Integrating full solid-state antimony microelectrode's intelligent digestive tract vibration capsule

Abstract

The application discloses a kind of integrated full solid-state antimony microelectrode's intelligent digestive tract vibration capsule, comprising: capsule shell, inner support, vibration motor, main control module, battery, pH sensor, data wireless transmission module, temperature sensor and pressure sensor;Inner support is arranged in capsule shell, vibration motor, main control module, battery, pH sensor, data wireless transmission module, temperature sensor and pressure sensor are all arranged in inner support;Vibration motor, pH sensor, data wireless transmission module, temperature sensor and pressure sensor are connected with main control module, and are controlled by main control module;Main control module is connected with battery, and is powered by battery;The sensing end of pH sensor is worn out capsule shell, and the cladding material of reference electrode is provided with microchannel.The application can detect in-vivo health condition, start vibration for abnormal area, solve the problem of constipation and other problems, solve the problem of transition treatment or insufficient treatment, and will not cause impact to healthy area.

Description

Technical Field

[0001] This invention relates to an intelligent digestive tract vibrating capsule integrating an all-solid-state antimony microelectrode, belonging to the field of capsule device technology. Background Technology

[0002] Vibrating capsule technology arose from the urgent need for the treatment of functional constipation. According to the "Chinese Expert Consensus on Chronic Constipation (2019)," the prevalence of chronic constipation among Chinese adults is 4%-10%, affecting approximately 50 million patients. Traditional treatments (such as laxatives) have problems such as dependence, the risk of diarrhea, and ineffectiveness in some patients, necessitating innovative therapies.

[0003] While some existing technologies have reported solutions for constipation and other problems through capsule vibration, the continuous internal vibration can disrupt the intestinal peristalsis rhythm, damage gut flora balance, and even harm the intestinal mucosa and nervous system regulation. To address these issues, the applicant has developed an intelligent digestive tract vibration capsule integrating an all-solid-state antimony microelectrode to solve the technical problems of excessive vibration found in existing technologies. Furthermore, improvements to the pH sensor reference electrode coating material and temperature and pressure compensation for pH detection have significantly improved the accuracy of pH detection. Summary of the Invention

[0004] This invention provides an intelligent digestive tract vibrating capsule integrating an all-solid-state antimony microelectrode, which can accurately test the ambient pH and has both pH testing and vibration functions; furthermore, it can compensate for pH in real time according to temperature and pressure, and the vibration motor can intelligently respond to changes in ambient pH; it can provide targeted treatment based on the actual situation in the body, avoiding overtreatment or undertreatment, and will not affect healthy areas.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0006] A smart digestive tract vibrating capsule integrating an all-solid-state antimony microelectrode includes: a capsule shell, an inner support, a vibration motor, a main control module, a battery, a pH sensor, a wireless data transmission module, a temperature sensor, and a pressure sensor;

[0007] The inner support is located inside the capsule shell, and the vibration motor, main control module, battery, pH sensor, wireless data transmission module, temperature sensor and pressure sensor are all located inside the inner support.

[0008] The vibration motor, pH sensor, wireless data transmission module, temperature sensor, and pressure sensor are all connected to and controlled by the main control module; the main control module is connected to and powered by the battery; the sensing end of the pH sensor protrudes from the capsule shell.

[0009] The pH sensor is an all-solid-state antimony pH microelectrode, including an antimony electrode and a reference electrode. The coating material of the reference electrode has micropores that lead from the outside to the reference electrode, forming microchannels. The microchannels are arranged in an interconnected mesh, and micropillar arrays, porous membrane interfaces, or necking structures are distributed on the microchannels. The total volume of the microchannels is nanoliters.

[0010] The aforementioned design of microchannels on the coating material of the reference electrode allows liquid to reach the internal reference electrode through micropores, significantly improving the detection effect; at the same time, micropillar arrays, porous membrane interfaces, or necking structures are used to trap mucus and particles.

[0011] To further improve detection accuracy, as one specific implementation scheme, microchannels are distributed across the surface of the reference electrode. The microchannels include inclined and parallel main channels with a necking structure. Adjacent main channels are connected by one or more connecting channels. One end of the main channel is connected to the outside world, and the other end is connected to the reference electrode.

[0012] The main flow channel and connecting flow channel mentioned above can be extended by using serpentine or spiral flow channels to ensure that the sample flows fully and the composition is uniform, while also facilitating the removal of air bubbles.

[0013] The aforementioned main channel is inclined so that the main channel and the reference electrode axis form an angle of 60-80°.

[0014] The coating material for the reference electrode is formed by pouring and curing adhesive onto a mold, creating a micropore arrangement on the mold. These micropores in the coating material interconnect the external environment with the reference electrode.

[0015] The aforementioned microchannels are arranged in an interconnected network, with the internal channels being interconnected, allowing the incoming liquid to freely switch between channels.

[0016] The aforementioned microchannels are equipped with micropillar arrays, porous membrane interfaces, or necking structures for the physical trapping of mucus and particles.

[0017] The aforementioned microchannels use serpentine or spiral flow paths to extend the flow path, ensuring that the collected samples flow fully and have uniform composition, while also facilitating the removal of air bubbles.

[0018] The preparation method of the all-solid-state antimony pH microelectrode includes:

[0019] 1. Wrap a thin copper wire around the electrode mold, with both ends exposed in the mold made of electrode coating material, ensuring that the wire runs through the height of the electrode and protrudes 2mm.

[0020] 2. Completely wrap the prepared reference electrode (antimony metal and silver oxide) and copper wire around the internal reference part and place it into the mold;

[0021] 3. Dissolve the tip of the antimony electrode with an 18% hydrofluoric acid solution, so that the antimony metal protrudes from the electrode tip, which has a diameter of approximately 0.8 mm.

[0022] 4. Pour the molten polytetrafluoroethylene into the electrode mold tooling groove to fill the gaps inside the electrode, and remove it after solidification;

[0023] 5. After cleaning and drying the exposed antimony electrode, immerse it vertically in sodium peroxide solution for 6-10 minutes to form an oxide film on the antimony surface. After cooling to room temperature, clean and dry it.

[0024] 6. Place the electrode in a 5wt% ferric chloride (FeCl3) solution and heat to 50°C for 30 minutes to mix and react. Remove the copper wires on the electrode from the previous step. Repeat this process about 10 times until the copper on the electrode is completely removed.

