Control system for in-vivo flushing of endoscope

Abstract

The invention discloses an endoscope in-vivo flushing control system, which comprises an endoscope, a surgery power device, a first information acquisition device, a second information acquisition device, an image processing device, a detection module and a data control module, and is characterized in that joint point potential and myoelectricity potential data are acquired through the first information acquisition device; a first information acquisition device on the endoscope acquires an image of a focus in the body through the endoscope and processes the image to obtain first diagnosis and treatment data, a second information acquisition device on the endoscope acquires an image of the focus in the body through the endoscope and processes the image to obtain second diagnosis and treatment data, and the data control module matches a first flushing parameter based on fusion of the first diagnosis and treatment data and the second diagnosis and treatment data. The endoscope operates at a particular irrigation flow rate and irrigation pressure based on the first irrigation parameter. Further, the data control module generates a second flushing parameter according to the change of the patient articulation point myoelectric potential data, and the endoscope works at a specific flushing flow rate and flushing pressure based on the first flushing parameter.

Description

Technical Field

[0001] The present invention relates to the technical field of endoscope in-vivo flushing, and in particular to a control system for endoscope in-vivo flushing. Background Art

[0002] With the development of minimally invasive medical technology, endoscopes have become an indispensable tool in modern surgical diagnosis and treatment. Endoscopic technology not only realizes real-time observation of internal tissues, but also can cooperate with surgical instruments to complete complex interventional operations. It is especially widely used in cavity surgeries such as the digestive system, urinary system and respiratory system. In order to improve the clarity of the endoscope's intraoperative field of view and the exposure rate of lesions, it is often necessary to clean, cool or locally perfuse the observation area through in vivo flushing. Traditional in vivo flushing control systems mostly rely on manual or simple mechanical settings, lacking the ability to dynamically respond to the patient's physiological state and real-time lesion images during surgery, resulting in hysteresis, uncertainty and insufficient adaptation to individual differences in the flushing process, which is not conducive to the development requirements of precision medicine.

[0003] Although some high-end endoscope systems on the market currently have a certain degree of image enhancement or automatic flushing function, their control logic is generally based on static image recognition or preset flushing mode, ignoring the changes in the patient's bioelectric signals during surgery, especially the dynamic changes in electromyographic response and joint potential. These physiological signals can often reflect the patient's potential pathological state or sensitive reaction during surgery, and are an important data source for building intelligent flushing control algorithms. In addition, the existing system is still a shallow matching mode in processing image information and physiological signal fusion decision-making, lacking deep integration and real-time feedback mechanism for multi-dimensional diagnosis and treatment information, and it is difficult to achieve personalized, closed-loop flushing control. In recent years, with the rapid progress of artificial intelligence, bioelectric detection technology and image processing algorithms, intelligent medical control systems driven by fusion diagnosis and treatment data have gradually become a research hotspot. Especially in the endoscope application scenario, how to combine high-frequency signals such as joint potential and electromyographic potential collected from the body surface with lesion images collected by the endoscope, analyze the coupling relationship between lesion status and physiological load through algorithm modeling, and then dynamically adjust the flushing parameters to achieve refined and responsive flushing control is a technical bottleneck that needs to be broken through. Summary of the invention

[0004] In view of the above problems, the present invention provides a control system for endoscopic internal flushing. The control system extracts the pathological features of the lesion images in the body through the endoscope to obtain the first diagnosis and treatment data, calculates the impedance in the body based on the detection electrode and the reference electrode to obtain the second diagnosis and treatment data, and matches the optimal first flushing parameters by fusing the first diagnosis and treatment data and the second diagnosis and treatment data. Further, by detecting the change in the myoelectric potential of the patient during the process of the endoscope flushing the body, it is judged whether the current flushing flow rate and pressure of the endoscope are too large, and the first flushing parameters are automatically adjusted to the second flushing parameters to ensure the best treatment effect.

