Hemorrhoidal fistula risk prediction and prevention system

Abstract

The present invention relates to the field of risk prediction technology, specifically to a hemorrhoid fistula risk prediction and prevention system, which includes an environmental factor monitoring module, an acoustic wave acquisition and analysis module, an anal pressure monitoring module, a signal feature analysis module, a risk rating algorithm module, a lesion prediction module, and an intervention measure recommendation module. In the present invention, modules such as environmental monitoring, acoustic wave and pressure data analysis, and signal feature analysis are integrated to provide a comprehensive solution strategy for hemorrhoid fistula risk assessment and prevention. A signal feature analysis module is introduced to deeply mine subtle changes in pressure fluctuation data through spectrum analysis, and linear discriminant analysis is used to accurately divide risk levels, which not only improves the accuracy of risk assessment, but also customizes preventive measures based on specific risk levels, realizes the refinement and personalization of disease management, and predicts lesion trends through time series analysis, reveals potential health threats in advance, and improves the timeliness and effectiveness of preventive measures.

Description

Technical Field

[0001] The present invention relates to the technical field of risk prediction, and in particular to a hemorrhoidal fistula risk prediction and prevention system. Background Art

[0002] The field of risk prediction technology involves using data analysis, machine learning, statistical methods, and other techniques to predict and assess potential risks. In healthcare, this aims to identify potential health risks at an early stage, enabling early intervention, disease prevention, and health management. By analyzing and learning from large amounts of data, disease development patterns and risk factors can be identified, providing scientific decision-making support for medical professionals and patients.

[0003] The hemorrhoidal fistula risk prediction and prevention system is a system specifically designed to predict and prevent the risk of hemorrhoidal fistulas. Its main purpose is to evaluate the probability of developing hemorrhoidal fistulas by analyzing multi-dimensional data such as an individual's medical history, lifestyle, genetic factors, etc. It can propose personalized preventive measures based on risk assessment to help individuals take effective changes in their lifestyle and eating habits, thereby reducing the incidence of hemorrhoidal fistulas, improving the initiative and preventive nature of hemorrhoidal fistula disease management, reducing the health burden and medical costs brought by hemorrhoidal fistulas, and improving people's quality of life.

[0004] Traditional technologies have significant shortcomings in the comprehensive utilization of multi-dimensional data, particularly in real-time and personalized prevention. They rely on overly generalized risk assessments and lack in-depth consideration of individual differences. This leads to interventions that are overly generalized and overly generalized, making it difficult to precisely match them to individual needs. Furthermore, the lack of monitoring and analysis of immediate health status changes makes disease prevention strategies sluggish, preventing them from fully realizing the potential of early intervention. This leads to increased uncertainty in the effectiveness of preventive measures, potentially exacerbating patient health risks, increasing medical costs, and reducing quality of life. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a hemorrhoidal fistula risk prediction and prevention system.

[0006] To achieve the above objectives, the present invention adopts the following technical solutions: the hemorrhoid fistula risk prediction and prevention system includes an environmental factor monitoring module, an acoustic wave acquisition and analysis module, an anal pressure monitoring module, a signal feature analysis module, a risk rating algorithm module, a lesion prediction module, and an intervention measure recommendation module;

[0007] The environmental factor monitoring module collects external temperature, humidity and individual activity data based on environmental sensors and individuals to obtain an environmental activity data set;

[0008] The sound wave acquisition and analysis module combines the environmental activity data set to capture the sound waves generated by intestinal activity, analyzes the correlation between the two, and obtains the sound wave activity feature set;

[0009] The anal pressure monitoring module records the pressure fluctuations in the anal area in combination with the acoustic wave activity feature set, and correlates the pressure fluctuations with the features in the acoustic wave activity feature set to obtain pressure fluctuation data;

[0010] The signal feature analysis module extracts the vibration frequency components and energy distribution based on the pressure fluctuation data using a spectrum analysis method to construct a feature parameter matrix;

[0011] The risk rating algorithm module analyzes the data in the characteristic parameter matrix, performs linear discriminant analysis, and obtains risk level indicators;

[0012] The lesion prediction module uses the risk level index and the autoregressive moving average method to predict the lesion development trend and generate a lesion development prediction record;

[0013] The intervention measure recommendation module proposes dietary adjustments and lifestyle improvement measures based on the lesion development prediction record and obtains a list of intervention measures.

[0014] As a further solution of the present invention, the environmental activity data set includes the average daily temperature, maximum humidity, and average number of active steps; the sound wave activity feature set includes the intestinal peristalsis sound wave frequency, the maximum sound wave intensity, and the sound wave duration; the pressure fluctuation data includes the maximum pressure value of anal contraction, the pressure release rate, and the pressure stability index; the characteristic parameter matrix includes the main frequency of anal pressure fluctuation, the energy correlation between sound waves and pressure fluctuations, and the time-frequency feature clustering index; the risk level index includes the probability score of hemorrhoidal fistula, the contribution of risk factors, and the emergency warning level; the lesion development prediction record includes the expected month of lesion occurrence, the lesion development rate, and the critical prevention time window; the intervention measures list includes necessary dietary fiber increase recommendations, recommended daily activity amount, and regular medical examination frequency.

[0015] As a further solution of the present invention, the environmental factor monitoring module includes an environmental data acquisition submodule, a data optimization submodule, and a result analysis submodule;

[0016] The environmental data acquisition submodule activates the temperature and humidity sensors and the activity monitoring device to continuously collect environmental temperature, humidity, and individual activity information within a set time interval, timestamps and preliminarily classifies the collected raw data, and generates raw data records;

[0017] The data optimization submodule receives the original data records, filters the data according to the preset threshold, removes outliers, performs linear interpolation on the data set, fills in missing values, segments the data, and performs normalization based on the time series to obtain a standardized data set;

[0018] The result analysis submodule analyzes the data change trend based on the standardized data set, identifies the correlation between ambient temperature and humidity and individual activity level by calculating the data change rate within the difference time window, and constructs an environmental activity data set.

