Bacterial pollution propagation simulation method and system based on colored aerosol

Abstract

The invention provides a bacterial contamination propagation simulation method and system based on colored aerosol, and belongs to the technical field of sanitation and safety. The method comprises the following steps: simulating a toilet scene; by releasing colored aerosol composed of colored small balls, a flushing experiment is executed under the conditions that a toilet lid is opened and closed respectively; recording the diffusion path, range and attachment position of the aerosol by using sampling equipment and a camera, and generating aerosol pollution data; hand pollution data is generated through contact between the hand model and the pollution surface, pollution transfer from the hand to food is further simulated, and food pollution data is generated; the polluted food is input into the intestinal tract simulation device, particle distribution and interaction with the inner wall of the intestinal tract are observed through an optical sensor, and a lesion simulation result is recorded; and integrating multi-source data and evaluating comprehensive risks through a machine learning algorithm. Quantitative simulation of the whole chain of bacterial transmission is realized, and an efficient technical means is provided for reducing the risk of cross infection and optimizing sanitary measures.

Description

Technical Field

[0001] This invention relates to the field of health and safety technology, and in particular to a method and system for simulating bacterial contamination transmission based on colored aerosols. Background Technology

[0002] With the continuous development of health-related fields, research on the transmission pathways of aerosols in enclosed spaces and their impact on human health has gradually become a hot topic. Particularly in bathroom settings, toilet flushing can lead to the generation and spread of aerosols, which is considered a significant route for the transmission of bacteria and viruses. Existing research has shown that the diffusion range and attachment sites of aerosol particles during toilet flushing may potentially contaminate environmental surfaces and human bodies, leading to pathogens entering the human body through hand contact and food ingestion, causing intestinal diseases.

[0003] Current research on aerosol transmission largely focuses on single stages, such as the diffusion behavior of aerosols in the air or their deposition on specific surfaces. Such studies lack simulations of the complex interactions along actual transmission paths, such as a complete chain analysis from environmental surfaces to hand contact, food contamination, and human ingestion. Existing technologies struggle to quantify the impact of aerosol pollution on human health, especially in enclosed spaces (such as bathrooms), where the risks of aerosol diffusion and multi-pathway transmission triggered by toilet flushing have not been systematically assessed.

[0004] Existing research has not adequately considered the impact of varying environmental conditions on aerosol propagation pathways. For example, the effect of the toilet seat's opening / closing position on the aerosol diffusion range, attachment location, and subsequent pollution transmission remains unclear, and quantifiable comparative experimental data are lacking. This makes it difficult to provide a scientific basis for optimizing hygiene habits (such as closing the toilet seat when flushing) and to assess differences in health risks under different conditions.

[0005] Current technologies generally lack comprehensive simulation methods for the transmission chain of contamination, especially regarding the transmission of hand contamination to food and the mechanisms of action of contaminants after entering the human gut. Systematic experimental models have not yet been established. Existing methods struggle to simultaneously analyze the correlation between the spread of aerosol contamination, the degree of hand contamination, the concentration of food contamination, and gut health risks, leading to biased and inaccurate risk assessments. Summary of the Invention

[0006] To overcome the shortcomings of the prior art, the purpose of this invention is to provide a method and system for simulating bacterial contamination transmission based on colored aerosols, which can comprehensively evaluate the bacterial contamination transmission effect under different conditions.

[0007] To achieve the above objectives, the present invention provides the following solution:

[0008] A method for simulating bacterial contamination transmission based on colored aerosols, comprising:

[0009] An aerosol sampling device, a hand model, and a camera were installed at the experimental toilet to simulate a bathroom scene. After adding liquid to the experimental toilet, the flushing device was controlled to perform a flushing operation. Experiments simulating bacterial contamination with colored aerosols were conducted under two experimental conditions: with the toilet lid open and closed. The experiment simulating bacterial contamination with colored aerosols was carried out by releasing colored aerosols into the experimental toilet. The colored aerosols consisted of multiple colored spheres.

[0010] The aerosol sampling device and the camera are used to record the diffusion path, range and attachment position of the colored balls during the flushing process to generate aerosol pollution distribution data. The hand model is then brought into contact with the contaminated surface to record the attachment distribution of the colored balls on the hand model, generating hand pollution data.

[0011] The contaminated hand model is brought into contact with simulated food material, and the range and concentration of contamination from the hand to the food surface are recorded to generate food contamination data. The contaminated simulated food material is then placed in a simulated intestinal device, and the distribution of the colored balls in the simulated intestinal device and the interaction between the colored balls and the intestinal wall are observed in real time through optical sensors to record the lesion simulation results.

[0012] The aerosol diffusion data, hand contamination distribution data, food contamination data, and lesion simulation results are integrated, and the risk simulation results are determined through a preset machine learning algorithm.

[0013] Preferably, it further includes:

[0014] The risk simulation results under two conditions—toilet seat open and closed—were compared and analyzed to output quantitative risk indicators and assess the differences in bacterial transmission pathways and health risk levels under different conditions.

[0015] Preferably, the aerosol sampling device is positioned 10-30 cm above the toilet seat of the experimental toilet, and is used to collect aerosols generated by flushing; the surface material of the hand model simulates the friction coefficient and hydrophobicity of human skin, and the hand model is placed on the side of the experimental toilet at a simulated sink position; the camera is fixed in front of and to the side of the experimental toilet to record the aerosol diffusion trajectory.