[0025] 7. After completion, the electrodes are cleaned in an ultrasonic oscillator in 0.7mol / L acetone and deionized water to remove chloride ions from the electrodes; finally, they are dried in a vacuum drying oven at 45±3℃.

[0026] For technologies not specifically described in this application, please refer to patent CN 208598410 U, CN 104007158 B or other existing technologies. Since there are no special improvements to the technologies not specifically described, they will not be elaborated further.

[0027] The main control module synchronously acquires the electrical signal V_pH output from the pH sensor, the ambient temperature T output from the temperature sensor, and the pressure P output from the pressure sensor at a preset sampling frequency, and performs dynamic compensation processing on the pH sensor signal based on the temperature T and pressure P.

[0028] Real-time temperature compensation is performed on temperature T, which is used to correct the actual pH value of the pH standard point (such as pH1 and pH2 used in two-point calibration) at the current temperature. The effective slope of the pH sensor at the current temperature is calculated in combination with the measured corresponding voltage, which is used to correct the voltage value of subsequent measurements, and the pH value after preliminary temperature compensation is obtained.

[0029] The main control module switches the pH data processing strategy based on changes in real-time pressure values.

[0030] When P < P_quiet (quiet threshold), the pH sensor measurement is considered stable, and the temperature-compensated pH value is directly used to enter the state machine for judgment;

[0031] When P ≥ P_quiet, it is considered that the pH sensor is subjected to mechanical disturbance. The main control module corrects the pH signal through a pre-established compensation model and then inputs it into the state machine for judgment.

[0032] When P ≥ P_max (safety threshold) lasts for ≥5 minutes, the state machine decision is paused until the pressure drops.

[0033] The main control module switches the pH data processing strategy based on the real-time pressure value changes: when the pressure value is below the preset quiet threshold, the pH sensor is considered to be in a stable measurement state, and the corresponding pH value is directly calculated based on the electrical signal output by the pH sensor and the state machine is entered for judgment; when the pressure value is equal to or higher than the preset quiet threshold, the pH sensor is considered to be subjected to mechanical disturbance, and the main control module corrects the point signal output by the pH sensor based on a pre-established compensation learning model trained by machine learning methods before calculating the pH value; when the pressure value is continuously higher than the preset safety threshold (when the duration of continuous higher than the preset safety threshold is greater than 5 minutes, it is considered continuous), the main control module pauses the judgment process of the state machine, and resumes judgment after detecting that the pressure value has fallen back below the preset quiet threshold, so as to avoid the influence of abnormal pressure state on the pH judgment result.

[0034] The main control module also constructs the response characteristics of the pH sensor under the current operating conditions based on at least two known pH reference points, and estimates the effective sensitivity parameters of the electrode accordingly. The effective sensitivity parameters are used as input to the pH calculation process or machine learning model to correct subsequent pH measurement results in real time.

[0035] The main control module, based on a machine learning model, comprehensively analyzes the pH value after temperature and pressure compensation, the trend of electrical signal changes, the characteristics of pressure changes, and historical measurement data, and outputs the final pH determination result for determining the location of the gastrointestinal tract, thereby improving the stability and accuracy of pH determination in complex in vivo environments.

[0036] The aforementioned compensation model can be implemented using linear or nonlinear regression, decision trees, or neural networks. Model training uses temperature, pressure, and historical pH data as input, and model parameters are obtained by optimizing the prediction error. In practical applications, the model parameters can be updated based on new measurement data.

[0037] To improve the measurement stability of pH sensors under mechanical disturbance environments, as a preferred implementation scheme, this application establishes a compensation model using machine learning methods to correct the pH signal under pressure influence. The specific modeling process is as follows:

[0038] 1. Data Acquisition and Feature Extraction:

[0039] In an in vitro simulated intestinal environment, the output voltage signal of the pH sensor, the output value of the temperature sensor, the output value of the pressure sensor, and the corresponding real pH value were collected under different pressure states (quiet, disturbed, high pressure) (synchronously recorded using a standard pH meter).

[0040] Extract the following features as model input:

[0041] The initial pH value after temperature compensation is calculated based on two-point calibration and temperature compensation.

[0042] Pressure value P and its first-order difference ΔP (reflecting the trend of pressure change);

[0043] Pressure status labels, such as: 0 indicates quiet, 1 indicates disturbance, and 2 indicates high pressure;

[0044] Historical pH sequence, pH values ​​at the 10 most recent sampling points;

[0045] The rate of change of electrode slope reflects the stability of electrode response.

[0046] 2. Model Selection and Training

[0047] Lightweight Gradient Boosting Decision Tree (LightGBM) is used as the compensation model, which features efficient training and low computational overhead, making it suitable for embedded main control modules.

[0048] The training objective is to minimize the mean square error (MSE) between the pH correction value output by the model and the true pH value.

[0049] The training set contains approximately 5,000 labeled samples, and the validation set is used to prevent overfitting. After training, the model parameters are fixed as lookup tables or lightweight inference code and burned into the main control module.

[0050] 3. Model Inference and pH Correction Process

[0051] In actual operation, when the pressure sensor detects that the current pressure value is greater than or equal to the quiet threshold P_quiet, the main control module initiates the compensation model inference process:

[0052] Input the feature vector at the current time step:

[0053] \[

[0054] > \mathbf{X} = [\text{pH}_{\text{temp\_comp}}, P, \Delta P, \text{state}, \text{pH}_{\text{hist}}, \text{slope\_rate}]

[0055] > \]

[0056] The model outputs a correction coefficient k (ranging from 0.95 to 1.05) and an offset b (unit: pH value), used to correct the pH value after initial temperature compensation.

[0057] \[

[0058] > \text{pH}_{\text{corrected}} = k \times \text{pH}_{\text{temp\_comp}} + b

[0059] > \]

[0060] 4. Model Validation Results

[0061] Under simulated intestinal peristalsis pressure disturbance (pressure fluctuation range ±5 kPa), the uncompensated pH measurement deviation can reach ±0.18 pH; after correction by this machine learning model, the deviation is reduced to within ±0.01 pH, which significantly improves the stability and accuracy of pH measurement under mechanical disturbance.