[0005] The object of the invention of the present application can be achieved by the following technical means:

[0006] A control system for endoscopic internal flushing, comprising an endoscope, a surgical power device, a first information collection device, a second information collection device, an image processing device, a detection module and a data control module. Among them, the surgical power device, the first information collection device, the second information collection device, the image processing device, the detection module and the data control module are all arranged on the endoscope, and the second information collection device is signal-connected to the image processing device;

[0007] The detection module is respectively signal-connected to the first information collection device and the image processing device, and the data control module is connected to the detection module;

[0008] Among them,

[0009] The endoscope is used to penetrate into the patient's body and flush drugs into the body;

[0010] The surgical power device is used to adjust the working parameters of the drugs flushed into the body by the endoscope;

[0011] The first information collection device is arranged at the detection point on the patient's skin surface and is used to collect joint point potential and myoelectric potential data;

[0012] The second information collection device is used to collect lesion images;

[0013] The image processing device is used to extract the pathological features of the lesion images;

[0014] The detection module is used to generate the first diagnosis and treatment data according to the joint point potential and generate the second diagnosis and treatment data according to the pathological features;

[0015] The data control module is used to generate the first flushing parameters according to the first diagnosis and treatment data and the second diagnosis and treatment data, and adjust the first flushing parameters to the second flushing parameters based on the myoelectric potential data.

[0016] In the present invention, the surgical power device adjusts the water flow rate and impact degree based on the first flushing parameters and the second flushing parameters.

[0017] In the present invention, the pathological feature of the lesion image is the gray value of the diseased tissue in the patient's body The texture feature C of the diseased tissue = α1·∑ i,j P(i,j) 2 +α2·[-∑ i,j P(i,j)·log2P(i,j)] and the shape feature of the diseased tissue where Q is the total number of pixels of the lesion image, l g is the gray value of the g-th pixel of the lesion image, where is the average gray value of each pixel of the lesion image, α1 and α2 are weight coefficients, i is the abscissa of the feature point pixel, j is the ordinate of the feature point pixel, P(i,j) is the gray level co-occurrence matrix of the pixel (i,j), x1 and x2 are the abscissas of two adjacent feature points, and y1 and y2 are the ordinates of two adjacent feature points.

[0018] In the present invention, at least 3 feature points in the lesion image are identified. The feature points include the lesion aggregation region, the lesion edge region, and the lesion color change region. The lesion image is converted into a single-channel grayscale image, the pixel windows of each feature point are selected, and the gray value K of the diseased tissue, the shape feature E of the diseased tissue, and the texture feature C of the diseased tissue of the pixel window are respectively extracted.

[0019] In the present invention, the detection module includes a detection electrode and a reference electrode; the detection module obtains an electric field with an intensity of I formed by the first information acquisition device, the detection module collects the potential difference M1 of the detection electrode, the detection module collects the potential difference M2 of the reference electrode, and the detection module calculates the body impedance Z = (M1 - M2) / I, and takes this body impedance Z as the first diagnosis and treatment data.

[0020] In the present invention, the detection electrode and the reference electrode are placed parallel to the patient's muscle fibers at the detection point.

[0021] In the present invention, the data processing unit further has a database unit and a data processing unit thereon, where

[0022] The database unit is used to store historical diagnosis and treatment data;

[0023] The data processing unit is used to generate the first flushing parameter and process the electromyogram potential data.

[0024] In the present invention, the historical diagnosis and treatment data includes at least one historical case. Any historical case includes patient information, first diagnosis and treatment data, second diagnosis and treatment data, first flushing parameter, and second flushing parameter. Any patient information is a one-to-one mapping of the first diagnosis and treatment data, the second diagnosis and treatment data, the first flushing parameter, and the second flushing parameter.

[0025] In the present invention, the data control module traverses the historical diagnosis and treatment data, selects and extracts the first flushing parameter from the historical diagnosis and treatment data based on the first diagnosis and treatment data and the second diagnosis and treatment data, and generates the first flushing parameter as the second diagnosis and treatment data.

[0026] In the present invention, the method for the data processing unit to process the electromyogram potential data includes the following steps:

[0027] Step 1: Convert the electromyogram potential data into an electromyogram signal x(t), and generate the root mean square of the electromyogram. N is the duration of the electromyogram signal x(t), and t is the moment;

[0028] Step 2: If the root mean square of the electromyogram T min <R<T max , then generate the second flushing parameter and send the second flushing parameter to the endoscope. Otherwise, send the first flushing parameter to the endoscope, where T min is the lower limit value of the root mean square of the electromyogram, and T max is the upper limit value of the root mean square of the electromyogram;

[0029] Step 3: Construct historical diagnosis and treatment data with the pathological features, the second diagnosis and treatment data, the first flushing parameter, and the second flushing parameter, and send the historical diagnosis and treatment data to the data control module.