[0019] As a further solution of the present invention, the sound wave acquisition and analysis module includes a sound wave capture submodule, a sound wave data screening submodule, and an acoustic environment feature mapping submodule;

[0020] The sound wave capture submodule configures a microphone array within the environment, starts capturing intestinal activity sound waves at regular intervals, records basic properties of the sound waves, including frequency range, intensity, and timestamp, organizes the information into a sequence, and generates a sound wave sequence record;

[0021] The acoustic wave data screening submodule performs preliminary screening based on the acoustic wave sequence records according to the intensity and frequency range of the acoustic waves, eliminates background noise interference items, performs time series analysis on the remaining acoustic waves, identifies the repeatability of the acoustic wave pattern, and obtains the screened acoustic wave pattern;

[0022] The acoustic environment feature mapping submodule matches and analyzes the filtered sound wave patterns with the data in the environmental activity dataset, analyzes the interaction between intestinal activity sound waves and environmental factors and individual activity levels, and constructs a sound wave activity feature set.

[0023] As a further solution of the present invention, the anal pressure monitoring module includes a pressure data acquisition submodule, a data integration submodule, and a feature matching submodule;

[0024] The pressure data acquisition submodule deploys sensors in the anus area to record pressure changes in real time, including the magnitude of the change, the frequency of occurrence, and the duration. The data is continuously collected to form a serialized record to generate a pressure change record.

[0025] The data integration submodule performs preliminary processing on the pressure change records, including removing interference signals introduced by body movement or external factors, reducing data fluctuations, ensuring signal continuity and analyzability, and obtaining a preprocessed pressure signal;

[0026] The feature matching submodule is based on the preprocessed pressure signal, analyzes the key features in the signal, including the pattern of pressure peaks, and the changing laws of frequency and amplitude. Combined with the acoustic activity feature set, it identifies the correlation pattern between pressure changes and intestinal activity and obtains pressure fluctuation data.

[0027] As a further solution of the present invention, the signal feature analysis module includes a frequency analysis submodule, an energy distribution analysis submodule, and a feature parameter construction submodule;

[0028] The frequency analysis submodule performs frequency domain conversion based on the pressure fluctuation data, identifies and records the occurrence frequency and intensity of the frequency components, extracts the key frequency components of the signal, and generates frequency characteristic data;

[0029] The energy distribution analysis submodule performs energy spectrum analysis on the frequency characteristic data, calculates the energy value of each frequency component using a spectrum analysis method, analyzes the energy distribution of the entire spectrum, and obtains energy distribution data;

[0030] The spectrum analysis method is based on the formula:

[0031]

[0032]

[0033] Calculate the energy value of each frequency component to obtain energy distribution data, where: is the signal in the frequency domain, is the window function adjustment coefficient, is the signal in the time domain, is the window function, is the sampling rate adjustment factor, Frequency The energy spectral density at is the frequency correction offset value, is the signal enhancement factor;

[0034] The characteristic parameter construction submodule comprehensively analyzes the energy distribution data and the identified frequency characteristics, combines data points according to specific rules, constructs a matrix that comprehensively describes the signal characteristics, and obtains a characteristic parameter matrix.

[0035] As a further solution of the present invention, the risk rating algorithm module includes a data preparation submodule, a linear discriminant analysis submodule, and a risk level generation submodule;

[0036] The data preparation submodule performs data standardization based on the characteristic parameter matrix to ensure the scale consistency of variables during the analysis process and obtain standardized characteristic data;

[0037] The linear discriminant analysis submodule uses standardized feature data to perform linear discriminant analysis, calculates the contribution of features to classification, extracts feature combinations that can distinguish different risk levels, and generates risk discrimination features;

[0038] The risk level generation submodule classifies data samples based on risk discrimination features, assigns each sample to a corresponding risk level, obtains the risk level of each sample, and acquires a risk level indicator.

[0039] As a further solution of the present invention, the lesion prediction module includes a risk level data analysis submodule, a time series modeling submodule, and a trend prediction submodule;

[0040] The risk level data analysis submodule collects risk level indicators, analyzes the changing patterns within a time span, identifies the changing trends and patterns of risk levels, and obtains risk level trend data;

[0041] The time series modeling submodule uses the autoregressive moving average method based on the risk level trend data to analyze the relationship between the risk level and time, determine the model parameters, establish a mathematical basis for the prediction of the lesion development trend, and generate a time series model;

[0042] The autoregressive moving average method is based on the formula:

[0043]

[0044] Generate a time series model where is the order of the autoregressive term, is the number of differences required for the original sequence, is the seasonal adjustment factor, is the order of the moving average term, is the autoregressive coefficient, is the trend smoothing coefficient, is the moving average coefficient, is the fluctuation adjustment coefficient, is the error adjustment margin, is the lag operator, is the observed value at time point t, is the error term at time point t;

[0045] The trend prediction submodule uses a time series model to predict the changing trend of the risk level in the future, predicts the development of the lesion based on the trend, and obtains the lesion development prediction record.

[0046] As a further solution of the present invention, the intervention measure recommendation module includes a lesion development analysis submodule, a lifestyle improvement suggestion submodule, and a diet adjustment suggestion submodule;

[0047] The lesion development analysis submodule analyzes the potential development trend and key influencing factors of the lesion based on the lesion development prediction record, identifies risk points, and obtains lesion risk analysis results;

[0048] The lifestyle improvement suggestion submodule proposes targeted lifestyle improvement measures based on the lesion risk analysis results, including improving sleep quality and regular physical activity, and generates lifestyle improvement suggestions;

[0049] The dietary adjustment suggestion submodule combines the lesion risk analysis results and lifestyle improvement suggestions to propose a dietary adjustment plan based on the individual's condition and development trend, including dietary structure optimization, specific nutrient intake, and obtaining a list of intervention measures.

