A colon electrophysiological monitoring system based on a lantern-shaped flexible array and an intelligent evaluation method for intestinal inflammation

Abstract

The application discloses a colon electrophysiological monitoring system based on a lantern-shaped flexible array and an intelligent evaluation method for intestinal inflammation, and relates to the technical field of medical treatment, and in particular to the technical field of colon electrophysiological monitoring. The colon electrophysiological monitoring system based on the lantern-shaped flexible array comprises a flexible electrode array, an inflatable balloon catheter and a signal acquisition module. The flexible electrode array comprises a plurality of electrode points and is formed into a plurality of longitudinal strips through laser cutting. The inflatable balloon catheter is used for winding and fixing the flexible electrode array on the outer surface of the balloon. In the inflated state of the balloon, the longitudinal strips are radially expanded, the whole is in the shape of a lantern, and each electrode point is conformally attached to the colon wall from multiple angles. The signal acquisition module is used for acquiring multi-channel colon electrophysiological signals. The signal analysis module is used for multi-dimensional analysis to obtain a preliminary judgment of the colon health degree. In combination with a machine learning model, a colon inflammation severity grading result is output. The application can stably adhere to the colon mucosa, realize high-quality monitoring of colon electrophysiological signals, identify intestinal inflammation by analyzing and extracting multi-dimensional features of the extracted intestinal electrical signals, and realize automatic evaluation of the colon inflammation degree in combination with artificial intelligence.

Description

Technical Field

[0001] This invention relates to the field of medical monitoring and intelligent diagnostic technology, specifically to a colonic electrophysiological monitoring system based on a lantern-shaped flexible array and an intelligent assessment method for enteritis. Background Technology

[0002] Inflammation of the colon is a core pathological change in many intestinal diseases, including ulcerative colitis and Crohn's disease, and is also a key indicator for assessing disease activity and treatment response. Currently, there are over 7 million people worldwide suffering from inflammatory bowel disease, with the incidence rate in children increasing annually. Clinical assessment of colon inflammation mainly relies on symptom scores, endoscopy, and fecal / serum biomarkers. While endoscopic biopsy is the gold standard, it is invasive and cannot be continuously monitored at the bedside; biomarkers lack site specificity and are generally difficult to reflect the dynamic functional changes of inflammation. More importantly, conventional methods often treat inflammation as a relatively static endpoint, failing to capture its real-time pathophysiological processes.

[0003] Gastrointestinal electrophysiological activities, such as slow waves and spike potentials, are direct manifestations of the functional state of smooth muscle cells and Cajal interstitial cells, providing real-time, objective assessments of gastrointestinal function. In recent years, significant progress has been made in minimally invasive procedures and surface electrophysiological mapping of the stomach, enabling the identification of specific electrophysiological phenotypes. However, research on colonic electrophysiological activity, especially its dynamic evolution under inflammatory conditions, remains unclear, particularly lacking suitable electrophysiological monitoring systems for the colon and methods to intelligently correlate electrophysiological signal characteristics with inflammatory status. Therefore, there is an urgent need to develop a non-invasive / minimally invasive, real-time, and quantitative functional diagnostic strategy for monitoring colonic inflammation. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a colonic electrophysiological monitoring system and an intelligent assessment method for enteritis based on a lantern-shaped flexible array. This invention can stably adhere to the colonic mucosa and achieve high-quality monitoring of colonic electrophysiological signals. By analyzing the multidimensional features of the extracted intestinal electrophysiological signals, it identifies intestinal inflammation and combines it with artificial intelligence to automatically assess the degree of colonic inflammation.

[0005] The technical solution of the present invention is as follows: A colonic electrophysiological monitoring system based on a lantern-shaped flexible array, comprising: The flexible electrode array is fabricated from a flexible printed circuit board and contains multiple electrode points. It is laser-cut into several longitudinal strips, which enables the electrode array to be transformed from a planar shape into a three-dimensional configuration. An inflatable balloon catheter, wherein the flexible electrode array is wound and fixed to the outer surface of the balloon, and when the balloon is inflated, the longitudinal strips are radially expanded, forming a lantern shape, so that each electrode point conformally adheres to the colon wall from multiple angles; The signal acquisition module, electrically connected to the electrode array, is used to amplify, filter, and perform analog-to-digital conversion on the multi-channel colonic electrophysiological signals, and to display and store them in real time. The signal analysis module performs multidimensional analysis on the collected electrophysiological data to make a preliminary judgment on the degree of colon health. Then, it combines artificial intelligence with an embedded trained machine learning model to perform more comprehensive feature extraction and health baseline normalization on the collected electrophysiological signals, and output the severity classification results of colon inflammation.

