Capsule detection method for gastrointestinal hemorrhage, server, medium and program product

Abstract

The invention provides a capsule detection method for alimentary canal hemorrhage, a server, a medium and a program product, and relates to the field of image data processing.The method comprises the steps that a plurality of magnetic field sensors are used for measuring changes of magnetic materials in a capsule to determine position information of the capsule in an alimentary canal, and then the information is fused with three-dimensional anatomical model data of the human alimentary canal to obtain a three-dimensional anatomical model; a three-dimensional visual model is generated, first occult blood data (determined by measuring hemoglobin concentration through an electrochemical method) and digestive tract fluid analysis data sent by an intelligent occult blood detection device in the capsule are obtained through wireless communication, and second occult blood data are obtained through correction of an occult blood detection compensation model; and finally, determining a display picture by using the three-dimensional model in combination with capsule positioning information and transmitting the display picture to a display end for display. By implementing the method, real-time, comprehensive and accurate detection and visual positioning of alimentary canal hemorrhage can be realized.

Description

Technical Field

[0001] The present application relates to the technical field of image data processing, and in particular to a capsule detection method, server, medium and program product for gastrointestinal bleeding. Background Art

[0002] Gastrointestinal bleeding is a common and potentially life-threatening disease with a mortality rate of up to 30%. Accurate detection of the location and extent of gastrointestinal bleeding is extremely critical for the diagnosis and treatment of the disease. Digestive endoscopy can diagnose gastrointestinal bleeding by directly observing lesions in the gastrointestinal cavity and has been widely used in clinical practice. However, there are still cases of unexplained gastrointestinal bleeding (Obscure Gastrointestinal Bleeding, OGIB) in clinical practice, and the source of bleeding cannot be determined by conventional digestive endoscopy (gastroscopy, colonoscopy). Common causes are vascular abnormalities, tumors, or diverticula in the small intestine. Due to the hidden location, long length, and high mobility of the small intestine, it is difficult to reach or find lesions during routine examinations, resulting in difficult diagnosis and affecting subsequent treatment and patient prognosis.

[0003] The commonly used examination method for unexplained gastrointestinal bleeding is capsule endoscopy. After entering the human digestive tract, the traditional capsule endoscopy will take pictures of the internal conditions of the digestive tract. The captured images will be transmitted to the sensor worn by the patient through the transmission method of wireless waves. After the image is transmitted to the sensor, the image recognition system will identify the bleeding-related images for diagnosis. In addition, when detecting whether there is bleeding in the digestive tract, fecal occult blood reagents or test strips are also commonly used tools in clinical practice. Among them, the colloidal gold method is widely used among many detection methods. The principle of this method is based on a specific chemical reaction. When the test strip comes into contact with hemoglobin in the feces, the color of the test strip will change, thereby roughly judging whether the patient has gastrointestinal bleeding.

[0004] However, the traditional capsule endoscopy has a high rate of missed diagnosis of unexplained gastrointestinal bleeding. This is because the traditional capsule endoscope has no power and moves forward passively by relying on gastrointestinal peristalsis. It cannot control the speed and is prone to blind spots in the field of vision or unclear images, or misses slight bleeding that cannot be observed by the naked eye. In addition, the traditional capsule endoscope cannot accurately locate, which affects the subsequent effective hemostasis treatment. On the other hand, the colloidal gold method for detecting occult blood is an in vitro test of the patient's stool specimens. It cannot locate the bleeding location and is easily affected by the digestive fluid. The acidity and alkalinity of the digestive fluid will interfere with the test paper test. Therefore, it is difficult for traditional capsule endoscopy and fecal occult blood detection methods to detect the location and extent of gastrointestinal bleeding in real time, comprehensively and accurately in vivo. Summary of the invention

[0005] The present application provides a capsule detection method, server, medium and program product for gastrointestinal bleeding, which are used to achieve comprehensive and accurate detection and positioning of gastrointestinal bleeding conditions.

[0006] In a first aspect, the present application provides a capsule detection method for gastrointestinal bleeding, which is applied to a server. The method comprises: after measuring the induction intensity and direction change of the magnetic material inside the capsule in the magnetic field through multiple magnetic field sensors, determining the capsule position information of the capsule in the digestive tract, the capsule position information including real-time positioning data and motion trajectory data; fusing the capsule position information of the capsule with pre-set three-dimensional anatomical model data of the human digestive tract to obtain a three-dimensional visualization model including the capsule position information and the digestive tract structure; acquiring first occult blood detection data information sent by an intelligent occult blood detection device pre-placed in the capsule through wireless communication, and recording the current capsule position information, the intelligent occult blood detection device being electrically connected to the capsule. The chemical detection method detects the change of current between electrodes to detect the hemoglobin concentration, and the first occult blood detection data information is determined by the hemoglobin concentration; the digestive tract fluid analysis data information is obtained by the intelligent occult blood detection device; the digestive tract fluid analysis data information and the first occult blood detection data information are input into the occult blood detection compensation model to obtain the corrected second occult blood detection data information, and the occult blood detection compensation model is constructed by training multiple experimental data on the influence of changes in different digestive fluid components on the hemoglobin concentration detection in the electrochemical detection method under a simulated digestive fluid environment; the second occult blood detection data information and the capsule positioning information are determined by using a three-dimensional visualization model to display a screen, and sent to a display terminal for display.