[0050] Compared with the prior art, the advantages and positive effects of the present invention are:

[0051] In this invention, modules such as environmental monitoring, sound wave and pressure data analysis, and signal feature analysis are integrated to provide a comprehensive solution for hemorrhoid and fistula risk assessment and prevention. The signal feature analysis module is introduced to deeply explore subtle changes in pressure fluctuation data through spectrum analysis, and linear discriminant analysis is used to accurately divide risk levels. This not only improves the accuracy of risk assessment, but also can customize preventive measures based on specific risk levels, thereby achieving refined and personalized disease management. At the same time, by predicting lesion trends through time series analysis, the system can reveal potential health threats in advance, provide a strong basis for intervention, and significantly improve the timeliness and effectiveness of preventive measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is a system flow chart of the present invention;

[0053] Figure 2 Schematic diagram of the system framework of the present invention;

[0054] Figure 3 This is a flow chart of the environmental factor monitoring module of the present invention;

[0055] Figure 4 This is a flow chart of the sound wave acquisition and analysis module of the present invention;

[0056] Figure 5 This is a flow chart of the anal pressure monitoring module of the present invention;

[0057] Figure 6 This is a flow chart of the signal feature analysis module of the present invention;

[0058] Figure 7 This is a flow chart of the risk rating algorithm module of the present invention;

[0059] Figure 8 This is a flow chart of the lesion prediction module of the present invention;

[0060] Figure 9 A modular flow chart is proposed for the intervention measures of the present invention. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0062] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0063] Example 1: Please refer to Figure 1 and Figure 2 The hemorrhoid fistula risk prediction and prevention system includes an environmental factor monitoring module, an acoustic wave acquisition and analysis module, anal pressure monitoring module, a signal feature analysis module, a risk rating algorithm module, a lesion prediction module, and an intervention measure recommendation module;

[0064] The environmental factor monitoring module collects external temperature, humidity and individual activity data based on environmental sensors and individuals to obtain environmental activity data sets;

[0065] The sound wave acquisition and analysis module combines the environmental activity dataset to capture the sound waves generated by intestinal activity, analyzes the correlation between the two, and obtains the sound wave activity feature set;

[0066] The anal pressure monitoring module combines the acoustic wave activity feature set to record the pressure fluctuations in the anal area, and correlates them with the features in the acoustic wave activity feature set to obtain pressure fluctuation data;

[0067] The signal feature analysis module uses spectrum analysis method to extract vibration frequency components and energy distribution based on pressure fluctuation data and construct a characteristic parameter matrix;

[0068] The risk rating algorithm module analyzes the data in the characteristic parameter matrix, performs linear discriminant analysis, and obtains risk level indicators;

[0069] The lesion prediction module uses the risk level index and the autoregressive moving average method to predict the lesion development trend and generate lesion development prediction records;

[0070] The intervention measure recommendation module proposes dietary adjustments and lifestyle improvement measures based on the lesion development prediction records and obtains a list of intervention measures.

[0071] The environmental activity data set includes the average daily temperature, maximum humidity, and average number of active steps. The sound wave activity feature set includes the intestinal peristalsis sound wave frequency, maximum sound wave intensity, and sound wave duration. The pressure fluctuation data includes the maximum pressure value of anal contraction, pressure release rate, and pressure stability index. The characteristic parameter matrix includes the main frequency of anal pressure fluctuation, the energy correlation between sound waves and pressure fluctuations, and the time-frequency feature clustering index. The risk level indicators include the probability score of hemorrhoidal fistula, the contribution of risk factors, and the emergency warning level. The lesion development prediction record includes the expected month of lesion occurrence, the lesion development rate, and the critical prevention time window. The intervention measures list includes recommendations for increasing necessary dietary fiber, recommended daily activity, and the frequency of regular medical examinations.

[0072] In the environmental factor monitoring module, data is collected through environmental sensors and individual activity monitoring devices, including information such as average daily temperature, maximum humidity, and average number of steps. The data first undergoes preliminary cleaning and normalization to ensure quality and consistency. Next, statistical analysis methods are used to analyze trends and identify patterns in the collected data, identifying potential correlations between environmental factors, individual activity levels, and health status. This ultimately generates an environmental activity dataset, providing foundational data support for subsequent module analyses.

[0073] In the sound wave acquisition and analysis module, a microphone array captures the sound waves generated by intestinal activity, including characteristics such as peristaltic frequency, maximum intensity, and duration. The collected sound wave signals undergo denoising and spectral analysis to extract key characteristic parameters. Subsequently, a machine learning algorithm correlates these sound wave characteristics with the environmental activity dataset to identify patterns of interaction between the two. This generates a set of sound wave activity features, providing crucial information for understanding intestinal activity patterns.

[0074] In the anal pressure monitoring module, pressure sensors record pressure fluctuations in the anal region. The resulting data includes information such as the peak anal contraction pressure, pressure release rate, and pressure stability index. This data undergoes signal processing techniques, such as filtering and signal enhancement, to enhance its analyzability. Subsequently, correlation analysis is performed between the acoustic activity signature set and the pressure fluctuation data, and statistical analysis methods are used to explore the relationship between the two. This generates pressure fluctuation data, providing precise quantitative indicators for studying physiological activity in the anal region.