[0006] Furthermore, the electrode array is designed and manufactured using flexible electronic circuitry, with several rows and columns of circular metal electrode points arranged on it, making it suitable for space-constrained cavities.

[0007] Furthermore, when the balloon is not inflated, it is used for non-invasive insertion into the colon. After the balloon is inflated, the longitudinal strips are stretched to form a lantern-shaped three-dimensional structure, and each electrode point closely adheres to the colonic mucosa from different angles.

[0008] Furthermore, the multi-channel colonic electrophysiological signals obtained by the signal acquisition module are preprocessed, including at least one of filtering, baseline drift removal, and downsampling processing on the acquired raw intestinal electrophysiological signals.

[0009] Furthermore, the multidimensional analysis includes one or more of time-domain feature analysis, frequency-domain feature analysis, time-frequency domain dynamic feature analysis, and information domain feature analysis; the feature parameters obtained from each analysis process are correlated with the intestinal inflammation status to construct an objective quantitative indicator system for characterizing the presence of colonic inflammation.

[0010] Furthermore, the time-domain feature analysis involves performing waveform analysis on the preprocessed signal to obtain time-domain feature parameters, which include, but are not limited to, at least one of signal-to-noise ratio, mean amplitude, peak value, coefficient of variation, and zero-crossing rate. The frequency domain feature analysis involves estimating the power spectral density of the preprocessed signal to obtain frequency domain feature parameters, including but not limited to at least one of the following: dominant frequency, peak power spectral density, and energy distribution of typical physiological frequency bands (such as slow waves, spikes, etc.). The time-frequency domain dynamic feature analysis involves performing wavelet transform and short-time Fourier transform on the preprocessed signal to obtain time-frequency feature parameters, including but not limited to at least one of energy concentration, time-frequency kurtosis, frequency range, and subband energy. The information domain feature analysis involves performing complexity and randomness analysis on the preprocessed signal and extracting feature parameters including at least one of the following: spectral entropy, permutation entropy, approximate entropy, and multi-scale entropy.

[0011] Furthermore, by analyzing the signal-to-noise ratio, the signal quality under different healthy colonic states is compared; in the maximum amplitude analysis, the continuous data of each channel is divided into multiple non-overlapping time windows, the absolute maximum amplitude of the signal within each time window is extracted, and the absolute maximum amplitude is mapped onto a two-dimensional grid corresponding to the electrode layout to generate a dynamic heatmap characterizing the signal intensity distribution of each channel in different time windows; time-frequency analysis is performed through continuous wavelet transform, and wavelet coefficients are calculated using Morlet wavelets to generate a scale map with time and scale as coordinate axes and the amplitude of the wavelet coefficients representing the spectral power, to track instantaneous oscillation events and the change of the dominant frequency over time; the power spectral density of the signal is calculated, the signal is divided into multiple sub-bands, and the relative power percentage of each sub-band relative to the total power is calculated; the approximate entropy analysis of the signal uses maximum likelihood estimation.

[0012] Furthermore, the machine learning model is an automatic inflammation assessment model. It segments the collected electrophysiological signal data through a sliding window, calculates the time domain, frequency domain, power, and energy characteristics of each window, aggregates the median and interquartile range of each feature value, constructs a multidimensional feature vector, establishes a population health baseline using a leave-one-out cross-validation strategy, calculates the population baseline median and interquartile range, and distinguishes between mild and severe inflammation based on the model.

[0013] On the other hand, the present invention also provides a method for intelligent assessment of enteritis based on a colonic electrophysiological monitoring system using a lantern-shaped flexible array, the method comprising the following steps: Step 1: Perform window data quality control and noise processing on the raw electrophysiological signal data collected by the monitoring system. Mark inappropriate data windows as invalid based on the criteria of all zeros, constant values, or unresolvable data. Step 2: Use machine learning methods to extract the time domain, frequency domain, power, and energy characteristics of the window data; Step 3: Establish a health baseline at the sample population level. Use the electrophysiological signals of the samples in a healthy state as a benchmark, compare them with the electrophysiological signals under different inflammatory conditions, and normalize the corresponding features to form a one-to-one sample feature value. Step 4: Using the severity of the given signal deviation from the population baseline as a construction strategy, an artificial intelligence algorithm is used to form a classification prediction model that distinguishes different degrees of inflammation.