[0016] Preferably, the parameters for the two experimental conditions are:

[0017] When the toilet seat is open, the angle between the toilet seat and the horizontal plane is 90°, and the flushing water flow rate is 10-15L / min.

[0018] When the toilet lid is closed, the angle between the toilet lid and the horizontal plane is 0°, and the flushing water flow intensity is 10-15L / min.

[0019] Preferably, the particle size of the colored microspheres ranges from 1 to 10 μm; the density of the colored microspheres is 1.0 to 1.2 g / cm³. 3 The colored spheres include red spheres and blue spheres; the red spheres represent aerosols directly generated by flushing water, and the blue spheres represent secondary diffusion particles.

[0020] Preferably, the aerosol sampling device and the camera are used to record the diffusion path, range, and attachment location of the colored spheres during the flushing process to generate aerosol pollution distribution data. The hand model is then brought into contact with the contaminated surface to record the attachment distribution of the colored spheres on the hand model, generating hand contamination data, including:

[0021] The aerosol sampling device collects aerosol samples generated during the flushing process in real time; the aerosol sampling device includes a particle counter and a filter, used to capture the concentration and spatial distribution of colored spheres;

[0022] The camera records the movement trajectory of the colored ball at a shooting frequency of 30-60 frames per second;

[0023] The image data captured by the camera is subjected to trajectory tracking analysis. Combined with aerosol sampling data, a three-dimensional diffusion path model is generated through image processing algorithms, and a heat map of the attachment location is marked.

[0024] Based on the three-dimensional diffusion path model and the heat map, the diffusion range and surface adhesion density of aerosols are quantified, and aerosol pollution distribution data are generated.

[0025] Preferably, the contaminated hand model is brought into contact with simulated food material, and the contamination range and concentration of colored spheres transferred from the hand to the food surface are recorded to generate food contamination data. The contaminated simulated food material is then placed in a simulated intestinal device, and the distribution of the colored spheres in the simulated intestinal device and the interaction between the colored spheres and the intestinal wall are observed in real time using an optical sensor to record the lesion simulation results, including:

[0026] The contaminated hand model is driven to contact the surface of the simulated food material at a preset contact angle and contact pressure; the contact time is 2-4 seconds.

[0027] The distribution of colored spheres was captured by scanning the surface of simulated food materials using a multispectral imager.

[0028] Based on image analysis algorithms, the boundaries of polluted areas in the distribution image are identified, and the proportion of polluted area and the number of particles per unit area are calculated.

[0029] Combining the density parameters of the colored balls, using the formula The contamination concentration is calculated, and the contamination concentration, the proportion of contaminated area, and the number of particles per unit area are determined as the food contamination data; where C is the contamination concentration, N is the number of particles, ρ is the density parameter of the colored spheres, and A is the contaminated area.

[0030] The simulated food material is fed into a simulated intestinal device made of a transparent flexible tube; the inner wall of the transparent flexible tube is coated with a viscoelastic material that simulates the intestinal mucosa; the elastic modulus of the viscoelastic material is in the range of 1-5 kPa and the viscosity is in the range of 0.1-0.5 Pa·s.

[0031] The movement trajectory of the colored ball in the simulated intestine and its collision and adsorption behavior with the inner wall were observed simultaneously using a fiber optic endoscope and the camera.

[0032] Fluorescent labeling technology was used to determine the dynamic migration path of the colored spheres, and a particle distribution model was generated using a three-dimensional reconstruction algorithm.

[0033] The particle distribution model was simulated and analyzed to obtain particle retention time, mucosal adhesion density, and local pressure peak.

[0034] The particle retention time, the mucosal adhesion density, and the local pressure peak are input into the trained lesion assessment model to obtain lesion simulation results; the lesion simulation results include lesion risk level and simulated pathology report.

[0035] Preferably, the preset machine learning algorithm is a CNN+LSTM neural network.

[0036] A bacterial contamination transmission simulation system based on colored aerosols, comprising:

[0037] The experimental setup unit is used to set up an aerosol sampling device, a hand model, and a camera at the experimental toilet to simulate a bathroom scene. After adding liquid to the experimental toilet, the unit controls the toilet's flushing device to perform a flushing operation. Experiments simulating bacterial contamination with colored aerosols are conducted under two conditions: the toilet lid is open and closed. The experiment simulating bacterial contamination with colored aerosols is carried out by releasing colored aerosols into the experimental toilet. The colored aerosols consist of multiple colored spheres.

[0038] The first data acquisition unit is used to record the diffusion path, range and attachment position of the colored balls during the flushing process using the aerosol sampling device and the camera, generate aerosol pollution distribution data, and bring the hand model into contact with the polluted surface to record the attachment distribution of the colored balls on the hand model to generate hand pollution data.

[0039] The second data acquisition unit is used to bring the contaminated hand model into contact with the simulated food material, record the contamination range and concentration of the colored balls transferred from the hand to the food surface, generate food contamination data, and place the contaminated simulated food material in a simulated intestinal device. The distribution of the colored balls in the simulated intestinal device and the interaction between the colored balls and the intestinal wall are observed in real time through an optical sensor to record the lesion simulation results.

[0040] The integration unit is used to integrate the aerosol diffusion data, the hand contamination distribution data, the food contamination data, and the lesion simulation results, and to determine the risk simulation results through a preset machine learning algorithm.

