Endoscope washer reuse liquid concentration detection method and system and endoscope washer

Abstract

This invention belongs to the field of peracetic acid disinfectant concentration detection technology, and relates to a method, system, and endoscope disinfection machine for detecting the concentration of reused solution. The method includes acquiring the spectral information of each source domain sample in a source domain dataset and each target domain sample in a target domain dataset; extracting features from the spectral information of each sample to obtain a spectral feature vector; mapping the spectral feature vectors of the source domain samples and the spectral feature vectors of the target domain samples to obtain a transformation matrix for the target domain samples; constructing a detection model and pre-training the detection model using each source domain sample in the source domain dataset to obtain a pre-trained detection model; transforming the spectral feature vectors of each target domain sample using the transformation matrix, and training the pre-trained detection model using the transformed spectral feature vectors to obtain a target detection model, thereby detecting the concentration of reused solution from the endoscope disinfection machine based on spectral information.

Description

Technical Field

[0001] This invention relates to the field of peracetic acid disinfectant concentration detection technology, and in particular to a method for detecting the concentration of reusable endoscope cleaning and disinfection solution, a system for detecting the concentration of reusable endoscope cleaning and disinfection solution, and an endoscope cleaning and disinfection machine. Background Technology

[0002] In recent years, the incidence of cancer has risen, becoming a major problem affecting people's health. Early detection and treatment of malignant tumors can greatly reduce mortality. However, many malignant tumors have mild early symptoms and are hidden within the body's internal walls, making them difficult to detect externally. Examples include stomach cancer and esophageal cancer. This necessitates the use of tools such as endoscopes for in-depth diagnosis and treatment. Consequently, the pressure on endoscope cleaning and disinfection is increasing. Traditional manual cleaning and disinfection methods, or those using dedicated cleaning stations, are insufficient to meet the growing demand for endoscopic examinations. Furthermore, training professional cleaning and disinfection personnel is extremely costly and requires a high level of expertise from operators. Therefore, researchers have designed an automated endoscope cleaning and disinfection machine.

[0003] The disinfection process of endoscope washing and sterilization machines typically uses peracetic acid as its disinfectant. A large amount of peracetic acid solution is used to soak the endoscope to kill harmful substances such as Proteus and Helicobacter pylori that may remain after the enzyme washing process. Because peracetic acid solution is relatively expensive, it is usually collected and reused during the disinfection process. That is, after cleaning the endoscope, the peracetic acid solution is stored in a reuse container, and the reused solution is used to disinfect the endoscope for the next cleaning. However, because peracetic acid is unstable and easily volatilizes and decomposes, the concentration of the peracetic acid solution gradually decreases during repeated use, directly affecting the disinfection results. Therefore, it is necessary to monitor the concentration of the reused solution regularly and replace it with fresh disinfectant when the concentration is insufficient to ensure the disinfection effect of the endoscope.

[0004] Currently, the detection methods for peracetic acid involve professional testing personnel sampling and analyzing samples. This requires preparing complex sample solvents (buffer solutions, potassium iodide, etc.), titrating the sample solvent with peracetic acid, and finally calculating the concentration of peracetic acid based on the stoichiometric relationship of the reaction equation. This process is cumbersome and time-consuming. Alternatively, non-quantitative methods can be used, where peracetic acid reacts with a specific colorimetric reagent to generate a colored substance. The concentration of peracetic acid is then determined by comparing the colorimetric reaction with a color chart. However, the color stability of the colorimetric reaction itself is poor, and the concentration gradient of the color chart is large, resulting in low accuracy in concentration detection.

[0005] In summary, existing methods for detecting the concentration of reclaimed solution in endoscope cleaning and disinfection machines suffer from problems such as cumbersome detection process, long detection and analysis time, and low detection accuracy. Summary of the Invention

[0006] Therefore, the technical problem to be solved by the present invention is to overcome the problems of cumbersome detection process, long detection and analysis time and low detection accuracy in the existing methods for detecting the concentration of reclaimed liquid in endoscope cleaning and disinfection machines.

[0007] To solve the above technical problems, the present invention provides a method for detecting the concentration of reusable endoscope cleaning and disinfection solution, and a controller for a system for detecting the concentration of reusable endoscope cleaning and disinfection solution, comprising: The spectral information of each source domain sample in the source domain dataset and the spectral information of each target domain sample in the target domain dataset are obtained. Among them, each source domain sample is a prepared peracetic acid disinfectant solution of different concentrations, and each target domain sample is a reuse solution of different concentrations obtained from the endoscope washing and disinfection machine reuse tank. Feature extraction is performed on the spectral information of each source domain sample and each target domain sample to obtain the spectral feature vectors of each source domain sample and each target domain sample. The spectral feature vectors of the source domain sample and the target domain sample are mapped to minimize the distance between the spectral feature vectors of the target domain sample and the spectral feature vectors of the source domain sample, and the transformation matrix of the target domain sample is obtained. A detection model is constructed with spectral feature vectors as input and recycled liquid concentration as output; the detection model is pre-trained using the spectral feature vectors of each source domain sample in the source domain dataset to obtain a pre-trained detection model; The spectral feature vectors of each target domain sample are transformed using a transformation matrix, and the pre-trained detection model is trained using the transformed spectral feature vectors to obtain the target detection model. The target detection model is then used to detect the concentration of the endoscope cleaning and disinfection machine reuse solution.