[0062] 5. Dynamic update mechanism

[0063] To adapt to differences in patients or environments, the model supports online fine-tuning: the main control module can make minor corrections to the model output using recursive least squares (RLS) based on pH measurement sequences under long-term stable conditions, further improving individual adaptability.

[0064] Through the aforementioned machine learning compensation model, the system achieves intelligent correction of pH signals under complex mechanical disturbance environments, providing a reliable data foundation for capsule position determination and vibration control.

[0065] This application abandons the approach of seeking static optimal compensation parameters and constructs a dynamic real-time deviation fusion mechanism, which significantly improves the accuracy of compensation.

[0066] To further improve the effect of the vibrating capsule, the aforementioned intelligent digestive tract vibrating capsule integrating all-solid-state antimony microelectrodes also includes a delayed start module. The delayed start module is connected to both the main control module and the vibration motor and is used to control the delayed start of the vibration motor.

[0067] When the pH value detected by the pH sensor transitions from strongly acidic (gastric environment, usually pH ≤ 3.5) to weakly acidic or alkaline (duodenal environment, usually pH ≥ 5.5), the activation module delays for 10-20 minutes. If the pH value remains above the threshold Y (e.g., Y = 4.5) during the delay, it is determined that the capsule has left the stomach and entered the intestine, and the vibration motor starts vibrating. If the pH value becomes strongly acidic again during the delay, it is determined that the capsule has not left the stomach and entered the intestine, and the vibration motor is not activated.

[0068] When the pH sensor detects a pH < 4.5, indicating an "in-stomach state," the vibration motor stops vibrating while sampling / transmitting data.

[0069] When the pH sensor detects a pH ≥ 4.5, it indicates a "non-gastric state" and triggers the vibration motor to start vibrating.

[0070] When the vibrating motor vibrates, it vibrates alternately at frequencies of 3Hz, 6Hz, and 9Hz.

[0071] During use, the pH sensor samples the pH data of bodily fluids in the gastrointestinal tract at a specific frequency (e.g., once every 5 seconds) to ensure the real-time and continuous nature of the data.

[0072] To further improve the effectiveness of the intelligent digestive tract vibrating capsule integrating all-solid-state antimony microelectrodes, the vibration motor operates in three modes:

[0073] Mode 1: No inner sub-period

[0074] Cyclic sequence: 3s working → 16s sleeping → 3s working → 7s sleeping → 3s working → 4s sleeping; During operation, it vibrates alternately at frequencies of 3Hz, 6Hz, and 9Hz.

[0075] Mode 2: Inner layer sub-cycle: 20 minutes of work + 10 minutes of rest

[0076] Sub-working cycle: 2s working → 3s rest; during working, it vibrates alternately at frequencies of 3Hz, 6Hz, and 9Hz.

[0077] Mode 3: Inner layer sub-cycle: 20 minutes of work + 10 minutes of rest

[0078] Sub-working cycle: 2s working → 3s rest; during working, it vibrates alternately at frequencies of 10Hz, 20Hz, and 30Hz.

[0079] Mode 1 is the default mode; if the pH value changes by less than 0.1 for more than 30 minutes, mode 2 is activated; if the pH value changes by less than 0.1 for more than 1 hour after mode 2 is activated, mode 3 is activated.

[0080] The three modes described above are programs embedded within the capsule. Mode switching depends on the pH detection value after the vibration function is activated. If the pH value change is less than 0.1 for more than 30 minutes and has not occurred or has disappeared, then mode one is used.

[0081] For ease of assembly, the capsule shell is formed by splicing a first shell and a second shell. The first shell includes a first cylinder and a first spherical surface at the bottom of the first cylinder, with a pH sensing hole at the bottom of the first spherical surface. The second shell includes a second cylinder and a second spherical surface at the top of the second cylinder, with the bottom of the second cylinder movably spliced ​​to the top of the first cylinder. Preferably, the bottom of the second cylinder and the top of the first cylinder are spliced ​​together by a threaded structure, which facilitates assembly while effectively ensuring stability.

[0082] The aforementioned temperature and pressure sensors are integrated into the main control module to sense the temperature and pressure of the digestive tract.

[0083] To improve the structural stability of the device, the inner support is a cavity structure with a spherical top, and the top of the outer side of the inner support is located inside the second spherical surface; the outer wall of the inner support and the inner wall of the first cylinder are engaged together by mutually matching groove protrusions; the inner side of the inner support includes a spherical cavity, a first cylindrical cavity and a second cylindrical cavity connected from top to bottom, and the diameter of the first cylindrical cavity is smaller than the diameter of the spherical cavity and the second cylindrical cavity;

[0084] The vibration motor adopts an eccentric wheel-spring damping structure. The drive end of the vibration motor is glued to the first cylindrical cavity. A gel layer is provided between the drive end of the vibration motor and the first cylindrical cavity. The eccentric end of the vibration motor is located in the spherical cavity. The vibration motor simulates intestinal peristalsis. With the help of the gel layer, the displacement of internal components caused by mechanical vibration is eliminated, ensuring that the vibration motor body (vibration motor body) sway amplitude is <0.1mm when the vibration motor is working. The aforementioned gel layer is relatively thin (≤0.2mm) and does not affect the vibration performance of the motor.

[0085] The main control module, battery, pH sensor, wireless data transmission module, temperature sensor and pressure sensor are all located in the second cylindrical cavity, and the sensing end of the pH sensor extends out from the pH sensing hole at the bottom of the capsule shell.

[0086] To improve the stability of the device, the sensing end of the pH sensor is fixed to the pH sensing hole by dispensing adhesive.

[0087] The capsule can be taken slowly with warm water. As it passes through the esophagus, the pH level can be monitored to determine if the patient is experiencing acid reflux, and the data is recorded. After activation, the capsule enters the body's natural cavities and moves downwards with the peristalsis of the organs. During this process, a pH sensor continuously monitors the data and records it, providing feedback on the patient's health. The capsule's built-in temperature and pressure sensors monitor the data in real time after entering the body and compensate for any errors in the pH sensor readings, ensuring accurate testing.

[0088] The location of the capsule in the human body can be determined by measuring the pH value: oral cavity (pH ~6.5-7.0); esophagus (pH ~6.0-7.0); stomach (pH 1.5-3.5); duodenum (first part of the small intestine) (pH ~6.0-7.0); jejunum and ileum (the rest of the small intestine) (pH ~7.0-7.5); colon (large intestine) (pH ~5.5-7.0).