[0007] By adopting the above technical solution, multiple magnetic field sensors measure the magnetic material inside the capsule, and use the induction intensity and direction changes in the magnetic field to determine the capsule position information, including real-time positioning and motion trajectory data, which can accurately track the path of the capsule in the digestive tract and fuse with the three-dimensional anatomical model data of the human digestive tract to construct a three-dimensional visualization model to intuitively present the relative position of the capsule and the digestive tract. The intelligent occult blood detection device electrochemically detects the hemoglobin concentration to obtain the first occult blood data, and combines the digestive tract fluid analysis data to input the compensation model to obtain the corrected second occult blood data, which can more accurately and in real time reflect the actual occult blood situation in the digestive tract, effectively solving the problem of false positives or false negatives in traditional occult blood detection methods (such as the colloidal gold method affected by the acidity and alkali of the digestive fluid), greatly improving the accuracy of occult blood detection, thereby realizing real-time, comprehensive and accurate qualitative positioning detection of gastrointestinal bleeding.

[0008] In combination with the content of the first aspect, further, after the step of obtaining the first occult blood detection data information sent by the intelligent occult blood detection device pre-placed in the capsule through wireless communication and recording the current capsule position information, it also includes: receiving a capsule control instruction sent by the user end, and controlling the controllable magnetic field source according to the capsule control instruction, and the controllable magnetic field source is used to control the position and posture of the capsule in the digestive tract by adjusting the magnetic field strength and direction parameters.

[0009] By adopting the above technical solution, on the basis of determining the capsule position information and obtaining occult blood detection data, the ability to actively control the capsule movement is added, which can enable the capsule to approach the suspected bleeding site more accurately and cooperate with other detection functions to further improve the inspection effect of specific areas of the digestive tract.

[0010] In combination with the content of the first aspect, further, after receiving the capsule control instruction sent by the user end and controlling the controllable magnetic field source according to the capsule control instruction, it also includes: capturing the current multi-angle image information of the digestive tract through multiple cameras in the capsule; acquiring the multi-angle image information of the digestive tract through wireless communication, and determining the location information of the bleeding point in combination with a pre-set digestive tract bleeding determination model, the digestive tract bleeding determination model is trained in advance based on multiple digestive tract image sample data of known bleeding conditions and non-bleeding conditions using a deep learning algorithm.

[0011] By adopting the above technical solutions, on the basis of existing magnetic field positioning, occult blood detection and related data processing, the addition of image information enriches the detection data source, and multi-angle images can more comprehensively reflect the internal conditions of the digestive tract. By analyzing the images with a model trained by a deep learning algorithm, bleeding points can be effectively identified, and they can be verified with other test results, thus improving the accuracy and reliability of bleeding point positioning, and helping doctors to more accurately judge the digestive tract bleeding situation.

[0012] In combination with the content of the first aspect, further, after the step of using the three-dimensional visualization model to determine the display screen of the second occult blood detection data information and the capsule positioning information, and sending it to the display end for display, it also includes: marking the bleeding points of the display screen with colors corresponding to different depths, and the depth is determined according to the hemoglobin concentration value in the second occult blood detection data information.

[0013] By adopting the above technical solution, the hemoglobin concentration of the bleeding point is intuitively reflected through color marking, so that the doctor can quickly and intuitively understand the severity of the bleeding when viewing the display screen.

[0014] In combination with the content of the first aspect, further, after measuring the induction intensity and direction change of the magnetic material inside the capsule in the magnetic field by multiple magnetic field sensors, before the step of determining the capsule position information in the digestive tract, it also includes: receiving an opening detection instruction sent by the user end, and detecting whether multiple magnetic field sensors are distributed in a circular array at a set part of the human body according to the opening detection instruction; if so, calibrating each other through multiple magnetic field sensors.

[0015] By adopting the above technical solution, before using the magnetic field sensor to determine the position of the capsule, its distribution and calibration are ensured. The magnetic field sensors distributed in a circular array can sense the changes in magnetic materials from multiple directions, thereby improving the accuracy and stability of position detection.

[0016] In combination with the content of the first aspect, further, after the step of using the three-dimensional visualization model to determine the display screen of the second occult blood detection data information and the capsule positioning information, and sending them to the display end for display, it also includes: after receiving a gesture operation request from the display end for the three-dimensional visualization model, performing corresponding transformation processing on the three-dimensional visualization model according to the gesture operation request, and the transformation processing includes determining the rotation operation of the three-dimensional visualization model around a set axis by calculating the new coordinates of the vertex coordinates of the three-dimensional visualization model under the action of the rotation matrix when processing the rotation transformation processing; when processing the scaling transformation processing, adjusting the relative distance of the coordinates of each vertex of the three-dimensional visualization model according to the scaling factor to change the overall size of the three-dimensional visualization model; when processing the translation transformation processing, modifying the coordinate positions of all vertices of the three-dimensional visualization model according to the translation vector; after completing the transformation processing, sending the updated three-dimensional visualization model data to the display end for real-time display.

[0017] By adopting the above technical solution, after the detection data is presented in a three-dimensional visualization model and sent to the display end, the display end is allowed to perform gesture operations, such as rotation to observe the position relationship between the digestive tract and the capsule from different angles, zooming to view local details, and translation for easy comprehensive browsing. Based on the previously constructed accurate three-dimensional model and detection data, these interactive operations enable doctors to explore model information more freely and comprehensively, better understand the internal structure of the digestive tract and the lesions detected by the capsule, assist in diagnosis and disease analysis, and improve doctors' analysis and utilization efficiency of the location and degree of digestive tract bleeding.

[0018] In combination with the content of the first aspect, further, before the step of obtaining the first occult blood detection data information sent by the intelligent occult blood detection device pre-placed in the capsule through wireless communication and recording the current capsule position information, it also includes: using low-power Bluetooth technology to transmit data with the intelligent occult blood detection device; encrypting the transmitted data according to a pre-set communication encryption module.

[0019] By adopting the above technical solution, the communication encryption module ensures the security of data transmission, prevents the detection data from being stolen or tampered with during the transmission process, and protects the patient's privacy and the integrity of the detection data.

[0020] In a second aspect, the present application provides a server, comprising: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the server to execute the method described in the first aspect and any possible implementation of the first aspect.

