Wireless detection device and method for enhancing early warning of CT infusion seepage and swelling

Abstract

The invention relates to a wireless detection device and method for enhancing early warning of CT infusion seepage and swelling, the wireless detection device comprises a laser-induced graphene strain sensor, a microcontroller and a mobile phone end upper computer, and the microcontroller is used for keeping a medical non-woven fabric adhesive tape attached to the skin in a shape-preserving manner; the laser-induced graphene strain sensor is obtained by transferring laser-induced graphene on a polyimide film through PDMS, the laser-induced graphene strain sensor is attached to the upper portion of an infusion needle opening through a medical non-woven fabric adhesive tape, the sensor is perpendicular to a blood vessel, and wires at the two ends of the strain sensor are connected with the microcontroller. The mobile phone end upper computer comprises a data visualization module, a data storage module and a user interaction module, and can receive the Bluetooth data sent by the microcontroller, convert the data into real-time waveforms and send out early warning information when infusion seepage occurs. According to the invention, the seepage bulge phenomenon occurring during enhanced CT infusion can be found in time, and monitoring and early warning can be carried out through the wireless terminal.

Description

Technical Field

[0001] The present invention belongs to the field of flexible sensor technology and medical diagnosis technology, and more particularly, relates to a wireless detection device and method for enhancing early warning of CT infusion exudate bulge. Background Art

[0002] Flexible sensors are a type of sensor. In addition to the basic function of converting non-electrical signals into electrical signals, they are also flexible, which gives them broad development prospects in the fields of wearable devices, medical diagnostic technology, and flexible electronics.

[0003] During enhanced CT scans, an iodine contrast agent is injected into the body, a process that typically takes only about 30 seconds. Due to the short duration of the injection, edema and swelling are common. Once this occurs, the contrast agent leaks into the subcutaneous tissue, potentially causing local pain, swelling, and inflammation, and in severe cases, even soft tissue necrosis. Furthermore, failure of the contrast agent to fully enter the vascular system can directly affect image clarity, leading to delayed diagnosis and the need for re-injection, increasing patient pain and medical costs. This not only impairs imaging quality but also delays the optimal timing for diagnosis and treatment. Monitoring this infusion process to promptly warn of edema and swelling and alert the operating physician is essential.

[0004] CN115591052A discloses an infusion needle bulging detection and call system based on intelligent infusion information. Its overall structure includes an infusion stand, an infusion pump, a weighing monitoring module for monitoring the weight of the infusion bag, a dripping rate monitoring module for monitoring the dripping rate of the infusion pump, an infusion patch for monitoring the pressure at the needle tip, and a controller. Its basic principle is to monitor the weight of the infusion bag through the weighing monitoring module to determine the infusion progress, and to monitor the dripping rate of the infusion pump through the dripping rate monitoring module. Combined with the pressure at the needle tip monitored by the infusion patch, the controller can give an alarm judgment based on whether the dripping rate and pressure change, thereby timely detecting the needle bulging phenomenon. The system adopts a rigid structure as a whole, which can easily make the subject feel uncomfortable during monitoring. The system uses the pressure at the needle tip to monitor infusion leakage, which will lead to insufficient detection accuracy and false alarms or missed alarms.

[0005] CN213335954U discloses a laser-induced graphene flexible strain-temperature dual-parameter sensor. It comprises a flexible, stretchable substrate and a graphene conductive layer fixed to the flexible, stretchable substrate via laser induction. The graphene conductive layer includes a central graphene layer and connecting electrodes at both ends of the graphene layer. The connecting electrodes are provided with a conductive coating and are connected and fixed with wires. The straight line containing the conductive graphene constitutes the sensor body. The width of the conductive graphene is 1.29 mm ± 1 mm, and the length is 17.62 mm ± 5 mm. The two connecting electrodes and the single graphene line form a bone-like structure. This sensor can reflect both strain and temperature changes. During the sensor's fabrication process, insufficient crosslinking of PDMS and LIG resulted in a failure to form a uniform LIG / PDMS composite porous foam, which compromised its sensitivity.

[0006] Furthermore, the sensing layer is not encapsulated for the second time, which may cause the sensing material to fall off in practical applications.

[0007] During enhanced CT scans, patients place their arms overhead. After inserting the IV needle, the operating physician often remotely controls the infusion from the CT operating room. During this process, neither the patient nor the operating physician can detect the specific progress of the infusion. If edema or swelling occurs and the physician fails to detect it in time, the enhanced CT scan will lose its effectiveness, significantly impacting diagnostic accuracy and the physician's work schedule. Current systems for detecting edema or swelling suffer from low sensitivity, the inability to monitor wirelessly, and excessive size, making them unsuitable for monitoring the infusion process during enhanced CT scans. Summary of the Invention

[0008] In order to solve the above technical problems and address the problems existing in the background technology, the present invention provides a wireless detection device and method for enhanced CT infusion exudate bulging warning. The wireless detection device provided by the present invention has high sensitivity, can realize wireless signal transmission, and can be remotely monitored. It is suitable for monitoring the process of contrast agent injection in enhanced CT.

[0009] In order to achieve the above-mentioned purpose, the present invention provides the following technical solutions: A wireless detection device for enhancing CT infusion exudate and bulging warning includes a laser-induced graphene strain sensor, a microcontroller, and a mobile phone host computer. The laser-induced graphene strain sensor includes a PDMS substrate, a LIG / PDMS composite porous foam material electrode, and connecting wires. The microcontroller includes an MCU main control circuit, a power circuit, a strain signal acquisition circuit, a filtering circuit, a Bluetooth module, a micro lithium battery, and a silicone shell. The mobile phone host computer is equipped with a data visualization function module, a data storage function module, and a user interaction function module. A wireless detection device for early warning of fluid leakage and bulging during enhanced CT infusion uses a highly sensitive laser-induced graphene strain sensor as the front-end sensing unit to detect skin surface deformation during enhanced CT infusion. A microcontroller collects, analyzes, and transmits sensor data, which is then received and visualized by a mobile phone host computer. The interactive function provides alerts when infusion anomalies occur. The design and fabrication of laser-induced graphene strain sensors involves the following steps: (1) The sensor structure was designed using the 2D modeling software AutoCAD, and the pattern was imported into a CO2 laser to ablate the LIG pattern on the surface of the PI film; (2) PDMS was evenly coated on the LIG surface and the PDMS was cured at 80°C to form a LIG / PDMS composite porous foam material; (3) Peeling off the PDMS from the PI film to obtain a strain sensor body with PDMS as the substrate and LIG / PDMS composite porous foam material electrodes as the sensing layer; (4) Use highly conductive silver glue to connect wires at both ends of the LIG / PDMS composite porous foam electrode, and then use PDMS as a flexible encapsulation layer; The microcontroller includes an MCU main control circuit, a power supply circuit, a strain signal acquisition circuit, a filtering circuit, a Bluetooth module, a micro lithium battery, and a silicone shell. The MCU main control circuit includes a main control chip, a clock circuit, a reset circuit, power management, a peripheral interface, and a storage module. The power supply circuit includes a voltage conversion module, a voltage stabilizing circuit, and a decoupling capacitor. The strain signal acquisition circuit includes a strain sensor and a voltage divider circuit. The filtering circuit uses a first-order low-pass filter. The Bluetooth module includes a decoupling capacitor and a peripheral power supply circuit. The micro lithium battery has an output voltage of 3.7V, which is used to power all circuits. The silicone shell is used to protect the electronic components of the microcontroller.