[0089] The capsule can be opened using existing methods such as magnetic control, light control, reed switch, or Bluetooth.

[0090] The capsule enters the human body and monitors the body's acid-base data in real time, moving downwards. The circuit determines the capsule's environment based on the signal changes of the pH electrode. When it is in an acidic environment in the stomach, the vibration motor stops; when it detects that the capsule has left the stomach and entered the intestines, the vibration motor is triggered to start.

[0091] The vibration motor stops when the stomach is in an acidic environment. The core reason for this is to avoid interfering with the stomach's normal physiological functions.

[0092] The stomach's core function is to grind, mix, and perform preliminary digestion of food; it's a powerful mechanical churning vessel. If a capsule begins vibrating at a high frequency inside the stomach, it can disrupt the stomach's normal peristaltic rhythm, theoretically potentially causing discomfort, nausea, or vomiting.

[0093] Potentially triggering incorrect gastric emptying: Gastric emptying (the movement of food into the small intestine) is a precisely regulated process. Inappropriate vibrations can be misinterpreted by the stomach's nerves as a signal that the contents have been properly ground, leading to premature gastric emptying. This disrupts the entire digestive sequence. Simultaneously, it avoids depleting the stomach's energy reserves, which typically remain in the stomach for 2-6 hours. If the process is initiated in the stomach, the precious energy reserves may be exhausted before reaching the target intestines, resulting in treatment failure.

[0094] Vibration must be within the intestines to be effective. The therapeutic purpose of vibrating capsules (such as constipation treatment) is to target the intestines, especially the colon. Stimulating intestinal peristalsis: The gentle, rhythmic vibrations are designed to mimic the physical stimulation of food residue or feces on the intestinal wall, thereby activating the intestines' own peristaltic reflex and propelling the contents forward.

[0095] After collecting patient usage data, the vibration capsule transmits it to the communication terminal in real time via Bluetooth.

[0096] The aforementioned wireless data transmission module can use the low-power Bluetooth chip nRF52832;

[0097] Bluetooth communication protocol:

[0098] 1) Role Positioning: Peripheral Equipment

[0099] Capsules will always serve as peripheral devices, while mobile terminals will serve as central devices.

[0100] The capsule is in a deep sleep state most of the time, and the Bluetooth radio frequency is completely turned off;

[0101] 2) Communication mode: Primarily connectionless broadcasting, supplemented by connection-based communication.

[0102] To achieve the lowest power consumption, try to avoid establishing a stable Bluetooth connection, as maintaining the connection requires periodic time synchronization with the phone, which is very power-consuming;

[0103] 3) Enhance Bluetooth security mechanisms

[0104] Simple pairing: This usually uses the "Just Works" mode because the device does not have a display screen for the user to confirm the pairing code, and the risk of data tampering is relatively low for therapeutic devices;

[0105] Data encryption: The data may be encrypted using simple XOR encryption or lightweight encryption algorithms at the application layer to prevent it from being easily parsed by ordinary Bluetooth scanners, but the strength is usually not as good as that of financial-grade devices.

[0106] Device authentication: Authentication is performed through a built-in unique device ID. The app only communicates with the capsule it is paired with and ignores other devices.

[0107] The above method ensures that multiple capsule frequency bands can work simultaneously without interference, and that signal matching is good within a 5m range.

[0108] To meet the complexities of the usage environment, this low-power Bluetooth chip, nrf52832, utilizes its powerful processing performance to simultaneously construct 100 vibration capsules that work and transmit their operating status. After powering on, the vibration capsules are in Bluetooth standby mode, passively receiving Bluetooth scans from mobile terminals for docking. Bluetooth communication is compatible with mainstream mobile terminals such as Android and Apple.

[0109] The capsule shell is made of biocompatible PC (polycarbonate).

[0110] The temperature sensor mentioned above uses an NTC thermistor.

[0111] The pressure sensor mentioned above is a MEMS piezoresistive pressure sensor.

[0112] Any techniques not mentioned in this invention are based on existing technologies.

[0113] Compared with the prior art, the present invention has the following beneficial effects:

[0114] 1) It can accurately test the pH of the environment and has both pH testing and vibration functions; furthermore, it can compensate for pH in real time according to temperature and pressure, and the vibration motor can intelligently respond to changes in the pH of the environment.

[0115] 2) It can provide targeted treatment based on the actual condition of the body, relieve colonic spasms, promote colonic motility, treat constipation, promote defecation, and achieve the effects of beauty and health preservation, avoiding overtreatment or undertreatment, and will not affect healthy areas.

[0116] 3) The measured pH value can also be used to determine the capsule's location, the time it takes to pass through the intestines, and to obtain human health data; Attached Figure Description

[0117] Figure 1 This is a schematic diagram of the structure of the intelligent digestive tract vibration capsule integrating all-solid-state antimony microelectrodes of the present invention;

[0118] Figure 2 This is a layout diagram of the microchannels;

[0119] Figure 3 This is a graph showing the long-term stability test of the pH sensor over 72 hours.

[0120] Figure 4 Here is the pH compensation logic diagram;

[0121] Figure 5 This is a schematic diagram of a pH-compensated state machine;

[0122] Figure 6 pH response calibration curve;

[0123] In the diagram, 1 is the capsule shell, 2 is the inner support, 3 is the vibration motor, 4 is the main control module, 5 is the battery, 6 is the pH sensor, 7 is the wireless data transmission module, 8 is the temperature sensor, 9 is the pressure sensor, 10 is the internal reference electrode, 11 is the coating material, 12 is the main channel, 13 is the necking structure, and 14 is the connecting channel. Detailed Implementation

[0124] To better understand the present invention, the following embodiments further illustrate the content of the present invention, but the content of the present invention is not limited to the following embodiments.

[0125] The directional terms in this application, such as "center," "vertical," "horizontal," "length," "width," "thickness," "upper," "lower," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," are based on the appendix. Figure 1 The orientations or positional relationships shown are for the purpose of describing this application only, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0126] Example 1

[0127] like Figure 1 As shown, a smart digestive tract vibrating capsule integrating an all-solid-state antimony microelectrode includes: a capsule shell, an inner support, a vibration motor, a main control module, a battery, a pH sensor, a wireless data transmission module, a temperature sensor, and a pressure sensor.

