact identification and quantification: Signal processing algorithms running inside the CPU (e.g., adaptive filter-based algorithms like RLS or LMS; blind source separation algorithms based on independent component analysis (ICA) or principal component analysis (PCA); or more advanced deep learning-based artifact removal models) use motion data as key input: Feature extraction: Extract features from motion data that can characterize the intensity, frequency, and type of motion, such as the norm, variance, short-term energy, and zero-crossing rate of acceleration signals; and integrate angular velocity signals to obtain posture changes.

[0074] Artifact modeling / estimation: The algorithm uses these motion features to establish a mathematical model of motion artifacts in the optical signal, or directly estimate the noise component in the optical signal caused by motion .

[0075] Artifact compensation / removal: Subtraction method: ,in is the estimated motion artifact.

[0076] Filtering method: The adaptive filter dynamically adjusts its coefficients according to the motion signal to filter out the frequency bands or components related to motion in the optical signal.

[0077] Physical compensation assistance (such as in combination with the aforementioned MEMS module): If the wristband integrates a MEMS active physical compensation module, the output of the motion sensing module is the direct input source for driving the MEMS platform to perform reverse motion to stabilize the optical sensor. Even with physical compensation, residual micro-motion still needs to be further processed by the algorithm, and the motion sensing module data is still important in this case.

[0078] High-quality vital signs calculations: Using motion artifact compensated optical signals To calculate vital signs , blood oxygen saturation SpO2, heart rate HR and other vital signs. These high-confidence vital signs (for example, and its changing trends Input into the "Multi-head Infusion Intelligent Switching Control System".

[0079] At the same time, these vital signs can be assigned a quality indicator or confidence score based on the intensity of the current movement. For example, when the IMU detects sustained intense movement, even after compensation, the output vital signs may be marked as "medium confidence." The switching system will be more cautious when making decisions based on this data, perhaps requiring a longer stabilization observation period or integrating other clinical information.

[0080] By integrating motion sensing modules and effectively utilizing their data, the Medical Monitoring Bracelet can significantly improve its monitoring accuracy and anti-interference capabilities. This directly translates into huge value for the "Multi-head Infusion Intelligent Switching System": The switching system makes judgments based on more realistic physiological data, greatly reducing the risk of erroneous intervention due to motion artifacts.

[0081] Even when the patient has a certain amount of activity, the system can still obtain relatively reliable monitoring data, expanding the applicable scenarios and populations of the bracelet and switching system.

[0082] For infusion solutions requiring precise closed-loop control, the artifact correction capabilities of the motion sensing module are essential for stable and efficient control. The switching system can more confidently make small, timely adjustments to the infusion rate.

[0083] Reduce invalid alarms and improve medical care efficiency: By distinguishing between true physiological changes and motion artifacts, it can significantly reduce false positive alarms triggered by artifacts, reduce alarm fatigue of medical staff, and allow them to focus more on clinical events that really need attention In the doctor's monitoring wristband, the multispectral sensing module is responsible for collecting optical signals related to the patient's physiological status. However, as mentioned above, daily patient activities, even minor body movements, can seriously interfere with this optical measurement process, resulting in motion artifacts. If these artifact-laden signals are directly used to guide the infusion system, it can cause certain damage.

[0084] Traditional motion artifact processing methods rely primarily on digital algorithms for compensation after signal acquisition. However, the MEMS active motion compensation module in this invention introduces a more proactive, physical solution that attempts to largely offset motion artifacts before they significantly contaminate the original optical signal. This is crucial for ensuring high-fidelity input data for switching systems.

[0085] The switching system accurately and safely regulates multiple infusions based on the patient's real-time physiological needs. The MEMS active motion compensation module directly supports this core mission through the following methods: By physically stabilizing the optical sensing unit, the energy and amplitude of motion artifacts entering the signal acquisition chain are significantly reduced. This means that subsequent digital signal processing algorithms can operate on the original signal with a higher signal-to-noise ratio.

[0086] Physiological optical signal changes caused by parameters such as vital sign values ​​are inherently very weak. If these weak signals are overwhelmed by strong motion artifacts, even advanced algorithms may have difficulty effectively extracting them.

[0087] Improve the reliability and responsiveness of switching system decisions: When a patient's optical characteristics change, the actual change may be obscured if the wristband is shaken violently. The MEMS module stabilizes the optical probe, allowing the CPU to more clearly display the actual downward trend of vital sign values, even under certain motion interference. This makes the CPU's calculated vital sign values ​​more reliable, and this information is transmitted to the central control console for reporting.

