A method and system for analyzing digestive health data

Abstract

This invention relates to the field of intelligent medical monitoring and health data analysis technology, specifically disclosing a method and system for analyzing digestive tract health data. It involves real-time acquisition of differential frequency and sound intensity data generated by intestinal peristalsis using a multi-source sensor array deployed on the abdominal surface. Empirical mode decomposition and Hilbert transform are combined to extract intestinal motility rhythm stability features. Normalization and Haar wavelet decomposition methods are used to calculate the variation features of bowel sound intensity, constructing an intestinal functional state feature vector. This feature vector is input into a health assessment model trained based on a random forest algorithm, outputting individualized digestive tract health scores and health level indicators, and triggering multiple types of abnormal warning signals. This constructs a closed-loop management system from data acquisition and intelligent analysis to risk warning, improving the intelligent level of early identification and health management of digestive tract diseases.

Description

Technical Field

[0001] This invention relates to the field of intelligent medical monitoring and health data analysis technology, specifically to a method and system for analyzing digestive tract health data. Background Technology

[0002] Currently, digestive health monitoring mainly relies on endoscopy, imaging examinations, or laboratory tests. These methods are typically invasive, costly, and difficult to implement continuous dynamic monitoring. In recent years, digestive function assessment technologies based on bowel sound analysis have gradually gained attention. These technologies collect sound signals generated by intestinal peristalsis and extract features using signal processing algorithms to assess intestinal health. However, existing technologies mostly employ single sensors or traditional spectral analysis methods, making it difficult to effectively capture the nonlinear and non-stationary characteristics of intestinal motility, resulting in insufficient sensitivity and specificity in feature extraction. Furthermore, existing methods lack adaptability to individual differences in data processing and have failed to establish effective multi-feature fusion assessment models, limiting their application in real-time health monitoring and early warning of abnormalities.

[0003] The existing technology has the following shortcomings:

[0004] A key but often overlooked problem with existing bowel sound analysis techniques is the lack of effective decoupling between intestinal motility fluctuations and rhythm disturbances. Traditional methods typically rely on single frequency or time domain features (such as average sound intensity and dominant frequency analysis), neglecting the dynamic correlation between frequency stability and intensity variation during intestinal peristalsis. This makes it difficult to distinguish between different pathological patterns such as "peristaltic rhythm disorder" and "abnormal motility fluctuation." Furthermore, existing techniques often use fixed thresholds or linear models for anomaly detection, failing to adapt to individual differences in intestinal motility (such as the influence of diet and menstrual cycles), easily leading to misdiagnosis. This invention achieves independent quantification and synergistic analysis of rhythm stability and motility fluctuations by combining empirical mode decomposition of differential frequency data with Hilbert instantaneous frequency analysis and wavelet energy ratio feature extraction from sound intensity data. This allows for accurate identification of coupled anomalies and enables personalized health assessment based on a random forest model, effectively overcoming the shortcomings of existing techniques. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for analyzing digestive health data to solve the problems mentioned above.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A method for analyzing digestive health data includes the following steps:

[0008] S1: Real-time acquisition of differential frequency data and sound intensity data of intestinal peristalsis sounds through a multi-source sensor array deployed on the abdominal surface;

[0009] The multi-source sensor array includes a wideband microphone sensor and a sound pressure sensor;

[0010] S2: Construct a differential frequency feature sequence from the collected differential frequency data, analyze the differential frequency feature sequence, and calculate the stability feature value of intestinal motility rhythm based on the analysis results to evaluate the regularity of intestinal peristalsis;

[0011] S3: Normalize the sound intensity data, construct an intensity fluctuation sequence, and calculate the characteristic value of bowel sound intensity variation based on the intensity fluctuation sequence to identify whether there are abnormal fluctuations in intestinal motility;

[0012] S4: Construct a feature vector of intestinal function status by combining the stability feature value of intestinal motility rhythm with the feature value of bowel sound intensity variation. Input the vector into the health assessment model for analysis. Generate a digestive health level index based on the analysis results and output a warning signal for abnormal digestive health.