[0041] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0042] This invention provides a method and system for simulating bacterial contamination transmission based on colored aerosols. The method includes: setting up an aerosol sampling device, a hand model, and a camera at an experimental toilet to simulate a bathroom scene; adding liquid to the experimental toilet and controlling the flushing device to perform a flushing operation; and conducting experiments simulating bacterial contamination with colored aerosols under two experimental conditions: the toilet lid is open and closed. The experiment simulating bacterial contamination with colored aerosols is conducted by releasing colored aerosols into the experimental toilet. The colored aerosols consist of multiple colored spheres. The aerosol sampling device and the camera are used to record the diffusion path, range, and attachment position of the colored spheres during flushing, generating aerosol... The invention generates hand contamination data by quantifying the distribution of aerosol contamination and then contacting the hand model with a contaminated surface to record the adhesion distribution of colored spheres on the hand model. Next, the contaminated hand model is contacted with simulated food material to record the range and concentration of colored spheres transferred from the hand to the food surface, generating food contamination data. The contaminated simulated food material is then placed in a simulated intestinal device, and the distribution of the colored spheres in the simulated intestinal device and their interaction with the intestinal wall are observed in real time using optical sensors to record the pathological simulation results. The aerosol diffusion data, hand contamination distribution data, food contamination data, and pathological simulation results are integrated, and a pre-set machine learning algorithm is used to determine the risk simulation results. This invention achieves quantitative simulation of the entire bacterial transmission chain, providing an efficient technical means to reduce the risk of cross-infection and optimize hygiene measures. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 A flowchart of the method provided in an embodiment of the present invention;

[0045] Figure 2 A first data acquisition flowchart provided for an embodiment of the present invention;

[0046] Figure 3 This is a second data acquisition flowchart provided in an embodiment of the present invention;

[0047] Figure 4 This is a schematic diagram of the system structure provided in an embodiment of the present invention. Detailed Implementation

[0048] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] The purpose of this invention is to provide a method and system for simulating bacterial contamination transmission based on colored aerosols, which realizes quantitative simulation of the entire bacterial transmission chain and provides an efficient technical means to reduce the risk of cross-infection and optimize hygiene measures.

[0050] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0051] Figure 1 The method flowchart provided in the embodiments of the present invention is as follows: Figure 1 As shown, this invention provides a method for simulating bacterial contamination transmission based on colored aerosols, comprising:

[0052] Step 100: An aerosol sampling device, a hand model, and a camera are set up at the experimental toilet to simulate a bathroom scene. After adding liquid to the experimental toilet, the flushing device is controlled to perform a flushing operation. The experiment simulating bacterial contamination with colored aerosols is conducted under two experimental conditions: with the toilet lid open and closed. The experiment simulating bacterial contamination with colored aerosols is conducted by releasing colored aerosols into the experimental toilet. The colored aerosols consist of multiple colored spheres.

[0053] Step 200: Use aerosol sampling equipment and a camera to record the diffusion path, range and attachment location of the colored balls during the flushing process to generate aerosol pollution distribution data. Then, put the hand model into contact with the contaminated surface and record the attachment distribution of the colored balls on the hand model to generate hand pollution data.

[0054] Step 300: The contaminated hand model is brought into contact with simulated food material, and the range and concentration of contamination from the hand to the food surface are recorded to generate food contamination data. The contaminated simulated food material is then placed in a simulated intestinal device, and the distribution of the colored balls in the simulated intestinal device and the interaction between the colored balls and the intestinal wall are observed in real time through optical sensors to record the lesion simulation results.

[0055] Step 400: Integrate aerosol diffusion data, hand contamination distribution data, food contamination data, and lesion simulation results, and determine the risk simulation results through a preset machine learning algorithm.

[0056] Preferably, it further includes:

[0057] The risk simulation results under two conditions—toilet seat open and closed—were compared and analyzed to output quantitative risk indicators and assess the differences in bacterial transmission pathways and health risk levels under different conditions.

[0058] Preferably, the aerosol sampling device is positioned 10-30 cm above the toilet seat of the experimental toilet, and is used to collect aerosols generated by flushing; the surface material of the hand model simulates the friction coefficient and hydrophobicity of human skin, and the hand model is placed on the side of the experimental toilet at a simulated sink position; the camera is fixed in front of and to the side of the experimental toilet to record the aerosol diffusion trajectory.

[0059] Specifically, in this embodiment, a simulated bathroom scene was constructed around the experimental toilet, including the toilet itself, a simulated sink, and related equipment. An aerosol sampling device was vertically installed 10-30cm above the toilet seat, with its sampling port facing inwards, to capture aerosol particles generated during flushing. A hand model was made of silicone with a specially treated surface to simulate the friction coefficient and hydrophobicity of human skin. The model was fixed to a pre-set position on the simulated sink to the side of the toilet, simulating hand contact behavior after use. Two high-definition cameras were fixed to the front and side of the toilet, respectively. The front camera covered the entire field of view of the toilet, while the side camera was installed at a 45-degree angle to ensure complete recording of the aerosol diffusion trajectory and the interaction process with the hand model.