[0088] Without MEMS physical compensation, the wristband might only obtain accurate readings when the patient is completely still. However, the presence of the MEMS module enables the wristband to provide clinically valuable monitoring data even when the patient is performing minor daily activities (such as turning over or adjusting arm posture). This greatly enhances the practicality and robustness of the switching system in real-world ward environments.

[0089] The MEMS active motion compensation module in this embodiment is usually integrated inside the bracelet, adjacent to the multispectral sensing module. Its core structure and workflow are as follows: Miniaturized high-precision six-degree-of-freedom (6-DOF) parallel platform Dynamic platform: A lightweight platform with extremely small dimensions (e.g., typically 5mm x 5mm, 1-2mm thick). The key optical components in the multispectral sensing module, namely the VCSEL light source array and quantum dot-enhanced photodetector array, are precisely fixed or integrated on this dynamic platform.

[0090] Static platform: relatively fixed on the PCB mainboard or internal structural parts of the bracelet.

[0091] Parallel drive chains: Six independent micro-drive chains connect the moving and static platforms. This structure often uses a miniature design of the Gough-Stewart platform (or similar parallel mechanism).

[0092] Microactuator: The core driving element of each branch chain is a microactuator. Commonly used ones are: Piezoelectric ceramic stack actuators utilize the inverse piezoelectric effect of piezoelectric materials to generate precise displacement when voltage is applied. They offer fast response speeds (kHz-level), high displacement resolution (nanometers), and high output force density, making them ideal for micro-displacement compensation applications. For example, each actuator can provide a travel range of ±30µm to ±60µm at a drive voltage of 0-100V.

[0093] Electrostatic comb drives or electromagnetic micromotors can also be used as alternatives, but piezoelectric actuators often perform better in terms of overall performance.

[0094] Micro displacement sensor (optional, used for closed-loop control): A micro displacement sensor (such as capacitive or optical) may be integrated on the platform or branch to accurately measure the actual displacement and posture of the dynamic platform, forming a closed-loop position control of the MEMS platform itself and further improving compensation accuracy.

[0095] Complete compensation cycle: Motion Sensing: "Motion Sensing Module (IMU)" on the bracelet "Real-time Detection" Three-dimensional acceleration of the wearer's wrist and three-dimensional angular velocity .sampling The frequency is usually 200 Hz or higher.

[0096] Disturbance prediction and modeling: The wristband’s central processing unit (CPU) receives real-time data from the IMU.

[0097] A pre-established motion-optical perturbation model runs inside the CPU. This model describes how specific wrist movements (represented by IMU data) are converted into relative displacements (translation and rotation) of the optical sensor relative to the skin contact point, as well as changes in contact pressure. This model can be established through experimental calibration (for example, by synchronously recording changes in IMU data and optical signals under controlled motion conditions) or based on a simplified biomechanical model.

[0098] Through the attitude solution algorithm (such as the extended Kalman filter EKF or complementary filter), the CPU estimates the disturbance displacement vector caused by the wrist movement that will act on the optical sensor: ; Compensation instruction calculation: The CPU's goal is to calculate a compensation displacement vector that is equal in magnitude and opposite in direction to the predicted disturbance: Then, through the inverse kinematics model of the MEMS platform, the CPU decomposes the desired six-degree-of-freedom compensation displacement into a sequence of driving voltages (or currents) required to be applied to each of the six piezoelectric actuators. .

[0099] This process may employ advanced control algorithms, such as the aforementioned fractional-order PID (FOPID) controller or model predictive control (MPC), to achieve fast, accurate, and smooth responses.

[0100] Driver execution: The CPU applies the calculated control voltage sequence to the six piezoelectric actuators of the MEMS platform through a digital-to-analog converter (DAC) and a high-voltage drive circuit.

[0101] The piezoelectric actuator precisely expands and contracts according to the voltage signal, driving the dynamic platform (and the optical components fixed on it) to perform tiny displacements and posture adjustments in the opposite direction of the wrist movement.

[0102] Optical stabilization: Ideally, this active counter-motion of the platform can offset, in real time, most of the relative motion of the optical sensor relative to the skin detection area caused by wrist movement. The goal is to keep the optical sensor's detection spot as stable as possible in inertial space (or more critically, relative to the target tissue area at a specific depth under the skin).