[0013] As a further aspect of the present invention: the assessment of the regularity of intestinal peristalsis specifically includes:

[0014] Differential frequency data of intestinal peristalsis sounds are acquired, and the acquired differential frequency data are constructed into a differential frequency feature sequence. The differential frequency feature sequence is analyzed, and based on the analysis results, the stability feature value of intestinal motility rhythm is calculated. It is determined whether the stability feature value of intestinal motility rhythm is greater than or equal to a preset threshold. If it is, the intestinal peristalsis is irregular; if not, the intestinal peristalsis is regular.

[0015] As a further aspect of the present invention: the process for obtaining the intestinal motility rhythm stability characteristic value is as follows:

[0016] Differential frequency data sequences collected during intestinal peristalsis are obtained, and empirical mode decomposition is performed on the differential frequency data sequences to obtain a set of intrinsic mode functions and a residual term;

[0017] Applying the Hilbert transform to each intrinsic mode function yields its corresponding analytic signal and instantaneous frequency, expressed as follows:

[0018] Based on the instantaneous frequency of each intrinsic mode function, the ratio of the variance to the mean of the instantaneous frequency over the entire time range is calculated to obtain the frequency stability index.

[0019] The frequency stability indices of all intrinsic mode functions are weighted and averaged to obtain the characteristic value of intestinal motility rhythm stability.

[0020] As a further aspect of the present invention: the identification of whether there are abnormal fluctuations in intestinal motility specifically includes:

[0021] Acquire sound intensity data of intestinal peristalsis sounds, normalize the sound intensity data, construct an intensity fluctuation sequence, calculate the characteristic value of bowel sound intensity variation based on the intensity fluctuation sequence, and determine whether the characteristic value of bowel sound intensity variation is greater than or equal to a preset threshold. If it is, there is abnormal fluctuation in intestinal motility; otherwise, there is no abnormal fluctuation in intestinal motility.

[0022] As a further aspect of the present invention: the characteristic values ​​of bowel sound intensity variation specifically include:

[0023] Acquire sound intensity data of intestinal peristalsis sounds and normalize the sound intensity data;

[0024] The normalized sound intensity data is decomposed into first-order Haar wavelet decomposition to obtain approximation coefficients and detail coefficients.

[0025] Calculate the energy of the approximation coefficients and the energy of the detail coefficients at each scale;

[0026] Calculate the ratio of detail energy to approximation energy at each scale, and construct an energy fluctuation sequence based on the obtained ratios in scale order;

[0027] The ratio of the average to the standard deviation of all ratios within the energy fluctuation sequence is calculated to obtain the characteristic value of bowel sound intensity variation.

[0028] As a further aspect of the present invention: the step of constructing an intestinal functional state feature vector from the intestinal motility rhythm stability feature value and the bowel sound intensity variation feature value, and inputting it into a health assessment model for analysis, specifically includes:

[0029] The stability features of intestinal motility rhythm and the variation features of bowel sound intensity of the digestive tract are obtained, and the stability features of intestinal motility rhythm and the variation features of bowel sound intensity are used to construct a feature vector of intestinal functional state. This vector is then input into a health assessment model. The training objective of the health assessment model is to minimize the error between the predicted digestive health score and the actual digestive health score. The health assessment model is trained, and a digestive health score is output based on the trained health assessment model. The health assessment model is a random forest model.

[0030] As a further aspect of the present invention: the training process of the health assessment model is as follows:

[0031] We obtained intestinal motility rhythm stability features, bowel sound intensity variation features, and digestive health scores from multiple groups of subjects and used them to form a training dataset, which was divided into training and validation sets. We used a random forest model to minimize the mean squared error between the predicted and actual digestive health scores as the objective function for model training. In each training iteration, we generated multiple decision trees through bootstrapping and randomly selected feature subsets when splitting nodes to enhance the model's generalization ability.

[0032] As a further aspect of the present invention: the generation of digestive tract health level indicators based on the analysis results specifically includes:

[0033] Determine whether the digestive health score of the digestive tract is greater than or equal to a preset threshold. If yes, the corresponding digestive health status is normal; otherwise, the corresponding digestive health status is abnormal.