[0014] Furthermore, the specific parameters obtained in step 2 include the mean, absolute value, root mean square, peak-to-peak difference, skewness, zero-crossing rate, three parameters, spectral entropy, total power, absolute power, and relative power, to describe the waveform amplitude, shape, spectrum, and complexity.

[0015] The beneficial effects of this invention are: ① The colonic electrophysiological monitoring system based on a lantern-shaped flexible array proposed in this invention has significant advantages over existing technologies, including flexibility, deformability, and expandability of its sensor array. Combined with an insertable balloon catheter, it can conform to the intestinal pathway to reduce tissue damage and achieve high-quality multi-site sensing measurements, thereby obtaining high-throughput colonic electrophysiological data.

[0016] ② The data processing and analysis method for intestinal electroencephalogram (EEG) signals proposed in this invention, compared with existing methods, constructs a multi-dimensional feature index system covering the time domain, frequency domain, energy domain, and information domain, offering a more objective and comprehensive advantage. It not only decouples inflammation-specific multi-dimensional feature combinations from complex, non-stationary EEG signals but also systematically reveals the multimodal electrophysiological fingerprint corresponding to the colonic state, providing a novel objective quantitative indicator for assessing colitis function.

[0017] ③ The artificial intelligence-based automatic assessment method for the degree of colonic inflammation proposed in this invention innovatively correlates intestinal electrophysiological characteristics with clinical inflammation levels, providing a more efficient and repeatable method for judging disease activity and overcoming the subjective limitations of existing clinical and endoscopic scoring systems. In the future, it can serve as a screening tool to assist in determining the necessity of endoscopic examinations for colitis patients, optimize the allocation of medical resources, and also as an objective and continuous endpoint for evaluating treatment response in drug clinical trials.

[0018] ④ In summary, this invention does not merely collect intestinal electrophysiological signals or improve electrode structures, but systematically integrates the high-throughput intestinal electrophysiological signals acquired by the array electrodes with feature extraction methods and artificial intelligence grading methods, realizing a complete closed-loop solution from intestinal electrophysiological signals to the determination of the degree of inflammation. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of a colonic electrophysiological monitoring system based on a lantern-shaped flexible array, provided in one embodiment of the present invention; Figure 2 This is a block diagram of an electrophysiological signal acquisition module provided in one embodiment of the present invention; Figure 3 This is a physical image of an inflatable lantern-shaped flexible array electrode provided in one embodiment of the present invention; Figure 4This is a schematic diagram of a method for acquiring inflammatory electrophysiological signals according to an embodiment of the present invention; Figure 5 This is a typical waveform diagram of multi-channel intestinal electrophysiology provided in one embodiment of the present invention; Figure 6 This is a flowchart of a data processing and analysis method for colonic electroencephalograms provided in one embodiment of the present invention; Figure 7 This is a graph showing the results of intestinal electrical time-frequency characteristic analysis under different inflammatory conditions, provided by an embodiment of the present invention; Figure 8 This is a flowchart of an automatic assessment method for the degree of colon inflammation based on artificial intelligence, provided in one embodiment of the present invention. Figure 9 This is a confusion matrix result diagram of an automatic assessment model for intestinal inflammation provided in an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0022] like Figure 1 As shown in the diagram, this embodiment provides a schematic diagram of a colonic electrophysiological monitoring system based on a lantern-shaped flexible array, including a lantern-shaped flexible array electrode, a signal acquisition module, and a signal analysis module. The corresponding signal acquisition module system block diagram is shown below. Figure 2 As shown, it consists of a signal conditioning circuit, a signal acquisition card, and a data display and storage host computer. The lantern-shaped flexible array is composed of a flexible multi-channel electrode array and a catheter with a balloon, as shown. Figure 3 As shown. The electrode array in this embodiment is designed and manufactured using a general-purpose flexible electronic circuit. It features 16 circular metal electrode points with a diameter of 0.5 mm, arranged in a 2×8 pattern with a row spacing of 2 mm and a column spacing of 0.91 mm. The total thickness of the array is <100 μm, and its overall size is compact, making it suitable for space-constrained cavities such as the rat colon. The initial planar flexible array is laser-cut to form multiple longitudinal strips, allowing the array to deform along the axial direction. The electrode array is then wound around the outer wall of a 3.5 Fr balloon catheter and fixed at both ends. When the balloon is not inflated, the outer diameter of the device is <3 mm, facilitating non-invasive insertion via the anus. Once the balloon is inflated, the longitudinal strips are stretched to form a lantern-shaped three-dimensional configuration, with the 16 electrodes closely adhering to the colonic mucosa from different angles.