[0010] The main control chip of the MCU main control circuit adopts Cortex-M4 as the core microcontroller processor, and its model is STM32F103C8T6; Furthermore, the PC14 and PC15 pins of the main control chip U1 (STM32F103C8T6) are connected to the 32.768kHz low-speed crystal oscillator circuit, and the OSC_IN and OSC_OUT pins are connected to the 8MHz high-speed crystal oscillator circuit to provide clock signals for the system. The VDD_1, VDD_2, VDD_3, and VDD_4 power supply pins of U1 are grounded through decoupling capacitors C5, C6, C7, and C8, and the VSS_1, VSS_2, VSS_3, and VSS_4 pins are directly grounded. VBAT is connected to the backup power supply through the decoupling capacitor C9, VDDA is grounded through the filter resistor R7 and the decoupling capacitor C10, and VSSA Directly grounded, U1's NRST pin is connected to 3.3V through a pull-up resistor R1, and is grounded through capacitor C11 to form a reset circuit. The BOOT0 pin is grounded through resistor R2 and set to Flash boot mode. PA13 and PA14 are connected to SWDI and SWCLK respectively, and are connected to the SWD burning interface for program burning and debugging. PD3, PD4, and PD5 are connected to the status indicator LED through current-limiting resistors R3, R4, and R5 respectively. The negative pole of the LED is grounded, and PB2 is grounded through resistor R6 to avoid floating interference. The power supply voltage regulator module U2 (RT9193) is responsible for converting 5V to 3.3V power supply, and uses filter capacitors at the input and output ends to improve the voltage regulation effect.

[0011] Furthermore, the power supply circuit includes: using the RT9193-33GB model power conversion chip U2, and using a rechargeable lithium battery with an output voltage of 3.7V and a battery capacity of 300mAh for power supply. The power conversion chip U2, VIN is connected to the battery power supply, the input side is grounded through the filter capacitors C6 and C8, VOUT outputs VCC3V3 power supply, and the output side is grounded through the filter capacitors C5 and C7; the chip EN and BP are grounded through the capacitor C8.

[0012] Furthermore, the strain signal acquisition circuit includes a fixed resistor R11, a precise resistor R12, a model G6K-2F-Y-5V relay, a model SS8050 transistor, a pull-down resistor R10, and a strain sensor to form a dynamic calibration voltage divider circuit. The voltage reference chip REF3030AIDBZR provides a stable 3V voltage for the ADC module, wherein the relay is in normally open mode, pin 1 is connected to VCC, pin 8 is connected to the PA8 pin of the MCU through the transistor Q2 and the pull-down resistor R10, pin 2 is connected to VCC3V3 and the fixed resistor R11, the sensor resistor R_sensor and the precise resistor R12 are connected in parallel, and the two ends are respectively connected to the relay pin 3 and PS1, the voltage reference chip U8, the input port IN is connected to VCC3V3 through the decoupling capacitor C16, the output port OUT is connected to the ADC module power supply port VDDA of the MCU, and the GAD port is grounded through the decoupling capacitor C17.

[0013] Furthermore, the hardware filtering circuit is a three-stage cascade filtering circuit, including a second-order active low-pass filter, a dual-T active notch filter circuit, and a first-order RC anti-aliasing filter. The 50HZ active notch filter circuit consists of two op amps and a dual-T notch filter. Negative feedback is introduced to improve the frequency selection effect. Op amp U11.1 not only provides the gain of the feedback loop, but also serves to isolate the dual-T network. The cutoff frequency of the second-order active low-pass filter is 106Hz, the center frequency of the dual-T active notch filter circuit is 50Hz, and the cutoff frequency of the first-order RC anti-aliasing filter is 150Hz. The input signal of the second-order active low-pass filter is the output signal PS1 of the strain acquisition circuit. The output signal of this circuit is output to the next-stage dual-T active notch filter circuit. The output signal of the dual-T active notch filter circuit is the input signal of the next-stage first-order RC anti-aliasing filter. The output signal of the RC anti-aliasing filter is the output signal of the entire hardware filtering circuit and is output to the ADC input pin PA0.

[0014] Furthermore, the Bluetooth module includes a Bluetooth chip RF-BM-BG22C3 and the peripheral circuits required for the Bluetooth module. Among them, pin PA06 of chip U5 (model RF-BM-BG22C3) is connected to PA9 of the MCU, pin 2 (PA05) is connected to PA10 of the MCU, pin 3 (PA04) is connected to PA11 of the MCU, pin 4 (PA03) is connected to PA12 of the MCU, pin 5 (PA02) is connected to SWDIO of the MCU, and pin 6 (PA01) is connected to SWDCLK of the MCU; power pin 14 (VCC) is connected to the VCC power supply, grounded and filtered through capacitors C25 (0.1μF) and C26 (100nF), and pin 15 (GND) is directly grounded; function pin 10 (PB00) is connected to switch SW3 (TS-1088R-02026) and is pulled up to VCC through resistor R21 (10K).

[0015] Furthermore, the strain signal acquisition circuit controls the opening and closing of the relay through the MCU to achieve the initial voltage calibration of the strain sensor. The specific implementation process is as follows: Power-on self-calibration: MCU controls the relay to close, R12 is connected in parallel with R_sensor, and the equivalent resistance is: , the reference voltage is provided by the chip REF3030, and the voltage division output at this time can be calculated by the ADC module, so the actual initial resistance of the sensor can be obtained as , Normal measurement: After the initial resistance calibration is completed, the MCU controls the relay to disconnect, and the system uses the calibrated R1 as the reference value.

[0016] Furthermore, the hardware filtering circuit takes into account that LIG-based strain sensors have excellent sensitivity but are also susceptible to external interference and noise. In this hardware filtering circuit, a second-order active low-pass filter is used to filter noise above 100 Hz. A dual-T active notch filter filter circuit is used to filter out the 50 Hz power frequency interference in the CT room. A first-order RC anti-aliasing filter with a cutoff frequency of 150 Hz is used to suppress high-frequency aliasing noise and match the ADC sampling rate.