[0034] As a further aspect of the present invention: based on the assessment results of abnormal digestive tract health status, it is determined that an individual is at risk of digestive tract functional abnormalities, and the system will automatically trigger a health abnormality warning signal; based on the normal digestive tract health status, it is considered that the digestive tract functional status is normal, and no health abnormality warning signal is issued.

[0035] A digestive health data analysis system, comprising:

[0036] The data acquisition module collects differential frequency data and sound intensity data of intestinal peristalsis sounds in real time through a multi-source sensor array deployed on the abdominal surface.

[0037] The rhythm stability analysis module constructs a differential frequency feature sequence from the collected differential frequency data, analyzes the differential frequency feature sequence, and calculates the intestinal motility rhythm stability feature value based on the analysis results to assess the regularity of intestinal peristalsis.

[0038] The intestinal motility abnormality analysis module normalizes the sound intensity data, constructs an intensity fluctuation sequence, and calculates the characteristic value of bowel sound intensity variation based on the intensity fluctuation sequence to identify whether there are abnormal fluctuations in intestinal motility.

[0039] The digestive tract health assessment and early warning module constructs a feature vector of intestinal function status by combining the stability feature value of intestinal motility rhythm and the variation feature value of bowel sound intensity. This vector is then input into the health assessment model for analysis. Based on the analysis results, a digestive tract health level index is generated, and an early warning signal for abnormal digestive tract health is output.

[0040] The beneficial effects of this invention are:

[0041] (1) This invention utilizes a multi-source sensor array deployed on the abdominal surface to collect real-time information on the frequency and intensity changes of sound generated during intestinal peristalsis. Combining empirical mode decomposition (EMD) with Hilbert transform, a high-precision and robust intestinal rhythm stability feature extraction mechanism is constructed. This mechanism adaptively decomposes nonlinear and non-stationary differential frequency signals into multiple intrinsic mode functions (IMFs). Furthermore, it obtains the instantaneous frequency information of each mode component through Hilbert transform, calculates the ratio of its variance to mean as a frequency stability index, and finally obtains the intestinal motility rhythm stability feature value through weighted averaging, thereby achieving a refined characterization of the regularity of intestinal peristalsis. Simultaneously, in terms of dynamic fluctuation detection, the system normalizes the collected sound intensity data and uses first-order Haar wavelet decomposition technology to extract approximation coefficients and detail coefficients. It then calculates the energy distribution and its ratio at different scales, forming an energy fluctuation sequence. The ratio of its mean to standard deviation is used as the intestinal sound intensity variation feature value to identify whether there are abnormal fluctuations in intestinal motility. This method not only possesses excellent time-frequency localization capabilities, enabling it to capture transient trends in bowel sound intensity, but also enhances the ability of features to discriminate pathological or functional abnormalities through multi-scale energy ratios. This improves the system's anti-interference performance and individual adaptability in complex environments, significantly outperforming traditional methods based on Fourier transform or simple time-domain statistics, and demonstrating higher sensitivity and clinical application value.

[0042] (2) This invention integrates the intestinal motility rhythm stability feature value and the bowel sound intensity variation feature value to construct an intestinal functional state feature vector, which is then input into a health assessment model trained based on the random forest algorithm, to achieve accurate modeling and dynamic assessment of an individual's digestive health status. During training, the model uses self-sampling to generate multiple decision trees and introduces a feature subset random selection mechanism when splitting nodes, effectively enhancing the model's generalization ability and anti-interference performance. Its optimization objective is to minimize the mean square error between the predicted score and the actual health score comprehensively assessed by clinicians, ensuring good clinical consistency of the output results. In practical applications, the system compares the health score output by the model with a preset threshold, automatically classifies health levels, and combines an abnormality type identification mechanism to determine whether it is a "peristaltic rhythm disorder," "dynamic fluctuation abnormality," or "comprehensive functional disorder" abnormality, thereby generating multi-dimensional warning information. Warning signals can be pushed to users or medical personnel in real time via wearable devices, smart terminals, or remote medical platforms, realizing a closed-loop management process from data collection, feature extraction, model prediction to risk feedback. This method not only improves the personalization and intelligence of digestive health assessment, but also provides efficient and reliable technical support for early intervention, continuous monitoring and remote health management of diseases. Attached Figure Description

[0043] The invention will now be further described with reference to the accompanying drawings.