[0030] Implementing a control method and system for endoscopic internal flushing of the present invention has the beneficial effects that: this control method for internal flushing can monitor the changes in the electromyogram potential of the patient during the internal flushing process in real time, reasonably adjust the flow rate and pressure of the endoscopic flushing, reduce the irritation to the body caused by the drug fluid during the internal flushing process, reduce the incidence of complications, and the safety and effectiveness of the treatment effect are better. In addition, in the complete technical solution disclosed by the present invention, the pathological features of the lesion image and the internal impedance are extracted to obtain the first diagnosis and treatment data and the second diagnosis and treatment data, and the first diagnosis and treatment data and the second diagnosis and treatment data can be used as a reference for the subsequent clinical treatment of patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] To more clearly illustrate the technical solutions of the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments recorded in the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0032] Figure 1 Hardware block diagram of the control system for endoscopic in-vivo irrigation of the present invention;

[0033] Figure 2 Method flowchart for the second information acquisition device of the present invention to identify at least three feature points in the lesion image;

[0034] Figure 3 Method flowchart for the data control module of the present invention to generate the first irrigation parameter based on the fusion of the first diagnosis and treatment data and the second diagnosis and treatment data;

[0035] Figure 4 Method flowchart for the data processing unit of the present invention to process electromyogram potential data. Specific embodiments

[0036] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention.

[0037] In a preferred embodiment of the technical solution disclosed in the present application, the endoscope is inserted into the patient's tissue, and the second information acquisition device on the endoscope is used to obtain the lesion situation in the patient's tissue, and the data is transmitted through wired connection. To solve the problem that the traditional in-vivo irrigation depends on the doctor's experience and the control of the irrigation flow rate and pressure is improper, the present application discloses a technical solution for an endoscopic in-vivo irrigation device.

[0038] Refer to Figure 1 , the control system for endoscopic in-vivo irrigation includes an endoscope, a surgical power device, a first information acquisition device, a second information acquisition device, an image processing device, a detection module, and a data control module. Among them, the endoscope is used to be inserted into the patient's body and inject drugs into the body. The drugs play a role in infiltrating the body and flushing the remaining lesion residues.

[0039] The surgical power device is used to adjust the working parameters of the drugs injected into the body by the endoscope. The purpose of adjusting the working parameters of the drugs injected into the body is to minimize the negative impact on the healthy tissues in the body caused by excessive irrigation flow rate and pressure of the drugs, and improve the effect of drug infiltration into the body and flushing of the remaining lesion residues.

[0040] The first information acquisition device is arranged at the detection point on the patient's skin surface and is used to acquire joint point electric potential and electromyogram potential data. In this embodiment, the detection point is the body position marking point of the in-vivo flexion angle axis. Preferably, according to the human body's electrical characteristics, the first information acquisition device selects a frequency of 50 Hz and a relative dielectric constant of 190. At this frequency and relative dielectric constant, the interference received by the first information acquisition device during operation is the smallest, and the best human body phase signs are obtained.

[0041] The second information acquisition device is arranged on the endoscope and is used to acquire lesion images. In this embodiment, the second information acquisition device is a micro camera and can acquire in-vivo lesion images. Optionally, the second information acquisition device can also transmit a video stream composed of multiple frames of lesion images to the data control module in the form of a video stream.

[0042] Furthermore, the control system for in-vivo flushing of the endoscope further includes an image processing device. The image processing device is connected to the second information acquisition device, and the image processing device extracts the pathological features of the lesion image based on preset feature points. In this embodiment, the feature points include the lesion aggregation area, the lesion edge area, and the lesion color change area. Lesion foci exist in the body in an aggregated form, forming a lesion area. Selecting the lesion aggregation area as a feature point can focus on the shape, size, and distribution of the lesion area; the edge of the lesion focus is the transition area between the lesion area and the surrounding normal tissue and contains important information about the degree of the lesion. Extracting the edge features of the lesion focus can quantitatively evaluate the edge contour of the lesion and analyze the morphological features of the lesion; color changes in the lesion image reflect the physiological state, metabolic activity, or degree of the lesion of the tissue.

[0043] The detection module is used to generate the first diagnosis and treatment data based on the joint point electric potential and the second diagnosis and treatment data based on the pathological features. In this embodiment, the detection module also has at least one detection electrode and at least one reference electrode. The detection electrode and the reference electrode are made of Ag / AgCl (silver / silver chloride) material. The detection electrode and the reference electrode are placed parallel to the patient's muscle fibers at the detection point. Among them, the detection electrode and the reference electrode generate an electric field with an intensity of I, a potential difference M1 is formed on the detection electrode, and a potential difference M2 is formed on the reference electrode.