[0128] The inner support is located inside the capsule shell, and the vibration motor, main control module, battery, pH sensor, wireless data transmission module, temperature sensor and pressure sensor are all located inside the inner support.

[0129] The vibration motor, pH sensor, wireless data transmission module, temperature sensor, and pressure sensor are all connected to and controlled by the main control module; the main control module is connected to and powered by the battery; the sensing end of the pH sensor protrudes from the capsule shell.

[0130] The pH sensor is an all-solid-state antimony pH microelectrode, comprising an antimony electrode and a reference electrode, such as... Figure 2 As shown, the coating material of the reference electrode has micropores leading from the outside to the reference electrode, forming microchannels. The microchannels are arranged in an interconnected mesh, and micropillar arrays, porous membrane interfaces, or necking structures are distributed on the microchannels. The total volume of the microchannels is nanoliters.

[0131] Example 2

[0132] Based on Example 1, the following improvements were made: Microchannels cover the surface of the reference electrode. Each microchannel includes inclined and parallel main channels with necking structures. Adjacent main channels are connected by one or more connecting channels. One end of each main channel connects to the outside environment, and the other end connects to the reference electrode. The axial direction of the main channel forms an angle of approximately 75° with the axial direction of the reference electrode. The total volume of the microchannels is 0.2 nL (nanoliters). Figure 3 As shown, the pH sensor in this example exhibits good stability during continuous monitoring in simulated body fluids, with a maximum drift of <1 mV / h and a response time of <2s.

[0133] Compared to not using microchannels (all other conditions being the same), the maximum drift was reduced by 94% and the response time was reduced by 80%.

[0134] Comparative Example 1

[0135] The difference from Example 2 is that the main channel axis is angled at approximately 90° with the reference electrode axis. Everything else is the same as in Example 2. The maximum drift was improved by 12.3%, and the response time was improved by 20.4%.

[0136] Comparative Example 2

[0137] The difference from Example 2 is that the main channel axis is angled at approximately 45° with the reference electrode axis. Everything else is the same as in Example 2. The maximum drift was improved by 18.6%, and the response time was improved by 26.4%.

[0138] Example 3

[0139] Based on Example 2, the following improvements were made: To facilitate assembly, the capsule shell is formed by splicing a first shell and a second shell; the first shell includes a first cylinder and a first spherical surface at the bottom of the first cylinder, with a pH sensing hole at the bottom of the first spherical surface; the second shell includes a second cylinder and a second spherical surface at the top of the second cylinder, with the bottom of the second cylinder movably spliced ​​to the top of the first cylinder. In this example, the bottom of the second cylinder and the top of the first cylinder are spliced ​​together by a threaded structure, which facilitates assembly while effectively ensuring stability.

[0140] Example 4

[0141] Based on Embodiment 3, the following improvements were made: In order to improve the structural stability of the device, the inner support is a cavity structure with a spherical top, and the top of the outer side of the inner support is located inside the second spherical surface; the outer wall of the inner support and the inner wall of the first cylinder are movably engaged together by mutually matching groove protrusions; the inner side of the inner support includes a spherical cavity, a first cylindrical cavity and a second cylindrical cavity connected sequentially from top to bottom, and the diameter of the first cylindrical cavity is smaller than the diameter of the spherical cavity and the second cylindrical cavity;

[0142] The vibration motor employs an eccentric wheel-spring damping structure. The drive end of the vibration motor is glued to the first cylindrical cavity. A 0.15mm thick gel layer is placed between the drive end of the vibration motor and the first cylindrical cavity. The eccentric end of the vibration motor is located within a spherical cavity. The vibration motor simulates intestinal peristalsis, and in conjunction with the gel layer, eliminates internal component displacement caused by mechanical vibration, ensuring that the vibration motor body's (vibration motor body's) sway amplitude is <0.1mm during operation. The main control module, battery, pH sensor, wireless data transmission module, temperature sensor, and pressure sensor are all located within a second cylindrical cavity. The sensing end of the pH sensor protrudes from the pH sensing hole at the bottom of the capsule shell. To improve the stability of the device, the sensing end of the pH sensor is fixed to the pH sensing hole using adhesive dispensing. The pH sensor uses an antimony electrode.

[0143] Example 5

[0144] Based on Example 4, the following improvements were made: real-time monitoring data from temperature and pressure sensors were used to compensate for the pH sensor data.

[0145] System components: pH sensor: used to measure the pH value of the body fluid at the location of the capsule; temperature sensor: used to collect the temperature of the body fluid at the location of the capsule; pressure sensor: used to collect the environmental pressure of the capsule; main control module: used to collect sensor data, perform pH compensation, and determine the status.

[0146] Data acquisition: The main control module simultaneously acquires pH voltage signal V_pH, temperature T and pressure P according to the preset sampling frequency; performs real-time temperature compensation on temperature T and uses it to correct the actual value of pH reference points (such as pH1 and pH2 obtained from two-point calibration) at the current temperature.

[0147] Two-point calibration and electrode slope calculation: The electrode response voltages E1 and E2 are measured using two known pH buffer solutions. Combined with the temperature-compensated pH1 and pH2, the effective slope of the electrode at the current temperature is calculated. This slope is used to correct the voltage values ​​of subsequent measurements to obtain the initial temperature-compensated pH value.

[0148] Pressure status assessment and compensation:

[0149] When P < P_quiet (quiet threshold), the pH sensor measurement is considered stable, and the temperature-compensated pH value is directly used to enter the state machine for judgment;

[0150] When P ≥ P_quiet, it is considered that the pH sensor is subjected to mechanical disturbance. The main control module corrects the pH signal through a pre-established compensation model and then inputs it into the state machine for judgment.

[0151] When P ≥ P_max (safety threshold) persists for a certain period of time, the state machine decision is paused until the pressure drops.

[0152] In this example, a compensation model is established using machine learning methods to correct the pH signal under the influence of pressure. The specific modeling process is as follows:

[0153] 1. Data Acquisition and Feature Extraction:

[0154] In an in vitro simulated intestinal environment, the output voltage signal of the pH sensor, the output value of the temperature sensor, the output value of the pressure sensor, and the corresponding real pH value were collected under different pressure states (quiet, disturbed, high pressure) (synchronously recorded using a standard pH meter).