[0008] Preferably, the endoscope disinfection machine reuse solution concentration detection system further includes a liquid level detection device and an optical detection device coupled to the controller. The liquid level detection device is used to detect the liquid level of the reuse solution in the endoscope disinfection machine reuse tank, and the optical detection device is used to acquire the spectral information of the reuse solution. The system uses a target detection model to detect the concentration of the endoscope disinfection machine reuse solution, including: Acquire liquid level information detected by the liquid level detection device; When the liquid level is greater than the preset threshold, a collection command is sent to the optical detection device so that the optical detection device can obtain the spectral information of the reused liquid in the endoscope cleaning and disinfection machine's reuse tank and obtain the spectral information obtained by the optical detection device. Feature extraction is performed on the spectral information of the recycled liquid to obtain the spectral feature vector of the recycled liquid; The spectral feature vector of the recycled liquid is transformed using a transformation matrix, and the transformed spectral feature vector of the recycled liquid is input into the target detection model to output the concentration of the recycled liquid.

[0009] Preferably, the optical detection device includes a transparent tubing, a light source, a probe, an optical fiber, and a spectrometer. The transparent tubing is connected to the endoscope washing and disinfection machine's reuse container to draw out the reused liquid from the container. The light source and spectrometer are both coupled to a controller, and the probe is connected to the spectrometer via an optical fiber. A data acquisition command is sent to the optical detection device to enable it to acquire the spectral information of the reused liquid in the endoscope washing and disinfection machine's reuse container. The acquired spectral information includes: Send a collection command to the light source to control the light source to turn on and irradiate the recycled liquid in the transparent tube, so that the probe can collect the optical information generated by the recycled liquid under the illumination of the light source and transmit the optical information to the spectrometer through optical fiber; The spectral information of the recycled liquid is obtained after the optical information is processed by the receiving spectrometer.

[0010] Preferably, after detecting the concentration of the endoscope cleaning and disinfection machine reuse solution using a target detection model, the method further includes: Support vector machines are used to detect and identify the spectral information of the reused liquid, determining whether the spectral information is normal or abnormal due to enzyme residue; and / or, The exponential smoothing method is used to calculate the predicted concentration of the recycled liquid at the next time step based on the concentration of the recycled liquid at the current time step, thereby predicting the number of times the recycled liquid can be used.

[0011] Preferably, the predicted concentration of the recycled liquid at the next time moment is calculated based on the current concentration of the recycled liquid using the exponential smoothing method, including: Calculate the total concentration of the reused liquid within a preset historical time period; The weight index decreases exponentially over time, and weight values ​​are assigned to the concentration of the recycled liquid at the current moment and the total concentration of the recycled liquid in the preset historical period, respectively, with the weight index decreasing exponentially over time. Calculate the product of the concentration of the recycled liquid at the current moment and its weight value to obtain the weighted data of the recycled liquid at the current moment. Calculate the product of the total concentration of the recycled liquid within a preset historical period and its weight value to obtain the weighted data of the total concentration of the recycled liquid within the preset historical period. The weighted data of the recycled liquid at the current moment is calculated and the weighted data of the total concentration of the recycled liquid in the preset historical period are added together to obtain the predicted concentration of the recycled liquid at the next moment.

[0012] Preferably, the endoscope cleaning and disinfection machine reuse solution concentration detection system further includes a PLC controller coupled to the controller, and the PLC controller is also connected to the solenoid valve and chemical pump of the endoscope cleaning and disinfection machine; after detecting the concentration of the endoscope cleaning and disinfection machine reuse solution using the target detection model, it further includes: When the concentration of the recycled liquid is lower than the preset minimum concentration, a recycled liquid discharge command is sent to the PLC controller so that the PLC controller controls the discharge solenoid valve of the endoscope washing and disinfection machine to open, thereby allowing the recycled liquid in the endoscope washing and disinfection machine's recycled tank to be discharged. Send a disinfectant addition command to the PLC controller so that the PLC controller can control the endoscope washing and disinfection machine to close the drain solenoid valve, open the inlet solenoid valve and the chemical pump, thereby adding new disinfectant to the endoscope washing and disinfection machine.

[0013] Preferably, the endoscope cleaning and disinfection machine reuse solution concentration detection system further includes a display device coupled to the controller; after detecting the concentration of the endoscope cleaning and disinfection machine reuse solution using the target detection model, it further includes: The concentration of the reused solution is transmitted to the display device so that the display device can indicate the concentration of the reused solution in the endoscope cleaning and disinfection machine's reuse tank; When the concentration of the reused liquid is lower than the preset minimum concentration, an alarm command is generated and transmitted to the display device so that the display device can show a warning that the concentration of the reused liquid in the endoscope cleaning and disinfection machine's reuse tank is too low.

[0014] Preferably, a detection model is constructed using the LASSO algorithm, with the input being a spectral feature vector and the output being the concentration of the recycled liquid, including: Based on the product of spectral feature vectors and coefficient matrices, a concentration prediction term for recycled liquid is constructed. Based on the sum of squared residuals between the actual concentration term and the predicted concentration term of the recycled liquid, a fitting error term for the detection model is constructed. Based on the product of the L1 norm of the coefficient matrix and its regularization strength, an L1 regularization term is constructed. The optimal coefficient matrix function is constructed with the goal of minimizing the sum of the fitting error term and the L1 regularization term of the detection model. The detection model is constructed based on the fact that the concentration of recycled liquid output by the detection model is equal to the product of the input spectral feature vector and the optimal coefficient matrix function.