[0023] Based on the aforementioned lantern-shaped flexible array-based electrophysiological monitoring system for colonic inflammation, a sampling frequency of 5 kHz was set to collect resting-state electrophysiological signals in animals with different intestinal inflammation conditions. To obtain electrophysiological signals corresponding to the progression of inflammation, the following methods were employed: Figure 4 The parallel experimental setup is shown. Male SD rats were used as experimental animals, divided into a pathological verification group and a signal acquisition group. All experimental animals were fed dextran sulfate sodium solution to induce enteritis. On days 0, 3, and 7 of enteritis induction, a lantern-shaped sensor array was inserted into the distal colon (approximately 2 cm from the anus) of the rats in the signal acquisition group via the anus. After inflation, the colonic electrical signals were recorded under resting conditions. Each recording lasted 100 seconds and was repeated 3 times. The typical intestinal electrophysiological signals are shown below. Figure 5 As shown in the figure. At the same time point, rats in the pathological group were sacrificed, and colon tissue was collected for H&E staining and serum metabolomics and proteomics analysis to verify the progression of inflammation.

[0024] The collected electrophysiological data are processed and analyzed using appropriate methods, the specific process of which is as follows: Figure 6 As shown. First, the raw data is bandpass filtered to obtain signals between 0.05Hz and 50Hz, then downsampled by a factor of 4 to retain effective information while improving computational efficiency. Feature analysis that can be performed on colonic electrophysiological signals includes one or more of the following: signal-to-noise ratio (SNR), maximum amplitude, time-frequency distribution, power spectral density, and approximate entropy. The SNR analysis uses the following formula: in and These represent the power of the signal and the power of the noise, respectively. and These correspond to the amplitudes of the signal and noise, respectively. By analyzing the signal-to-noise ratio, the signal quality under different health conditions of the colon can be compared. In the maximum amplitude analysis, the continuous data of each channel is divided into multiple non-overlapping time windows. The absolute maximum amplitude of the signal within each time window is extracted and mapped onto a two-dimensional grid corresponding to the electrode layout to generate a dynamic heatmap characterizing the signal intensity distribution of each channel in different time windows. Time-frequency analysis is performed through continuous wavelet transform. Wavelet coefficients are calculated using Morlet wavelets to generate a scale map with time and scale as coordinate axes and the amplitude of the wavelet coefficients representing spectral power, in order to track instantaneous oscillation events and the change of the dominant frequency over time. The power spectral density of the signal is calculated, and the signal within 0~10Hz is divided into multiple sub-bands, including Delta (0.01~0.5 Hz), Theta (0.5~1 Hz), Alpha (1~2 Hz), Beta (2~6 Hz), and Gamma (6~10 Hz). The relative power percentage of each sub-band compared to the total power of the 0~10Hz band is calculated. The approximate entropy analysis of the signal uses maximum likelihood estimation. By correlating and comparing the above characteristic parameters with the validation status of experimental animals, the electrophysiological differences in different physiological states of the colon can be clearly observed, thereby realizing the judgment of colon inflammation based on objective indicators. Figure 7 The results of the time-frequency characteristic analysis of intestinal electrophysiology under different inflammatory conditions are shown in the heat map. As can be seen from the heat map, as the inflammation intensifies, the slow wave part of the electrophysiological signal gradually disappears, and the high-frequency part begins to appear. The frequency distribution shows a trend of shifting towards higher frequencies.