[0155] Extract the following features as model input:

[0156] The initial pH value after temperature compensation is calculated based on two-point calibration and temperature compensation.

[0157] Pressure value P and its first-order difference ΔP (reflecting the trend of pressure change);

[0158] Pressure status labels, such as: 0 indicates quiet, 1 indicates disturbance, and 2 indicates high pressure;

[0159] Historical pH sequence, pH values ​​at the 10 most recent sampling points;

[0160] The rate of change of electrode slope reflects the stability of electrode response.

[0161] 2. Model Selection and Training

[0162] Lightweight Gradient Boosting Decision Tree (LightGBM) is used as the compensation model, which features efficient training and low computational overhead, making it suitable for embedded main control modules.

[0163] The training objective is to minimize the mean square error (MSE) between the pH correction value output by the model and the true pH value.

[0164] The training set contains approximately 5,000 labeled samples, and the validation set is used to prevent overfitting. After training, the model parameters are fixed as lookup tables or lightweight inference code and burned into the main control module.

[0165] 3. Model Inference and pH Correction Process

[0166] In actual operation, when the pressure sensor detects that the current pressure value is greater than or equal to the quiet threshold P_quiet, the main control module initiates the compensation model inference process:

[0167] Input the feature vector at the current time step:

[0168] \[

[0169] > \mathbf{X} = [\text{pH}_{\text{temp\_comp}}, P, \Delta P, \text{state}, \text{pH}_{\text{hist}}, \text{slope\_rate}]

[0170] > \]

[0171] The model outputs a correction coefficient k (ranging from 0.95 to 1.05) and an offset b (unit: pH value), used to correct the pH value after initial temperature compensation.

[0172] \[

[0173] > \text{pH}_{\text{corrected}} = k \times \text{pH}_{\text{temp\_comp}} + b

[0174] > \]

[0175] 4. Model Validation Results

[0176] Under simulated intestinal peristalsis pressure disturbance (pressure fluctuation range ±5 kPa), the uncompensated pH measurement deviation can reach ±0.18 pH; after correction by this machine learning model, the deviation is reduced to within ±0.01 pH, which significantly improves the stability and accuracy of pH measurement under mechanical disturbance.

[0177] 5. Dynamic update mechanism

[0178] To adapt to differences in patients or environments, the model supports online fine-tuning: the main control module can make minor corrections to the model output using recursive least squares (RLS) based on pH measurement sequences under long-term stable conditions, further improving individual adaptability.

[0179] Through the aforementioned machine learning compensation model, the system achieves intelligent correction of pH signals under complex mechanical disturbance environments, providing a reliable data foundation for capsule position determination and vibration control.

[0180] Gastrointestinal position determination and vibration control: The main control module uses the corrected pH value and state machine output to determine whether the capsule has entered the intestine, and can further trigger the vibration module to adjust its working mode to achieve automated response.

[0181] The aforementioned temperature and pressure compensation combined with a machine learning model can effectively offset the effects of electrode aging, mechanical disturbance, and environmental temperature changes on pH measurement, improving the stability of pH measurement and the accuracy of gastrointestinal location determination of the capsule in complex in vivo environments. This data-driven adaptive compensation strategy, unlike existing single formulas or fixed compensation methods, demonstrates the inventiveness of this invention.

[0182] Before any calibration, the total pH measurement deviation was ±0.18~0.3. After applying the temperature-pressure co-calibration method in this example, the pH measurement deviation was reduced to ≤±0.01. Furthermore, if only temperature calibration was performed, under the same conditions, the pH measurement deviation was reduced to ±0.02~0.03.

[0183] The capsule can be taken slowly with warm water. As it passes through the esophagus, the pH level can be monitored to determine if the patient is experiencing acid reflux, and the data is recorded. After activation, the capsule enters the body's natural cavities and moves downwards with the peristalsis of the organs. During this process, a pH sensor continuously monitors the data and records it, providing feedback on the patient's health. The capsule's built-in temperature and pressure sensors monitor the data in real time after entering the body and compensate for any errors in the pH sensor readings, ensuring accurate testing.

[0184] The location of the capsule in the human body can be determined by measuring the pH value: oral cavity (pH ~6.5-7.0); esophagus (pH ~6.0-7.0); stomach (pH 1.5-3.5); duodenum (first part of the small intestine) (pH ~6.0-7.0); jejunum and ileum (the rest of the small intestine) (pH ~7.0-7.5); colon (large intestine) (pH ~5.5-7.0).

[0185] Example 6

[0186] Based on Example 5, the following improvements were made: In order to reduce the impact on the normal area, the above-mentioned intelligent digestive tract vibration capsule integrating all-solid-state antimony microelectrodes also includes a delayed start module (SOT-23-5). The delayed start module is connected to both the main control module and the vibration motor and is used to control the delayed start of the vibration motor.

[0187] When the pH value detected by the pH sensor transitions from strongly acidic (gastric environment, usually pH ≤ 3.5) to weakly acidic or alkaline (duodenal environment, usually pH ≥ 5.5), the activation module delays for 10-20 minutes. If the pH value remains above the threshold Y (e.g., Y = 4.5) during the delay, it is determined that the capsule has left the stomach and entered the intestine, and the vibration motor starts vibrating. If the pH value becomes strongly acidic again during the delay, it is determined that the capsule has not left the stomach and entered the intestine, and the vibration motor is not activated.

[0188] Example 7

[0189] Based on Example 6, the following improvements were made: when the pH sensor detects pH < 4.5, it is in the "in-stomach state", the vibration motor stops vibrating, and sampling / transmission is performed simultaneously; when the pH sensor detects pH ≥ 4.5, it is in the "out-of-stomach state", and the vibration motor is triggered to start vibrating.

[0190] The vibration of the vibratory motor is divided into three modes:

[0191] Mode 1: No inner sub-period

[0192] Cyclic sequence: 3s working → 16s sleeping → 3s working → 7s sleeping → 3s working → 4s sleeping; During operation, it vibrates alternately at frequencies of 3Hz, 6Hz, and 9Hz.

[0193] Mode 2: Inner layer sub-cycle: 20 minutes of work + 10 minutes of rest

[0194] Sub-working cycle: 2s working → 3s rest; during working, it vibrates alternately at frequencies of 3Hz, 6Hz, and 9Hz.