[0015] The present invention also provides a system for detecting the concentration of reused solution in an endoscope cleaning and disinfection machine, comprising: The controller is used to implement the steps of the above-mentioned method for detecting the concentration of reclaimed solution in the endoscope cleaning and disinfection machine; A liquid level detection device, coupled to the controller, is used to detect the liquid level of the reuse liquid in the reuse tank of the endoscope washing and disinfection machine; An optical detection device, coupled to a controller, is used to acquire spectral information of the recycled liquid.

[0016] The present invention also provides an endoscope cleaning and disinfection machine, including the above-mentioned endoscope cleaning and disinfection machine reuse solution concentration detection system.

[0017] The method for detecting the concentration of reclaimed solution from an endoscope cleaning and disinfection machine provided in this application has the following advantages: Because hydrogen-containing groups in peracetic acid molecules exhibit combination and overtone absorption characteristics, and the total number of hydrogen-containing molecules varies in peracetic acid solutions of different concentrations, the absorption intensity of peracetic acid solutions at specific wavelengths differs. Therefore, the spectral information of peracetic acid solutions at different concentrations varies. To address this, this application considers constructing a detection model to predict the concentration of the recycled solution using its spectral data, eliminating the need for chemical treatment and simplifying the concentration detection process. Furthermore, since recycled solution samples are difficult to collect and are few in number during actual use, making it difficult to directly train a high-precision concentration detection model, this application introduces the concept of transfer learning. The detection model is pre-trained by obtaining source domain samples through the preparation of disinfectant solution. Considering the differences between the prepared disinfectant solution and the reused solution in the endoscope cleaning and disinfection machine, a transformation matrix is ​​constructed to eliminate the domain offset between the source and target domain samples. The pre-trained detection model is then retrained using the transformed target domain samples. The anti-interference capability of domain adaptation and domain transfer learning improves the detection accuracy of the model. This application establishes a detection model based on peracetic acid spectral analysis, achieving rapid, non-destructive, and accurate online detection of reused solution concentration. This overcomes the problems of frequent manual intervention, cumbersome process, and low efficiency and accuracy in detecting the reused solution concentration of endoscope cleaning and disinfection machines. Attached Figure Description

[0018] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings, wherein: Figure 1 Flowchart of the method for detecting the concentration of reclaimed solution from an endoscope cleaning and disinfection machine provided in this application; Figure 2 A schematic diagram of the endoscope cleaning and disinfection machine reuse solution concentration detection system provided in this application; Figure 3 This is a schematic diagram of the display page of the display device provided in the embodiments of this application; Figure 4 This is a schematic diagram of the endoscope cleaning and disinfection machine provided in this application; Explanation of reference numerals in the accompanying drawings: 1. Controller; 2. Liquid level detection device; 3. Optical detection device; 31. Transparent pipeline; 32. Light source; 33. Probe; 34. Fiber optic cable; 35. Spectrometer; 36. Spectrum acquisition chamber; 4. Endoscope washing and disinfection machine reuse container; 5. PLC controller; 6. Display device; 7. Endoscope washing and disinfection machine reuse solution concentration detection system; 8. Robotic arm; 9. Enzyme washing tank; 10. Disinfection tank; 11. Drying tank; 12. Exhaust fan; 13. Reuse solution manual sampling port. Detailed Implementation

[0019] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.

[0020] Because hydrogen-containing groups in peracetic acid molecules exhibit combination and overtone absorption characteristics, and the total number of hydrogen-containing molecules varies in peracetic acid solutions of different concentrations, the absorption intensity of peracetic acid solutions at specific wavelengths differs. Based on this theory, the spectral information of peracetic acid solutions of different concentrations under illumination will also differ. Therefore, this application, for the first time, considers utilizing the correlation between spectral information and peracetic acid concentration to construct a mapping relationship between the two. This allows for the prediction of peracetic acid concentration information simply by detecting the spectral information of the peracetic acid solution under a light source, eliminating the need for chemical treatment of the peracetic acid solution or the preparation of complex detection solutions to determine its concentration.

[0021] Please refer to the following: Figure 1 and Figure 2 , Figure 1 The diagram shown is a flowchart of the method for detecting the concentration of reusable solution in an endoscope cleaning and disinfection machine provided in this application. Figure 2 The diagram shows a schematic of the endoscope disinfection machine reuse solution concentration detection system used to implement the endoscope disinfection machine reuse solution concentration detection method. The endoscope disinfection machine reuse solution concentration detection method provided in this application is applied to the controller 1 in the endoscope disinfection machine reuse solution concentration detection system; the method specifically includes: S10: Obtain the spectral information of each source domain sample in the source domain dataset and the spectral information of each target domain sample in the target domain dataset; wherein, each source domain sample is a prepared peracetic acid disinfectant solution of different concentrations, and each target domain sample is a reuse solution of different concentrations obtained from the endoscope washing and disinfection machine reuse tank.