[0103] The central processing unit (CPU) and its corresponding storage module are the core components of the doctor's monitoring bracelet. For advanced applications that require precise linkage with the "multi-head infusion intelligent switching control system" (hereinafter referred to as the "switching system"), the CPU's performance, real-time processing capabilities, and the complex algorithms running on it directly determine whether the entire bracelet can provide timely, accurate, and reliable decision-making basis for the switching system.

[0104] The core function of the switching system is to intelligently manage and switch multiple infusion routes based on the patient's real-time, dynamically changing physiological state (for example, selecting, starting, stopping, or adjusting the rate of different pathways such as basal nutrient solution, high-concentration glucose, and electrolyte solution). The CPU provides this intelligent decision-making capability for the switching system by performing the following key tasks: The CPU is responsible for receiving raw photoelectric signals from the multispectral sensor module and acceleration and angular velocity data from the motion sensor module (IMU). These data streams have different characteristics and sampling rates.

[0105] It requires strict time synchronization and alignment of these data, which is the basis for the effectiveness of all subsequent advanced algorithms.

[0106] After receiving the IMU data, the CPU estimates the real-time position and motion trend of the wrist through complex attitude solution algorithms (such as the Extended Kalman Filter - EKF).

[0107] Based on this estimate and a preset motion-disturbance model, the CPU executes a fractional-order PID (FOPID) control law or a similar advanced control algorithm to calculate the precise voltage sequence required to drive the six piezoelectric actuators in the MEMS active motion compensation module.

[0108] This process requires extremely high real-time performance (typically millisecond or even sub-millisecond response) to ensure that the MEMS platform's counter-movement effectively offsets the physical disturbances to the optical sensor caused by wrist motion. This is crucial for the switching system, as the quality of the physical compensation directly impacts the purity of the raw signal input to the vital sign calculation module. A sluggish or inaccurate CPU will cause MEMS compensation to fail, contaminating all subsequent data.

[0109] Even after MEMS physical compensation, subtle motion artifacts and noise may still remain in the original optical signal. The CPU is responsible for executing multi-stage digital filtering and signal enhancement algorithms: Dynamic fractional-order filtering: The CPU calculates the "pollution index (Pl)" in real time based on IMU data and optical signal characteristics, and dynamically adjusts the order of the fractional-order filter accordingly This requires the CPU to have floating-point computing capabilities (or efficient fixed-point analog) to quickly calculate the Siqmoid function and GL differential coefficients.

[0110] Signal reconstruction based on compressed sensing: The CPU executes an iterative optimization algorithm (such as FISTA) to solve the L1 norm minimization problem to separate the physiological signal component and the motion artifact component from the mixed signal. This involves comparing with a pre-trained large dictionary matrix ( ) to perform calculations, which places high demands on the CPU's computing power and memory bandwidth.

[0111] These complex signal processing steps are a prerequisite for providing high-fidelity vital signs to the switching system. If the CPU lacks computing power to complete these calculations in real time, or if algorithm accuracy is sacrificed to simplify calculations, the blood glucose and blood oxygen data ultimately delivered to the switching system will contain more noise and artifacts, directly threatening the accuracy of switching decisions. The CPU uses the pure spectrum signal obtained through the above processing , calculate key physiological parameters.

[0112] Vital sign values ​​( ) Computation: This usually involves running a pre-trained machine learning model such as Support Vector Regression (SVR). The CPU needs to load the parameters of the SVR model (weight vector w, bias vector , kernel function parameters The CPU also calculates other parameters such as SpO2, HR, HRV, and RR simultaneously.

[0113] The CPU not only calculates instantaneous values, but also performs trend analysis on these parameters (especially blood glucose), such as calculating their first-order derivative (rate of change). ) and the second derivative (variation of acceleration This trend information is crucial for the switching system to make proactive decisions. For example, the switching system is not only concerned with whether the current blood glucose level is below the threshold, but also whether the blood glucose level is rapidly decreasing, so that intervention measures (such as switching to pathway B for glucose infusion) can be initiated in advance.

[0114] This is the core function of the CPU that directly serves the switching system. The CPU runs a set of rule engines or decision trees based on clinical guidelines and personalized settings: Compare the real-time calculated vital signs and their trends with the preset thresholds (such as the hypoglycemia alarm threshold , high blood sugar warning threshold , Hyperglycemia intervention threshold , duration ) for comparison.