[0044] Figure 1 This is a flowchart of a digestive health data analysis method according to the present invention;

[0045] Figure 2 This is a flowchart of a digestive health data analysis system according to the present invention. Detailed Implementation

[0046] 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.

[0047] Please see Figure 1 As shown, this invention is a method for analyzing digestive health data, comprising the following steps:

[0048] S1: Real-time acquisition of differential frequency data and sound intensity data of intestinal peristalsis sounds through a multi-source sensor array deployed on the abdominal surface;

[0049] The multi-source sensor array includes a wideband microphone sensor and a sound pressure sensor;

[0050] S2: Construct a differential frequency feature sequence from the collected differential frequency data, analyze the differential frequency feature sequence, and calculate the stability feature value of intestinal motility rhythm based on the analysis results to evaluate the regularity of intestinal peristalsis;

[0051] S3: Normalize the sound intensity data, construct an intensity fluctuation sequence, and calculate the characteristic value of bowel sound intensity variation based on the intensity fluctuation sequence to identify whether there are abnormal fluctuations in intestinal motility;

[0052] S4: Construct a feature vector of intestinal function status by combining the stability feature value of intestinal motility rhythm with the feature value of bowel sound intensity variation. Input the vector into the health assessment model for analysis. Generate a digestive health level index based on the analysis results and output a warning signal for abnormal digestive health.

[0053] In S1, differential frequency data and sound intensity data of intestinal peristalsis sounds are collected in real time through a multi-source sensor array deployed on the abdominal surface, specifically including:

[0054] Real-time acquisition of intestinal peristalsis sound signals is achieved through a multi-source sensor array deployed on the abdominal surface of the subject. The multi-source sensor array includes at least one broadband microphone sensor and multiple sound pressure sensors. The broadband microphone sensor is used to capture full-frequency bowel sound signals generated during intestinal peristalsis, and it has a high frequency response range (e.g., 20Hz to 20kHz), which can effectively capture sound changes caused by intestinal movement. The sound pressure sensors are used to measure local abdominal pressure fluctuations caused by intestinal peristalsis, and assist in extracting differential frequency information related to bowel sounds.

[0055] In the specific implementation process, the subject attaches the sensor array to the abdominal region, such as the periumbilical area and the left lower abdomen, typical areas for bowel sounds. After the acquisition system is activated, each sensor synchronously records the time-domain waveform of the sound generated by intestinal peristalsis and the corresponding pressure change signal. By performing spectral analysis on the raw audio signal acquired by the broadband microphone sensor and combining it with the pressure change curve obtained by the sound pressure sensor, the differential frequency data of the intestinal peristalsis sound is calculated. At the same time, the peak and fluctuation amplitude of the sound intensity in the audio signal are extracted to construct a sound intensity data sequence, which serves as the input basis for the subsequent feature extraction module. The entire acquisition process is non-invasive, continuous, and has good temporal resolution, accurately reflecting the dynamic changes of individual intestinal motility.

[0056] In S2, the collected differential frequency data is constructed into a differential frequency feature sequence. This sequence is then analyzed, and based on the results, the stability characteristic value of intestinal motility rhythm is calculated to assess the regularity of intestinal peristalsis. Specifically, this includes:

[0057] Differential frequency data of intestinal peristalsis sounds are acquired, and the acquired differential frequency data are constructed into a differential frequency feature sequence. The differential frequency feature sequence is analyzed, and based on the analysis results, the stability feature value of intestinal motility rhythm is calculated. It is determined whether the stability feature value of intestinal motility rhythm is greater than or equal to a preset threshold. If it is, the intestinal peristalsis is irregular; if not, the intestinal peristalsis is regular.