[0075] In the signal feature analysis module, spectral analysis methods are used to perform a detailed analysis of the vibration frequency components and energy distribution in the pressure fluctuation data. This involves applying a Fourier transform to extract the signal's frequency characteristics, followed by energy spectrum analysis to calculate the energy values ​​of different frequency components. This not only reveals the signal's primary frequency components but also analyzes the overall pattern of its energy distribution. Finally, combining the frequency characteristics and energy distribution, a characteristic parameter matrix is ​​constructed, providing a detailed signal description for risk assessment.

[0076] In the risk rating algorithm module, linear discriminant analysis is used to process the data in the feature parameter matrix, aiming to identify the feature combination that best represents the risk level. The algorithm first calculates the contribution of each feature to the classification result, then selects the features with the highest contribution and linearly combines them to effectively distinguish different risk levels. Ultimately, the algorithm generates risk level indicators, including the probability score of hemorrhoidal fistula occurrence, the contribution of risk factors, and the emergency warning level, providing a scientific basis for accurate assessment of lesion risk.

[0077] In the lesion prediction module, an autoregressive moving average method is used to analyze risk level indicators over time series to predict lesion development trends. This analysis details information such as the expected month of lesion onset, the rate of lesion development, and the critical prevention window. Through in-depth analysis of past and current risk level data, the model can predict the likely future trends of lesions and generate a lesion development forecast, providing a valuable reference for developing timely and effective preventive measures.

[0078] The intervention recommendation module proposes targeted dietary adjustments and lifestyle improvements based on lesion progression predictions and individual circumstances. These recommendations include recommendations for increasing dietary fiber, recommended daily activity levels, and the frequency of regular medical checkups. Personalized interventions aim to reduce the incidence of hemorrhoidal fistulas. Through scientific lifestyle adjustments and dietary management, these interventions provide a comprehensive health protection plan and a list of interventions, serving as a valuable reference for individuals and medical professionals in guiding health management and disease prevention.

[0079] See also Figure 2 and Figure 3 ,The environmental factor monitoring module includes the environmental data ,acquisition submodule, the data optimization submodule, and the result ,analysis submodule;

[0080] The environmental data acquisition submodule activates the temperature and humidity sensors and the activity monitoring device to continuously collect environmental temperature, humidity, and individual activity information within a set time interval, timestamps and preliminarily classifies the collected raw data, and generates raw data records;

[0081] In the environmental data collection submodule, the temperature and humidity sensors and activity monitoring device were activated to regularly collect data on ambient temperature, humidity, and individual activity levels. Data preprocessing was performed using the Python Pandas library, including adding a timestamp (pd.to_datetime()) to each data entry, performing preliminary classification based on data type, and grouping the data using the .groupby() method to generate the raw data records.

[0082] The data optimization submodule receives the original data records, filters the data according to the preset threshold, removes outliers, performs linear interpolation on the data set, fills missing values, segments the data, and normalizes it based on the time series to obtain a standardized data set;

[0083] In the data optimization submodule, raw data records are received and filtered using the NumPy library. Outlier thresholds are set using the np.where() function to remove data. Missing data is filled with linear interpolation using np.interp(). To standardize the data, the StandardScaler() function is used to normalize the time series data, and the .reshape() function is used to adjust the data shape to meet the needs of subsequent analysis, generating a standardized dataset.

[0084] The result analysis submodule uses the standardized data set as the basis to analyze the data change trend. By calculating the data change rate within the difference time window, it identifies the correlation between environmental temperature and humidity and individual activity, and constructs an environmental activity data set.

[0085] In the results analysis submodule, based on a standardized dataset, the SciPy library's stats.pearsonr() statistical function was used to analyze the correlation between ambient temperature and humidity and individual activity levels. Matplotlib was used to plot data trends, and plt.plot() was used to display the rate of change within different time windows. This allowed for the identification of trends and ultimately the construction and generation of an environmental activity dataset. This step further explored the interplay between environmental factors and individual activity levels, laying the foundation for further health risk analysis.

[0086] See also Figure 2 and Figure 4 ,The sound wave acquisition and analysis module includes a sound wave capturing submodule, a sound wave data screening submodule, and a sound environment feature mapping submodule;

[0087] The sound wave capture submodule configures the microphone array in the environment, starts to capture intestinal activity sound waves at a fixed time, records the basic properties of the sound waves, including frequency range, intensity and timestamp, organizes the information into a sequence, and generates a sound wave sequence record;

[0088] In the sound wave capture submodule, the microphone array within the environment is configured, and Audacity software automatically initiates recording, periodically collecting sound waves generated by intestinal activity. The recorded sound wave data is read using the Python wave library, using the wave.open() function to obtain the sound wave file. The readframes() method then extracts basic sound wave properties, such as frequency range and intensity. Each sound wave sample is timestamped, and the acquisition time is recorded using datetime.now(), organizing the data into a sound wave sequence.

[0089] The acoustic wave data screening submodule performs preliminary screening based on the acoustic wave sequence records according to the intensity and frequency range of the acoustic waves, removes background noise interference, performs time series analysis on the remaining acoustic waves, identifies the repeatability of the acoustic wave pattern, and obtains the filtered acoustic wave pattern;

[0090] In the acoustic wave data filtering submodule, the signal.spectrogram() function from the SciPy library is used to perform spectral analysis on the acoustic wave sequence records to identify the frequency range. Next, an intensity threshold is set and numpy.where() is used to filter out acoustic signals above background noise to eliminate invalid data. The pandas.Series.rolling() method is used to perform time series analysis on the filtered acoustic wave data, identifying and marking repetitive patterns. Finally, the filtered acoustic wave patterns are obtained.