[0115] According to the comparison results, corresponding recommended instructions are generated, such as: "It is recommended to suspend infusion", "It is recommended to start glucose infusion, initial rate XXml / h", "It is recommended to adjust the infusion rate to YYU / h".

[0116] The CPU also assesses the confidence level of physiological data based on the current state of motion. If it detects intense exercise that could degrade data quality, the CPU can temporarily suspend the linkage suggestion or issue a "low data quality, please note" prompt to avoid making erroneous decisions based on unreliable data from the system backbone.

[0117] The CPU must be able to execute these logical judgments quickly and accurately to ensure that when critical changes in physiological status occur (such as before hypoglycemia occurs), it can promptly issue warnings and operational recommendations to the switching system (or medical staff).

[0118] The CPU is responsible for securely transmitting monitoring data, alarm information, and recommended instructions to the switching system's control unit, medical staff's monitoring terminal, or hospital information system (HIS) via wireless modules such as BLE, Wi-Fi, or NB-IoT. This includes data packaging (e.g., following the HL7FHIR standard), encryption (e.g., AES-128), and management of the communication protocol stack.

[0119] It also needs to manage onboard memory to store configuration information, historical data, algorithm models, etc.

[0120] System overall scheduling and power consumption management: The CPU runs a real-time operating system (RTOS) or an efficient bare-metal scheduler to coordinate all the above concurrent tasks and ensure the real-time and priority of critical tasks (such as MEMS control, vital sign calculation, and alarm judgment).

[0121] At the same time, the CPU needs to cooperate with the power management unit (PMU) to precisely control the power consumption of each module to maximize the battery life of the bracelet while meeting performance requirements.

[0122] The CPU typically uses a microcontroller (MCU) based on the ARM Cortex-M4F or the higher-performance Cortex-M7F core. The F suffix indicates an integrated hardware floating-point unit (FPU), which is crucial for accelerating tasks involving large amounts of floating-point operations, such as attitude resolution, FOPID calculations, fractional-order filtering, and machine learning model inference in MEMS control.

[0123] Clock frequency: Typical values ​​range from 80MHz to 200MHz or even higher. A higher clock frequency can shorten instruction execution time and improve real-time response capabilities.

[0124] DSP instruction set support: The Cortex-M4F / M7F cores typically include DSP extension instructions to accelerate signal processing tasks such as filtering and FFT.

[0125] Memory: RAM (Random Access Memory): At least 256KB to 1MB. Sufficient RAM is required to store real-time sensor data buffers, intermediate algorithm variables, machine learning model runtime parameters, operating system stacks, etc. In particular, the dictionary matrix and iterative processes in compressed sensing, as well as the feature vectors of the SVR model, have certain RAM requirements.

[0126] Flash memory: At least 1MB to 4MB. Used to store firmware code, operating system, pre-trained machine learning model parameters (SVR model, compressed sensing dictionary), configuration information, and a certain amount of historical data.

[0127] Peripheral interface: Rich SPI, I 2 The C interface is used to connect multispectral sensors, IMUs, MEMS driver chips, wireless modules, etc.

[0128] Multiple high-precision ADCs (for auxiliary sensors or internal condition monitoring) and DACs (for driving MEMS).

[0129] DMA (Direct Memory Access) controller, used to efficiently transfer sensor data to memory without CPU intervention, reducing the CPU burden.

[0130] Multiple timers / counters for precise control of light source modulation, sampling synchronization, task scheduling, etc.

[0131] Low power consumption features: Supports multiple low-power modes (Sleep, DeepSleep, Stop, etc.) and can wake up quickly.

[0132] Dynamic voltage and frequency scaling (DVFS) capability.

[0133] Low leakage current process.

[0134] Real-time operating system (RTOS) support: Good compatibility with RTOS (such as FreeRTOS and ZephyrOS) facilitates multi-tasking management, priority scheduling, and resource synchronization, ensuring deterministic execution of critical tasks.

[0135] The wireless communication module detects physiological information, analyzes results, and provides warnings of potential risks in real time, delivering this critical task to the switching system, medical staff, and the hospital information system. In intelligent switching scenarios involving multiple infusions, the real-time, reliable, secure, and interoperable nature of communication directly impacts the efficiency of the switching system's decision-making and patient safety.