[0058] The process for obtaining the stability feature values ​​of the intestinal motility rhythm is as follows:

[0059] Differential frequency data sequences collected during intestinal peristalsis are obtained. Empirical mode decomposition (EMD) is performed on these sequences to obtain a set of intrinsic mode functions and a residual term. The calculation expression is as follows: ;in, This indicates the number of intrinsic mode functions. This represents the total number of intrinsic mode functions. Indicates the first One intrinsic mode function, Indicates the final residual. Indicates the time collection point. This represents a differential frequency data sequence;

[0060] Applying the Hilbert transform to each intrinsic mode function yields its corresponding analytic signal, calculated as follows: ;

[0061] Applying the Hilbert transform to each intrinsic mode function yields its corresponding instantaneous frequency, expressed as: ;

[0062] in, Indicates an analytical signal. Indicates instantaneous frequency. Indicates instantaneous amplitude. Represents the phase function. Represents the imaginary unit;

[0063] Based on the instantaneous frequency of each intrinsic mode function, the ratio of the variance to the mean of the instantaneous frequency over the entire time range is calculated to obtain the frequency stability index.

[0064] The frequency stability indices of all intrinsic mode functions are weighted and averaged to obtain the characteristic value of intestinal motility rhythm stability. The calculation expression is as follows: ;

[0065] in, Indicates the first Frequency stability index of an intrinsic mode function This represents the characteristic value of intestinal motility rhythm stability.

[0066] It should be noted that this invention overcomes the limitations of traditional spectral analysis methods in processing nonlinear and non-stationary signals by adaptively decomposing complex differential frequency data sequences into multiple intrinsic mode functions and a residual term. Subsequently, the instantaneous frequency of each intrinsic mode function component is obtained using Hilbert transform, and the ratio of its variance to mean is calculated as a frequency stability index. Furthermore, the information from each component is fused through weighted averaging to construct a stable feature value for intestinal motility rhythm with clear physiological significance and strong anti-interference ability. This innovation not only improves the sensitivity of feature extraction to minor abnormal fluctuations in intestinal rhythm but also enhances the model's robustness to individual differences and environmental noise, providing reliable input for subsequent health assessment models and demonstrating significant technological advancement and clinical application value.

[0067] In S3, the sound intensity data is normalized to construct an intensity fluctuation sequence. Based on this sequence, the characteristic value of bowel sound intensity variation is calculated to identify any abnormal fluctuations in intestinal motility. Specifically, this includes:

[0068] Acquire sound intensity data of intestinal peristalsis sounds, normalize the sound intensity data, construct an intensity fluctuation sequence, calculate the characteristic value of bowel sound intensity variation based on the intensity fluctuation sequence, and determine whether the characteristic value of bowel sound intensity variation is greater than or equal to a preset threshold. If it is, there is abnormal fluctuation in intestinal motility; otherwise, there is no abnormal fluctuation in intestinal motility.

[0069] The bowel sound intensity variation characteristic values ​​specifically include:

[0070] The sound intensity data of intestinal peristalsis sounds were obtained, and the sound intensity data was normalized. The calculation expression is as follows: ;

[0071] in, This represents the normalized sound intensity data. Indicates the time collection point. This represents the sound intensity data collected in real time. This represents the maximum value in the sound intensity data. This represents the minimum value in the sound intensity data;

[0072] The normalized sound intensity data is decomposed into first-order Haar wavelet decomposition to obtain approximation coefficients and detail coefficients.

[0073] The energy of the approximation coefficients at each scale is calculated using the following expression: ;

[0074] in, The energy representing the approximation coefficient. Indicates the quantity of scale, Indicates the first Approximation coefficients at various scales;

[0075] The energy of the detail factor at each scale is calculated using the following expression: ;

[0076] in, Energy representing detail coefficients Indicates the first Detail factor at each scale;

[0077] Calculate the ratio of detail energy to approximation energy at each scale, and construct an energy fluctuation sequence based on the obtained ratios in scale order;

[0078] The ratio of the average to the standard deviation of all ratios within the energy fluctuation sequence is calculated to obtain the characteristic value of bowel sound intensity variation.