[0044] In this embodiment, the first information acquisition device acquires the electric field with an intensity of I generated by the detection electrode and the reference electrode, the potential difference M1 of the detection electrode, and the potential difference M2 of the reference electrode, and sends them to the detection module. The detection module calculates the in-vivo impedance Z = (M1 - M2) / I and uses this in-vivo impedance Z as the first diagnosis and treatment data.

[0045] The in-vivo impedance is related to the state of internal tissues, the degree of lesions, and physiological characteristics in the body. When there are lesions in the human body, the electrical properties of the internal tissues will change, leading to a change in the in-vivo impedance. By obtaining the in-vivo impedance Z and using the in-vivo impedance Z as the first diagnostic data, it serves as a factor affecting the determination of the first flushing parameter.

[0046] In this embodiment, the second information acquisition device identifies at least three feature points in the lesion image. The feature points include the lesion aggregation area, the lesion edge area, and the lesion color change area. The lesion image is converted into a single-channel grayscale image, and pixel windows of each feature point are selected, and the grayscale value K of the lesion tissue, the shape feature E of the lesion tissue, and the texture feature C of the lesion tissue are respectively extracted.

[0047] Specifically, referring to Figure 3 , the method for the second information acquisition device to identify at least three feature points in the lesion image includes the following steps:

[0048] Step 101: Smooth the lesion image through the Gaussian filter G(x,y), where τ is the standard deviation of the Gaussian filter;

[0049] Step 102: Calculate the gradients of the lesion image in the x direction and the y direction. Among them, The gradient amplitude is obtained The gradient direction r = arctan(G y / G x ). For each pixel in the lesion image, only the local maximum value in its gradient direction is retained;

[0050] Step 103: Set the Gaussian threshold interval [T h , T l , classify the gradient amplitude f. If f is greater than T h , then this pixel point is a strong edge. If f ∈ [T h , T l , then this pixel point is a weak edge.

[0051] Step 104: Screen the set U of all pixel points of the strong edge U = {u1, u2,..., u n}, perform convolution processing on the set U of pixel points. If at least one pixel value is 1 in the area covered by any one element in the set U of pixel points, then the central pixel of this element is marked as 1. Otherwise, the central pixel of this element is marked as 0.

[0052] Step 105: The elements in the set U of pixel points where all central pixels are marked as 1 form the lesion aggregation region; the elements in the set U of pixel points where all central pixels are marked as 0 form the edge region of the lesion aggregation region.

[0053] Step 106: Classify each element in the set U of pixel points into RGB channels, and calculate the squared difference D of the RGB channels of each element: D = 1 / 3[(R - G) 2 + (G - B) 2 + (B - R) 2 . Traverse the squared difference D of the RGB channels of each element, and the elements with D > T o form the lesion color change region.

[0054] In the lesion image, different types of internal tissues usually have different gray values. By collecting the gray values of the lesion tissues at the feature points, density data of the lesion tissues can be obtained; by collecting the shape features of the lesion tissues at the feature points, the size and shape information of the lesion tissues can be obtained; by collecting the texture features of the lesion tissues at the feature points, the aggregation of the lesion tissues in the body and the junction with the surrounding tissues can be obtained.

[0055] In this embodiment, the gray value K of the lesion tissue, the shape feature E of the lesion tissue, and the texture feature C of the lesion tissue are extracted. The above pathological features are all quantified to improve the speed of the data control module for matching the first flushing parameter based on historical diagnosis and treatment data, reduce the redundancy of the data control module, and improve the robustness of the data matching process.

[0056] Furthermore, during the quantification process of the pathological features, the pathological features of the lesion image are the gray values of the lesion tissues in the patient's body The texture feature C of the lesion tissue = α1·∑ i,j P(i,j) 2 + α2·[-∑ i, j P(i,j)·log2P(i,j)] and the shape feature of the lesion tissue where Q is the total number of pixels in the lesion image, l g is the gray value of the g-th pixel, where is the mean gray value, α1 and α2 are weight coefficients, i is the abscissa of the feature point pixel, j is the ordinate of the feature point pixel, P(i,j) is the gray level co-occurrence matrix of the pixel (i,j), x1 and x2 are the abscissas of two adjacent feature points, and y1 and y2 are the ordinates of two adjacent feature points.

[0057] It should be understood that the pathological feature quantification process provided in this embodiment is a technical standard applicable to the complete technical solution of this application. The actual pathological feature quantification process depends on different image processing means adopted. When other pathological feature quantification processes or methods are applicable, they are equivalent to the technical standard proposed in this embodiment.