[0017] Furthermore, the circuit board of the microcontroller is a flexible printed circuit board (FPCB), and the microcontroller itself is covered by a soft silicone shell, which can achieve a small degree of bending.

[0018] Furthermore, the size structure of the laser-induced graphene strain sensor is a barbell shape that is thin in the middle and wide at both ends. The thinner part is the sensing part of the sensor, with dimensions of width: 1.2 mm ± 0.5 mm and length: 15 mm ± 5 mm; the wider part is the wire connection part, with dimensions of width: 3 mm ± 0.5 mm and length: 2 mm ± 1 mm.

[0019] Furthermore, during the fabrication of the laser-induced graphene strain sensor, components A and B of the PDMS solution were mixed in a ratio of 10.6:1, stirred evenly, and then evenly coated on the LIG surface of the PI membrane. The mixture was then placed in a drying oven and dried and cured at 70-100°C for 80-150 minutes. After the PDMS was completely cured, a LIG / PDMS composite porous foam material was obtained.

[0020] Furthermore, its strain-sensitive material is a porous foam-like LIG / PDMS composite material, which has excellent strain sensitivity. Under a stretching of less than 30%, the sensitivity GF1 is as high as 240.16, and under a stretching of 30%-40%, the sensitivity is as high as 2700.

[0021] Furthermore, after connecting the wires at both ends of the electrode, uncured PDMS is poured on the surface of the LIG / PDMS composite porous foam electrode, and the encapsulation layer is formed after the PDMS is cured.

[0022] Furthermore, when coating the PDMS solution, the thickness of the PDMS film is controlled by controlling the PDMS mass per unit area. The thickness of the PDMS film affects the transfer quality of graphene and the detection range of the sensor. The specific parameters are: 0.064-0.08g / cm 2 The thickness of the obtained PDMS film is 300µm-500µm.

[0023] Furthermore, the device includes a microcontroller body, a flexible cable connection line, and a strain sensor body. The microcontroller body includes a flexible circuit board (FPCB), a 200mAh rechargeable lithium battery, and a soft silicone shell. The ADC pin of the microcontroller extends from the inside of the shell through a flexible cable to connect to the sensor. The positive and negative power lines of the battery are connected to the power interface on the circuit board, and two additional power lines are extended to the outside of the shell through the Type_c adapter to charge the battery. The specific structure of the soft silicone shell is a flexible wristband-type shell, including a shell body and wristbands extending outward at both ends. The wristband is provided with a buckle for adjusting the tightness. The strain sensor body includes a strain sensor and medical non-woven tape, wherein the medical non-woven tape can be cut according to actual needs.

[0024] Furthermore, the method comprises the following steps: Step 1: Attach and fix the strain sensor. When performing enhanced CT, use medical non-woven tape to apply the strain sensor flatly to the skin 3-5 cm above the patient's infusion port. The characteristics are: the stretching direction of the strain sensor should be perpendicular to the blood vessel, and the stretchable direction of the medical non-woven tape should be consistent with the stretching direction of the strain sensor. Step 2: Turn on the power switch of the microcontroller. When the power circuit is turned on, the battery voltage is converted to 3.3V. The 3.3V voltage is used to power the MCU main control circuit, strain signal acquisition circuit, hardware filter circuit and Bluetooth module. Step 3: The strain signal acquisition circuit filters the data collected by the laser-induced graphene strain sensor through the hardware filter circuit and transmits it to the MCU main control circuit. The MCU main control circuit processes this data and sends it to the mobile phone host computer through the Bluetooth module. Step 4: Open the mobile host computer, set the specified UUID, find the name of the microcontroller's Bluetooth module and connect it. The mobile host computer can convert the data sent by the slave computer into a waveform, and when the patient has leakage or bulging during the infusion process, the mobile phone APP will issue an early warning to prompt the doctor.

[0025] Furthermore, an algorithm is used to detect infusion bulging during enhanced CT. The algorithm acquires preprocessed strain signals at a 100Hz sampling rate, calculates the baseline mean (μ) and standard deviation (σ) in real time using a 60-second sliding window, and updates the baseline every 10 minutes using exponential weighting (α=0.1) to eliminate environmental drift. A three-level dynamic threshold is used for judgment: when the instantaneous rate of change of the strain value ΔS>3σ or the sliding variance>2σ², an early warning is triggered and the system switches to 100Hz high-frequency sampling; if the strain value exceeds μ+5σ (corresponding to 0.5-1mL of exudate) for 3 seconds, a primary alarm is activated; and when the strain value exceeds μ+8σ (approximately 2mL of exudate) or increases at a positive slope for 10 consecutive seconds (dS / dt>0.1σ / s), a high-level alarm is triggered.

[0026] Compared with the existing infusion seepage bulge detection system, the beneficial effects of the present invention are: The sensing unit of the present invention is a laser-induced graphene strain sensor, which directly forms graphene on a polyimide film by ablation using a CO2 laser, and then transfers the graphene to a flexible substrate by transfer printing to form a LIG-PDMS composite material.

[0027] The sensor has a simple structure and does not require complex chemical processes. Its preparation process is green, safe, pollution-free, simple to operate, and low-cost. A single chip takes only 15 seconds to produce at a cost of just 1.5 yuan. It can be used as a medical consumable, eliminating the need for disinfection and sterilization, saving time. The sensor has high sensitivity and a fast response time, capable of detecting small strains below 1%. It maintains its sensing performance at strains of 45%, and performs well below 30%.

[0028] The sensor also has excellent structural stability. Cyclic tensile tests have shown that the sensor can maintain its strain performance after 18,000 cycles of stretching. These excellent characteristics make the sensor possible for detecting infusion seepage and bulging in enhanced CT.

[0029] The sensor of this invention is secured to the patient using a medical stretchable nonwoven tape. The tape stretches as the skin surface deforms due to infusion fluid, causing the strain sensor to strain and respond. When applying the sensor, care should be taken to ensure that the sensor strains perpendicular to the blood vessels and that the medical stretchable tape stretches in the same direction as the strain sensor. The application site is 3-5 cm above the infusion site.

[0030] This invention provides a wireless detection device for enhanced CT scans to warn of infusion leakage and bulging. The data acquisition and analysis unit is a FPCB (Flat Printed Circuit Board) containing an MCU main control circuit, power supply circuit, strain signal acquisition circuit, filter circuit, Bluetooth module, miniature lithium battery, and silicone housing. The FPCB measures 27mm by 49mm, boasting a compact structure and low power consumption. It can operate stably for 8 hours when powered by a lithium battery. Encapsulated in a soft silicone housing, it adheres conformably to the skin.