[0025] To further differentiate the severity of intestinal inflammation, artificial intelligence methods were introduced to construct an automatic inflammation severity assessment model. The flowchart of the model construction is as follows: Figure 8 As shown. The specific steps are as follows: For each 100-second record, a sliding window is used with a 4-second window and a 2-second step size to obtain 49 signal segments. After removing invalid windows that are all zero, constant, or unresolvable, the time domain, frequency domain, power, and energy characteristics of each window are calculated in the 1~100Hz frequency band, including absolute mean, root mean square, peak-to-peak value, skewness, zero-crossing rate, Hjorth parameter, spectral entropy, total power, as well as the absolute and relative power of the 1~10, 10~20, 20~50, 1~50, and 50~100 Hz frequency bands, and the logarithmic energy, totaling 23 window-level features. Subsequently, the median and interquartile range of each feature value of all valid windows in each 100-second record are aggregated to form a 46-dimensional feature vector. To eliminate inter-individual variation, a leave-one-out cross-validation strategy was adopted. In each fold, only the day 0 sample from the training set mice was used to establish the population health baseline. The median and interquartile range of the population baseline were calculated, and all sample features were normalized using the following formula: Where Xj represents the value of a single 100 s signal on the j-th feature (i.e., the j-th vector of the aforementioned feature matrix); Medianhealthy_j and IQRhealthy_j are estimates of the Day0 samples from the training set in the feature space, representing the population health baseline of mice; where ε is 1×10 -8 Using normalized 46-dimensional features as input, an XGBoost binary classification model was constructed to distinguish between mild and severe inflammation. The model hyperparameters were determined through grid search: learning rate 0.2, maximum depth 3, subsampling rate 0.8, column sampling rate 0.9, and number of trees 100. A globally optimal cutoff point of 0.42 was also fixed. Model performance was evaluated using the overall AUC, sensitivity, specificity, and confusion matrix of the prediction results from each fold test set. The model was interpreted using the average feature importance across folds and the SHAP value. In this embodiment, the overall AUC reached 0.752, sensitivity 0.672, and specificity 0.734. Feature importance was determined by the interquartile range of relative power from 1 to 50 Hz and the median of relative power from 50 to 100 Hz as key discriminant indicators. The model confusion matrix results are shown below. Figure 9 As shown, the results indicate that the model can effectively classify mild and severe inflammation, demonstrating the model's value.

[0026] The above results demonstrate that the colonic electrophysiological monitoring system and intelligent assessment method for enteritis based on a lantern-shaped flexible array, as described in this invention, can be used to extract high-performance electrophysiological signals within the colon. Furthermore, by combining multidimensional feature extraction and artificial intelligence algorithms, it can achieve intelligent assessment of colonic inflammation. This assessment method, when combined with drug evaluation in the future, holds promise for enabling personalized and precise prognostic management of enteritis patients.

[0027] The above embodiments are used to explain and illustrate the present invention, not to limit it. Any modifications and alterations made to the present invention within the spirit and scope of the claims fall within the protection scope of the present invention. Therefore, the protection scope of this patent should be determined by the appended claims.

Claims

1. A colonic electrophysiological monitoring system based on a lantern-shaped flexible array, characterized in that, include: The flexible electrode array is fabricated from a flexible printed circuit board and contains multiple electrode points. It is laser-cut into several longitudinal strips, which enables the electrode array to be transformed from a planar shape into a three-dimensional configuration. An inflatable balloon catheter, wherein the flexible electrode array is wound and fixed to the outer surface of the balloon, and when the balloon is inflated, the longitudinal strips are radially expanded, forming a lantern shape, so that each electrode point conformally adheres to the colon wall from multiple angles; The signal acquisition module, electrically connected to the electrode array, is used to amplify, filter, and perform analog-to-digital conversion on the multi-channel colonic electrophysiological signals, and to display and store them in real time. The signal analysis module performs multidimensional analysis on the collected electrophysiological data to make a preliminary judgment on the degree of colon health. Then, it combines artificial intelligence with an embedded trained machine learning model to perform more comprehensive feature extraction and health baseline normalization on the collected electrophysiological signals, and output the severity classification results of colon inflammation.

2. The colonic electrophysiological monitoring system based on a lantern-shaped flexible array according to claim 1, characterized in that, The electrode array is designed and manufactured using flexible electronic circuits, and features several rows and columns of circular metal electrode points, making it suitable for space-constrained cavities.

3. The colonic electrophysiological monitoring system based on a lantern-shaped flexible array according to claim 1, characterized in that, When the balloon is not inflated, it is used for non-invasive placement into the colon. After the balloon is inflated, the longitudinal strips are stretched to form a lantern-shaped three-dimensional structure, and each electrode point closely adheres to the colonic mucosa from different angles.

4. The colonic electrophysiological monitoring system based on a lantern-shaped flexible array according to claim 1, characterized in that, The multi-channel colonic electrophysiological signals obtained by the signal acquisition module are preprocessed, including at least one of filtering, baseline drift removal, and downsampling of the acquired raw intestinal electrophysiological signals.