[0058] A data control module, configured to generate a first flushing parameter according to the first diagnosis and treatment data and the second diagnosis and treatment data, and adjust the first flushing parameter to a second flushing parameter based on the electromyogram potential data.

[0059] In this embodiment, the data processing unit further has a database unit and a data processing unit thereon. Among them, the database unit is used to store historical diagnosis and treatment data; the data processing unit is used to generate a first flushing parameter and process the electromyogram potential data.

[0060] The data control module traverses the historical diagnosis and treatment data, selects and extracts the first flushing parameter in the historical diagnosis and treatment data based on the first diagnosis and treatment data and the second diagnosis and treatment data, and generates the first flushing parameter as the second diagnosis and treatment data. Optionally, the fusion of the first diagnosis and treatment data and the second diagnosis and treatment data is performed in the data control module. Refer to Figure 3 and the method for generating the first flushing parameter involves the following steps:

[0061] Step 101: Extract the in-vivo impedance Z in the first diagnosis and treatment data, and extract the gray value K, the shape feature E, and the texture feature C of the lesion tissue in the second diagnosis and treatment data;

[0062] Step 102: Construct a feature vector X = [Z, K, E, C], and normalize the feature vector X, X n =(X - μ) / σ, where μ is the mean value of the feature vector X, and σ is the standard deviation of the feature vector X;

[0063] Step 103: Traverse the historical diagnosis and treatment data. For any historical case v, calculate the feature vector Y of the historical case v, and normalize the feature vector Y to obtain Y n ;

[0064] Step 105: Calculate the feature vector X n and the feature vector Y n The cosine similarity Sim = (X n ·Y v ) / (||X n ||·||Y v ||);

[0065] Step 106: Select the historical case u with the largest cosine similarity Sim, extract the first flushing parameter of the historical case u, and output the first flushing parameter to the data processing unit.

[0066] In this embodiment, the historical diagnosis and treatment data includes at least one historical case. Any historical case includes patient information, first diagnosis and treatment data, second diagnosis and treatment data, first flushing parameter, and second flushing parameter. Any patient information is a one-to-one mapping of the first diagnosis and treatment data, the second diagnosis and treatment data, the first flushing parameter, and the second flushing parameter.

[0067] In this embodiment, the first flushing parameter and the second flushing parameter include the flow rate and flushing pressure of endoscope flushing. The surgical power device has an adjustable valve. The opening degree of the valve is adjusted based on the first flushing parameter and the second flushing parameter to adjust the water flow speed and flushing pressure.

[0068] In this embodiment, the method for the data processing unit to process the electromyogram potential data refers to Figure 4 , and includes the following steps:

[0069] Step 201: Filter, amplify, and denoise the original signal of the electromyogram potential data, convert it into an electromyogram signal x(t), and generate the root mean square of the electromyogram N is the duration of the electromyogram signal x(t), and t is the time;

[0070] Step 2: If the root mean square of the electromyogram T min <R<T max , then generate the second flushing parameter and send the second flushing parameter to the endoscope. Otherwise, send the first flushing parameter to the endoscope. Among them, T min is the lower limit value of the root mean square of the electromyogram, and T max is the upper limit value of the root mean square of the electromyogram;

[0071] Step 3: Construct historical diagnosis and treatment data from the first diagnosis and treatment data, the second diagnosis and treatment data, the first flushing parameter, and the second flushing parameter, and send the historical diagnosis and treatment data to the data control module.

[0072] In this embodiment, the current first diagnosis and treatment data, second diagnosis and treatment data, first flushing parameter, and second flushing parameter are combined into a historical case. After being sent to the data control module, they are input into the database unit, and the database unit updates the historical diagnosis and treatment data.

[0073] It should be understood that in this embodiment, the historical diagnosis and treatment data includes at least one historical case. The more the number of historical cases, the more accurate the result obtained by the method for the data processing unit to process the electromyogram potential data through continuous accumulation of historical data.

[0074] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A control system for endoscopic in-vivo irrigation, characterized in that It includes an endoscope, a surgical power device, a first information acquisition device, a second information acquisition device, an image processing device, a detection module, and a data control module. Among them, the surgical power device, the first information acquisition device, the second information acquisition device, the image processing device, the detection module, and the data control module are all arranged on the endoscope. The second information acquisition device is signal-connected to the image processing device; The detection module is respectively signal-connected to the first information acquisition device and the image processing device, and the data control module is connected to the detection module; Among them, The endoscope is used to penetrate into the patient's body and inject drugs into the body; The surgical power device is used to adjust the working parameters of the drugs injected into the body by the endoscope; The first information acquisition device is arranged at the detection point on the patient's skin surface and is used to collect joint point potential and electromyogram potential data; The second information acquisition device is used to collect lesion images; The image processing device is used to extract the pathological features of the lesion images; The detection module is used to generate first diagnosis and treatment data based on the joint point potential and generate second diagnosis and treatment data based on the pathological features; The data control module is used to generate first flushing parameters based on the first diagnosis and treatment data and the second diagnosis and treatment data, and adjust the first flushing parameters to second flushing parameters based on the electromyogram potential data.