[0021] In a third aspect, the present application provides a computer-readable storage medium comprising instructions, which, when executed on a server, causes the server to execute the method described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, the present application provides a computer program product, which, when executed on a server, enables the server to execute the method described in the first aspect and any possible implementation manner of the first aspect.

[0023] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. Due to the adoption of multi-sensor fusion positioning, electrochemical occult blood detection and three-dimensional visualization model construction technical means, that is, using multiple magnetic field sensors to determine the capsule position information, the intelligent occult blood detection device electrochemically detects the hemoglobin concentration and combines the digestive tract fluid analysis data with the compensation model for correction, and then integrates it with the three-dimensional anatomical model of the human digestive tract into a visualization model. Therefore, it effectively solves the technical problems in the prior art that the capsule positioning is inaccurate and the occult blood detection is easily disturbed by the digestive fluid, resulting in inaccurate detection results. It then realizes the real-time and accurate detection and positioning of gastrointestinal bleeding, and the technical effect of intuitively presenting the detection results, thereby improving the accuracy, sensitivity and specificity of gastrointestinal bleeding detection.

[0024] 2. Due to the use of a controllable magnetic field source to control the position and posture of the capsule, on the basis of determining the capsule position information and obtaining occult blood detection data, the controllable magnetic field source is controlled to adjust the magnetic field strength and direction parameters by receiving user-side instructions. Therefore, the technical problem that the existing technology cannot actively control the movement of the capsule to approach the suspected lesion site for accurate inspection is effectively solved, thereby achieving the technical effect of improving the inspection effect of specific areas of the digestive tract, enhancing the flexibility and effectiveness of detection, and assisting doctors to more accurately judge the bleeding points and lesions.

[0025] 3. Due to the adoption of technical means for visually marking bleeding points based on hemoglobin concentration, on the basis of constructing and displaying a three-dimensional visualization model including capsule positioning information and occult blood detection data, the bleeding points on the display screen are marked with different shades of colors according to the hemoglobin concentration value in the second occult blood detection data information. Therefore, the technical problem in the prior art that the test results are not intuitive and it is difficult to quickly judge the severity of bleeding is effectively solved, thereby achieving the technical effect of assisting doctors to quickly and intuitively understand the severity of bleeding, efficiently convey key information, and improve diagnostic efficiency and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of a capsule detection method for gastrointestinal bleeding in an embodiment of the present application; Figure 2 This is another schematic flow chart of the capsule detection method for gastrointestinal bleeding in the embodiment of the present application; Figure 3 It is a schematic diagram of a physical device structure of a server in an embodiment of the present application. DETAILED DESCRIPTION

[0027] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification and appended claims of the present application, the singular expressions "one", "a kind of", "said", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to and includes any or all possible combinations of one or more listed items.

[0028] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more.

[0029] For ease of understanding, the following is a description of the process of the method provided by this implementation. Figure 1 , is a flow chart of the capsule detection method for gastrointestinal bleeding in the embodiment of the present application.

[0030] S101, after measuring the induction intensity and direction change of the magnetic material inside the capsule in the magnetic field through multiple magnetic field sensors, determine the capsule position information in the digestive tract, the capsule position information including real-time positioning data and motion trajectory data; The server is the core data processing and control center of the entire gastrointestinal bleeding capsule detection system. The server must first ensure the normal operation and precise layout of the magnetic field sensor. After receiving the start detection command sent by the user, the server will detect whether multiple magnetic field sensors are distributed in a circular array at the set part of the human body (such as around the abdomen). The specific detection method is: first control each magnetic field sensor to transmit a signal of a specific frequency and intensity, and then detect the intensity and phase of the signal received by other sensors. In the case of an ideal circular array distribution, the signal strength and phase between adjacent sensors should show a specific law to determine whether it conforms to the expected law of the circular array distribution. If it is detected that it conforms to the expected law, it is determined that the magnetic field sensor is distributed in a circular array. The server will start multiple magnetic field sensors for mutual calibration. In actual operation, each magnetic field sensor will receive the signal emitted by other sensors. By comparing the intensity, phase and other parameters of the received signal, the multi-sensor data fusion algorithm (such as weighted average method, Kalman filter algorithm, etc.) is used to process the measurement data of each magnetic field sensor, thereby realizing mutual calibration between magnetic field sensors and ensuring the accuracy and consistency of the measurement data. After the calibration is completed, multiple magnetic field sensors begin to measure the induction intensity and direction changes of the magnetic material inside the capsule in the magnetic field in real time.

[0031] The magnetic material inside the capsule will change its state in the magnetic field as the capsule moves in the digestive tract. The magnetic field sensor obtains the position information of the capsule by measuring this change. Specifically, the server knows that the magnetic material in the capsule has a certain amount of charge and speed (these parameters that have been determined when the capsule is designed are equivalent to known conditions). According to the relationship between the Lorentz force and the magnetic field strength, charge, and speed in the electromagnetic principle, the server can calculate the force exerted on the magnetic material in the magnetic field. Based on the calculated force, according to Newton's second law (the relationship between force and acceleration), the server can calculate the acceleration of the magnetic material. Acceleration reflects the speed of an object. Knowing the acceleration, the server can further infer how the speed of the magnetic material changes and how its position in space moves. By repeating the above calculation and analysis process, the server can continuously track the position changes of the magnetic material over time, thereby determining the specific position of the capsule at each moment, which is the real-time positioning data. Connecting the position points of the capsule at each moment forms the motion trajectory data of the capsule in the digestive tract, which can clearly show how the capsule moves in the digestive tract.

[0032] S102, fusing the capsule position information of the capsule with pre-set three-dimensional anatomical model data of the human digestive tract to obtain a three-dimensional visualization model including the capsule position information and the digestive tract structure; When the capsule position information of the capsule is fused with the pre-set three-dimensional anatomical model data of the human digestive tract to obtain a three-dimensional visualization model including the capsule position information and the digestive tract structure, the server needs to call the pre-stored three-dimensional anatomical model data of the human digestive tract.