[0079] It should be noted that this invention normalizes the sound intensity data to eliminate the influence of individual differences and equipment acquisition biases, thereby improving data comparability. Subsequently, a first-order Haar wavelet transform is introduced to decompose the normalized sound intensity signal into approximation coefficients and detail coefficients at multiple scales. By calculating the ratio of detail energy to approximation energy at each scale, an energy fluctuation sequence is constructed. Furthermore, the ratio of the mean to the standard deviation of this sequence is used as the characteristic value of bowel sound intensity variation, achieving a highly sensitive characterization of intestinal motility fluctuations. This method not only possesses excellent time-frequency localization capabilities, capturing the transient change trend of bowel sound intensity, but also enhances the feature's ability to discriminate pathological or functional abnormalities through wavelet energy ratios. It exhibits strong robustness and clinical applicability, significantly improving the accuracy and intelligence level of intestinal motility abnormality detection.

[0080] In S4, the intestinal motility rhythm stability feature value and the bowel sound intensity variation feature value are used to construct an intestinal functional state feature vector, which is then input into the health assessment model for analysis. Based on the analysis results, a digestive tract health level index is generated, and a digestive tract health abnormality warning signal is output, specifically including:

[0081] The stability features of intestinal motility rhythm and the variation features of bowel sound intensity of the digestive tract are obtained, and the stability features of intestinal motility rhythm and the variation features of bowel sound intensity are used to construct a feature vector of intestinal functional state. This vector is then input into a health assessment model. The training objective of the health assessment model is to minimize the error between the predicted digestive health score and the actual digestive health score. The health assessment model is trained, and a digestive health score is output based on the trained health assessment model. The health assessment model is a random forest model.

[0082] The training process of the health assessment model is as follows:

[0083] We obtained intestinal motility stability features, bowel sound intensity variation features, and digestive health scores from multiple groups of subjects, and compiled them into a training dataset, which was divided into training and validation sets. A random forest model was used to minimize the mean squared error between the predicted and actual digestive health scores as the objective function for model training. In each training iteration, multiple decision trees were generated through bootstrapping, and a subset of features was randomly selected during node splitting to enhance the model's generalization ability. After training, the optimal model parameters were saved and used to predict health scores on newly input intestinal functional state feature vectors, thereby achieving a quantitative assessment of individual digestive health status.

[0084] Determine whether the digestive health score of the digestive tract is greater than or equal to a preset threshold. If yes, the corresponding digestive health status is normal; otherwise, the corresponding digestive health status is abnormal.

[0085] If the assessment results of abnormal digestive health status are used, it is determined that the individual is at risk of digestive dysfunction, and the system will automatically trigger a health abnormality warning signal; if the digestive health status is normal, the digestive function is considered to be normal, and no warning will be issued.

[0086] Upon triggering the warning, the system further combines the intestinal motility stability feature value and bowel sound intensity variation feature value in the input intestinal function status feature vector, and compares them with their respective preset thresholds to identify the specific type of abnormality.

[0087] If the rhythm stability characteristic value is greater than or equal to its corresponding preset threshold while the intensity variation characteristic value is less than its preset threshold, it is judged as "peristaltic rhythm disorder" abnormality;

[0088] If the rhythm stability characteristic value is less than its preset threshold while the intensity variation characteristic value is greater than or equal to its threshold, it is judged as an "abnormal dynamic fluctuation type" anomaly; if both are greater than or equal to their respective preset thresholds, it is judged as a "comprehensive functional disorder type" anomaly.

[0089] Ultimately, the system delivers warning information to users or medical staff via wearable devices, smart terminals, or telemedicine platforms through audio-visual prompts, screen displays, and mobile push notifications. It also records the timestamp of the warning event, relevant characteristic data, and abnormality type classification for subsequent health trend analysis, assisted diagnosis, and the development of personalized intervention recommendations. The entire warning process achieves closed-loop management from data collection, feature extraction, model prediction to risk feedback, possessing real-time performance, intelligence, and clinical application value.

[0090] Please see Figure 2 As shown, a digestive health data analysis system includes:

[0091] The data acquisition module collects differential frequency data and sound intensity data of intestinal peristalsis sounds in real time through a multi-source sensor array deployed on the abdominal surface.