[0022] Specifically, the source domain dataset used in this application consists of 50 peracetic acid disinfectant solutions with concentrations ranging from 1.14 g / L to 183.6 g / L, prepared in the laboratory; the target domain dataset consists of 24 reused solution samples collected in an endoscope cleaning and disinfection machine, and their actual concentrations were determined by a professional testing institution.

[0023] Optionally, to ensure the accuracy of the spectral information of the samples, samples with abnormal spectral information can be removed by Mahalanobis distance. At the same time, extensive Savitzky-Golay convolution can be used to smooth the spectral information of each sample, filtering out noise interference in the spectral information as much as possible and highlighting the effective information of the spectrum.

[0024] S20: Extract features from the spectral information of each source domain sample and each target domain sample to obtain the spectral feature vectors of each source domain sample and each target domain sample.

[0025] In a specific example of this application, a competitive adaptive reweighted sampling algorithm is used to extract features from spectral information. Based on Darwin's "survival of the fittest" principle, the competitive adaptive reweighted sampling algorithm uses Monte Carlo sampling and adaptive reweighting to continuously select feature subsets from variables (spectral information) and calculate the weight of each feature subset according to its contribution to concentration detection. In each sampling, the feature subsets are reweighted, so that the weight of feature subsets that contribute more to concentration detection increases and the weight of feature subsets that contribute less decreases. After multiple iterations of sampling and reweighting, the spectral information is filtered to obtain the spectral feature vector.

[0026] Alternatively, other feature extraction networks can be used to extract features from spectral information, such as convolutional neural networks, recurrent neural networks, etc.

[0027] S30: Map the spectral feature vectors of the source domain sample and the spectral feature vectors of the target domain sample, with the goal of minimizing the distance between the spectral feature vectors of the target domain sample and the spectral feature vectors of the source domain sample, and obtain the transformation matrix of the target domain sample.

[0028] Specifically, considering the differences between the reused solution in actual use and the disinfectant solution prepared in the laboratory, i.e., the marginal probability distribution of the spectral feature vector of the source domain sample. Marginal probability distribution of samples in the target domain Unlike other methods, this application employs migration component analysis to map the spectral feature vectors of the source and target domain samples to a high-dimensional reproducing kernel Hilbert space. By minimizing the distance between the spectral feature vectors of the source and target domain samples in this space, while preserving the original internal features of each domain sample, the distributions of the two domain samples are approximated. This results in a transformation matrix that satisfies the marginal probability distribution of the mapped target domain sample's spectral feature vector. This allows the detection model trained on source domain samples to also be applicable to target domain samples.

[0029] S40: Construct a detection model with spectral feature vectors as input and recycled liquid concentration as output; pre-train the detection model using the spectral feature vectors of each source domain sample in the source domain dataset to obtain a pre-trained detection model.

[0030] Specifically, this application utilizes the LASSO algorithm to construct a detection model with spectral feature vectors as input and recycled liquid concentration as output, including: Step 1-1: Based on spectral feature vectors sum coefficient matrix The product of these terms is used to construct a concentration prediction term for the recycled liquid. .

[0031] Steps 1-2: Based on the actual concentration term Y of the recycled liquid and the predicted concentration term of the recycled liquid The sum of squared residuals between them is used to construct the fitting error term of the detection model. .

[0032] Steps 1-3: Based on the coefficient matrix L1 norm and its regularization strength The product of these terms is used to construct the L1 regularization term. .

[0033] Steps 1-4: Fitting the error term based on the detection model and L1 regularization term To minimize the sum, construct the optimal coefficient matrix function. .

[0034] Steps 1-5: Reclaimed liquid concentration based on the output of the detection model Equal to the input spectral feature vector With the optimal coefficient matrix function The product of these factors is used to construct the detection model. .

[0035] S50: The spectral feature vectors of each target domain sample are transformed using a transformation matrix, and the pre-trained detection model is trained using the transformed spectral feature vectors to obtain the target detection model. The target detection model is then used to detect the concentration of the endoscope cleaning and disinfection machine reuse solution.

[0036] Specifically, the object detection model is represented as .

[0037] Specifically, this application verified through multiple experiments that there is indeed a mapping relationship between the concentration of peracetic acid solution and its spectral information, and the experimental data are shown in Table 1: Table 1

[0038] The data in Table 1 show that for electromagnetic waves of various wavenumbers, changes in the concentration of peracetic acid solution alter its absorbance for electromagnetic waves, indicating a correlation between peracetic acid solution and its spectral information.

[0039] Furthermore, such as Figure 2 As shown, the endoscope cleaning and disinfection machine reuse liquid concentration detection system also includes a liquid level detection device 2 and an optical detection device 3 coupled to the controller 1. The liquid level detection device 2 is used to detect the liquid level of the reuse liquid in the endoscope cleaning and disinfection machine reuse tank 4, and the optical detection device 3 is used to acquire the spectral information of the reuse liquid.

[0040] Specifically, the liquid level detection device 2 uses a 0-24V photoelectric switch type sensor, and its output is connected to the I / O port of the controller 1.

[0041] Based on the aforementioned liquid level detection device 2 and optical detection device 3, step S50 involves using a target detection model to detect the concentration of the endoscope cleaning and disinfection machine's reused liquid, including: Step 2-1: Obtain the liquid level information detected by the liquid level detection device 2.