[0031] It can realize the function of wireless data transmission, remote monitoring and judgment, with a judgment success rate of up to 94.8%, meeting the working conditions of the operating doctor: the operating room and the CT room are separated by a lead door, and the doctor cannot understand the specific situation of the patient during infusion in real time. The infusion situation is detected by a high-sensitivity LIG strain sensor, and the data is sent to the CT operating room via a Bluetooth module, which is convenient for the doctor's operation and timely detection of infusion leakage. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic diagram of the overall structure of the present invention; Figure 2 Schematic diagram of the structure of the laser-induced graphene strain sensor of the present invention; Figure 3 is the resistance change of the sensor under 10% strain after multiple tests; Figure 4 is the response of the sensor under 1% strain; Figure 5 This is the schematic diagram of the MCU main control circuit; Figure 6 It is a strain signal acquisition circuit; Figure 7 It is a hardware filter circuit; Figure 8 This is the Bluetooth module circuit. DETAILED DESCRIPTION

[0033] To make the objectives, features, and advantages of the present invention more readily apparent, the following detailed description of the present invention is provided with reference to the accompanying drawings. The accompanying drawings illustrate several embodiments of the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein.

[0034] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0035] A wireless detection device for enhancing CT infusion exudate and bulging warning includes a laser-induced graphene strain sensor, a microcontroller, and a mobile phone host computer. The laser-induced graphene strain sensor includes a PDMS substrate, a LIG / PDMS composite porous foam material electrode, and connecting wires. The microcontroller includes an MCU main control circuit, a power circuit, a strain signal acquisition circuit, a filtering circuit, a Bluetooth module, a micro lithium battery, and a silicone shell. The mobile phone host computer is equipped with a data visualization function module, a data storage function module, and a user interaction function module. A wireless detection device for early warning of fluid leakage and bulging during enhanced CT infusion uses a highly sensitive laser-induced graphene strain sensor as the front-end sensing unit to detect skin surface deformation during enhanced CT infusion. A microcontroller collects, analyzes, and transmits sensor data, which is then received and visualized by a mobile phone host computer. The interactive function provides alerts when infusion anomalies occur. The design and fabrication of laser-induced graphene strain sensors involves the following steps: (1) The sensor structure was designed using the 2D modeling software AutoCAD, and the pattern was imported into a CO2 laser to ablate the LIG pattern on the surface of the PI film; (2) PDMS was evenly coated on the LIG surface and the PDMS was cured at 80°C to form a LIG / PDMS composite porous foam material; (3) Peeling off the PDMS from the PI film to obtain a strain sensor body with PDMS as the substrate and LIG / PDMS composite porous foam material electrodes as the sensing layer; (4) Use highly conductive silver glue to connect wires at both ends of the LIG / PDMS composite porous foam electrode, and then use PDMS as a flexible encapsulation layer; The microcontroller includes an MCU main control circuit, a power supply circuit, a strain signal acquisition circuit, a filtering circuit, a Bluetooth module, a micro lithium battery, and a silicone shell. The MCU main control circuit includes a main control chip, a clock circuit, a reset circuit, power management, a peripheral interface, and a storage module. The power supply circuit includes a voltage conversion module, a voltage stabilizing circuit, and a decoupling capacitor. The strain signal acquisition circuit includes a strain sensor and a voltage divider circuit. The filtering circuit uses a first-order low-pass filter. The Bluetooth module includes a decoupling capacitor and a peripheral power supply circuit. The micro lithium battery has an output voltage of 3.7V, which is used to power all circuits. The silicone shell is used to protect the electronic components of the microcontroller.

[0036] The main control chip of the MCU main control circuit adopts Cortex-M4 as the core microcontroller processor, and its model is STM32F103C8T6; Preferably, the PC14 and PC15 pins of the main control chip U1 (STM32F103C8T6) are connected to a 32.768kHz low-speed crystal oscillator circuit, and the OSC_IN and OSC_OUT pins are connected to an 8MHz high-speed crystal oscillator circuit to provide a clock signal for the system. The VDD_1, VDD_2, VDD_3, and VDD_4 power supply pins of U1 are grounded through decoupling capacitors C5, C6, C7, and C8, and the VSS_1, VSS_2, VSS_3, and VSS_4 pins are directly grounded. VBAT is connected to the backup power supply through a decoupling capacitor C9, VDDA is grounded through a filter resistor R7 and a decoupling capacitor C10, and VSSA is grounded. Directly grounded, U1's NRST pin is connected to 3.3V through a pull-up resistor R1, and is grounded through capacitor C11 to form a reset circuit. The BOOT0 pin is grounded through resistor R2 and set to Flash boot mode. PA13 and PA14 are connected to SWDI and SWCLK respectively, and are connected to the SWD burning interface for program burning and debugging. PD3, PD4, and PD5 are connected to the status indicator LED through current-limiting resistors R3, R4, and R5 respectively. The negative pole of the LED is grounded, and PB2 is grounded through resistor R6 to avoid floating interference. The power supply voltage regulator module U2 (RT9193) is responsible for converting 5V to 3.3V power supply, and uses filter capacitors at the input and output ends to improve the voltage regulation effect.

[0037] Preferably, the power supply circuit includes: a power conversion chip U2 of the RT9193-33GB model, a rechargeable lithium battery with an output voltage of 3.7V and a battery capacity of 300mAh for power supply, the power conversion chip U2, VIN is connected to the battery power supply, the input side is grounded through the filter capacitors C6 and C8, VOUT outputs VCC3V3 power supply, and the output side is grounded through the filter capacitors C5 and C7; the chips EN and BP are grounded through the capacitor C8.

[0038] Preferably, the strain signal acquisition circuit includes a fixed resistor R11, a precise resistor R12, a model G6K-2F-Y-5V relay, a model SS8050 transistor, a pull-down resistor R10, and a strain sensor to form a dynamic calibration voltage divider circuit, and the voltage reference chip REF3030AIDBZR provides a stable 3V voltage for the ADC module, wherein the relay is in normally open mode, pin 1 is connected to VCC, pin 8 is connected to the PA8 pin of the MCU through the transistor Q2 and the pull-down resistor R10, pin 2 is connected to VCC3V3 and the fixed resistor R11, the sensor resistor R_sensor and the precise resistor R12 are connected in parallel, and the two ends are respectively connected to the relay pin 3 and PS1, the voltage reference chip U8, the input port IN is connected to VCC3V3 through the decoupling capacitor C16, the output port OUT is connected to the ADC module power supply port VDDA of the MCU, and the GAD port is grounded through the decoupling capacitor C17.