[0092] The rhythm stability analysis module constructs a differential frequency feature sequence from the collected differential frequency data, analyzes the differential frequency feature sequence, and calculates the intestinal motility rhythm stability feature value based on the analysis results to assess the regularity of intestinal peristalsis.

[0093] The intestinal motility abnormality analysis module normalizes the sound intensity data, constructs an intensity fluctuation sequence, and calculates the characteristic value of bowel sound intensity variation based on the intensity fluctuation sequence to identify whether there are abnormal fluctuations in intestinal motility.

[0094] The digestive tract health assessment and early warning module constructs a feature vector of intestinal function status by combining the stability feature value of intestinal motility rhythm and the variation feature value of bowel sound intensity. This vector is then input into the health assessment model for analysis. Based on the analysis results, a digestive tract health level index is generated, and an early warning signal for abnormal digestive tract health is output.

[0095] The working principle of this invention is as follows: A multi-source sensor array deployed on the abdominal surface of the subject, including a broadband microphone sensor and a sound pressure sensor, synchronously collects sound signals and pressure fluctuation data generated during intestinal peristalsis. Differential frequency and sound intensity information are then extracted to construct highly sensitive physiological characteristic parameters. At the feature extraction level, this invention innovatively introduces a method combining Empirical Mode Decomposition (EMD) and Hilbert Transform to accurately extract intestinal motility rhythm stability feature values, thereby quantitatively assessing the regularity of intestinal peristalsis. Simultaneously, normalization processing and Haar wavelet decomposition strategies are employed to calculate the variation feature values ​​of bowel sound intensity to identify whether there are abnormal fluctuations in intestinal motility. These two key features are fused into a unified intestinal functional state feature vector, which is then fed as input into a health assessment model trained based on the random forest algorithm. This model is optimized to minimize the mean squared error between the predicted score and the clinical health score, exhibiting good generalization ability and prediction accuracy. Ultimately, the system generates digestive tract health level indicators based on the model output and, combined with a preset threshold judgment mechanism, outputs multi-category health abnormality warning signals, covering various types such as "peristaltic rhythm disorder," "dynamic fluctuation abnormality," and "comprehensive functional disorder," supporting personalized medical intervention and long-term health trend analysis. Overall, this invention constructs a complete closed-loop management process from data collection to intelligent analysis and early warning feedback, demonstrating high clinical applicability, technological advancement, and application expansion potential.