[0042] Step 2-2: When the liquid level is greater than the preset threshold, a collection command is sent to the optical detection device 3 so that the optical detection device 3 can acquire the spectral information of the reuse liquid in the endoscope cleaning and disinfection machine reuse tank 4 and acquire the spectral information acquired by the optical detection device 3.

[0043] Steps 2-3: Extract features from the spectral information of the recycled liquid to obtain the spectral feature vector of the recycled liquid.

[0044] Steps 2-4: Use the transformation matrix to transform the spectral feature vector of the recycled liquid, and input the transformed spectral feature vector of the recycled liquid into the target detection model to output the concentration of the recycled liquid.

[0045] Furthermore, such as Figure 2 As shown, the optical detection device 3 includes a transparent tube 31, a light source 32, a probe 33, an optical fiber 34, and a spectrometer 35. The transparent tube 31 is introduced into the side of the endoscope cleaning and disinfection machine's reuse container 4 to draw out the reused liquid from the container 4. The light source 32 is coupled to the controller 1 and is used to irradiate the reused liquid in the transparent tube 31. The probe 33 is connected to the spectrometer 35 via the optical fiber 34 to collect the optical information generated by the reused liquid when irradiated by the light source 32 and transmits the optical information to the spectrometer 35. The spectrometer 35 is coupled to the controller 1.

[0046] Specifically, a halogen lamp is selected as the light source 32, allowing the recycled liquid to be excited and display a continuous spectrum in the near-infrared spectral range. The spectrometer 35 is an industrial-grade fast Fourier transform near-infrared spectrometer with a wavenumber range of 4000 cm⁻¹. -1-12800cm -1 To ensure that the spectral resolution is at least 4 cm⁻¹ across the entire near-infrared spectral region. -1 This ensures the accuracy of spectral measurements. The spectrometer 35 is connected to the controller 1 via a USB interface for serial communication.

[0047] To ensure that the spectral acquisition is not affected by external natural light, the light source 32 and the probe 33 are located in the spectral acquisition chamber 36, such as... Figure 2 As shown.

[0048] Based on the aforementioned optical detection device 3, the controller 1 sends a data acquisition command to the optical detection device 3, enabling the optical detection device 3 to acquire the spectral information of the reused liquid in the endoscope cleaning and disinfection machine's reuse tank 4, and to acquire the spectral information acquired by the optical detection device 3, including: Step 3-1: Send a collection command to the light source 32 to control the light source 32 to turn on and irradiate the recycled liquid in the transparent tube 31, so that the probe 33 can collect the optical information generated by the recycled liquid under the illumination of the light source 32, and transmit the optical information to the spectrometer 35 through the optical fiber 34.

[0049] Step 3-2: Receive the spectral information of the recycled liquid obtained after the optical information is processed by the spectrometer 35.

[0050] Furthermore, since the enzyme solution is used to decompose organic matter (blood, mucus, etc.) during the cleaning and disinfection of endoscopes, it needs to be thoroughly removed by rinsing after disinfection. If the enzyme solution accumulates excessively in the reuse solution, it may adhere to the delicate structures such as the endoscope tubes due to excessive concentration, thus disrupting the concentration balance of the reuse solution. Therefore, in addition to checking whether the concentration of the reuse solution can still be used to achieve the disinfection purpose, it is also necessary to check whether there is excessive enzyme solution in the reuse solution.

[0051] Specifically, after using the target detection model to detect the concentration of the endoscope cleaning and disinfection machine reuse solution, the method also includes: using a support vector machine to detect and identify the spectral information of the reuse solution, and determining whether the spectral information of the reuse solution is normal spectral data or abnormal spectral data caused by enzyme residue.

[0052] When the spectral information of the reused solution is abnormal due to enzyme residue, it can promptly indicate that the concentration of the current reused solution is unbalanced and cannot continue to be used for endoscope disinfection and cleaning, thereby replacing the reused solution in the endoscope cleaning and disinfection machine's reuse tank 4.

[0053] Furthermore, controller 1 can also use the exponential smoothing method to calculate the predicted concentration of the recycled liquid at the next moment based on the concentration of the recycled liquid at the current moment, thereby predicting the number of times the recycled liquid can be used.

[0054] The core idea of ​​exponential smoothing is to perform a weighted average of historical data, and the weight decreases exponentially as the data time progresses, that is, recent data has a larger weight and older data has a smaller weight.

[0055] Specifically, the predicted concentration of the recycled liquid at the next time step is calculated using the exponential smoothing method based on the current concentration of the recycled liquid, including: Step 4-1: Calculate the total concentration of the reused liquid within the preset historical time period.

[0056] Step 4-2: Assign weight values ​​to the concentration of the recycled liquid at the current moment and the total concentration of the recycled liquid in the preset historical period, with the weight index decreasing exponentially over time as a constraint.

[0057] Step 4-3: Calculate the product of the concentration of the recycled liquid and its weight value at the current moment to obtain the weighted data of the recycled liquid at the current moment. Calculate the product of the total concentration of the recycled liquid and its weight value within the preset historical period to obtain the weighted data of the total concentration of the recycled liquid within the preset historical period.