[0060] Before the experiment, a predetermined volume of liquid was injected into the toilet to simulate actual use. After the flushing device was activated, flushing was performed with the toilet lid fully open and closed. During flushing, a colored aerosol composed of small colored spheres was instantaneously released into the water through a release device located inside the toilet, simulating the physical diffusion behavior of bacterial contamination. An aerosol sampling device collected particle samples in real time, while a camera recorded the movement path, diffusion range, and attachment position of the colored spheres on environmental surfaces at a shooting frequency of 30-60 frames per second. The experiment was repeated multiple times to ensure data stability. After each experiment, the equipment was cleaned and the initial conditions were reset.

[0061] After particle samples captured by aerosol sampling equipment are collected through a filter membrane, the concentration and size distribution of colored spheres are analyzed using a particle counter. Video data captured by a camera is imported into image processing software, and a three-dimensional diffusion model of the aerosol is reconstructed using a trajectory tracking algorithm, marking the density thermograms at the attachment sites. After a hand model comes into contact with a contaminated surface, the surface adhesion is captured by a high-resolution imaging system, and image segmentation technology is used to count the number of particles and the coverage area in different regions. All data are integrated into an analysis platform, and combined with preset algorithms, aerosol pollution distribution map, hand pollution thermogram, and quantitative indicators are generated, providing basic data support for subsequent risk assessment.

[0062] Preferably, the parameters for the two experimental conditions are:

[0063] When the toilet seat is open, the angle between the toilet seat and the horizontal plane is 90°, and the flushing water flow rate is 10-15L / min.

[0064] When the toilet lid is closed, the angle between the toilet lid and the horizontal plane is 0°, and the flushing water flow intensity is 10-15L / min.

[0065] Preferably, the particle size of the colored microspheres ranges from 1 to 10 μm; the density of the colored microspheres is 1.0 to 1.2 g / cm³. 3 The colored spheres include red spheres and blue spheres; the red spheres represent aerosols directly generated by flushing water, and the blue spheres represent secondary diffusion particles.

[0066] Preferably, the aerosol sampling device and the camera are used to record the diffusion path, range, and attachment location of the colored spheres during the flushing process to generate aerosol pollution distribution data. The hand model is then brought into contact with the contaminated surface to record the attachment distribution of the colored spheres on the hand model, generating hand contamination data, including:

[0067] The aerosol sampling device collects aerosol samples generated during the flushing process in real time; the aerosol sampling device includes a particle counter and a filter, used to capture the concentration and spatial distribution of colored spheres;

[0068] The camera records the movement trajectory of the colored ball at a shooting frequency of 30-60 frames per second;

[0069] The image data captured by the camera is subjected to trajectory tracking analysis. Combined with aerosol sampling data, a three-dimensional diffusion path model is generated through image processing algorithms, and a heat map of the attachment location is marked.

[0070] Based on the three-dimensional diffusion path model and the heat map, the diffusion range and surface adhesion density of aerosols are quantified, and aerosol pollution distribution data are generated.

[0071] Specifically, the aerosol sampling device uses a particle counter to monitor the concentration of colored spheres released during flushing in real time, while a filter collects aerosol samples to analyze the spatial distribution of the particles. The sampling device is installed directly above the toilet seat to ensure coverage of the aerosol diffusion range generated by flushing. Cameras are fixed at the front and side of the toilet, continuously recording the trajectory of the colored spheres at a shooting frequency of 30 to 60 frames per second. During recording, the light intensity and lens focal length are adjusted to ensure that the particle trajectory is clearly visible and to avoid environmental interference affecting data accuracy.

[0072] Video data captured by the camera is imported into image processing software. A trajectory tracking algorithm is used to analyze the movement path of the colored spheres frame by frame. Combined with concentration data from a particle counter, a three-dimensional diffusion path model of the aerosol is reconstructed. The model uses color gradients to mark particle density in different regions, generating a heat map of attachment sites to visually display the pollution distribution of aerosols on environmental surfaces. The algorithm further extracts the length of the diffusion path, the coverage area, and the coordinates of attachment points, providing structured data for quantitative analysis.

[0073] Based on a three-dimensional diffusion model and thermograms, the diffusion radius, maximum adhesion density, and contaminated area percentage of aerosols are calculated to generate aerosol pollution distribution data. After a hand model comes into contact with a contaminated surface, the surface adhesion is captured using a high-resolution imaging system. Image segmentation technology is used to count the number of particles and the coverage area, and the pollution concentration per unit area is calculated by combining the particle size parameters of colored spheres. All data are integrated into an analysis platform to generate a comprehensive report including spatial distribution, dynamic paths, and quantitative indicators, providing a reliable basis for subsequent risk assessment.

[0074] Preferably, the contaminated hand model is brought into contact with simulated food material, and the contamination range and concentration of colored spheres transferred from the hand to the food surface are recorded to generate food contamination data. The contaminated simulated food material is then placed in a simulated intestinal device, and the distribution of the colored spheres in the simulated intestinal device and the interaction between the colored spheres and the intestinal wall are observed in real time using an optical sensor to record the lesion simulation results, including:

[0075] The contaminated hand model is driven to contact the surface of the simulated food material at a preset contact angle and contact pressure; the contact time is 2-4 seconds.

[0076] The distribution of colored spheres was captured by scanning the surface of simulated food materials using a multispectral imager.

[0077] Based on image analysis algorithms, the boundaries of polluted areas in the distribution image are identified, and the proportion of polluted area and the number of particles per unit area are calculated.