[0136] The CPU of the bracelet obtains the patient's real-time vital signs through complex processing. , blood sugar change trend , SpO2, HR and other parameters, and determines the need to adjust the infusion strategy based on the built-in logic.

[0137] The wireless communication module must immediately and accurately send out these physiological data, analysis results, and specific infusion recommendation instructions.

[0138] Recipients may include: Local control unit of the multi-head infusion intelligent switching device: This device is directly connected to the infusion pump head and can perform specific switching actions after receiving instructions (which may require confirmation from medical staff).

[0139] Medical staff's mobile terminal (such as a dedicated PDA, smart phone or display screen of a central monitoring station): Medical staff can see the patient's physiological data, alarms and suggestions issued by the bracelet in real time, and perform manual intervention or confirm system suggestions accordingly.

[0140] Hospital Information System (HIS) / Electronic Medical Record System (EMR): The data monitored by the wristband can be automatically uploaded and integrated into the patient's electronic medical record for subsequent review and analysis or as clinical research data. Telemedicine platform (optional): In some cases, expert physicians can view patient data and guide treatment through the remote platform; Establishment of real-time data stream: Provides continuous, near real-time input of the patient's physiological status to the switching system, enabling it to dynamically adjust the infusion strategy.

[0141] Rapid transmission of emergency alarms and suggestions: When the bracelet detects a critical situation (such as severe hypoglycemia or a sharp drop in SpO2), the communication module must send the alarm information and intervention suggestions with the highest priority and lowest latency to buy precious time for treatment.

[0142] Two-way communication capability (partial scenarios): Downlink configuration and control: Medical staff or switching systems may need to send configuration parameters to the wristband (such as personalized alarm thresholds, algorithm parameter adjustments, and firmware upgrade instructions).

[0143] Instruction confirmation and feedback: After the bracelet sends a suggestion, it may need to receive confirmation information or operation result feedback from the switching device or medical terminal.

[0144] Data recording and tracing: Important physiological data and alarm event records are transmitted to the back-end system to provide a basis for medical quality control and responsibility tracing.

[0145] To meet the different requirements for data transmission rate, power consumption, coverage, and real-time performance in different scenarios, the wireless communication module in this embodiment generally adopts a combination of multiple communication technologies: Bluetooth Low Energy (BLE): Continuous communication with close-range switching devices: If the control unit of a multi-head infusion intelligent switching device is located at the patient's bedside or within close range, BLE is an ideal low-power, continuous connection method for transmitting real-time physiological data streams (such as blood glucose and SpO2 values ​​updated every few seconds or tens of seconds) and low-priority status information.

[0146] Connection with medical staff's near-field mobile terminals: When medical staff are making rounds, their handheld devices can quickly connect to the patient's wristband via BLE to view detailed data, receive instant alarms, or configure parameters.

[0147] Initial configuration and firmware update of the wristband (small data volume).

[0148] Technical features: extremely low power consumption (suitable for battery-powered bracelets), simple pairing, relatively low cost, short-range transmission (typical range 10-100 meters).

[0149] Version selection: BLE 5.0 or higher is preferred because it has a longer transmission distance (up to hundreds of meters line of sight), a higher data transmission rate (up to 2Mbps), and stronger broadcast capabilities (suitable for alarm information dissemination in an unconnected state).

[0150] Value to the switching system: It ensures a continuous, low-power "heartbeat" connection between the wristband and the local switching device, ensuring that the switching device can always obtain the latest physiological data baseline.

[0151] Wi-Fi (IEEE802.11n / ac / ax): Fast upload of high-priority alarms: When a critical alarm event occurs (for example, a wristband predicts impending severe hypoglycemia), if the patient is within the hospital's Wi-Fi network, the alarm information, including detailed context, can be quickly uploaded to the central monitoring system or directly pushed to the designated terminal of the responsible medical staff via Wi-Fi. Wi-Fi has a much higher bandwidth than BLE, allowing for the transmission of richer information.

[0152] Batch historical data upload: For example, at a specific time point (such as every hour, or before the patient is discharged), the detailed physiological data records for a period of time stored in the bracelet are uploaded to the HIS / EMR via Wi-Fi in batches.

[0153] Large data firmware updates (OTA - Over-The-Air).

[0154] Technical features: high transmission rate, medium coverage (depends on AP deployment), and higher power consumption compared to BLE.