[0033] The three-dimensional anatomical model data of the human digestive tract can be reconstructed through medical imaging data (such as CT, MRI, etc.) of a large number of human samples, with high-precision and detailed anatomical structure information. The server first converts the format of the acquired capsule position information and preprocesses the data to match the coordinate system and data format of the three-dimensional anatomical model data of the human digestive tract. Then, the capsule position information is accurately mapped to the three-dimensional anatomical model of the human digestive tract using a spatial registration algorithm (such as a registration algorithm based on feature points, a registration algorithm based on surfaces, etc.). During the registration process, the server will search for matching of the capsule position information with feature points or feature areas in the three-dimensional anatomical model of the human digestive tract, and achieve precise alignment of the two by calculating transformation matrices (such as translation, rotation, and scaling matrices).

[0034] The fused model will present the position and movement trajectory of the capsule in the human digestive tract, as well as its relative relationship with the surrounding digestive tract structures, in an intuitive 3D visualization. To improve the visualization effect, the server can also render the model, such as setting different materials and lighting effects, so that doctors can observe the situation of the capsule in the digestive tract more clearly, assisting doctors in diagnosis and analysis.

[0035] S103, acquiring first occult blood detection data information sent by an intelligent occult blood detection device pre-placed in the capsule through wireless communication, and recording current capsule position information, wherein the intelligent occult blood detection device detects hemoglobin concentration by detecting current changes between electrodes through an electrochemical detection method, and the first occult blood detection data information is determined by the hemoglobin concentration; The intelligent occult blood detection device detects the hemoglobin concentration by detecting the current change between electrodes through electrochemical detection. Its core principle is based on the redox characteristics of hemoglobin. In a specific electrochemical environment, hemoglobin will undergo redox reactions, resulting in changes in the current between electrodes. Here, the electrodes refer to the working electrode and the counter electrode. The intelligent occult blood detection device contains key components such as the working electrode, the counter electrode and the reference electrode. The surface of the working electrode is modified with biological sensitive materials (such as specific enzymes or nanomaterials) that have specific adsorption and catalytic effects on hemoglobin. When the capsule moves in the digestive tract and contacts substances containing hemoglobin, hemoglobin will be adsorbed to the surface of the working electrode and react, causing electron transfer between the electrodes and generating current changes. The microprocessor inside the intelligent occult blood detection device monitors and processes the current signal between the electrodes in real time, converts the current change into the hemoglobin concentration value according to the pre-set algorithm (the algorithm is based on a mathematical model between current change and hemoglobin concentration established based on a large amount of experimental data), and sends it to the server as the first occult blood detection data information through low-power Bluetooth technology wireless communication.

[0036] At the same time, when the server receives the first occult blood detection data information, it will immediately obtain the current capsule location information (including real-time positioning data and motion trajectory data) from the real-time data stored in it. The server associates and stores the first occult blood detection data information with the corresponding capsule location information for subsequent comprehensive analysis and processing. To ensure the accuracy and reliability of the data, the server will perform verification and error detection processing on the received data. For example, the cyclic redundancy check (CRC) algorithm is used to verify the data. If the verification result finds that the data is wrong or incomplete, the server will send a retransmission request to the intelligent occult blood detection device to ensure that the accurate first occult blood detection data information and the corresponding capsule location information are obtained. In this way, when the doctor checks the test results, he can not only understand the occult blood situation at different locations in the digestive tract, but also clarify its specific location, providing a strong basis for accurately determining the location of the bleeding point.

[0037] This solution uses low-power Bluetooth technology (BLE) to achieve data transmission. BLE technology has the characteristics of low power consumption and moderate transmission distance, which is very suitable for application scenarios such as capsule detection, which has strict requirements on power consumption and relatively close distance between devices. During the communication process, the server encrypts the transmitted data according to the pre-set communication encryption module. The specific encryption algorithm can adopt the Advanced Encryption Standard (AES) encryption algorithm, which has the advantages of high encryption strength and fast speed. Before data transmission, the server and the intelligent occult blood detection device share the encryption key in advance. The data sender (intelligent occult blood detection device) uses the key to encrypt the first occult blood detection data information to be transmitted, and the data receiver (server) uses the same key for decryption to ensure data transmission security, prevent the detection data from being stolen or tampered with during the transmission process, and protect the patient's privacy and the integrity of the detection data.

[0038] S104, obtaining digestive tract fluid analysis data information through the intelligent occult blood detection device; In addition to the function of detecting hemoglobin concentration, the intelligent occult blood detection device also integrates a variety of sensors for analyzing the composition of digestive tract fluid. Among them, the pH sensor is used to detect the pH of digestive tract fluid. Its principle is based on a specific electrochemical sensing principle, and the pH value of the solution is determined by measuring the potential difference between the electrode and the solution. The pH value of digestive tract fluid in different parts is different. For example, gastric fluid is usually acidic, while intestinal fluid is alkaline. By detecting the pH value, the position of the capsule in the digestive tract can be preliminarily determined, and a reference can be provided for the subsequent correction of occult blood detection data. In addition, it is also equipped with ion-selective electrodes for detecting the concentration of common ions in digestive tract fluid, such as sodium ions, potassium ions, chloride ions, etc. The concentration changes of these ions in digestive tract fluid are closely related to the physiological state and digestion process of the human body, and will also affect the electrochemical detection of hemoglobin concentration. The ion-selective electrode converts the ion concentration into an electrical signal output by selectively responding to specific ions.