[0096] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0097] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0098] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0099] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0100] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for analyzing digestive tract health data, characterized in that, Includes the following steps: S1: Real-time acquisition of differential frequency data and sound intensity data of intestinal peristalsis sounds through a multi-source sensor array deployed on the abdominal surface; The multi-source sensor array includes a wideband microphone sensor and a sound pressure sensor; S2: The collected differential frequency data are constructed into a differential frequency feature sequence. The differential frequency feature sequence is analyzed, and based on the analysis results, the stability characteristic value of intestinal motility rhythm is calculated to assess the regularity of intestinal peristalsis. Specifically, this includes: Differential frequency data of intestinal peristalsis sounds are acquired, and the acquired differential frequency data is constructed into a differential frequency feature sequence. The differential frequency feature sequence is analyzed, and based on the analysis results, the stability feature value of intestinal motility rhythm is calculated. It is determined whether the stability feature value of intestinal motility rhythm is greater than or equal to a preset threshold. If it is, the intestinal peristalsis is irregular; if not, the intestinal peristalsis is regular. The process for obtaining the stability feature values ​​of the intestinal motility rhythm is as follows: Differential frequency data sequences collected during intestinal peristalsis are obtained, and empirical mode decomposition is performed on the differential frequency data sequences to obtain a set of intrinsic mode functions and a residual term; Applying the Hilbert transform to each intrinsic mode function yields its corresponding analytic signal and instantaneous frequency, expressed as follows: Based on the instantaneous frequency of each intrinsic mode function, the ratio of the variance to the mean of the instantaneous frequency over the entire time range is calculated to obtain the frequency stability index. The frequency stability indices of all intrinsic mode functions are weighted and averaged to obtain the characteristic value of intestinal motility rhythm stability. S3: Normalize the sound intensity data, construct an intensity fluctuation sequence, and calculate the characteristic value of bowel sound intensity variation based on the intensity fluctuation sequence to identify whether there are abnormal fluctuations in intestinal motility. Specifically, this includes: Acquire sound intensity data of intestinal peristalsis sounds, normalize the sound intensity data, construct an intensity fluctuation sequence, calculate the characteristic value of intestinal sound intensity variation based on the intensity fluctuation sequence, and determine whether the characteristic value of intestinal sound intensity variation is greater than or equal to a preset threshold. If it is, there is abnormal fluctuation in intestinal motility; otherwise, there is no abnormal fluctuation in intestinal motility. The bowel sound intensity variation characteristic values ​​specifically include: Acquire sound intensity data of intestinal peristalsis sounds and normalize the sound intensity data; The normalized sound intensity data is decomposed into first-order Haar wavelet decomposition to obtain approximation coefficients and detail coefficients. Calculate the energy of the approximation coefficients and the energy of the detail coefficients at each scale; Calculate the ratio of detail energy to approximation energy at each scale, and construct an energy fluctuation sequence based on the obtained ratios in scale order; The ratio of the average to the standard deviation of all ratios within the energy fluctuation sequence is calculated to obtain the characteristic value of bowel sound intensity variation; S4: Construct a feature vector of intestinal functional status by combining the characteristic values ​​of intestinal motility rhythm stability and the characteristic values ​​of bowel sound intensity variation. Input this vector into the health assessment model for analysis. Based on the analysis results, generate a digestive health level index and output a warning signal for abnormal digestive health, specifically including: The stability features of intestinal motility rhythm and the variation features of bowel sound intensity of the digestive tract are obtained, and these features are used to construct an intestinal functional state feature vector. This vector is then input into a health assessment model. The training objective of the health assessment model is to minimize the error between the predicted and actual digestive health scores. The health assessment model is trained, and a digestive health score is output based on the trained model. The health assessment model is a random forest model. The training process of the health assessment model is as follows: We obtained the intestinal motility rhythm stability feature value, bowel sound intensity variation feature value and digestive health score of multiple groups of subjects, and formed them into a training dataset, which was divided into training set and validation set. We used a random forest model to minimize the mean square error between the predicted digestive health score and the actual digestive health score as the objective function for model training. In each training iteration, we generated multiple decision trees through bootstrap sampling, and randomly selected feature subsets when splitting nodes to enhance the model's generalization ability. Determine whether the digestive health score of the digestive tract is greater than or equal to a preset threshold. If yes, the corresponding digestive health status is normal; otherwise, the corresponding digestive health status is abnormal.

2. The method for analyzing digestive health data according to claim 1, characterized in that, The output of the early warning signal for abnormal digestive tract health specifically includes: Based on the assessment results of abnormal digestive health status, the system will determine that the individual is at risk of digestive dysfunction and will automatically trigger a health abnormality warning signal; if the digestive health status is normal, the individual is considered to have normal digestive function and no health abnormality warning signal will be issued.

3. A digestive tract health data analysis system, characterized in that, A method for analyzing digestive health data as described in any one of claims 1-2, comprising: The data acquisition module collects differential frequency data and sound intensity data of intestinal peristalsis sounds in real time through a multi-source sensor array deployed on the abdominal surface. The rhythm stability analysis module constructs a differential frequency feature sequence from the collected differential frequency data, analyzes the differential frequency feature sequence, and calculates the intestinal motility rhythm stability feature value based on the analysis results to assess the regularity of intestinal peristalsis. The intestinal motility abnormality analysis module normalizes the sound intensity data, constructs an intensity fluctuation sequence, and calculates the characteristic value of bowel sound intensity variation based on the intensity fluctuation sequence to identify whether there are abnormal fluctuations in intestinal motility. The digestive tract health assessment and early warning module constructs a feature vector of intestinal function status by combining the stability feature value of intestinal motility rhythm and the variation feature value of bowel sound intensity. This vector is then input into the health assessment model for analysis. Based on the analysis results, a digestive tract health level index is generated, and an early warning signal for abnormal digestive tract health is output.

Images