[0039] Preferably, the hardware filtering circuit is a three-stage cascade filtering circuit, including a second-order active low-pass filter, a dual-T active notch filter circuit, and a first-order RC anti-aliasing filter. The 50HZ active notch filter circuit is composed of two operational amplifiers and a dual-T notch filter. Negative feedback is introduced to improve the frequency selection effect. The operational amplifier U11.1 not only provides the gain of the feedback loop, but also isolates the dual-T network. The cutoff frequency of the second-order active low-pass filter is 106Hz, the center frequency of the dual-T active notch filter circuit is 50Hz, and the cutoff frequency of the first-order RC anti-aliasing filter is 150Hz. The input signal of the second-order active low-pass filter is the output signal PS1 of the strain acquisition circuit. The output signal of the circuit is output to the next-stage dual-T active notch filter circuit. The output signal of the dual-T active notch filter circuit is the input signal of the next-stage first-order RC anti-aliasing filter. The output signal of the RC anti-aliasing filter is the output signal of the overall hardware filtering circuit and is output to the ADC input pin PA0.

[0040] Preferably, the Bluetooth module includes a Bluetooth chip RF-BM-BG22C3 and peripheral circuits required for the Bluetooth module, wherein pin PA06 of chip U5 (model RF-BM-BG22C3) is connected to PA9 of the MCU, pin 2 (PA05) is connected to PA10 of the MCU, pin 3 (PA04) is connected to PA11 of the MCU, pin 4 (PA03) is connected to PA12 of the MCU, pin 5 (PA02) is connected to SWDIO of the MCU, and pin 6 (PA01) is connected to SWDCLK of the MCU; power pin 14 (VCC) is connected to the VCC power supply, grounded and filtered through capacitors C25 (0.1μF) and C26 (100nF), and pin 15 (GND) is directly grounded; function pin 10 (PB00) is connected to switch SW3 (TS-1088R-02026) and is pulled up to VCC through resistor R21 (10K).

[0041] Preferably, the strain signal acquisition circuit controls the opening and closing of the relay through the MCU to realize the initial voltage calibration of the strain sensor. The specific implementation process is as follows: Power-on self-calibration: MCU controls the relay to close, R12 is connected in parallel with R_sensor, and the equivalent resistance is: , the reference voltage is provided by the chip REF3030, and the voltage division output at this time can be calculated by the ADC module, so the actual initial resistance of the sensor can be obtained as , Normal measurement: After the initial resistance calibration is completed, the MCU controls the relay to disconnect, and the system uses the calibrated R1 as the reference value.

[0042] Preferably, the hardware filtering circuit uses a second-order active low-pass filter to filter out noise above 100 Hz, taking into account that LIG-based strain sensors have excellent sensitivity but are also susceptible to external interference and noise. A dual-T active notch filter circuit is used to filter out 50 Hz power frequency interference in the CT room. A first-order RC anti-aliasing filter has a cutoff frequency of 150 Hz to suppress high-frequency aliasing noise and match the ADC sampling rate.

[0043] Preferably, the circuit board of the microcontroller is a flexible printed circuit board (FPCB), and the microcontroller itself is enclosed by a soft silicone shell, which can achieve a small degree of bending.

[0044] Preferably, the size structure of the laser-induced graphene strain sensor is a barbell shape that is thin in the middle and wide at both ends. The thinner part is the sensing part of the sensor, with dimensions of width: 1.2 mm ± 0.5 mm and length: 15 mm ± 5 mm; the wider part is the wire connection part, with dimensions of width: 3 mm ± 0.5 mm and length: 2 mm ± 1 mm.

[0045] Preferably, during the fabrication of the laser-induced graphene strain sensor, component A and component B of the PDMS solution are mixed in a ratio of 10.6:1, stirred evenly, and then uniformly coated on the LIG surface on the PI film. The mixture is then placed in a drying oven and dried and cured at 70-100°C for 80-150 minutes. After the PDMS is completely cured, a LIG / PDMS composite porous foam material is obtained.

[0046] Preferably, the strain-sensitive material is a porous foam-like LIG / PDMS composite material, which has excellent strain sensitivity. Under a stretching of less than 30%, the sensitivity GF1 is as high as 240.16, and under a stretching of 30%-40%, the sensitivity is as high as 2700.

[0047] Preferably, after connecting the wires at both ends of the electrode, uncured PDMS is poured onto the surface of the LIG / PDMS composite porous foam electrode, and the encapsulation layer is formed after the PDMS is cured.

[0048] Preferably, when coating the PDMS solution, the thickness of the PDMS film is controlled by controlling the PDMS mass per unit area. The thickness of the PDMS film affects the transfer quality of the graphene and the detection range of the sensor. The specific parameters are: 0.064-0.08g / cm 2 The thickness of the obtained PDMS film is 300µm-500µm.

[0049] Preferably, the device includes a microcontroller body, a flexible cable connection line, and a strain sensor body. The microcontroller body includes a flexible circuit board (FPCB), a 200mAh rechargeable lithium battery and a soft silicone shell. The ADC pin of the microcontroller extends from the inside of the shell through a flexible cable to connect to the sensor. The positive and negative power lines of the battery are connected to the power interface on the circuit board, and two additional power lines are extended to the outside of the shell through the Type_c adapter to charge the battery. The specific structure of the soft silicone shell is a flexible wristband-type shell, including a shell body and wristbands extending outward at both ends. The wristband is provided with a buckle for adjusting the tightness. The strain sensor body includes a strain sensor and medical non-woven tape, wherein the medical non-woven tape can be cut according to actual needs.

[0050] Preferably, the method comprises the following steps: Step 1: Attach and fix the strain sensor. When performing enhanced CT, use medical non-woven tape to apply the strain sensor flatly to the skin 3-5 cm above the patient's infusion port. The characteristics are: the stretching direction of the strain sensor should be perpendicular to the blood vessel, and the stretchable direction of the medical non-woven tape should be consistent with the stretching direction of the strain sensor. Step 2: Turn on the power switch of the microcontroller. When the power circuit is turned on, the battery voltage is converted to 3.3V. The 3.3V voltage is used to power the MCU main control circuit, strain signal acquisition circuit, hardware filter circuit and Bluetooth module. Step 3: The strain signal acquisition circuit filters the data collected by the laser-induced graphene strain sensor through the hardware filter circuit and transmits it to the MCU main control circuit. The MCU main control circuit processes this data and sends it to the mobile phone host computer through the Bluetooth module. Step 4: Open the mobile host computer, set the specified UUID, find the name of the microcontroller's Bluetooth module and connect it. The mobile host computer can convert the data sent by the slave computer into a waveform, and when the patient has leakage or bulging during the infusion process, the mobile phone APP will issue an early warning to prompt the doctor.