[0039] The microprocessor in the intelligent occult blood detection device processes and integrates the data collected by these sensors in real time to form digestive fluid analysis data information, and then sends the digestive fluid analysis data information to the server through BLE wireless communication. After receiving the data, the server will also perform data verification and error processing to ensure the accuracy and integrity of the data. These digestive fluid analysis data information will be used together with the first occult blood detection data information as an important basis for subsequent correction calculations, providing more dimensional data support to improve the accuracy of occult blood detection.

[0040] S105, inputting the digestive tract fluid analysis data information and the first occult blood detection data information into an occult blood detection compensation model to obtain corrected second occult blood detection data information, wherein the occult blood detection compensation model is constructed by training multiple experimental data on the effects of changes in different digestive fluid components on hemoglobin concentration detection in an electrochemical detection method under a simulated digestive fluid environment; The occult blood detection compensation model is constructed by training multiple experimental data on the effects of changes in different digestive fluid components on hemoglobin concentration detection in the electrochemical detection method under a simulated digestive fluid environment. In the early stage of model construction, researchers can conduct a large number of simulation experiments, including accurately preparing simulated digestive fluid environments with different pH values ​​and different ion concentration combinations (covering the concentration range of common sodium ions, potassium ions, chloride ions, etc. that may appear in digestive tract fluids). These environments simulate the digestive fluid characteristics of different parts of the human digestive tract from the stomach to the intestines as realistically as possible. Then, hemoglobin samples with known precise concentrations are added to these simulated environments, and electrochemical detection equipment with the same principle and specifications as the intelligent occult blood detection device in the capsule is used for detection. The digestive fluid component parameters (such as pH value, various ion concentrations, etc.) and the corresponding detected hemoglobin concentration values ​​are recorded in detail for each test. By changing the digestive fluid composition and hemoglobin concentration many times, rich experimental data have been accumulated to form a large and comprehensive data set. In the model training stage, based on this data set, a suitable machine learning algorithm is selected to build the model. For example, a neural network algorithm is used, and the structural design of the neural network includes an input layer, a hidden layer, and an output layer. The number of nodes in the input layer corresponds to the number of digestive fluid component parameters (such as pH value, various ion concentrations, etc.), the number of nodes in the output layer is 1, representing the corrected hemoglobin concentration value, and the number of layers and nodes in the hidden layer are determined through multiple experiments and optimizations to ensure that the model has good fitting and generalization capabilities. The neural network is trained using the back propagation algorithm. During the training process, the weights and biases of the neural network are continuously adjusted to minimize the error between the corrected hemoglobin concentration value predicted by the model and the actual hemoglobin concentration value.

[0041] When the server receives the digestive fluid analysis data information and the first occult blood test data information in the actual test, it first pre-processes the data, checks the integrity and rationality of the data, and identifies and processes the abnormal values ​​(such as using mean substitution, interpolation, etc.). Then, the digestive fluid component data (including pH value, ion concentration, etc.) are sorted according to the format required by the model and input into the trained occult blood test compensation model. The model calculates the corresponding hemoglobin concentration correction value based on the input digestive fluid component data and its internal complex mathematical relationship and algorithm logic. Finally, the hemoglobin concentration value in the first occult blood test data information is added to the correction value to obtain the corrected second occult blood test data information. This process fully considers the impact of digestive fluid components on hemoglobin concentration detection, can effectively improve the accuracy of occult blood detection, provide doctors with more reliable test results, and help to more accurately judge gastrointestinal bleeding.

[0042] S106: Determine a display screen for the second occult blood detection data information and the capsule positioning information using a three-dimensional visualization model, and send the screen to a display terminal for display.

[0043] The server obtains the previously constructed three-dimensional visualization model containing the capsule position information and digestive tract structure from the storage. This model has been accurately integrated with the capsule's real-time positioning data and motion trajectory data and the three-dimensional anatomical model data of the human digestive tract. It can intuitively present the capsule's position and motion trajectory in the human digestive tract, as well as its relative relationship with the surrounding digestive tract structures.

[0044] The server then integrates the second occult blood detection data information and capsule positioning information into the three-dimensional visualization model. For the capsule positioning information, the current position of the capsule and its movement trajectory in the digestive tract are accurately marked in the model, and specific markings (such as bright spots or lines) are used to highlight it so that the doctor can clearly observe the movement of the capsule in the digestive tract. For the second occult blood detection data information, visual markings are made at the corresponding positions in the model according to the size of the hemoglobin concentration value. For example, the hemoglobin concentration value can be divided into different intervals, and each interval corresponds to a specific color or marking method (such as low concentration is represented by light color and high concentration is represented by dark color), so as to intuitively reflect the severity of occult blood in different parts of the digestive tract.

[0045] Next, the server determines the final display screen according to the preset display rules and the doctor's personalized settings (such as display angle, zoom ratio, etc.). When determining the display screen, the server will consider how to best display the relationship between the capsule position, occult blood situation, and digestive tract structure to facilitate the doctor's observation and diagnosis. For example, the display angle is automatically adjusted to make the capsule position in the center of the screen or highlight the suspected bleeding area; the model's zoom ratio is adjusted as needed to display the overall digestive tract situation and clearly display local details. Finally, the server sends the determined display screen to the display end through network transmission.