2. The control system for in-vivo irrigation of an endoscope according to claim 1, characterized in that, The surgical power device adjusts the flow rate and impact degree of the water flow based on the first flushing parameters and the second flushing parameters.

3. The control system for in-vivo irrigation of an endoscope according to claim 1, characterized in that, The pathological features of the lesion image are the gray values of the diseased tissues in the patient's body The texture feature C of the diseased tissue = α1·∑ i,j P(i,j) 2 +α2·[-∑ i,j P(i,j)·log2P(i,j)] and the shape feature of the diseased tissue where Q is the total number of pixels of the lesion image, l g is the gray value of the g-th pixel of the lesion image, where is the average gray value of each pixel of the lesion image, α1 and α2 are weight coefficients, i is the abscissa of the feature point pixel, j is the ordinate of the feature point pixel, P(i,j) is the gray-level co-occurrence matrix of the pixel (i,j), x1 and x2 are the abscissas of two adjacent feature points, and y1 and y2 are the ordinates of two adjacent feature points.

4. The control system for in-vivo irrigation of an endoscope according to claim 3, characterized in that, Identify at least 3 feature points in the lesion image. The feature points include the lesion aggregation area, the lesion edge area, and the lesion color change area. Convert the lesion image into a single-channel grayscale image. Select the pixel windows of each feature point, and respectively extract the gray value K of the lesion tissue, the shape feature E of the lesion tissue, and the texture feature C of the lesion tissue.

5. The control system for endoscopic in-vivo irrigation according to claim 1, characterized in that, The detection module includes a detection electrode and a reference electrode; the detection module obtains an electric field with an intensity of I formed by the first information acquisition device. The detection module collects the potential difference M1 of the detection electrode, the detection module collects the potential difference M2 of the reference electrode, and the detection module calculates the body impedance Z = (M1 - M2) / I, and takes this body impedance Z as the first diagnosis and treatment data.

6. The control system for in-vivo irrigation of an endoscope according to claim 5, characterized in that, The detection electrode and the reference electrode are placed parallel to the patient's muscle fibers at the detection point.

7. The control system for in-vivo irrigation of an endoscope according to claim 1, wherein The data processing unit also has a database unit and a data processing unit thereon. Among them, The database unit is used to store historical diagnosis and treatment data; The data processing unit is used to generate first flushing parameters and process electromyogram potential data.

8. The control system for endoscopic in-vivo irrigation according to claim 7, characterized in that, The historical diagnosis and treatment data contains at least one historical case. Any historical case includes patient information, first diagnosis and treatment data, second diagnosis and treatment data, first flushing parameters, and second flushing parameters. Any patient information is a one-to-one mapping of the first diagnosis and treatment data, the second diagnosis and treatment data, the first flushing parameters, and the second flushing parameters.

9. The control system for in-vivo irrigation of an endoscope according to claim 7, characterized in that, The data control module traverses the historical diagnosis and treatment data, selects and extracts the first flushing parameters in the historical diagnosis and treatment data based on the first diagnosis and treatment data and the second diagnosis and treatment data, and generates the first flushing parameters as the second diagnosis and treatment data.

10. The control system for in-vivo irrigation of an endoscope according to claim 7, characterized in that, The method for the data processing unit to process electromyogram potential data includes the following steps: Step 1: Convert the electromyogram potential data into an electromyogram signal x(t), and generate the root mean square of the electromyogram N is the duration of the electromyogram signal x(t), and t is the time instant; Step 2: If the root mean square of electromyogram T min <R<T max , then generate the second flushing parameter and send the second flushing parameter to the endoscope. Otherwise, send the first flushing parameter to the endoscope, where T min is the lower limit value of the root mean square of electromyogram, and T max is the upper limit value of the root mean square of electromyogram; Step 3: Construct historical diagnosis and treatment data from the pathological features, second diagnosis and treatment data, first flushing parameters, and second flushing parameters, and send the historical diagnosis and treatment data to the data control module.

Images