[0195] Mode 3: Inner layer sub-cycle: 20 minutes of work + 10 minutes of rest

[0196] Sub-working cycle: 2s working → 3s rest; during working, it vibrates alternately at frequencies of 10Hz, 20Hz, and 30Hz.

[0197] Mode 1 is the default mode; if the pH value changes by less than 0.1 for more than 30 minutes, mode 2 is activated; if the pH value changes by less than 0.1 for more than 1 hour after mode 2 is activated, mode 3 is activated.

[0198] The capsule enters the human body and monitors the body's pH levels in real time, moving downwards. The circuit determines the capsule's environment based on changes in the pH electrode signal. When the capsule is in an acidic environment in the stomach, the vibration motor stops; when it detects that the capsule has left the stomach and entered the intestines, the vibration motor is triggered to start. Compared to existing timer-based control methods, this invention achieves intelligent activation of the vibration function through real-time pH monitoring, significantly improving the consistency and reliability of the intestinal activation location and ensuring the effectiveness of treatment or sampling.

[0199] The aforementioned wireless data transmission module uses the low-power Bluetooth chip nRF52832; the vibration motor is a 408 coreless motor; the battery is a CR777; the temperature sensor is a TI TMP117; the pressure sensor is a Bosch BMP380; and the main control module is ICP-20100.

[0200] Bluetooth communication protocol:

[0201] 1) Role Positioning: Peripheral Equipment

[0202] Capsules will always serve as peripheral devices, while mobile terminals will serve as central devices.

[0203] The capsule is in a deep sleep state most of the time, and the Bluetooth radio frequency is completely turned off;

[0204] 2) Communication mode: Primarily connectionless broadcasting, supplemented by connection-based communication.

[0205] To achieve the lowest power consumption, try to avoid establishing a stable Bluetooth connection, as maintaining the connection requires periodic time synchronization with the phone, which is very power-consuming;

[0206] 3) Enhance Bluetooth security mechanisms

[0207] Simple pairing: This usually uses the "Just Works" mode because the device does not have a display screen for the user to confirm the pairing code, and the risk of data tampering is relatively low for therapeutic devices;

[0208] Data encryption: The data may be encrypted using simple XOR encryption or lightweight encryption algorithms at the application layer to prevent it from being easily parsed by ordinary Bluetooth scanners, but the strength is usually not as good as that of financial-grade devices.

[0209] Device authentication: Authentication is performed through a built-in unique device ID. The app only communicates with the capsule it is paired with and ignores other devices.

[0210] The above method ensures that multiple capsule frequency bands can work simultaneously without interference, and that signal matching is good within a 5m range.

[0211] To meet the complexities of the usage environment, this low-power Bluetooth chip, nrf52832, utilizes its powerful processing performance to simultaneously construct 100 vibration capsules that work and transmit their operating status. After powering on, the vibration capsules are in Bluetooth standby mode, passively receiving Bluetooth scans from mobile terminals for docking. Bluetooth communication is compatible with mainstream mobile terminals such as Android and Apple.

[0212] In this example, the capsule shell is made of biocompatible PC (polycarbonate); the temperature sensor is an NTC thermistor; and the pressure sensor is a MEMS piezoresistive pressure sensor.

[0213] The aforementioned examples of intelligent digestive tract vibrating capsules integrating all-solid-state antimony microelectrodes offer highly accurate pH testing, combining pH measurement with vibration functionality. The capsule's position can be determined by the measured pH value, providing the patient's health data. The vibration capsule's transit time through the intestines can be estimated. By vibrating against the colon wall, the capsule can relieve colonic spasms, promote colonic motility, treat constipation, and facilitate defecation, achieving beauty, health, and wellness benefits. Furthermore, the vibrating capsule can promote small intestinal peristalsis and reduce the absorption of food, thus aiding in weight loss. It can compensate for pH changes in real time based on temperature and pressure, with the vibration motor intelligently responding to pH variations. It can also provide targeted treatment based on the patient's actual condition, avoiding overtreatment or undertreatment. Antimony electrodes are more stable and reliable than glass electrodes, enabling stable operation within the complex acid-base environment of the human digestive tract.

Claims

1. A smart digestive tract vibration capsule integrating an all-solid-state antimony microelectrode, characterized in that: include: Capsule shell, inner support, vibration motor, main control module, battery, pH sensor, wireless data transmission module, temperature sensor and pressure sensor; The inner support is located inside the capsule shell, and the vibration motor, main control module, battery, pH sensor, wireless data transmission module, temperature sensor and pressure sensor are all located inside the inner support. The vibration motor, pH sensor, wireless data transmission module, temperature sensor, and pressure sensor are all connected to and controlled by the main control module; the main control module is connected to and powered by the battery; the sensing end of the pH sensor protrudes from the capsule shell. The pH sensor is an all-solid-state antimony pH microelectrode, including an antimony electrode and a reference electrode. The coating material of the reference electrode has micropores that lead from the outside to the reference electrode, forming microchannels. The microchannels are arranged in an interconnected mesh, and micropillar arrays, porous membrane interfaces, or necking structures are distributed on the microchannels. The total volume of the microchannels is nanoliters. The main control module synchronously acquires the electrical signal V_pH output from the pH sensor, the ambient temperature T output from the temperature sensor, and the pressure P output from the pressure sensor at a preset sampling frequency, and performs dynamic compensation processing on the pH sensor signal based on the temperature T and pressure P. Real-time temperature compensation is performed on temperature T, which is used to correct the actual pH value of the pH standard point at the current temperature. The effective slope of the pH sensor at the current temperature is calculated in combination with the measured voltage, which is used to correct the voltage value of subsequent measurements, and the pH value after preliminary temperature compensation is obtained. The main control module switches the pH data processing strategy based on changes in real-time pressure values. When P < the quiet threshold P_quiet, the pH sensor measurement is considered stable, and the temperature-compensated pH value is directly used to enter the state machine for judgment; When P ≥ the quiet threshold P_quiet, the pH sensor is considered to be subjected to mechanical disturbance. The main control module corrects the pH signal through a pre-established compensation model and then inputs it into the state machine for judgment. When P ≥ safety threshold P_max for ≥5 minutes, pause state machine judgment until pressure drops.