[0046] In some embodiments, when the server receives a rotation transformation gesture operation request from the display terminal, it first determines the rotation axis. It can be based on preset rules, such as the default vertical screen direction as the rotation axis, or it can allow users to customize the rotation axis according to a specific gesture combination. After determining the rotation axis, the rotation angle is determined according to the amplitude and direction of the gesture operation, and then the rotation matrix is ​​calculated. For each vertex coordinate in the three-dimensional visualization model, a specific calculation with the rotation matrix is ​​performed to obtain a new coordinate. Specifically, the original coordinate of each vertex is multiplied by the rotation matrix to determine its new position after rotation. Finally, the server updates the positions of all vertices of the model with the new coordinates to realize the rotation operation of the model around the set axis, so that the doctor can observe the model from different angles. When a zoom gesture operation request is received, the server obtains the zoom scale factor, which is calculated according to the amplitude of the gesture operation. The larger the finger pinching amplitude, the smaller the zoom scale factor, and the larger the finger opening amplitude, the larger the zoom scale factor. For each vertex coordinate of the three-dimensional visualization model, its coordinate value is multiplied by the zoom scale factor. For example, if the scale factor is less than 1, the vertex coordinate value becomes smaller, and the model is reduced as a whole; if it is greater than 1, the coordinate value becomes larger, and the model is enlarged as a whole. After completing the calculation, the server updates the coordinates of the model vertices, changes the relative distance between each vertex, and realizes the scaling of the overall size of the model, which is convenient for doctors to view the details of the model or the overall overview. When receiving a translation gesture operation request, the server determines the translation vector, whose direction and size are determined by the direction and distance of the gesture operation. Horizontal dragging corresponds to translation in the x-axis direction, vertical dragging corresponds to translation in the y-axis direction, and oblique dragging is decomposed into each axis. For each vertex coordinate of the model, the corresponding component of the translation vector is added respectively. After the server updates the vertex coordinates, the overall position of the model is translated, and the doctor can browse the different areas of the model through the translation operation. After completing the above transformation processing, the server organizes and packages the updated three-dimensional visualization model data to ensure that the data is complete and accurate, which is convenient for the display end to analyze. After receiving it, the display end redraws the model according to the preset rendering rules and display parameters. Due to the use of real-time communication and efficient processing methods, doctors can immediately see the changes in the model, interact smoothly, observe the relationship between the digestive tract and the capsule from different angles and scales, and assist in diagnosing the condition.

[0047] In the embodiments of the present application, the use of multi-sensor fusion positioning technology, electrochemical occult blood detection technology and three-dimensional visualization model construction and display technology effectively solves the problem in the prior art that capsule occult blood detection and positioning are easily interfered by digestive fluid and the detection results are not intuitive, thereby achieving the technical effect of precise positioning of gastrointestinal bleeding, accurate detection and intuitive presentation of detection results, thereby improving the accuracy of gastrointestinal bleeding detection.

[0048] After combining the above content, the following is a more detailed description of the process of the method provided by this implementation. Figure 2 , is another flow chart of the capsule detection method for gastrointestinal bleeding in the embodiments of the present application.

[0049] S201, receiving a capsule control instruction sent by a user terminal, and controlling the controllable magnetic field source according to the capsule control instruction, wherein the controllable magnetic field source is used to control the position and posture of the capsule in the digestive tract by adjusting the magnetic field strength and direction parameters; A stable and efficient communication link is established between the server and the user end, and the user end can be a console operated by a doctor or a special control software. When the user issues a capsule control instruction at the user end, the instruction is sent to the server through network transmission (such as a hospital internal LAN or a secure wireless network). After receiving the instruction, the server identifies the control intention contained in the instruction, such as forward, backward, left turn, right turn, rise, fall, and other operations, as well as the corresponding control parameters (such as the adjustment value of the magnetic field strength and direction parameters). The controllable magnetic field source is usually composed of multiple electromagnetic coils, which are distributed in specific positions around the human body (such as around the examination bed) and can generate a precise and controllable magnetic field. According to the parsed instructions, the server calculates the magnetic field strength and direction parameters that need to be adjusted, and sends these parameters to the controller of the controllable magnetic field source. After receiving the parameters, the controller changes the distribution and intensity of the magnetic field by adjusting the current size and direction in the electromagnetic coil, thereby applying precise magnetic force to the magnetic material inside the capsule. When the magnetic material inside the capsule is affected by the external magnetic field, it will undergo corresponding displacement and posture changes according to the direction and size of the magnetic field force, thereby realizing the position and posture adjustment of the capsule in the digestive tract. For example, when the capsule needs to move forward in the digestive tract, the server controls the controllable magnetic field source to generate a strong magnetic field attraction area in front of the capsule, and weakens the magnetic field in the rear, so that the capsule moves due to the forward force; when the posture of the capsule in the digestive tract needs to be adjusted, magnetic fields of different intensities are applied in different directions to make the capsule rotate or tilt.

[0050] During the control process, the server will receive feedback information from the magnetic field sensor in real time, including the current position and posture data of the capsule. The server will compare and analyze the feedback data with the target position and posture of the control command, and adjust the output parameters of the controllable magnetic field source through the closed-loop control algorithm to ensure that the capsule can accurately move to the target position and maintain the desired posture according to the command requirements. At the same time, the server will also monitor the working status of the controllable magnetic field source, including the temperature of the electromagnetic coil, current stability and other parameters. Once an abnormal situation is found, the server will take corresponding measures in time (such as adjusting the current size, suspending operation, etc.) to ensure the safe and stable operation of the system.