[0091] The acoustic environment feature mapping submodule matches and analyzes the filtered sound wave patterns with the data in the environmental activity dataset, analyzes the interaction between intestinal activity sound waves and environmental factors and individual activity levels, and constructs a sound wave activity feature set;

[0092] In the acoustic environment feature mapping submodule, the filtered acoustic patterns were correlated with data from the environmental activity dataset. The NumPy library's numpy.corrcoef() function was used to calculate correlation coefficients between acoustic features and environmental factors (such as temperature and humidity) and individual activity levels, analyzing the interaction between intestinal activity acoustic waves and these environmental factors. Based on these analysis results, a Pandas DataFrame was used to construct an acoustic activity feature set. This feature set includes correlation indicators for intestinal motility acoustic wave frequency, maximum acoustic wave intensity, acoustic wave duration, and environmental activity data, providing comprehensive information for further analysis.

[0093] See also Figure 2 and Figure 5 ,Anal pressure monitoring module includes pressure data acquisition submodule, data integration submodule, and feature matching submodule;

[0094] The pressure data acquisition submodule deploys sensors in the anal area to record pressure changes in real time, including the magnitude of the change, frequency of occurrence, and duration. The data is continuously collected to form a serialized record, generating a pressure change record.

[0095] In the pressure data acquisition submodule, sensors deployed in the anal area collect real-time data using Arduino programming, recording pressure changes, including amplitude, frequency, and duration. The collected data is transmitted to a computer via serial communication. A Python script automatically reads the data, using the Serial.read() method in the serial library to capture sensor data. Each data entry is timestamped using the datetime library, and a DataFrame is constructed using the pandas library to serialize the data and generate a record of pressure changes.

[0096] The data integration submodule performs preliminary processing on the pressure change records, including removing interference signals introduced by body movement or external factors, reducing data fluctuations, ensuring signal continuity and analyzability, and obtaining preprocessed pressure signals;

[0097] In the data integration submodule, the pressure change records are initially processed. A sliding average is performed on the data using the pandas.DataFrame.rolling(window=3).mean() method to reduce interference and data fluctuations introduced by body movement or external factors. Furthermore, the numpy.clip() method is used to set the signal threshold range and remove outliers, ensuring signal continuity and analyzability. The processed data is then standardized using sklearn.preprocessing.StandardScaler() to obtain the preprocessed pressure signal.

[0098] The feature matching submodule analyzes the key features of the preprocessed pressure signal, including the pattern of pressure peaks, and the changing patterns of frequency and amplitude. Combined with the acoustic activity feature set, it identifies the correlation between pressure changes and intestinal activity, and obtains pressure fluctuation data.

[0099] In the feature matching submodule, a fast Fourier transform (FFT) is performed on the preprocessed pressure signal using numpy.fft.fft() to extract the frequency components of the signal. Combined with the acoustic activity feature set, the correlation coefficient between the pressure signal and the acoustic wave features is calculated using scipy.stats.pearsonr() to analyze the association pattern between pressure peaks and intestinal activity. Cluster analysis is performed using the sklearn.cluster.KMeans(n_clusters=3) method to identify different pressure change patterns and generate pressure fluctuation data. This process deeply explores the interaction between pressure signals and intestinal activity sound waves, providing key data support for further health status assessment and risk prediction.

[0100] See also Figure 2 and Figure 6 ,The signal feature analysis module includes a frequency analysis submodule, an energy distribution analysis submodule, and a feature parameter construction submodule;

[0101] The frequency analysis submodule performs frequency domain conversion based on pressure fluctuation data, identifies and records the frequency and intensity of frequency components, extracts the key frequency components of the signal, and generates frequency characteristic data;

[0102] In the frequency analysis submodule, based on the pressure fluctuation data, the Fast Fourier Transform (FFT) method is first used to convert the signal into the frequency domain using the numpy.fft.fft() function. The frequency and intensity of each frequency component are then identified. In this step, the number of sampling points for the FFT transform is set to the signal length to ensure that subtle frequency variations are captured. After the analysis is complete, the key frequency components are extracted, including the main vibration frequencies and their corresponding intensities. The numpy.argmax() function is used to find the frequency point with the highest energy, thus forming the frequency signature data.

[0103] The energy distribution analysis submodule performs energy spectrum analysis on the frequency characteristic data, uses spectrum analysis methods to calculate the energy value of each frequency component, analyzes the energy distribution of the entire spectrum, and obtains energy distribution data;

[0104] Spectrum analysis method, according to the formula:

[0105]

[0106]

[0107] Calculate the energy value of each frequency component to obtain energy distribution data, where: is the signal in the frequency domain, is the window function adjustment coefficient, is the signal in the time domain, is the window function, is the sampling rate adjustment factor, Frequency The energy spectral density at is the frequency correction offset value, is the signal enhancement factor;

[0108] The execution process is as follows:

[0109] Signal preprocessing is to optimize the signal by applying window functions and adjusting coefficients. It will be multiplied by a window function and window function adjustment coefficients , to reduce spectrum leakage and improve spectrum smoothness;

[0110] Sampling rate adjustment, Used to adjust the sampling rate according to the frequency domain characteristics of the signal, helping to find the best balance between improving frequency resolution and reducing computational complexity;

[0111] Perform an improved FFT to transform the preprocessed signal Passed to the FFT algorithm to perform frequency domain conversion, taking into account signal enhancement and frequency correction, which directly affects the accuracy of subsequent energy spectrum calculation;

[0112] Energy spectrum calculation and enhancement, calculation of the signal after FFT conversion The energy spectrum of , combined with signal enhancement factors and frequency correction offset value , to ensure the accuracy and reliability of the energy spectrum.

[0113] By introducing improved parameters, the energy distribution analysis submodule can more accurately calculate the energy value of each frequency component and analyze the energy distribution of the entire spectrum. This not only improves the accuracy and reliability of spectrum analysis, but also increases the flexibility and adaptability of the analysis process, making it more capable of meeting the needs of complex signal processing.