2. The intelligent digestive tract vibration capsule integrating an all-solid-state antimony microelectrode according to claim 1, characterized in that: The microchannels cover the surface of the reference electrode. The microchannels include inclined and parallel main channels with a necking structure. Adjacent main channels are connected by one or more connecting channels. One end of the main channel is connected to the outside and the other end is connected to the reference electrode.

3. The intelligent digestive tract vibration capsule integrating an all-solid-state antimony microelectrode according to claim 1 or 2, characterized in that: The compensation model is established using machine learning methods to correct the pH signal under pressure. The modeling process is as follows: 1) Data Acquisition and Feature Extraction: In an in vitro simulated intestinal environment, the output voltage signal of the pH sensor, the output value of the temperature sensor, the output value of the pressure sensor, and the corresponding real pH value were collected under different pressure conditions. Extract the following features as model input: Initial pH value after temperature compensation; Pressure value P and its first-order difference ΔP; Pressure status labels: 0 indicates quiet, 1 indicates disturbance, and 2 indicates high pressure. Historical pH sequence, pH values ​​at the 10 most recent sampling points; The rate of change of electrode slope reflects the stability of electrode response; 2) Model selection and training: A lightweight gradient boosting decision tree was used as the compensation model for training. The training objective is to minimize the mean square error between the pH correction value output by the model and the true pH value; The training set contains 4,000 to 10,000 labeled samples. The validation set is used for model performance evaluation and hyperparameter optimization during training to suppress model overfitting. After the model training converges, its parameters are solidified into lookup tables or lightweight inference code and burned into the main control module for deployment. 3) Model inference and pH correction: When the pressure sensor detects that the current pressure value is greater than or equal to the quiet threshold P_quiet, the main control module starts the compensation model and outputs the correction coefficient k and offset b to correct the pH value after the initial temperature compensation. When P < the quiet threshold P_quiet, the pH sensor measurement is considered stable, and the temperature-compensated pH value is directly used to enter the state machine for judgment; When P ≥ safety threshold P_max for ≥5 minutes, pause state machine judgment until pressure drops.

4. The intelligent digestive tract vibration capsule integrating an all-solid-state antimony microelectrode according to claim 1 or 2, characterized in that: It also includes a delayed start module, which is connected to both the main control module and the vibration motor, and is used to control the delayed start of the vibration motor; When the pH value detected by the pH sensor transitions from strongly acidic to weakly acidic or alkaline, the activation module delays for 10-20 minutes. If the pH value remains above the threshold Y during the delay period, it is determined that the capsule has left the stomach and entered the intestine, and the vibration motor starts vibrating. If the pH value becomes highly acidic again during the delay period, it is determined that the capsule has not left the stomach and entered the intestines, and the vibration motor will not be activated.

5. The intelligent digestive tract vibration capsule integrating an all-solid-state antimony microelectrode according to claim 4, characterized in that: When the pH sensor detects that the pH is less than 4.5, it indicates a "gastric state," the vibration motor stops vibrating, and sampling / transmission is performed simultaneously. When the pH sensor detects a pH ≥ 4.5, it indicates a "non-gastric state" and triggers the vibration motor to start vibrating. When the vibrating motor vibrates, it vibrates alternately at frequencies of 3Hz, 6Hz, and 9Hz.

6. The intelligent digestive tract vibration capsule integrating an all-solid-state antimony microelectrode according to claim 1 or 2, characterized in that: The vibration of the vibratory motor is divided into three modes: Mode 1: No inner sub-period Cyclic sequence: 3s working → 16s sleeping → 3s working → 7s sleeping → 3s working → 4s sleeping; During operation, it vibrates alternately at frequencies of 3Hz, 6Hz, and 9Hz. Mode 2: Inner layer sub-cycle: 20 minutes of work + 10 minutes of rest Sub-working cycle: 2s working → 3s rest; during working, it vibrates alternately at frequencies of 3Hz, 6Hz, and 9Hz. Mode 3: Inner layer sub-cycle: 20 minutes of work + 10 minutes of rest Sub-working cycle: 2s working → 3s rest; during working, it vibrates alternately at frequencies of 10Hz, 20Hz, and 30Hz. Mode 1 is the default mode; if the pH value changes by less than 0.1 for more than 30 minutes, mode 2 is activated; if the pH value changes by less than 0.1 for more than 1 hour after mode 2 is activated, mode 3 is activated.

7. The intelligent digestive tract vibration capsule integrating an all-solid-state antimony microelectrode according to claim 1 or 2, characterized in that: A smart digestive tract vibration capsule integrating an all-solid-state antimony microelectrode is characterized in that: the capsule shell is composed of a first shell and a second shell spliced ​​together; the first shell includes a first cylinder and a first spherical surface at the bottom of the first cylinder, and a pH sensing hole is provided at the bottom of the first spherical surface; the second shell includes a second cylinder and a second spherical surface at the top of the second cylinder, and the bottom of the second cylinder is movably spliced ​​together with the top of the first cylinder.

8. The intelligent digestive tract vibration capsule integrating an all-solid-state antimony microelectrode according to claim 7, characterized in that: The inner support is a cavity structure with a spherical top, and the top of the outer side of the inner support is located inside the second spherical surface; the outer wall of the inner support and the inner wall of the first cylinder are engaged together by mutually matching groove protrusions; the inner side of the inner support includes a spherical cavity, a first cylindrical cavity and a second cylindrical cavity connected from top to bottom, and the diameter of the first cylindrical cavity is smaller than the diameter of the spherical cavity and the second cylindrical cavity. The vibration motor adopts an eccentric wheel-spring damping structure. The drive end of the vibration motor is glued to the first cylindrical cavity. A gel layer is provided between the drive end of the vibration motor and the first cylindrical cavity. The eccentric end of the vibration motor is located in the spherical cavity. The main control module, battery, pH sensor, wireless data transmission module, temperature sensor and pressure sensor are all located in the second cylindrical cavity. The sensing end of the pH sensor protrudes from the pH sensing hole at the bottom of the capsule shell. The sensing end of the pH sensor is fixed to the pH sensing hole by dispensing glue.

9. The intelligent digestive tract vibration capsule integrating an all-solid-state antimony microelectrode according to claim 1 or 2, characterized in that: The capsule shell is made of PC material; the temperature sensor is an NTC thermistor; and the pressure sensor is a MEMS piezoresistive pressure sensor.

Images