[0058] Step 4-4: Calculate the weighted data of the recycled liquid at the current moment and the weighted data of the total concentration of recycled liquid in the preset historical period to obtain the predicted concentration of recycled liquid at the next moment.

[0059] The specific formula for calculating the predicted concentration of the reused liquid at the next moment is as follows: , in, This represents the predicted concentration of the recycled liquid at time t+1; This represents the concentration of the recycled liquid at time t; This indicates the total concentration of the recycled liquid within the preset historical time period; The weighted value representing the concentration of the recycled liquid at the current moment, also known as the smoothing coefficient, is used to balance the weights of near-term and long-term data. This represents the weighted value of the total concentration of the recycled liquid within a preset historical period.

[0060] For example, using this concentration prediction method, the concentration change of the reuse liquid in the endoscope cleaning and disinfection machine reuse tank 4 can be calculated within a future preset time period. If the concentration of the reuse liquid is lower than the preset minimum concentration (i.e., the reuse liquid concentration that can meet the endoscope disinfection requirements) within the future preset time period, the number of times the reuse liquid can be reused can be obtained. If the predicted concentrations for the next three times are all greater than the preset minimum concentration, and the predicted concentration for the fourth time is less than the preset minimum concentration, it means that the reuse liquid can be reused 3 more times.

[0061] Furthermore, such as Figure 2As shown, the endoscope cleaning and disinfection machine's reuse solution concentration detection system also includes a PLC controller 5 coupled to controller 1. The PLC controller 5 is also connected to the solenoid valve and chemical pump (not shown in the figure) of the endoscope cleaning and disinfection machine. Specifically, the PLC controller 5 is connected to the solenoid valve and chemical pump via its I / O ports, and data can be transmitted between controller 1 and PLC controller 5 via a network cable using the TCP / IP protocol.

[0062] After the controller 1 uses a target detection model to detect the concentration of the endoscope cleaning and disinfection machine's reuse solution, it also includes: Step 5-1: When the concentration of the recycled liquid is lower than the preset minimum concentration, a recycled liquid discharge command is sent to the PLC controller 5 so that the PLC controller 5 controls the discharge solenoid valve of the endoscope washing and disinfection machine to open, thereby allowing the recycled liquid in the endoscope washing and disinfection machine's recycled tank 4 to be discharged.

[0063] Step 5-2: Send a disinfectant addition command to PLC controller 5 so that PLC controller 5 controls the endoscope washing and disinfection machine to close the drain solenoid valve, open the inlet solenoid valve and the chemical pump, thereby adding new disinfectant to the endoscope washing and disinfection machine.

[0064] By designing the PLC controller 5, fully automatic control can be achieved from the detection of the reuse liquid concentration to the replacement. There is no need to manually take samples from the endoscope cleaning and disinfection machine reuse tank 4 to detect the reuse liquid concentration, and manually drain the reuse liquid from the endoscope cleaning and disinfection machine reuse tank 4 when the reuse liquid concentration is insufficient to achieve the disinfection effect, and then add new disinfectant to it.

[0065] Furthermore, such as Figure 2 As shown, the endoscope cleaning and disinfection machine reuse solution concentration detection system also includes a display device 6 coupled to the controller. Specifically, the display device 6 and the controller 1 can transmit data via a network cable using the TCP / IP protocol.

[0066] Specifically, after the controller 1 uses a target detection model to detect the concentration of the endoscope cleaning and disinfection machine's reuse solution, it also includes: Step 6-1: Transmit the concentration of the reused liquid to the display device 6 so that the display device 6 can indicate the concentration of the reused liquid in the endoscope cleaning and disinfection machine reuse tank 4.

[0067] Step 6-2: When the concentration of the reused liquid is lower than the preset minimum concentration, an alarm command is generated and transmitted to the display device 6 so that the display device 6 can display a warning that the concentration of the reused liquid in the endoscope cleaning and disinfection machine reuse tank 4 is too low.

[0068] Optionally, when the spectral information of the reused liquid is abnormal spectral data caused by enzyme residue, the display device 6 can also promptly indicate the current concentration imbalance of the reused liquid.

[0069] like Figure 3 The diagram shown is a schematic of the display interface of the display device 6 provided in a specific embodiment of this application. The display interface has two view windows for observing the concentration change trend of the recycled liquid and the current spectrum of the recycled liquid. It can also set a "Drainage Concentration" text box for users to set the required recycled liquid concentration, "Current Concentration" displays the current concentration value of the recycled liquid, "Estimated Number of Cleanings" box predicts the number of times the recycled liquid can be used for repeated cleaning, "Status Bar" indicates the current usage status, and "Alarm Window" is used to indicate abnormal status.

[0070] Based on the above-mentioned method for detecting the concentration of reusable endoscope cleaning and disinfection liquid, this application embodiment also provides a system for detecting the concentration of reusable endoscope cleaning and disinfection liquid, which includes a controller 1, a liquid level detection device 2, and an optical detection device 3.

[0071] The controller 1 is used to implement the steps of the above-mentioned method for detecting the concentration of the reused solution in the endoscope cleaning and disinfection machine.

[0072] The liquid level detection device 2 is coupled to the controller 1 and is used to detect the liquid level of the reuse liquid in the endoscope cleaning and disinfection machine reuse tank 4.