[0114] The characteristic parameter construction submodule comprehensively analyzes the energy distribution data and the identified frequency characteristics, combines the data points according to specific rules, constructs a matrix that fully describes the signal characteristics, and obtains the characteristic parameter matrix;

[0115] In the feature parameter construction submodule, energy distribution data is combined with identified frequency feature data for analysis. A data frame is constructed using pandas.DataFrame(), combining the energy and frequency values ​​of each frequency component. These data points are combined by setting specific rules, such as the proportion of energy value to total energy and the cumulative contribution of frequency components. Ultimately, a feature parameter matrix is ​​constructed that comprehensively describes the signal characteristics. This matrix includes the dominant frequencies of anal pressure fluctuations, their energy correlations with acoustic wave features, and a time-frequency feature clustering index, providing a detailed signal description for risk assessment and lesion prediction.

[0116] See also Figure 2 and Figure 7 ,The risk rating algorithm module includes a data preparation submodule, a linear discriminant analysis submodule, and a risk level generation submodule;

[0117] The data preparation submodule performs data standardization based on the characteristic parameter matrix to ensure the scale consistency of variables during the analysis process and obtain standardized characteristic data;

[0118] In the data preparation submodule, we first extract data from the feature parameter matrix and standardize the features using sklearn.preprocessing.StandardScaler() to ensure consistent scaling across all variables in subsequent analyses. This step calculates the mean and standard deviation of each feature and applies the transformation (data - mean) / standard deviation to normalize the raw data, generating standardized feature data. This eliminates the impact of different scales on model training, improving model accuracy and generalization.

[0119] The linear discriminant analysis submodule uses standardized feature data to perform linear discriminant analysis. By calculating the contribution of features to classification, it extracts feature combinations that can distinguish different risk levels and generates risk discrimination features.

[0120] In the Linear Discriminant Analysis submodule, linear discriminant analysis (LDA) is performed on the standardized feature data. LDA calculates the statistical contribution of each feature to the classification, thereby identifying the most discriminative feature combinations. The ratio of intra-class divergence to inter-class divergence is calculated, and features that maximize inter-class differences are selected to form risk discriminant features. LDA aims to reduce the dimensionality of the feature space while retaining key information that distinguishes between classes.

[0121] The risk level generation submodule classifies data samples based on risk discrimination features, assigns each sample to a corresponding risk level, obtains the risk level of each sample, and obtains the risk level indicator;

[0122] In the risk level generation submodule, samples are classified based on risk-discriminating features using classification algorithms from the sklearn library, such as LogisticRegression. Each sample is assigned a corresponding risk level based on its characteristics. This classification process is based on a model that models the relationship between features and risk levels. Once trained, the model automatically identifies the risk level of new samples and outputs a risk level indicator for each sample. This effectively transforms complex feature data into an intuitive risk level assessment, providing a basis for subsequent prevention and intervention measures.

[0123] See also Figure 2 and Figure 8 ,The lesion prediction module includes the risk level data analysis submodule, the time series modeling submodule, and the trend prediction submodule;

[0124] The risk level data analysis submodule collects risk level indicators, analyzes the changing patterns within a time span, identifies the changing trends and patterns of risk levels, and obtains risk level trend data;

[0125] In the risk level data analysis submodule, after collecting risk level indicators, the pandas library is used to perform time series analysis on the data. The .resample() method is used to resample the data according to the time span to identify the changing patterns of risk levels over different time periods. Next, the .diff() method is used to calculate the change in risk level between adjacent time points. Combined with the .rolling(window=3) .mean() method, the trend is smoothed to identify the changing trends and patterns of risk levels. This generates risk level trend data, revealing the evolution of risk levels over time and providing a foundation for building a time series model.

[0126] The time series modeling submodule uses the autoregressive moving average method based on risk level trend data to analyze the relationship between risk levels and time, determine model parameters, establish a mathematical basis for predicting lesion development trends, and generate a time series model;

[0127] Autoregressive moving average method, according to the formula:

[0128]

[0129] Generate a time series model where is the order of the autoregressive term, is the number of differences required for the original sequence, is the seasonal adjustment factor, is the order of the moving average term, is the autoregressive coefficient, is the trend smoothing coefficient, is the moving average coefficient, is the fluctuation adjustment coefficient, is the error adjustment margin, is the lag operator, is the observed value at time point t, is the error term at time point t;

[0130] The execution process is as follows:

[0131] Seasonal analysis, analyzing the seasonal characteristics of time series data to determine seasonal adjustment factors The value of helps to accurately reflect seasonal changes in the model;

[0132] Trend analysis, by calculating the long-term trend of the time series and determining the trend smoothing coefficient , in order to adjust the effect of trend in the autoregressive term;

[0133] Volatility analysis, analyzing short-term fluctuations in time series and determining the volatility adjustment coefficient , helps to adjust the impact of volatility in the moving average term;

[0134] Error analysis, assessing the volatility of time series forecast errors to determine the magnitude of error adjustments , thus reflecting the error volatility at different time points in the error term;

[0135] Parameter estimation and model fitting, based on the above analysis, estimate the model parameters and fit the improved ARIMA model;

[0136] Model testing: Use statistical test methods (such as AIC, residual analysis, etc.) to verify the accuracy and predictive ability of the model.

[0137] By applying this improved formula, the autoregressive integrated moving average (ARIMA) method can more accurately analyze and predict the relationship between risk levels and time, providing a solid mathematical foundation for predicting lesion development trends and effectively generating a time series model that reflects the dynamic changes in risk levels.

[0138] The trend prediction submodule uses a time series model to predict the changing trend of risk levels over a period of time in the future, predicts the development of lesions based on the trend, and obtains lesion development prediction records;

[0139] In the trend prediction submodule, the trained time series model is used to predict the risk level trend for N future time points using the .forecast(steps=N) method. This process calculates the likely trend of the risk level over a period of time based on the model parameters and historical risk level data, thereby predicting the progression of the lesion. Ultimately, a lesion progression prediction record is generated, detailing the expected time of lesion occurrence, progression rate, and critical prevention time window, providing important information for developing effective intervention measures.