[0155] Value to the switching system: It provides a faster and more reliable "highway" for emergency alerts, ensuring that critical information reaches decision-makers quickly. It also facilitates the integration of detailed monitoring logs into hospital information systems, providing data support for efficacy evaluation and algorithm optimization of the switching system.

[0156] Narrowband IoT (NB-IoT) or LTE Cat-M1 (eMTC): Wide-area alarm and data transmission: When patients may be outside of hospital Wi-Fi coverage (e.g., walking within the hospital or being transported), or in home monitoring scenarios, NB-IoT / LTE-M can leverage cellular networks to achieve wide-area connectivity. This is crucial for ensuring timely reporting of critical alarms regardless of location.

[0157] Low-frequency, small-data-volume status reporting: It can serve as a backup communication link even when there is a Wi-Fi or BLE connection.

[0158] Technical features: wide coverage (leveraging existing cellular base stations), strong penetration (suitable for deep indoor locations), relatively low power consumption (optimized for IoT), and low data rate (suitable for small data packet transmission).

[0159] It greatly expands the effective operating range of the wristband and switching system (through the cloud platform), ensuring that even if the patient is not in an "ideal" network environment, key vital signs monitoring and alarm functions are still effective, which is crucial for infusion therapy that requires continuous and seamless monitoring.

[0160] Near Field Communication (NFC): Fast identity authentication and pairing: Medical staff's equipment or switching devices can be "touched" with the wristband through NFC to quickly complete secure pairing or patient identification.

[0161] Convenient parameter reading / writing (very small amounts of data).

[0162] Technical features: extremely close distance (centimeter level), low power consumption, and intuitive operation.

[0163] Intelligent selection and switching logic of communication modules: The wristband's CPU will intelligently select the most appropriate communication method based on the current network environment, data type, event priority, and power consumption strategy. For example, regular data prioritizes BLE, emergency alarms try Wi-Fi or NB-IoT, and large amounts of data wait for Wi-Fi connection, etc. To ensure that the data transmitted by the wristband can be understood by different medical information systems and devices, a standardized medical data exchange format is recommended. HL7 FHIR is an ideal choice. Physiological data (such as blood glucose and SpO2), observation results, device information, alarm events, etc. can all be mapped to FHIR resources.

[0164] When communicating internally or with specific devices, it is also possible to use a more lightweight custom protocol (such as based on JSON or Protocol Buffers), but it should ensure that there is a clear specification and good extensibility.

[0165] Communication security: Patient physiological data is highly sensitive personal privacy information, and strict security measures must be taken: Transport layer encryption: All wireless communication links should use strong encryption protocols.

[0166] BLE: Use AES-CCM encryption (based on LESecureConnections pairing).

[0167] Wi-Fi: Use WPA2 / WPA3 Enterprise-level encryption (such as EAP-TLS).

[0168] NB-IoT / LTE-M: Relies on the encryption mechanisms provided by the cellular network itself (such as encryption at the NAS and AS layers).

[0169] Application-layer encryption (optional enhancement): For particularly sensitive data, end-to-end application-layer encryption can be added to standard transport-layer encryption (e.g., using AES-128 / 256 to encrypt FHIR resource packages or custom data packages).

[0170] Authentication and authorization: Ensure that only authorized devices or users can access wristband data or send commands to the wristband. For example, using certificates, tokens, or secure key exchange mechanisms.

[0171] Data integrity verification: Use a message authentication code (MAC) or digital signature to ensure that data has not been tampered with during transmission.

[0172] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A doctor monitoring wristband for receiving patient infusion information, characterized in that: include: Multispectral sensing module, used to collect the original physiological light intensity signal from the patient's wrist; Motion sensing module, used to collect motion signals; a micro-electromechanical motion compensation module, configured in the multi-spectral sensing module, for actively adjusting the posture of the multi-spectral sensing module according to a motion compensation control instruction; a processor electrically connected to the multispectral sensing module, the motion sensing module, and the micro-electromechanical motion compensation module, the processor being configured to analyze the motion signal and generate the motion compensation control instruction; Using a dynamic filtering algorithm to process the physiological light intensity signal after compensation by the micro-electromechanical motion compensation module to filter out residual noise; Based on a preset signal reconstruction algorithm and light-tissue interaction model, a pure physiological spectral signal is extracted from the filtered physiological light intensity signal and at least one vital sign value is calculated, the vital sign value including but not limited to blood sugar, respiration, blood lipids, and body temperature; generating linkage control instructions for the infusion device based on the comparison result of the vital sign value and the preset threshold value; The communication module is electrically connected to the processor and is used to wirelessly transmit the vital sign values ​​to the medical monitoring terminal and send the linkage control instructions.