[0073] The optical detection device 3 is coupled to the controller 1 and is used to acquire the spectral information of the recycled liquid.

[0074] Furthermore, embodiments of this application also provide an endoscope cleaning and disinfection machine, which includes the aforementioned endoscope cleaning and disinfection machine reuse solution concentration detection system, such as... Figure 4 The diagram shown is a schematic of the endoscope washing and disinfection machine provided in this application. The endoscope washing and disinfection machine includes an endoscope washing and disinfection machine reuse solution concentration detection system U, a robotic arm 7, an enzyme washing tank 8, a disinfection tank 9, a drying tank 10, an exhaust fan 11, and a reuse solution manual sampling port 12. The endoscope washing and disinfection machine reuse solution concentration detection system U is located inside the endoscope washing and disinfection machine, and its display device 6 can be located at the reuse solution manual sampling port 12 to facilitate users to check the status of the reuse solution.

[0075] The endoscope washing and disinfection machine mainly consists of three functional tanks, corresponding to the three stages of enzyme washing, disinfection, and drying in the washing and disinfection process. The robotic arm 7 grasps the load rack carrying the endoscope and moves it between different stages. Since the washing speed of the enzyme washing stage is much shorter than the disinfection time, three disinfection tanks are provided to improve operating efficiency. An exhaust fan 11 is provided at the top for heat dissipation of the washing and disinfection machine and to remove water vapor and evaporating disinfectant gas generated during the washing process. The display device 6 can monitor the operation progress of the washing and disinfection machine in real time and the status of the reused liquid detected by the endoscope washing and disinfection machine reused liquid concentration detection system U. A manual sampling port 12 for the reused liquid is provided at the display device 6. After each endoscope is disinfected, the internal diaphragm pump will draw a part of the reused liquid from the reuse tank to the sampling port, so that the obtained sample can be tested by the test strip method and the result can be compared with the result detected by the endoscope washing and disinfection machine reused liquid concentration detection system U.

[0076] When the endoscope washing and disinfection machine is running, simply place the fixed endoscope into the pop-out enzyme washing tank drawer 8. After the machine passes the airtightness test, the drawer retracts and automatically executes the enzyme washing, disinfection, and drying process. During the process, the disinfectant properties need to be checked in time. After the cleaning is completed, the drying tank drawer 10 pops out, and the endoscope can be taken out. When an endoscope is disinfected, the disinfectant will be discharged into the reuse tank by gravity through the solenoid valve. When it needs to be reused, it will be discharged into the disinfection tank 9 through the chemical pump.

[0077] As described above, the endoscope disinfection machine reuse solution concentration detection system provided in this application utilizes near-infrared spectroscopy online analysis to achieve automatic detection of the reuse solution. Through reasonable hardware construction and software programming, it achieves rapid, non-destructive, and accurate detection of the reuse solution concentration. Furthermore, this method and system are compatible with the automatic control system of fully automated endoscope disinfection machines, achieving seamless system integration, effectively improving the automation level of endoscope disinfection machines, and bringing great convenience to technicians engaged in endoscope disinfection work.

[0078] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0079] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0080] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0081] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0082] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. An endoscope washer-disinfector reuse fluid concentration detection method, characterized by, A controller for detecting the concentration of reused solution in an endoscope cleaning and disinfection machine includes: The spectral information of each source domain sample in the source domain dataset and the spectral information of each target domain sample in the target domain dataset are obtained. Among them, each source domain sample is a prepared peracetic acid disinfectant solution of different concentrations, and each target domain sample is a reuse solution of different concentrations obtained from the endoscope washing and disinfection machine reuse tank. Feature extraction is performed on the spectral information of each source domain sample and each target domain sample to obtain the spectral feature vectors of each source domain sample and each target domain sample. The spectral feature vectors of the source domain sample and the target domain sample are mapped to minimize the distance between the spectral feature vectors of the target domain sample and the spectral feature vectors of the source domain sample, and the transformation matrix of the target domain sample is obtained. A detection model is constructed with spectral feature vectors as input and recycled liquid concentration as output; the detection model is pre-trained using the spectral feature vectors of each source domain sample in the source domain dataset to obtain a pre-trained detection model; The spectral feature vectors of each target domain sample are transformed using a transformation matrix, and the pre-trained detection model is trained using the transformed spectral feature vectors to obtain a target detection model. This target detection model is then used to detect the concentration of the endoscope cleaning and disinfection machine's reused solution. Specifically, this includes: Support vector machines are used to detect and identify the spectral information of the reused liquid, determining whether the spectral information is normal or abnormal due to enzyme residue; and / or, The exponential smoothing method is used to calculate the predicted concentration of the recycled liquid at the next time step based on the current concentration, thereby predicting the number of times the recycled liquid can be reused. Specifically, this includes: Calculate the total concentration of the reused liquid within a preset historical time period; The weight index decreases exponentially over time, and weight values ​​are assigned to the concentration of the recycled liquid at the current moment and the total concentration of the recycled liquid in the preset historical period, respectively, with the weight index decreasing exponentially over time. Calculate the product of the concentration of the recycled liquid at the current moment and its weight value to obtain the weighted data of the recycled liquid at the current moment. Calculate the product of the total concentration of the recycled liquid within a preset historical period and its weight value to obtain the weighted data of the total concentration of the recycled liquid within the preset historical period. The weighted data of the recycled liquid at the current moment is calculated and the weighted data of the total concentration of the recycled liquid in the preset historical period are added together to obtain the predicted concentration of the recycled liquid at the next moment, thereby predicting the number of times the recycled liquid can be used.