[0140] See also Figure 2 and Figure 9 ,The intervention measure recommendation module includes the lesion development analysis submodule, the lifestyle improvement suggestion submodule, and the diet adjustment suggestion submodule;

[0141] The lesion development analysis submodule analyzes the potential development trends and key influencing factors of lesions based on lesion development prediction records, identifies risk points, and obtains lesion risk analysis results;

[0142] The lesion progression analysis submodule first conducts an in-depth analysis of lesion progression prediction records. The Python pandas library is used to read the recorded data and the .describe() method is applied to perform statistical analysis of key indicators, revealing potential lesion development trends. Subsequently, logical functions in the numpy library are used to identify high-risk areas. For example, samples with high emergency warning levels are screened based on lesion progression speed and expected occurrence month, generating lesion risk analysis results. This analysis helps identify key factors that may lead to accelerated lesions, providing a scientific basis for developing targeted intervention measures.

[0143] The lifestyle improvement suggestion submodule proposes targeted lifestyle improvement measures based on the results of the pathology risk analysis, including improving sleep quality and regular physical activity, and generates lifestyle improvement suggestions;

[0144] In the lifestyle improvement recommendations submodule, a rule engine approach is used to develop improvement measures based on the lesion risk analysis results. This process establishes a series of logical rules. If the lesion progression rate exceeds a threshold, an increase in physical activity frequency is recommended, implemented using if-else statements. This approach generates personalized lifestyle improvement recommendations based on different risk levels and key influencing factors, including improving sleep quality and regular physical activity. These recommendations aim to slow lesion progression and improve an individual's quality of life by adjusting lifestyle habits.

[0145] The dietary adjustment recommendation submodule combines the results of disease risk analysis and lifestyle improvement suggestions to propose a dietary adjustment plan based on the individual's condition and development trend, including dietary structure optimization, specific nutrient intake, and a list of intervention measures;

[0146] The Dietary Adjustment Recommendation submodule combines disease risk analysis results with lifestyle improvement recommendations, comprehensively considering the individual's condition and development trends, to design a dietary adjustment plan. This process utilizes the MinMaxScaler function in sklearn.preprocessing to quantify individual nutritional intake needs. Using pandas data processing capabilities, this process matches individual needs with the nutritional database, develops a plan for optimizing dietary structure and specific nutrient intake, and generates a list of intervention measures. Through precise dietary adjustments, the goal is to provide individuals with scientific nutritional support, assisting in disease prevention and health management.

[0147] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A system for predicting and preventing hemorrhoid fistula risk, characterized by: The hemorrhoid fistula risk prediction and prevention system includes an environmental factor monitoring module, an acoustic wave acquisition and analysis module, an anal pressure monitoring module, a signal feature analysis module, a risk rating algorithm module, a lesion prediction module, and an intervention measure recommendation module; The environmental factor monitoring module collects external temperature, humidity and individual activity data based on environmental sensors and individuals to obtain an environmental activity data set; The sound wave collection and analysis module combines the environmental activity data set to capture the sound waves generated by intestinal activities, analyzes the relationship between the two, and obtains the sound wave activity feature set; The anal pressure monitoring module records the pressure fluctuations in the anal area in combination with the acoustic wave activity feature set, and associates the pressure fluctuations with the features in the acoustic wave activity feature set to obtain pressure fluctuation data; The signal feature analysis module uses a spectrum analysis method to extract vibration frequency components and energy distribution based on pressure fluctuation data and construct a feature parameter matrix; The signal feature analysis module includes a frequency analysis submodule, an energy distribution analysis submodule, and a feature parameter construction submodule; The frequency analysis submodule performs frequency domain conversion based on the pressure fluctuation data, identifies and records the occurrence frequency and intensity of the frequency components, extracts the key frequency components of the signal, and generates frequency characteristic data; The energy distribution analysis submodule performs energy spectrum analysis on the frequency characteristic data, uses spectrum analysis method to calculate the energy value of each frequency component, analyzes the energy distribution of the entire spectrum, and obtains energy distribution data; The spectrum analysis method is according to the formula: ； ； Calculate the energy value of each frequency component to obtain energy distribution data, where: is the signal in the frequency domain, is the window function adjustment coefficient, is the signal in the time domain, is the window function, is the sampling rate adjustment factor, For frequency The energy spectral density at is the frequency correction offset value, is the signal enhancement factor; The characteristic parameter construction submodule comprehensively analyzes the energy distribution data and the identified frequency characteristics, combines the data points according to specific rules, constructs a matrix that fully describes the signal characteristics, and obtains the characteristic parameter matrix; The risk rating algorithm module analyzes the data in the characteristic parameter matrix, performs linear discriminant analysis, and obtains risk level indicators; The lesion prediction module uses the risk level index and the autoregressive moving average method to predict the lesion development trend and generate a lesion development prediction record; The intervention measure recommendation module proposes dietary adjustment and lifestyle improvement measures based on the lesion development prediction record and obtains an intervention measure list.

2. The hemorrhoid fistula risk prediction and prevention system according to claim 1, characterized in that: The environmental activity data set includes the average daily temperature, maximum humidity, and average number of active steps; the sound wave activity feature set includes the intestinal peristalsis sound wave frequency, the maximum sound wave intensity, and the sound wave duration; the pressure fluctuation data includes the maximum pressure value of anal contraction, the pressure release rate, and the pressure stability index; the characteristic parameter matrix includes the main frequency of anal pressure fluctuation, the correlation between sound waves and pressure fluctuation energy, and the time-frequency feature clustering index; the risk level index includes the probability score of hemorrhoidal fistula, the contribution of risk factors, and the emergency warning level; the lesion development prediction record includes the expected lesion occurrence month, the lesion development rate, and the critical prevention time window; the intervention measures list includes necessary dietary fiber increase recommendations, recommended daily activity, and regular medical examination frequency.