[0051] In some embodiments, each patient of different ages and weights responds differently to the results of magnetic field control of the digestive tract. Human tissue has certain electrical conductivity and magnetic permeability. Although fat or other tissues have relatively weak electrical conductivity, they can still be regarded as a medium in a magnetic field environment. When the magnetic field passes through the human body, the human tissue will shield and attenuate the magnetic field. In this case, a large number of relevant data of patients of different ages and weights (the degree of weight can be quantified by indicators such as body mass index BMI) can be collected in advance. These data include basic information of patients (age, height, weight, gender, etc.), physiological structural characteristics of the digestive tract (such as digestive tract length, diameter, curvature, etc., which can be obtained through medical imaging technology such as CT, MRI, etc.), and actual motion response data of the capsule in its digestive tract under the control of different magnetic field strength and direction parameters (such as motion speed, trajectory, posture change, etc.). Then select a suitable machine learning algorithm to construct a digestive tract magnetic field control parameter recommendation model. Taking a neural network as an example, its structural design includes an input layer, a hidden layer, and an output layer. The number of nodes in the input layer corresponds to the number of information features of the patient (such as age, BMI, digestive tract structural features, etc.), and the number of nodes in the output layer is the number of magnetic field intensity and direction parameters required by the controllable magnetic field source (such as magnetic field intensity value, components of magnetic field direction vector, etc.). The number of layers and nodes in the hidden layer are determined through multiple experiments and optimization to ensure that the model has good fitting and generalization capabilities. The preprocessed patient data is divided into a training set, a validation set, and a test set. The model is trained using the training set data. During the training process, the model continuously adjusts the internal weights and parameters based on the input patient information and the corresponding actual motion response data, and learns the relationship between different patient characteristics and the optimal magnetic field control parameters. When a new patient undergoes capsule testing, first obtain their basic information (such as age, height, weight, etc.) and possible digestive tract structural feature data (such as CT or MRI examination results performed in advance). These patient information is input into the trained digestive tract magnetic field control parameter recommendation model. The model predicts the magnetic field intensity and direction parameters of the controllable magnetic field source suitable for the patient based on the input patient information and the algorithm logic learned internally. For example, for young and thin patients (low BMI), the model may recommend a relatively low-intensity magnetic field because their tissues have relatively weak shielding and attenuation effects on the magnetic field; while for older and obese patients (higher BMI), a higher-intensity magnetic field may be required to achieve the same capsule control effect, and the magnetic field direction parameters may also need to be adjusted according to the structural characteristics of their digestive tract to ensure that the capsule can accurately reach the target location and maintain a stable posture. The server sends the magnetic field control parameters recommended by the model to the controller of the controllable magnetic field source, and the controller adjusts the current size and direction in the electromagnetic coil according to these parameters, thereby generating a suitable magnetic field to control the movement of the capsule in the patient's digestive tract.

[0052] S202, capturing the current multi-angle image information of the digestive tract through multiple cameras in the capsule; The capsule is equipped with multiple micro cameras, which are usually distributed in different positions of the capsule (such as the front end, rear end, side, etc.) to obtain image information from different angles of the digestive tract. The optical lens of the camera adopts a high-resolution, wide-angle design, which can clearly capture the details and pathological characteristics of the digestive tract mucosa. Each camera is connected to a built-in image acquisition module, which is responsible for converting the optical signal captured by the camera into a digital signal and performing preliminary image processing (such as noise reduction, contrast enhancement, etc.). The image acquisition module sends the processed image data to the server through a wireless communication module (such as the same low-power Bluetooth technology as in step S103 or other suitable wireless communication methods). After receiving the image data, the server first verifies the integrity and correctness of the data. After the verification is passed, the server organizes and numbers the image data from different cameras according to pre-set rules so that the image information of each angle can be accurately identified and processed later.

[0053] S203. Acquire multi-angle image information of the digestive tract through wireless communication, and determine the location information of the bleeding point in combination with a pre-set digestive tract bleeding determination model. The digestive tract bleeding determination model is trained in advance using a deep learning algorithm based on a plurality of digestive tract image sample data of known bleeding conditions and non-bleeding conditions.

[0054] The gastrointestinal bleeding determination model is obtained by training with a deep learning algorithm based on multiple known bleeding and non-bleeding gastrointestinal image sample data. During the model training phase, a large number of gastrointestinal images from actual clinical cases were collected, covering bleeding conditions of different degrees and different parts as well as normal gastrointestinal images. These images were annotated in detail, including the location, size, morphology and corresponding pathological characteristics of the bleeding points, and then a suitable deep learning algorithm framework was selected to build the model structure. Common model structures can include convolutional neural networks (CNNs) and their variants. The convolution layer of CNN can automatically extract features from images, the pooling layer is used to reduce the data dimension, and the fully connected layer is used to classify and predict the extracted features. By inputting the annotated image samples into the model for training, the model continuously adjusts the internal weights and parameters to learn how to identify the characteristic patterns of bleeding points from the image. When the server receives the multi-angle image information of the gastrointestinal tract in the actual detection, the image is first preprocessed to meet the requirements of the model input. Then, the preprocessed image is input into the trained gastrointestinal bleeding determination model. The model extracts and analyzes the features of the image, determines whether there are bleeding points in the image based on the learned feature patterns, and predicts the location and related feature information of the bleeding points. The server associates and fuses the bleeding point location information output by the model with the previously recorded capsule location information to determine the exact location of the bleeding point in the three-dimensional space of the human digestive tract.

[0055] In the embodiments of the present application, the use of active capsule control technology and multi-angle image acquisition and analysis technology effectively solves the problems in the prior art of difficulty in accurately inspecting specific areas of the digestive tract, inaccurate and unreliable positioning of bleeding points, and insufficiently comprehensive observation of lesions, thereby achieving the technical effect of improving the inspection effect of specific areas of the digestive tract and improving the accuracy and reliability of bleeding point positioning.

[0056] The following describes the server in the embodiment of the present invention from the perspective of hardware processing. Figure 3 , is a schematic diagram of a physical device structure of a server in an embodiment of the present application.

[0057] It should be noted that Figure 3 The structure of the server shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0058] like Figure 3As shown, the server includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 302 or the program loaded from the storage part 308 to the random access memory (RAM) 303, such as executing the method described in the above embodiment. In RAM 303, various programs and data required for system operation are also stored. CPU 301, ROM 302 and RAM 303 are connected to each other through bus 304. Input / output (I / O) interface 305 is also connected to bus 304.

[0059] The following components are connected to the I / O interface 305: an input section 306 including an audio input device, a button switch, etc.; an output section 307 including a liquid crystal display (LCD) and an audio output device, an indicator light, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. A removable medium 311, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 310 as needed so that a computer program read therefrom is installed into the storage section 308 as needed.