5. A colonic electrophysiological monitoring system based on a lantern-shaped flexible array according to claim 1, characterized in that, The multidimensional analysis includes one or more of the following: time-domain feature analysis, frequency-domain feature analysis, time-frequency domain dynamic feature analysis, and information domain feature analysis; the feature parameters obtained from each analysis process are correlated with the intestinal inflammation status to construct an objective quantitative indicator system for characterizing the presence of colonic inflammation.

6. A colonic electrophysiological monitoring system based on a lantern-shaped flexible array according to claim 5, characterized in that, The time-domain feature analysis involves performing waveform analysis on the preprocessed signal to obtain time-domain feature parameters, which include at least one of signal-to-noise ratio, mean amplitude, peak value, coefficient of variation, and zero-crossing rate. The frequency domain feature analysis involves estimating the power spectral density of the preprocessed signal to obtain frequency domain feature parameters, including at least one of the following: dominant frequency, peak power spectral density, and energy distribution of typical physiological frequency bands. The time-frequency domain dynamic feature analysis involves performing wavelet transform and short-time Fourier transform on the preprocessed signal to obtain time-frequency feature parameters, including at least one of energy concentration, time-frequency kurtosis, frequency range, and subband energy. The information domain feature analysis involves performing complexity and randomness analysis on the preprocessed signal and extracting at least one of spectral entropy, permutation entropy, approximate entropy, and multi-scale entropy.

7. A colonic electrophysiological monitoring system based on a lantern-shaped flexible array according to claim 6, characterized in that, By analyzing the signal-to-noise ratio, the signal quality under different health conditions of the colon is compared; in the maximum amplitude analysis, the continuous data of each channel is divided into multiple non-overlapping time windows, the absolute maximum amplitude of the signal in each time window is extracted, and the absolute maximum amplitude is mapped onto a two-dimensional grid corresponding to the electrode layout to generate a dynamic heat map characterizing the signal intensity distribution of each channel in different time windows. Time-frequency analysis is performed using continuous wavelet transform. Wavelet coefficients are calculated using Morlet wavelets to generate a scale map with time and scale as coordinate axes and the amplitude of the wavelet coefficients representing the spectral power, in order to track instantaneous oscillation events and the change of the dominant frequency over time. The power spectral density of the signal is calculated, the signal is divided into multiple sub-bands, and the relative power percentage of each sub-band relative to the total power is calculated. The approximate entropy analysis of the signal uses maximum likelihood estimation.

8. The colonic electrophysiological monitoring system based on a lantern-shaped flexible array according to claim 1, characterized in that, The machine learning model is an automatic assessment model for the degree of inflammation. It uses a sliding window to segment the collected electrophysiological signal data, calculates the time domain, frequency domain, power and energy characteristics of each window, aggregates the median and interquartile range of each feature value, constructs a multidimensional feature vector, establishes a population health baseline using a leave-one-out cross-validation strategy, calculates the population baseline median and interquartile range, and distinguishes between mild and severe inflammation based on the model.

9. A method for intelligent assessment of enteritis based on the colonic electrophysiological monitoring system of the lantern-shaped flexible array described in claim 1, characterized in that, The method includes the following steps: Step 1: Perform window data quality control and noise processing on the raw electrophysiological signal data collected by the monitoring system. Mark inappropriate data windows as invalid based on the criteria of all zeros, constant values, or unresolvable data. Step 2: Use machine learning methods to extract the time domain, frequency domain, power, and energy characteristics of the window data; Step 3: Establish a health baseline at the sample population level. Use the electrophysiological signals of the samples in a healthy state as a benchmark, compare them with the electrophysiological signals under different inflammatory conditions, and normalize the corresponding features to form a one-to-one sample feature value. Step 4: Using the severity of the given signal deviation from the population baseline as a construction strategy, an artificial intelligence algorithm is used to form a classification prediction model that distinguishes different degrees of inflammation.

10. The intelligent assessment method for enteritis according to claim 1, characterized in that, The specific parameters obtained in step 2 include the single-mean value, absolute value, root mean square, peak-to-peak difference, skewness, zero-crossing rate, three parameters, spectral entropy, total power, absolute power, and relative power, to describe the waveform amplitude, shape, spectrum, and complexity.

Images