[0078] Combining the density parameters of the colored balls, using the formula The contamination concentration is calculated, and the contamination concentration, the proportion of contaminated area, and the number of particles per unit area are determined as the food contamination data; where C is the contamination concentration, N is the number of particles, ρ is the density parameter of the colored spheres, and A is the contaminated area.

[0079] The simulated food material is fed into a simulated intestinal device made of a transparent flexible tube; the inner wall of the transparent flexible tube is coated with a viscoelastic material that simulates the intestinal mucosa; the elastic modulus of the viscoelastic material is in the range of 1-5 kPa and the viscosity is in the range of 0.1-0.5 Pa·s.

[0080] The movement trajectory of the colored ball in the simulated intestine and its collision and adsorption behavior with the inner wall were observed simultaneously using a fiber optic endoscope and the camera.

[0081] Fluorescent labeling technology was used to determine the dynamic migration path of the colored spheres, and a particle distribution model was generated using a three-dimensional reconstruction algorithm.

[0082] The particle distribution model was simulated and analyzed to obtain particle retention time, mucosal adhesion density, and local pressure peak.

[0083] The particle retention time, the mucosal adhesion density, and the local pressure peak are input into the trained lesion assessment model to obtain lesion simulation results; the lesion simulation results include lesion risk level and simulated pathology report.

[0084] Specifically, in this embodiment, a robotic arm drives a contaminated hand model to contact the surface of a simulated food material at a preset contact angle (30° to 60°) and contact pressure (0.3 to 0.8 N) for 2 to 4 seconds, simulating the behavior of a real hand touching food. During the contact process, the movement trajectory and force of the robotic arm are precisely controlled by a program to ensure consistent experimental conditions for each experiment. The simulated food material is selected from materials with surface characteristics similar to common food ingredients, such as silicone or agar substrates, to realistically reflect the contamination transfer effect.

[0085] After contact, a multispectral imager is used to scan the simulated food surface. The imager's wavelength covers the range of 400 to 1000 nm to enhance the distinguishability of different colored spheres. The scanned images are transmitted in real time to image processing software, which automatically identifies the boundaries of contaminated areas using edge detection algorithms and calculates the proportion of contaminated area to the total surface area. Simultaneously, the software counts the number of colored spheres per unit area and, combined with preset particle size parameters, makes a preliminary assessment of the degree of contamination.

[0086] Based on the proportion of contaminated area and the number of particles per unit area, and further combined with the density parameter of the colored spheres, the contamination concentration is calculated through a data processing platform. The calculation process employs a linear weighting method, integrating the contaminated area, particle density, and distribution uniformity to generate standardized food contamination data. All data is output in the form of heatmaps and statistical tables, intuitively displaying the contamination distribution characteristics and key quantitative indicators.

[0087] Simulated food materials are fed into a simulated intestinal device composed of transparent, flexible tubes. The inner walls of the tubes are coated with a viscoelastic material with an elastic modulus controlled between 1 and 5 kPa and a viscosity ranging from 0.1 to 0.5 Pa·s, to simulate the physical properties of the real intestinal mucosa. Peristaltic pumps are connected to both ends of the device to simulate the rhythmic contractions of the intestine, ensuring that the flow of particles within the tubes closely approximates physiological conditions.

[0088] The movement trajectory of colored spheres within a simulated intestine was simultaneously observed using a fiber optic endoscope and a high-speed camera. The fiber optic endoscope, with a resolution of at least 4K, captured real-time collisions and adhesion between the particles and the intestinal wall; the high-speed camera recorded the rapid dynamic process at a frame rate exceeding 1000 frames per second. The observation data was time-stamped to ensure the synchronization and accuracy of multi-source information.

[0089] Fluorescent labeling technology was used to distinguish the migration paths of different colored spheres, and a 3D reconstruction algorithm was combined to generate a distribution model of the particles in the intestine. The model analyzed particle retention time, mucosal adhesion density, and local pressure peaks, and these parameters were input into a pre-trained machine learning lesion assessment model. Based on the correlation between historical experimental data and clinicopathology, the model outputs lesion risk levels and simulated pathology reports, ultimately forming a complete health risk assessment system for bacterial transmission.

[0090] Furthermore, this embodiment utilizes a lesion assessment model based on the random forest algorithm. The first step in constructing the random forest lesion assessment model is to collect relevant data. This includes all experimental results obtained from aerosol diffusion experiments, hand contamination data, food contamination data, and intestinal simulation experiments. After all data is collected, it needs to be preprocessed, such as handling missing values ​​and outliers, and normalized to ensure that different features (such as contamination concentration, residence time, and adhesion density) are on the same order of magnitude. Next, the data needs to be divided into a feature set (input variables) and a target variable (lesion risk level), and further divided into training and test sets to adapt to the machine learning process.

[0091] After preparing clean data, the next step is to choose the random forest algorithm as the model for lesion assessment. Random forests consist of multiple decision trees, using ensemble learning to improve the model's accuracy and stability. During the training phase, the training set is input into the random forest model. By building multiple decision trees, each tree independently evaluates the input data, and the final output, i.e., the predicted risk level of the lesion, is determined through voting. During training, cross-validation is used to validate the model, ensuring the reliability and consistency of the results obtained from each training iteration to avoid overfitting or underfitting.