[0060] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication part 309, and / or installed from a removable medium 311. When the computer program is executed by the central processing unit (CPU) 301, various functions defined in the present invention are performed.

[0061] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device.

[0062] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings.

[0063] Specifically, the server of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, the capsule detection method for gastrointestinal bleeding provided in the above embodiment is implemented.

[0064] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the server described in the above embodiment; or may exist independently without being assembled into the server. The above storage medium carries one or more computer programs, and when the above one or more computer programs are executed by a processor of the server, the server implements the capsule detection method for gastrointestinal bleeding provided in the above embodiment.

[0065] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0066] As used in the above embodiments, the term "when..." may be interpreted to mean "if..." or "after..." or "in response to determining..." or "in response to detecting...", depending on the context. Similarly, the phrases "upon determining..." or "if (the stated condition or event) is detected" may be interpreted to mean "if determining..." or "in response to determining..." or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)", depending on the context.

[0067] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by computer programs to instruct related hardware, and the programs can be stored in computer-readable storage media. When the programs are executed, they can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: ROM or random access memory RAM, magnetic disk or optical disk and other media that can store program codes.

Claims

1. A capsule detection method for gastrointestinal bleeding, applied to a server, characterized in that: The method comprises: After measuring the induction intensity and direction change of the magnetic material inside the capsule in the magnetic field through multiple magnetic field sensors, the capsule position information of the capsule in the digestive tract is determined, and the capsule position information includes real-time positioning data and motion trajectory data; fusing the capsule position information of the capsule with pre-set three-dimensional anatomical model data of the human digestive tract to obtain a three-dimensional visualization model including the capsule position information and the digestive tract structure; Acquiring first occult blood detection data information sent by an intelligent occult blood detection device pre-placed in a capsule through wireless communication, and recording current capsule position information, wherein the intelligent occult blood detection device detects hemoglobin concentration by detecting current changes between electrodes through an electrochemical detection method, and the first occult blood detection data information is determined by the hemoglobin concentration; Obtaining digestive tract fluid analysis data information through the intelligent occult blood detection device; Inputting the digestive tract fluid analysis data information and the first occult blood detection data information into an occult blood detection compensation model to obtain corrected second occult blood detection data information, wherein the occult blood detection compensation model is constructed by training multiple experimental data on the effects of changes in different digestive fluid components on hemoglobin concentration detection in an electrochemical detection method under a simulated digestive fluid environment; The second occult blood detection data information and the capsule positioning information are used to determine a display screen using a three-dimensional visualization model, and sent to a display terminal for display.

2. The method according to claim 1, characterized in that After the steps of acquiring the first occult blood detection data information sent by the intelligent occult blood detection device pre-placed in the capsule through wireless communication and recording the current capsule position information, the method further includes: The capsule control instruction sent by the user end is received, and the controllable magnetic field source is controlled according to the capsule control instruction. The controllable magnetic field source is used to control the position and posture of the capsule in the digestive tract by adjusting the magnetic field strength and direction parameters.

3. The method according to claim 2, characterized in that After receiving the capsule control instruction sent by the user end and controlling the controllable magnetic field source according to the capsule control instruction, the method further includes: Multiple cameras in the capsule are used to capture multi-angle image information of the current digestive tract; The multi-angle image information of the digestive tract is acquired through wireless communication, and the location information of the bleeding point is determined by combining with a pre-set digestive tract bleeding determination model. The digestive tract bleeding determination model is trained in advance using a deep learning algorithm based on a plurality of digestive tract image sample data of known bleeding and non-bleeding conditions.

4. The method according to claim 1, characterized in that: After the step of determining a display screen using the second occult blood detection data information and the capsule positioning information by using a three-dimensional visualization model and sending the screen to a display terminal for display, the method further includes: The bleeding points on the display screen are marked with colors corresponding to different shades, and the shades are determined according to the hemoglobin concentration value in the second occult blood detection data information.

5. The method according to claim 1, characterized in that After measuring the induction intensity and direction change of the magnetic material inside the capsule in the magnetic field by multiple magnetic field sensors, before determining the capsule position information in the digestive tract, the method also includes: receiving an opening detection instruction sent by a user terminal, and detecting whether a plurality of magnetic field sensors are distributed in a ring array at a set part of a human body according to the opening detection instruction; If so, multiple magnetic field sensors are calibrated with each other.

6. The method according to claim 1, characterized in that After the step of determining a display screen using the second occult blood detection data information and the capsule positioning information by using a three-dimensional visualization model and sending the screen to a display terminal for display, the method further includes: After receiving a gesture operation request for a three-dimensional visualization model from a display terminal, performing corresponding transformation processing on the three-dimensional visualization model according to the gesture operation request, wherein the transformation processing includes determining a rotation operation of the three-dimensional visualization model around a set axis by calculating new coordinates of each vertex coordinate of the three-dimensional visualization model under the action of a rotation matrix when processing a rotation transformation; When processing the scaling transformation, adjusting the relative distances of the coordinates of each vertex of the three-dimensional visualization model according to the scaling factor to change the overall size of the three-dimensional visualization model; When processing the translation transformation, modifying the coordinate positions of all vertices of the three-dimensional visualization model according to the translation vector; After the transformation process is completed, the updated three-dimensional visualization model data is sent to the display end for real-time display.

7. The method according to claim 1, characterized in that Before the step of acquiring the first occult blood detection data information sent by the intelligent occult blood detection device pre-placed in the capsule through wireless communication and recording the current capsule position information, the method further includes: Using low-power Bluetooth technology to transmit data with the intelligent occult blood detection device; The transmitted data is encrypted according to the pre-set communication encryption module.

8. A server, characterized in that: The server includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the server to execute the method described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instructions are executed on a server, the server is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on a server, the server is caused to execute the method according to any one of claims 1 to 7.

Images