2. A doctor monitoring wristband for receiving patient infusion information according to claim 1, characterized in that: The multispectral sensing module includes a quantum dot enhanced optical sensor array, which is configured with light sources of at least three different wavelengths and corresponding photoelectric detection units.

3. A doctor monitoring wristband for receiving patient infusion information according to claim 1, characterized in that: The micro-electromechanical motion compensation module is an active stabilization platform based on a six-degree-of-freedom parallel mechanism, and the motion compensation control instruction drives the active stabilization platform to perform sub-micron displacement compensation.

4. A doctor monitoring wristband for receiving patient infusion information according to claim 1, characterized in that: The dynamic filtering algorithm adopted by the processor is a fractional-order filtering algorithm, wherein the filtering order of the fractional-order filtering algorithm is adaptively adjusted according to the real-time signal-to-noise ratio of the physiological light intensity signal after motion compensation.

5. A doctor monitoring wristband for receiving patient infusion information according to claim 1, characterized in that: The filter order of the fractional filter algorithm The adaptive adjustment method is: When the real-time signal-to-noise ratio is higher than a first preset signal-to-noise ratio threshold, the filter order is adjusted to a first preset order range, where the first preset order range corresponds to or is adjacent to an integer order, so as to facilitate retaining low-frequency physiological signal characteristics; When the real-time signal-to-noise ratio is lower than a second preset signal-to-noise ratio threshold, the filter order is adjusted to a second preset order range, which corresponds to a preset lower fractional order value interval to enhance the suppression of high-frequency noise.

6. A doctor monitoring wristband for receiving patient infusion information according to claim 1, characterized in that: The signal reconstruction algorithm adopted by the processor is a compressed sensing reconstruction algorithm, which adopts a joint dictionary including physiological signal characteristic basis functions and motion artifact characteristic basis functions to reconstruct the pure physiological spectrum signal by solving a sparse optimization problem.

7. A doctor monitoring wristband for receiving patient infusion information according to claim 1, characterized in that: The light-tissue interaction model adopted by the processor is a fractional-order light transmission model, which expresses the absorption and scattering characteristics of light by tissue in the form of a fractional-order differential equation related to the concentration of the vital sign.

8. A doctor monitoring wristband for receiving patient infusion information according to claim 1, characterized in that: The communication module uses the low-power Bluetooth protocol to perform periodic data transmission when the vital sign value is within the normal range, and switches to the long-distance low-energy wide area network protocol to perform immediate alarm data transmission when the vital sign value triggers an early warning condition.

9. A doctor monitoring wristband for receiving patient infusion information according to claim 1, characterized in that: The linkage control instructions generated by the processor include: when the vital sign value continues to be higher than the first preset vital sign value threshold for a preset time period, generating an instruction for reducing the infusion rate; when the vital sign value is lower than the second preset vital sign value threshold, generating an instruction for pausing the infusion and triggering an alarm.

10. A monitoring method for receiving patient infusion information, according to a doctor monitoring bracelet for receiving patient infusion information according to any one of claims 1 to 9, characterized in that: The following steps are involved: Collecting original physiological light intensity signals from the patient's wrist through the multispectral sensing module and collecting motion signals through the motion sensing module; The processor analyzes the motion signal and generates a motion compensation control instruction; The micro-electromechanical motion compensation module actively adjusts the posture of the multi-spectral sensing module according to the motion compensation control instruction to obtain a physiological light intensity signal after motion compensation; The processor processes the motion-compensated physiological light intensity signal using the dynamic filtering algorithm to filter out residual noise and obtain a filtered physiological light intensity signal; The processor extracts a pure physiological spectrum signal from the filtered physiological light intensity signal based on the preset signal reconstruction algorithm and the light-tissue interaction model and calculates at least one vital sign value, the vital sign value including but not limited to blood sugar, respiration, blood lipids, and body temperature; The communication module wirelessly transmits the vital sign values ​​to the medical monitoring terminal; The processor generates a linkage control instruction for the infusion device based on the comparison result of the vital sign value and the preset threshold value, and sends it to the infusion device or the medical monitoring terminal through the communication module.

Images