[0092] After training, the model needs to be evaluated. Using the previously segmented test set, the model's performance is tested, and evaluation metrics such as prediction accuracy, recall, and F1 score are calculated. These metrics help quantify the model's reliability in actual predictions. If the evaluation results do not meet the expected performance, the parameters of the random forest, such as the number of trees and maximum depth, can be further adjusted to optimize the model. Simultaneously, feature selection can be performed to evaluate the impact of different features on model performance, thereby focusing on the most useful features and improving the model's effectiveness and efficiency. After multiple rounds of training and optimization, the final model can accurately assess the risk of disease under specific conditions and provide scientific evidence to guide subsequent public health measures.

[0093] Preferably, the preset machine learning algorithm is a CNN+LSTM neural network.

[0094] Specifically, this embodiment standardizes aerosol diffusion data (diffusion range, adhesion density), hand contamination distribution data (particle count per unit area, contamination concentration), food contamination data (contaminated area percentage, transfer concentration), and lesion simulation results (particle residence time, mucosal adhesion density, local pressure peak) to eliminate dimensional differences. Through data fusion technology, multi-source data are mapped to the same spatiotemporal coordinate system, and key feature parameters, such as aerosol diffusion index, hand contamination transfer rate, food contamination accumulation, and intestinal lesion risk factors, are extracted to form a structured feature matrix. Outliers are removed during data cleaning, and missing data is filled in using interpolation to ensure the integrity of the input model.

[0095] Optionally, the machine learning model in this embodiment employs a hybrid architecture of Convolutional Neural Network (CNN) and Long Short-Term Memory Network (LSTM). First, spatial data such as aerosol diffusion heatmaps and hand contamination distribution images are input into the CNN module. Multi-layer convolution and pooling operations are used to extract local features of contamination distribution (e.g., high-density regions, diffusion directionality). Subsequently, temporal data of particle movement within the intestine (e.g., retention time series, pressure fluctuation curves) are input into the LSTM module to capture dynamic changes over time (e.g., periodicity of particle adsorption, cumulative effect of lesion risk). The output features of the CNN and LSTM are concatenated and then fused through a fully connected layer to integrate spatial and temporal information. An attention mechanism is introduced to adaptively allocate weights (e.g., intestinal data has a higher impact on the final risk than aerosol diffusion), ultimately outputting a comprehensive risk score. The CNN module design uses a ResNet-18 pre-trained model as its basic architecture, fine-tuned for aerosol heatmaps and hand contamination images. Shallow convolutional kernels are retained to capture detailed textures, while deeper networks extract global distribution patterns. Bidirectional LSTM layers are used to handle the forward and reverse temporal correlations of intestinal particle movement. The number of neurons in the hidden layers is dynamically adjusted based on the temporal length of historical experimental data to avoid information loss. After concatenating the CNN and LSTM features, a Dropout layer (with a dropout rate of 0.3) is added to prevent overfitting, and batch normalization is used to improve training stability.

[0096] The training data comes from historical experimental records, covering multiple sets of experimental data under different toilet opening and closing conditions, flushing intensity, and environmental parameters. The dataset is divided into training and test sets in a 7:3 ratio. During training, a stochastic gradient descent optimizer is used, and the loss function combines cross-entropy (for classification tasks) and mean squared error (for regression tasks). The network parameters are adjusted through backpropagation. The model uses five-fold cross-validation to prevent overfitting, and performance is evaluated using the confusion matrix, ROC curve, and root mean square error (RMSE). The final model is deployed to an analytics platform, supporting real-time risk assessment of new experimental data and iteratively optimizing parameters based on validation results to ensure prediction accuracy and generalization ability.

[0097] As an example, the training data included aerosol heatmaps (images), hand contamination distribution maps (images), intestinal time-series data (time series), and labeled risk level tags. During training, the CNN part used a transfer learning strategy, freezing some pre-trained layer parameters; the LSTM part used random initialization and learned intestinal dynamics through time-step unfolding. The loss function combined classification cross-entropy (risk level) and mean squared error (regression score), and the optimizer used was AdamW (learning rate 1e-4, weight decay 1e-5). In the validation phase, the CNN's region of interest was visualized using gradient-weighted class activation mapping (Grad-CAM) based on heatmaps, and the response of the LSTM hidden states to key time nodes was analyzed to ensure that the model logic conformed to the physical laws of contamination propagation.

[0098] Optionally, in this embodiment, flushing experiments are performed in the same experimental environment with the toilet lid fully open and closed to ensure that parameters such as flushing intensity, liquid volume, and ambient temperature and humidity are strictly consistent. After each experiment, aerosol diffusion data, hand contamination data, food contamination data, and intestinal lesion simulation results are collected simultaneously, and timestamps are used to ensure data time sequence alignment. The experiment is repeated at least 10 times to eliminate random errors. After each experiment, the equipment is thoroughly cleaned and the initial conditions are reset to avoid residual contamination interfering with subsequent data.

[0099] The risk simulation results of toilet seat opening and closing states are input into the analysis platform to calculate core indicators such as the mean aerosol diffusion range, median hand contamination concentration, proportion of food contamination area, and intestinal lesion score. Paired t-tests or nonparametric statistical methods (such as Wilcoxon signed-rank tests) are used to analyze the significance of differences between the two sets of data, generating the percentage of risk difference and confidence intervals. Simultaneously, based on the aerosol diffusion path heatmap and hand contamination distribution map, image differencing algorithms are used to extract the spatial differences in contamination paths under opening and closing states, generating quantitative parameters such as path offset and overlapping area proportion.