[0051] Preferably, an algorithm is used to determine whether an infusion bulge is present during enhanced CT scans. The algorithm acquires preprocessed strain signals at a sampling rate of 100 Hz, calculates the baseline mean (μ) and standard deviation (σ) in real time using a 60-second sliding window, and updates the baseline every 10 minutes using exponential weighting (α=0.1) to eliminate environmental drift. A three-level dynamic threshold is used for judgment: when the instantaneous rate of change of the strain value ΔS>3σ or the sliding variance>2σ², an early warning is triggered and the system switches to 100 Hz high-frequency sampling; if the strain value exceeds μ+5σ (corresponding to 0.5-1 mL of exudate) for 3 seconds, a primary alarm is activated; and when the strain value exceeds μ+8σ (approximately 2 mL of exudate) or increases at a positive slope for 10 consecutive seconds (dS / dt>0.1σ / s), a high-level alarm is triggered.

[0052] Example 1 The present invention provides a wireless detection device for enhancing CT infusion leakage and bulging warning, such as Figure 1-8 As shown, it includes: a stretchable medical non-woven tape 1, a laser-induced graphene sensor 2, a wireless detection device body 4 for enhancing the early warning of CT infusion exudate bulge and a wristband 3 for fixation, an FPCB 5 with an integrated microcontroller, a rechargeable lithium battery attached under the FPCB, and a strain sensor connected to the microcontroller via a flexible cable 6.

[0053] Specifically, before the patient undergoes enhanced CT infusion, a stretchable medical non-woven tape is used to flatly fix the laser-induced graphene sensor 3-5 cm above the infusion port, and then a flexible cable is used to connect the two ends of the sensor to the microcontroller.

[0054] The microcontroller includes an MCU main control circuit, a power supply circuit, a strain signal acquisition circuit, a filter circuit, a Bluetooth module, and a micro lithium battery.

[0055] The strain signal acquisition circuit filters the data collected by the laser-induced graphene strain sensor through a hardware filter circuit and transmits it to the MCU main control circuit. The MCU main control circuit processes this data and sends it to the mobile phone host computer via the Bluetooth module.

[0056] Among them, such as Figure 5As shown, the PC14 and PC15 pins of the main control chip U1 (STM32F103C8T6) are connected to a 32.768kHz low-speed crystal oscillator circuit, while the OSC_IN and OSC_OUT pins are connected to an 8MHz high-speed crystal oscillator circuit, providing the system clock signal. U1's VDD_1, VDD_2, VDD_3, and VDD_4 power supply pins are connected to ground via decoupling capacitors C5, C6, C7, and C8. The VSS_1, VSS_2, VSS_3, and VSS_4 pins are directly grounded. VBAT is connected to the backup power supply via decoupling capacitor C9. VDDA is grounded via filter resistor R7 and decoupling capacitor C10, and VSSA is directly grounded. U1's NRST pin is pulled up to 3.3V via resistor R1 and grounded via capacitor C11, forming a reset circuit. The BOOT0 pin is grounded via resistor R2, setting the system to flash boot mode. PA13 and PA14 are connected to SWDIO and SWCLK, respectively, and are connected to the SWD programming interface for program programming and debugging. PD3, PD4, and PD5 are connected to status indicator LEDs through current-limiting resistors R3, R4, and R5, respectively, with the negative terminals of the LEDs grounded. PB2 is grounded through resistor R6 to prevent floating interference. Power supply regulator module U2 (RT9193) converts 5V to 3.3V and uses filter capacitors at the input and output to improve voltage regulation.

[0057] like Figure 6 As shown in the figure, the strain signal acquisition circuit includes a fixed resistor R11, a precision resistor R12, a G6K-2F-Y-5V relay, a SS8050 transistor, a pull-down resistor R10, and a strain sensor, forming a dynamic calibration voltage divider circuit. The voltage reference chip REF3030AIDBZR provides a stable 3V voltage for the ADC module. The relay is in normally open mode, with pin 1 connected to VCC, pin 8 connected to the MCU's PA8 pin via transistor Q2 and pull-down resistor R10, and pin 2 connected to VCC3V3 and fixed resistor R11. The sensor resistor R_sensor and precision resistor R12 are connected in parallel, with their ends connected to relay pin 3 and PS1, respectively. The voltage reference chip U8 has its input IN connected to VCC3V3 via decoupling capacitor C16, its output OUT connected to the MCU's ADC module power supply port VDDA, and its GAD pin connected to ground via decoupling capacitor C17.

[0058] like Figure 7As shown, the hardware filtering circuit is a three-stage cascaded filter circuit, consisting of a second-order active low-pass filter, a twin-T active notch filter circuit, and a first-order RC anti-aliasing filter. The 50Hz active notch filter circuit consists of two op amps and a twin-T notch filter. Negative feedback is introduced to improve frequency selection. Op amp U11.1 provides gain in the feedback loop and also isolates the twin-T network. The second-order active low-pass filter has a cutoff frequency of 106Hz, the twin-T active notch filter circuit has a center frequency of 50Hz, and the first-order RC anti-aliasing filter has a cutoff frequency of 150Hz. The input signal of the second-order active low-pass filter is the output signal PS1 of the strain acquisition circuit. The output signal of this circuit is fed into the next-stage twin-T active notch filter circuit, which in turn feeds into the next-stage first-order RC anti-aliasing filter. The output signal of the RC anti-aliasing filter is the output signal of the entire hardware filtering circuit and is then fed into ADC input pin PA0.

[0059] like Figure 8 As shown in the figure, the Bluetooth module includes the Bluetooth chip RF-BM-BG22C3 and the peripheral circuits required by the Bluetooth module. Among them, pin PA06 of chip U5 (model RF-BM-BG22C3) is connected to PA9 of the MCU, pin 2 (PA05) is connected to PA10 of the MCU, pin 3 (PA04) is connected to PA11 of the MCU, pin 4 (PA03) is connected to PA12 of the MCU, pin 5 (PA02) is connected to SWDIO of the MCU, and pin 6 (PA01) is connected to SWDCLK of the MCU; power pin 14 (VCC) is connected to the VCC power supply, grounded and filtered through capacitors C25 (0.1μF) and C26 (100nF), and pin 15 (GND) is directly grounded; function pin 10 (PB00) is connected to switch SW3 (TS-1088R-02026) and is pulled up to VCC through resistor R21 (10K).

[0060] Example 2 The preparation method of the laser-induced graphene strain sensor comprises the following steps: (1) A 75µm thick polyimide (PI) film was flatly adhered to a glass plate using a hydrosol. A pattern was designed using 2D modeling software and then imported into a CO2 laser. The CO2 laser was then used to ablate the PI film surface, inducing a specified graphene pattern. The pattern was a barbell shape with wide ends and a narrow middle. The graphene pattern at the wider end facilitated connection to an external circuit using copper wires, while the graphene pattern at the thinner middle end provided more sensitive strain performance.