2. The method according to claim 1, wherein The endoscope cleaning and disinfection machine reuse liquid concentration detection system also includes a liquid level detection device and an optical detection device coupled to the controller. The liquid level detection device is used to detect the liquid level of the reuse liquid in the endoscope cleaning and disinfection machine reuse tank, and the optical detection device is used to acquire the spectral information of the reuse liquid. The concentration of the endoscope cleaning and disinfection system's reused solution was detected using a target detection model, including: Acquire liquid level information detected by the liquid level detection device; When the liquid level is greater than the preset threshold, a collection command is sent to the optical detection device so that the optical detection device can obtain the spectral information of the reused liquid in the endoscope cleaning and disinfection machine's reuse tank and obtain the spectral information obtained by the optical detection device. Feature extraction is performed on the spectral information of the recycled liquid to obtain the spectral feature vector of the recycled liquid; The spectral feature vector of the recycled liquid is transformed using a transformation matrix, and the transformed spectral feature vector of the recycled liquid is input into the target detection model to output the concentration of the recycled liquid.

3. The method according to claim 2, wherein The optical detection device includes a transparent tubing, a light source, a probe, an optical fiber, and a spectrometer. The transparent tubing is connected to the endoscope washing and disinfection machine's reuse container to draw out the reused liquid from the container. The light source and spectrometer are both coupled to a controller, and the probe is connected to the spectrometer via an optical fiber. A data acquisition command is sent to the optical detection device to enable it to acquire the spectral information of the reused liquid in the endoscope washing and disinfection machine's reuse container. The device also acquires the spectral information obtained by the optical detection device, including: Send a collection command to the light source to control the light source to turn on and irradiate the recycled liquid in the transparent tube, so that the probe can collect the optical information generated by the recycled liquid under the illumination of the light source and transmit the optical information to the spectrometer through optical fiber; The spectral information of the recycled liquid is obtained after the optical information is processed by the receiving spectrometer.

4. The method according to claim 1, wherein The endoscope cleaning and disinfection machine reuse solution concentration detection system also includes a PLC controller coupled to the controller, which is also connected to the solenoid valves and chemical pump of the endoscope cleaning and disinfection machine; after using a target detection model to detect the concentration of the endoscope cleaning and disinfection machine reuse solution, it also includes: When the concentration of the recycled liquid is lower than the preset minimum concentration, a recycled liquid discharge command is sent to the PLC controller so that the PLC controller controls the discharge solenoid valve of the endoscope washing and disinfection machine to open, thereby allowing the recycled liquid in the endoscope washing and disinfection machine's recycled tank to be discharged. Send a disinfectant addition command to the PLC controller so that the PLC controller can control the endoscope washing and disinfection machine to close the drain solenoid valve, open the inlet solenoid valve and the chemical pump, thereby adding new disinfectant to the endoscope washing and disinfection machine.

5. The method according to claim 1, wherein The endoscope cleaning and disinfection machine reuse solution concentration detection system also includes a display device coupled to the controller; After using a target detection model to detect the concentration of the endoscope cleaning and disinfection machine reuse solution, the following steps are also included: The concentration of the reused solution is transmitted to the display device so that the display device can indicate the concentration of the reused solution in the endoscope cleaning and disinfection machine's reuse tank; When the concentration of the reused liquid is lower than the preset minimum concentration, an alarm command is generated and transmitted to the display device so that the display device can show a warning that the concentration of the reused liquid in the endoscope cleaning and disinfection machine's reuse tank is too low.

6. The endoscope reprocessor reuse fluid concentration detection method according to claim 1, wherein, A detection model was constructed using the LASSO algorithm, with spectral feature vectors as input and the concentration of recycled liquid as output, including: Based on the product of spectral feature vectors and coefficient matrices, a concentration prediction term for recycled liquid is constructed. Based on the sum of squared residuals between the actual concentration term and the predicted concentration term of the recycled liquid, a fitting error term for the detection model is constructed. Based on the product of the L1 norm of the coefficient matrix and its regularization strength, an L1 regularization term is constructed. The optimal coefficient matrix function is constructed with the goal of minimizing the sum of the fitting error term and the L1 regularization term of the detection model. The detection model is constructed based on the fact that the concentration of recycled liquid output by the detection model is equal to the product of the input spectral feature vector and the optimal coefficient matrix function.

7. A system for detecting the concentration of reused solution in an endoscope cleaning and disinfection machine, characterized in that, include: The controller is used to implement the steps of the endoscope cleaning and disinfection machine reuse solution concentration detection method according to any one of claims 1 to 6; A liquid level detection device, coupled to the controller, is used to detect the liquid level of the reuse liquid in the reuse tank of the endoscope washing and disinfection machine; An optical detection device, coupled to a controller, is used to acquire spectral information of the recycled liquid.

8. An endoscope cleaning and disinfection machine, characterized in that, Includes the endoscope cleaning and disinfection machine reuse solution concentration detection system as described in claim 7.

Images