[0100] Based on quantitative risk indicators and pathway difference parameters, health risks were classified under two conditions using a predefined risk level classification table (e.g., low risk: comprehensive score ≤ 0.3, medium risk: 0.3 < score ≤ 0.7, high risk: score > 0.7). The results were output in an interactive visual report format, including comparative bar charts, pathway difference heatmaps, and risk radar charts, intuitively demonstrating the impact of toilet seat opening and closing on bacterial transmission. The final report, combined with experimental data, recommended hygiene optimization measures (e.g., closed flushing can reduce risk levels), and expert verification ensured the scientific validity and operability of the conclusions.

[0101] Corresponding to the above methods, such as Figure 4 As shown, this embodiment also provides a bacterial contamination transmission simulation system based on colored aerosols, including:

[0102] The experimental setup unit is used to set up an aerosol sampling device, a hand model, and a camera at the experimental toilet to simulate a bathroom scene. After adding liquid to the experimental toilet, the unit controls the toilet's flushing device to perform a flushing operation. Experiments simulating bacterial contamination with colored aerosols are conducted under two conditions: the toilet lid is open and closed. The experiment simulating bacterial contamination with colored aerosols is carried out by releasing colored aerosols into the experimental toilet. The colored aerosols consist of multiple colored spheres.

[0103] The first data acquisition unit is used to record the diffusion path, range and attachment position of the colored balls during the flushing process using the aerosol sampling device and the camera, generate aerosol pollution distribution data, and bring the hand model into contact with the polluted surface to record the attachment distribution of the colored balls on the hand model to generate hand pollution data.

[0104] The second data acquisition unit is used to bring the contaminated hand model into contact with the simulated food material, record the contamination range and concentration of the colored balls transferred from the hand to the food surface, generate food contamination data, and place the contaminated simulated food material in a simulated intestinal device. The distribution of the colored balls in the simulated intestinal device and the interaction between the colored balls and the intestinal wall are observed in real time through an optical sensor to record the lesion simulation results.

[0105] The integration unit is used to integrate the aerosol diffusion data, the hand contamination distribution data, the food contamination data, and the lesion simulation results, and to determine the risk simulation results through a preset machine learning algorithm.

[0106] The beneficial effects of this invention are as follows:

[0107] (1) This invention systematically quantifies the health risks of the entire chain of bacterial transmission by comparing data on aerosol diffusion, hand contamination, food contamination and intestinal lesions under the conditions of toilet seat opening and closing, and provides a scientific basis for optimizing hygiene habits.

[0108] (2) This invention uses colored aerosols (colored spheres) combined with optical sensors, image processing algorithms and machine learning models to achieve visualized tracking and dynamic risk assessment of pollution paths, thereby improving data reliability.

[0109] (3) This invention directly links environmental pollution and human health effects by simulating intestinal devices and lesion evaluation indicators, providing technical support for public health prevention and control strategies (such as food processing and medical environment).

[0110] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0111] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for simulating bacterial contamination transmission based on colored aerosols, characterized in that, include: An aerosol sampling device, a hand model, and a camera were installed at the experimental toilet to simulate a bathroom scene. After adding liquid to the experimental toilet, the flushing device was controlled to perform a flushing operation. Experiments simulating bacterial contamination with colored aerosols were conducted under two experimental conditions: with the toilet lid open and closed. The experiment simulating bacterial contamination with colored aerosols was carried out by releasing colored aerosols into the experimental toilet. The colored aerosols consisted of multiple colored spheres. The aerosol sampling device and the camera are used to record the diffusion path, range and attachment position of the colored balls during the flushing process to generate aerosol pollution distribution data. The hand model is then brought into contact with the contaminated surface to record the attachment distribution of the colored balls on the hand model, generating hand pollution data. The contaminated hand model is brought into contact with simulated food material, and the range and concentration of contamination from the hand to the food surface are recorded to generate food contamination data. The contaminated simulated food material is then placed in a simulated intestinal device, and the distribution of the colored balls in the simulated intestinal device and the interaction between the colored balls and the intestinal wall are observed in real time through optical sensors to record the lesion simulation results. The aerosol diffusion data, hand contamination distribution data, food contamination data, and lesion simulation results are integrated, and the risk simulation results are determined through a preset machine learning algorithm.

2. The method for simulating bacterial contamination transmission based on colored aerosols according to claim 1, characterized in that, Also includes: The risk simulation results under two conditions—toilet seat open and closed—were compared and analyzed to output quantitative risk indicators and assess the differences in bacterial transmission pathways and health risk levels under different conditions.

3. The method for simulating bacterial contamination transmission based on colored aerosols according to claim 1, characterized in that, The aerosol sampling device is positioned 10-30 cm above the toilet seat of the experimental toilet and is used to collect aerosols generated during flushing. The surface material of the hand model simulates the friction coefficient and hydrophobicity of human skin, and the hand model is placed on the side of the experimental toilet at a simulated sink position. The camera is fixed in front of and to the side of the experimental toilet to record the aerosol diffusion trajectory.