[0061] In step (1), the size of the PI film can be selected according to the number of strain sensors required. The size and structure of the strain sensor are: width: 1.2mm ± 0.5mm, length: 15mm ± 5mm; the wider part is the copper wire connection, the size is: width: 3mm ± 0.5mm, length: 2mm ± 1mm; select the size of 25mm × 75mm to obtain four sensors, and so on. The strain sensor has a small size and structure and can be produced in batches at one time, which greatly reduces the production cost of the device and makes it possible to become a medical consumable.

[0062] (2) Mix components A and B of the PDMS solution in a ratio of 10:1 in a culture dish and stir for 5 minutes using a stainless steel hook. After stirring evenly, apply the mixture evenly to the LIG surface in (1), then place it in a drying oven at 70-100°C for 80-150 minutes for curing. After the PDMS is completely cured, a multilayer structure of PDMS-LIG-PI-hydrosol-glass plate is obtained from top to bottom. Using a PDMS with a mass ratio of 10:1, a strain-sensitive and tensile-stable PDMS / LIG composite porous foam material is produced, which makes the strain sensor have high sensitivity (capable of monitoring strain within 1%) and a wide detection range (45% strain). This makes it suitable for detecting the phenomenon of exudate bulging that occurs during the infusion process of enhanced CT.

[0063] (3) Soak the multilayer structure of PDMS-LIG-PI-hydrosol-glass plate in (2) in water for 5-10 minutes. After the hydrosol dissolves, tear off the PDMS-LIG-PI multilayer film from the glass plate and gradually separate the PDMS film and PI film. During this process, ensure that the speed and force are uniform. When peeling off the PI film, ensure that the separation direction of the PI film is consistent with the direction of the LIG circuit to prevent the PDMS film from tearing or the LIG circuit from breaking. After the two are completely separated, the LIG circuit originally formed by ablation on the surface of the PI film is completely transferred to the surface of the PDMS film. The LIG circuit transferred to the PDMS film has excellent flexibility (bending radius <1mm) and good tensile properties (0-55% stretching), and can be used in strain sensors and flexible electronics.

[0064] The wire is connected to the electrodes at both ends of the graphene through conductive silver glue. After the wire is fixed with the conductive silver glue, it is placed in an 80°C drying oven for heating and curing for 30-60 minutes.

[0065] The specific laser parameters in this example are: maximum power of 30W, maximum scanning speed of 1270mm / s. Laser induction parameters in this example are a laser power of 25% of the maximum power and a scanning speed of 20% of the maximum scanning speed, using a rastering method for laser induction. The LIG produced with these parameters has a smooth surface, sharp edges, and a high degree of carbonization, making it ideal for use as an electrode material. By adjusting the LIG parameters, the width (100-200μm), thickness (50-150μm), and internal microstructure (three-dimensional porous or dense) of the LIG can be controlled.

[0066] Example 3 The laser-induced graphene strain sensor according to one embodiment of the present invention has excellent response under small strain, such as Figure 3 As shown in FIG. 1 , the resistance changes after multiple stretching at a strain of 10%. It can be seen from the figure that the resistance of the sensor in the present invention increases with the increase of the stretching strain, and good response can be guaranteed even after multiple cycles at a small stretching strain.

[0067] Figure 4 This is the change in tensile resistance of the laser-induced graphene strain sensor of an embodiment of the present invention under a small deformation of less than 1%. As can be seen from the figure, the sensor in the present invention still has a very good response under a small deformation within 1%, reflecting its excellent sensitivity and can be used to detect exudate bulging during enhanced CT infusion.

[0068] Comparative Example 1: During PCB circuit design, a first-order low-pass hardware filter was selected as the filter circuit. The filter consisted of a capacitor C20 and a resistor R20, with a set cutoff frequency of 100 Hz. When this filter circuit was used for signal acquisition, the collected signal had excessive motion artifacts due to the power frequency interference of the CT room environment and the disturbance generated when the sensor was attached, affecting the judgment accuracy.

[0069] Comparative Example 2: In fabricating a laser-induced graphene flexible strain sensor, the laser parameters selected were 4W power, 200mm / s speed, 500 PPI, and a Z-axis height corresponding to the total thickness of the glass plate, hydrosol, and PI film, 3.2mm. These low laser power and high scanning speed resulted in insufficient laser energy applied to the PI film, failing to produce uniform and excellent carbonization of LIG. This resulted in incomplete LIG peeling during transfer.

[0070] Comparative Example 3: When collecting strain sensor models, a voltage divider circuit is used for collection. The voltage divider circuit is connected in parallel with the sensor through a resistor R30. The voltage divider formula is used. The response value of the sensor is calculated. Since the sensor will be pre-stretched during use and the initial resistance values ​​of different sensors are different, while the voltage divider resistance value is fixed, it will lead to voltage divider resistance mismatch, which greatly affects the acquisition accuracy and calculation range.

[0071] Comparative Example 4: After the laser-induced graphene strain sensor is manufactured, it is used directly without secondary packaging. During use, the laser-induced graphene circuit falls off, causing the strain sensor to fail due to short circuit.

[0072] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are protected by the present invention.

Claims

1. A wireless detection device for enhancing CT infusion seepage and bulging warning, characterized in that: A laser-induced graphene strain sensor is used as the front-end sensing unit to detect skin surface deformation during enhanced CT infusion. The microcontroller collects, analyzes, and sends sensor data, which is then received and visualized by the mobile host computer. The interactive function provides reminders when infusion anomalies occur. Includes laser-induced graphene strain sensor, microcontroller and mobile phone host computer; The fabrication of laser-induced graphene strain sensors involves the following steps: (1) Design the sensor structure using 2D modeling software, import the pattern into a CO2 laser, and ablate the LIG pattern on the PI film surface by laser. (2) PDMS was evenly coated on the LIG surface and the PDMS was cured at 80°C to form a LIG / PDMS composite porous foam material; (3) Peeling off the PDMS from the PI film to obtain a strain sensor body with PDMS as the substrate and LIG / PDMS composite porous foam material electrodes as the sensing layer; (4) Use highly conductive silver glue to connect wires at both ends of the LIG / PDMS composite porous foam material electrode, and then use PDMS as a flexible encapsulation layer.

2. The wireless detection device for enhancing CT infusion seepage bulge warning according to claim 1 is characterized in that: The laser-induced graphene strain sensor includes a PDMS substrate, LIG / PDMS composite porous foam material electrodes and connecting wires. The microcontroller includes an MCU main control circuit, a power circuit, a strain signal acquisition circuit, a filter circuit, a Bluetooth module, a micro lithium battery and a silicone shell. The mobile phone host computer is equipped with a data visualization function module, a data storage function module and a user interaction function module. The microcontroller includes the MCU main control circuit, power supply circuit, strain signal acquisition circuit, filtering circuit, Bluetooth module, micro lithium battery and silicone shell. Among them, the filtering circuit adopts a first-order low-pass filter. The output voltage of the micro lithium battery is 3.7V, which is used to power all circuits. The silicone shell is used to protect the electronic components of the microcontroller.