3. The hemorrhoid fistula risk prediction and prevention system according to claim 1, characterized in that: The environmental factor monitoring module includes an environmental data acquisition submodule, a data optimization submodule, and a result analysis submodule; The environmental data acquisition submodule activates the temperature and humidity sensor and the activity monitoring device to continuously collect environmental temperature, humidity and individual activity information within a set time interval, timestamps and preliminarily classifies the collected raw data, and generates raw data records; The data optimization submodule receives the original data records, screens the data according to a preset threshold, removes outliers, performs linear interpolation on the data set, fills in missing values, segments the data, and performs normalization based on the time series to obtain a standardized data set; The result analysis submodule analyzes the data change trend based on the standardized data set, identifies the correlation between the ambient temperature and humidity and the individual activity amount by calculating the data change rate within the difference time window, and constructs the environmental activity data set.

4. The hemorrhoid fistula risk prediction and prevention system according to claim 1, characterized in that: The acoustic wave collection and analysis module includes an acoustic wave capture submodule, an acoustic wave data screening submodule, and an acoustic environment feature mapping submodule; The sound wave capture submodule configures a microphone array in the environment, starts capturing intestinal activity sound waves at a fixed time, records basic properties of the sound waves, including frequency range, intensity and timestamp, organizes the information into a sequence, and generates a sound wave sequence record; The sound wave data screening submodule performs preliminary screening based on the sound wave sequence record according to the intensity and frequency range of the sound wave, removes background noise interference items, performs time series analysis on the remaining sound waves, identifies the repeatability of the sound wave pattern, and obtains the sound wave pattern after screening; The acoustic environment feature mapping submodule matches and analyzes the screened sound wave patterns with the data in the environmental activity data set, analyzes the interaction between the intestinal activity sound waves and environmental factors and the amount of individual activity, and constructs a sound wave activity feature set.

5. The hemorrhoid fistula risk prediction and prevention system according to claim 1, characterized in that: The anal pressure monitoring module includes a pressure data acquisition submodule, a data integration submodule, and a feature matching submodule; The pressure data acquisition submodule deploys sensors in the anal area to record pressure changes in real time, including the magnitude of change, frequency of occurrence, and duration. Data is continuously acquired to form a serialized record to generate a pressure change record. The data integration submodule performs preliminary processing on the pressure change record, including removing interference signals introduced by body movement or external factors, reducing data fluctuations, ensuring signal continuity and analyzability, and obtaining a preprocessed pressure signal; The feature matching submodule is based on the preprocessed pressure signal, analyzes the key features in the signal, including the pattern of pressure peaks, the changing laws of frequency and amplitude, and identifies the correlation pattern between pressure changes and intestinal activities in combination with the acoustic wave activity feature set to obtain pressure fluctuation data.

6. The hemorrhoid fistula risk prediction and prevention system according to claim 1, characterized in that: The risk rating algorithm module includes a data preparation submodule, a linear discriminant analysis submodule, and a risk level generation submodule; The data preparation submodule performs data standardization based on the characteristic parameter matrix to ensure the scale consistency of the variables during the analysis process and obtain standardized characteristic data; The linear discriminant analysis submodule uses the standardized feature data to perform linear discriminant analysis, calculates the contribution of features to classification, extracts feature combinations that can distinguish different risk levels, and generates risk discrimination features; The risk level generation submodule classifies data samples based on risk discrimination features, assigns each sample to a corresponding risk level, obtains the risk level of each sample, and acquires a risk level indicator.

7. The hemorrhoid fistula risk prediction and prevention system according to claim 1, characterized in that: The lesion prediction module includes a risk level data analysis submodule, a time series modeling submodule, and a trend prediction submodule; The risk level data analysis submodule collects risk level indicators, analyzes the changing rules within a time span, identifies the changing trends and patterns of risk levels, and obtains risk level trend data; The time series modeling submodule uses the autoregressive moving average method based on the risk level trend data to analyze the relationship between the risk level and time, determine the model parameters, establish a mathematical basis for the prediction of the lesion development trend, and generate a time series model; The autoregressive moving average method is based on the formula: ； Generate a time series model where is the order of the autoregressive term, is the number of differences required for the original sequence, is the seasonal adjustment factor, is the order of the moving average term, is the autoregression coefficient, is the trend smoothing coefficient, is the moving average coefficient, is the fluctuation adjustment coefficient, is the error adjustment margin, is the lag operator, is the observed value at time point t, is the error term at time point t; The trend prediction submodule uses a time series model to predict the changing trend of the risk level in the future, predicts the development of the lesion based on the trend, and obtains the lesion development prediction record.

8. The hemorrhoid fistula risk prediction and prevention system according to claim 1, characterized in that: The intervention measure recommendation module includes a lesion development analysis submodule, a lifestyle improvement suggestion submodule, and a diet adjustment suggestion submodule; The lesion development analysis submodule analyzes the potential development trend and key influencing factors of the lesion based on the lesion development prediction record, identifies risk points, and obtains lesion risk analysis results; The lifestyle improvement suggestion submodule proposes targeted lifestyle improvement measures based on the lesion risk analysis results, including improving sleep quality and regular physical activities, and generates lifestyle improvement suggestions; The dietary adjustment suggestion submodule combines the lesion risk analysis results and lifestyle improvement suggestions to propose a dietary adjustment plan based on individual conditions and development trends, including dietary structure optimization, specific nutrient intake, and obtaining a list of intervention measures.

Images