4. The method for simulating bacterial contamination transmission based on colored aerosols according to claim 1, characterized in that, The parameters for the two experimental conditions are as follows: When the toilet seat is open, the angle between the toilet seat and the horizontal plane is 90°, and the flushing water flow rate is 10-15L / min. When the toilet lid is closed, the angle between the toilet lid and the horizontal plane is 0°, and the flushing water flow intensity is 10-15L / min.

5. The method for simulating bacterial contamination transmission based on colored aerosols according to claim 1, characterized in that, The colored microspheres have a particle size range of 1-10 μm; the density of the colored microspheres is 1.0-1.2 g / cm³. 3 The colored spheres include red spheres and blue spheres; the red spheres represent aerosols directly generated by flushing water, and the blue spheres represent secondary diffusion particles.

6. The method for simulating bacterial contamination transmission based on colored aerosols according to claim 1, characterized in that, The aerosol sampling device and the camera are used to record the diffusion path, range, and attachment location of the colored spheres during flushing, generating aerosol pollution distribution data. The hand model is then placed in contact with the contaminated surface, and the attachment distribution of the colored spheres on the hand model is recorded to generate hand contamination data, including: The aerosol sampling device collects aerosol samples generated during the flushing process in real time; the aerosol sampling device includes a particle counter and a filter, used to capture the concentration and spatial distribution of colored spheres; The camera records the movement trajectory of the colored ball at a shooting frequency of 30-60 frames per second; The image data captured by the camera is subjected to trajectory tracking analysis. Combined with aerosol sampling data, a three-dimensional diffusion path model is generated through image processing algorithms, and a heat map of the attachment location is marked. Based on the three-dimensional diffusion path model and the heat map, the diffusion range and surface adhesion density of aerosols are quantified, and aerosol pollution distribution data are generated.

7. The method for simulating bacterial contamination transmission based on colored aerosols according to claim 1, characterized in that, The contaminated hand model is brought into contact with simulated food material, and the extent and concentration of contamination from the colored spheres transferred from the hand to the food surface are recorded to generate food contamination data. The contaminated simulated food material is then placed in a simulated intestinal device, and the distribution of the colored spheres within the device and their interaction with the intestinal wall are observed in real time using optical sensors to record the lesion simulation results, including: The contaminated hand model is driven to contact the surface of the simulated food material at a preset contact angle and contact pressure; the contact time is 2-4 seconds. The distribution of colored spheres was captured by scanning the surface of simulated food materials using a multispectral imager. Based on image analysis algorithms, the boundaries of polluted areas in the distribution image are identified, and the proportion of polluted area and the number of particles per unit area are calculated. Combining the density parameters of the colored balls, using the formula The contamination concentration is calculated, and the contamination concentration, the proportion of contaminated area, and the number of particles per unit area are determined as the food contamination data; where C is the contamination concentration, N is the number of particles, ρ is the density parameter of the colored spheres, and A is the contaminated area. The simulated food material is fed into a simulated intestinal device made of a transparent flexible tube; the inner wall of the transparent flexible tube is coated with a viscoelastic material that simulates the intestinal mucosa; the elastic modulus of the viscoelastic material is in the range of 1-5 kPa and the viscosity is in the range of 0.1-0.5 Pa·s. The movement trajectory of the colored ball in the simulated intestine and its collision and adsorption behavior with the inner wall were observed simultaneously using a fiber optic endoscope and the camera. Fluorescent labeling technology was used to determine the dynamic migration path of the colored spheres, and a particle distribution model was generated using a three-dimensional reconstruction algorithm. The particle distribution model was simulated and analyzed to obtain particle retention time, mucosal adhesion density, and local pressure peak. The particle retention time, the mucosal adhesion density, and the local pressure peak are input into the trained lesion assessment model to obtain lesion simulation results; the lesion simulation results include lesion risk level and simulated pathology report.

8. The method for simulating bacterial contamination transmission based on colored aerosols according to claim 1, characterized in that, The preset machine learning algorithm is a CNN+LSTM neural network.

9. A bacterial contamination transmission simulation system based on colored aerosols, characterized in that, include: The experimental setup unit is used to set up an aerosol sampling device, a hand model, and a camera at the experimental toilet to simulate a bathroom scene. After adding liquid to the experimental toilet, the unit controls the toilet's flushing device to perform a flushing operation. Experiments simulating bacterial contamination with colored aerosols are conducted under two conditions: the toilet lid is open and closed. The experiment simulating bacterial contamination with colored aerosols is carried out by releasing colored aerosols into the experimental toilet. The colored aerosols consist of multiple colored spheres. The first data acquisition unit is used to record the diffusion path, range and attachment position of the colored balls during the flushing process using the aerosol sampling device and the camera, generate aerosol pollution distribution data, and bring the hand model into contact with the polluted surface to record the attachment distribution of the colored balls on the hand model to generate hand pollution data. The second data acquisition unit is used to bring the contaminated hand model into contact with the simulated food material, record the contamination range and concentration of the colored balls transferred from the hand to the food surface, generate food contamination data, and place the contaminated simulated food material in a simulated intestinal device. The distribution of the colored balls in the simulated intestinal device and the interaction between the colored balls and the intestinal wall are observed in real time through an optical sensor to record the lesion simulation results. The integration unit is used to integrate the aerosol diffusion data, the hand contamination distribution data, the food contamination data, and the lesion simulation results, and to determine the risk simulation results through a preset machine learning algorithm.

Images