3. The wireless detection device for enhancing CT infusion seepage bulge warning according to claim 1 is characterized in that: The hardware filtering circuit is a three-stage cascade filtering circuit, including a second-order active low-pass filter, a dual-T active notch filter circuit, and a first-order RC anti-aliasing filter. The 50HZ active notch filter circuit consists of two operational amplifiers and a dual-T notch filter, and negative feedback is introduced at the same time. The cutoff frequency of the second-order active low-pass filter is 106Hz, the center frequency of the dual-T active notch filter circuit is 50Hz, and the cutoff frequency of the first-order RC anti-aliasing filter is 150Hz. The input signal of the second-order active low-pass filter is the output signal PS1 of the strain acquisition circuit. The output signal of this circuit is output to the next-stage dual-T active notch filter circuit. The output signal of the dual-T active notch filter circuit is the input signal of the next-stage first-order RC anti-aliasing filter. The output signal of the RC anti-aliasing filter is the output signal of the entire hardware filtering circuit and is output to the ADC input pin PA0. The second-order source low-pass filter in the hardware filtering circuit is used to filter noise above 100 Hz. The double-T active notch filter filter circuit is used to filter the 50 Hz power frequency interference in the CT room. The first-order RC anti-aliasing filter has a cutoff frequency of 150 Hz and is used to suppress high-frequency aliasing noise and match the ADC sampling rate.

4. The wireless detection device for enhancing CT infusion seepage bulge warning according to claim 1 is characterized in that: The strain signal acquisition circuit controls the opening and closing of the relay through the MCU to achieve the initial voltage calibration of the strain sensor. The specific implementation process is as follows: Power-on self-calibration: MCU controls the relay to close, R12 is connected in parallel with R_sensor, and the equivalent resistance is: , the reference voltage is provided by the chip REF3030, and the voltage division output at this time can be calculated by the ADC module, so the actual initial resistance of the sensor can be obtained as , Normal measurement: After the initial resistance calibration is completed, the MCU controls the relay to disconnect, and the system uses the calibrated R1 as the reference value.

5. The wireless detection device for enhancing CT infusion seepage bulge warning according to claim 1 is characterized in that: The size structure of the laser-induced graphene strain sensor is a barbell shape with a thin middle and wide ends. The thin part is the sensing part of the sensor, with dimensions of width: 1.2 mm ± 0.5 mm, length: 15 mm ± 5 mm; the wide part is the wire connection, with dimensions of width: 3 mm ± 0.5 mm, length: 2 mm ± 1 mm.

6. The wireless detection device for enhancing CT infusion seepage bulge warning according to claim 1, characterized in that: During the production of the laser-induced graphene strain sensor, components A and B of the PDMS solution are mixed in a ratio of 10.6:1, stirred evenly, and evenly coated on the LIG surface of the PI film. The mixture is then placed in a drying oven at 70-100°C and cured for 80-150 minutes. After the PDMS is completely cured, a LIG / PDMS composite porous foam material is obtained.

7. The wireless detection device for enhancing CT infusion seepage bulge warning according to claim 1, characterized in that: The strain-sensitive material of the laser-induced graphene strain sensor is a porous foam-like LIG / PDMS composite material. After connecting wires at both ends of the electrode, uncured PDMS is poured onto the surface of the LIG / PDMS composite porous foam electrode. After the PDMS solidifies, this forms an encapsulation layer. When applying the PDMS solution, the thickness of the PDMS film is controlled by controlling the PDMS mass per unit area. The thickness of the PDMS film affects the transfer quality of the graphene and the detection range of the sensor. The specific parameters are: 0.064-0.08g / cm 2 The thickness of the obtained PDMS film is 300µm-500µm.

8. The wireless detection device for enhancing CT infusion seepage bulge warning according to claim 1 is characterized in that: The device also includes a microcontroller body, a flexible cable connection line and a strain sensor body. The microcontroller body includes a flexible circuit board, a rechargeable lithium battery and a soft silicone shell. The ADC pin of the microcontroller extends from the shell through a flexible cable to connect to the sensor. The positive and negative power lines of the battery are connected to the power interface on the circuit board, and two power lines are extended to the outside of the shell through the Type-c adapter to charge the battery. The soft silicone shell is a flexible wristband shell, including a shell body and wristbands extending outward at both ends. The wristband is provided with a buckle for adjusting the tightness. The strain sensor body includes a strain sensor and a medical non-woven tape, wherein the medical non-woven tape is cut according to actual needs.

9. A wireless detection method for enhancing CT infusion seepage bulge warning, characterized in that: The following steps are involved: Step 1: Attach and fix the strain sensor. During enhanced CT scan, use medical non-woven tape to apply the strain sensor evenly to the skin 3-5 cm above the patient's infusion port. The strain sensor should be stretched perpendicular to the blood vessels, and the stretchable direction of the medical non-woven tape should be consistent with the stretching direction of the strain sensor. Step 2: Turn on the power switch of the microcontroller. When the switch is turned on, the power circuit converts the battery voltage to 3.3V. The 3.3V voltage powers the MCU main control circuit, strain signal acquisition circuit, hardware filter circuit, and Bluetooth module. Step 3: The strain signal acquisition circuit filters the data collected by the laser-induced graphene strain sensor through the hardware filter circuit and transmits it to the MCU main control circuit. The MCU main control circuit processes the data and sends it to the mobile phone host computer through the Bluetooth module; Step 4: Open the mobile host computer, set the specified UUID, find the name of the microcontroller's Bluetooth module and connect it. The mobile host computer can convert the data sent by the slave computer into a waveform, and when the patient has leakage or bulging during the infusion process, the mobile phone APP will issue an early warning to prompt the doctor.

10. The wireless detection method for enhancing CT infusion seepage bulge warning according to claim 9, characterized in that: An algorithm is used to detect infusion bulges during enhanced CT scans. The algorithm acquires preprocessed strain signals at a 100Hz sampling rate and calculates the baseline mean and standard deviation in real time using a 60-second sliding window. The baseline is updated every 10 minutes using exponential weighting to eliminate environmental drift. A three-level dynamic threshold is used: When the instantaneous rate of change of strain value ΔS exceeds 3σ or the sliding variance exceeds 2σ², an alert is triggered and the system switches to 100Hz high-frequency sampling. A primary alarm is activated if the strain value exceeds μ+5σ for three consecutive seconds. A high-level alarm is triggered if the strain value exceeds μ+8σ or increases at a positive slope for 10 consecutive seconds.

Images