Multi-dimensional evaluation method for bacteriostatic effect of moso bamboo leaf volatile oil

Abstract

The invention relates to the technical field of bacteriostatic effect evaluation, in particular to a phyllostachys pubescens leaf volatile oil bacteriostatic effect multi-dimensional evaluation method which comprises the following steps: setting a control environment with phyllostachys pubescens leaf volatile oil and a control environment without phyllostachys pubescens leaf volatile oil in a solid culture medium or a biological membrane culture device, and controlling culture temperature and time parameters to evaluate the bacteriostatic effect of phyllostachys pubescens leaf volatile oil. The bacteria grow to form bacterial colonies or biological membranes, and visible light and near-infrared bands are selected by utilizing multispectral imaging equipment. A moso bamboo leaf volatile oil treatment and control environment is set in a solid culture medium or a biological membrane culture device, and multi-channel image data from visible light to a near-infrared band is captured by using a multispectral imaging device, so that the micromorphological change of a bacterial colony or a biological membrane is captured; the limitation that the traditional method only depends on naked eye observation or single wave band imaging is overcome, and the physiological state differences, such as thallus density, pigment generation and metabolite accumulation, reflected by different spectral information can be revealed.

Description

Technical Field

[0001] The present invention relates to the technical field of antibacterial effect evaluation, and in particular to a multi-dimensional evaluation method for the antibacterial effect of moso bamboo leaf volatile oil. Background Art

[0002] The technical field of antibacterial effect evaluation involves systematic methods and technical systems for measuring and quantifying the ability of various substances or treatment methods to inhibit the growth, reproduction or cause the death of microorganisms.

[0003] Existing antibacterial effect evaluation technologies have some inherent limitations in actual operation. For example, the traditional agar diffusion method and liquid dilution method mainly rely on the macroscopic judgment of the presence or degree of microbial growth, such as the size of the inhibition zone or the minimum inhibitory concentration. Although these methods are classic and simple, they are difficult to provide detailed information about how antibacterial agents affect the morphological structure and physiological metabolic state of microorganisms, and are not deep enough to reveal the mechanism of action of antibacterial agents. When antibacterial agents are not directly lethal but work by interfering with the normal physiological processes of microorganisms or destroying their group behavior structures (such as biofilms), it is often difficult to fully reflect their true effectiveness based solely on growth inhibition indicators. Therefore, improvements are needed. Summary of the Invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a multi-dimensional evaluation method for the antibacterial effect of volatile oil from moso bamboo leaves.

[0005] In order to achieve the above objectives, the present invention adopts the following technical solution, a multi-dimensional evaluation method for the antibacterial effect of bamboo leaf volatile oil, comprising the following steps:

[0006] By setting up control environments with and without moso bamboo leaf volatile oil in a solid culture medium or biofilm culture device, controlling the culture temperature and time parameters, the bacteria are allowed to grow and form colonies or biofilms, and using a multispectral imaging device to select visible light and near-infrared bands, each culture sample is scanned one by one, and the reflectivity or transmittance value of each pixel in different bands is recorded to obtain multi-channel image data;

[0007] Based on the multi-channel image data, a reference point alignment method is used to complete image registration between multispectral channels. Radiation correction is performed based on a standard white plate and dark current data. The colony or biofilm area is segmented as the region of interest. The grayscale co-occurrence matrix of each channel image is calculated for the region of interest, and the contrast, energy and homogeneity values ​​are extracted. At the same time, a multi-scale and multi-directional filtering transformation is performed on the image to extract the response amplitude and phase information. The colony edge sharpness, internal uniformity and surface roughness are quantified to obtain a morphological damage index.

[0008] Preferably, the method further comprises:

[0009] By setting up different concentration gradients of bamboo leaf volatile oil treatment groups and a control group without volatile oil, the target strain was cultured in a constant temperature shaking incubator, gas samples were regularly extracted from the top space of the culture bottle, and volatile organic compounds were adsorbed by a solid phase microextraction probe at a set temperature and time. The probe was inserted into the gas chromatography inlet for thermal desorption. After separation on the chromatographic column, the retention time and mass spectrum of each component were recorded by the mass spectrometer detector. The obtained chromatographic peaks were integrated and baseline corrected, and the compounds were identified with reference to the standard spectral library. The data of each sample were normalized by total ion current, and the mean peak area of ​​each compound in the treatment group and the control group were compared. The compounds with changed peak area were screened to obtain a list of differential volatile organic compounds.

[0010] Based on the differential volatile organic compound list, the total number of volatile organic compounds in the list is counted, the change multiple of the content of each differential volatile organic compound between the treatment group and the control group is calculated and the absolute value sum is obtained, the disease is mapped to a known metabolic pathway, and the change amount of the associated volatile organic compounds on the specific pathway is weighted and summed to obtain a metabolic perturbation index. Based on the compound concentration data in the differential volatile organic compound list, the correlation coefficient between each compound is calculated, and a threshold is set to screen the associated compound pairs to construct a network. The number of connections, clustering coefficient and module division of each compound node in the network are analyzed to obtain the metabolic network topology change assessment results.

[0011] Preferably, the step of acquiring the multi-channel image data is specifically as follows:

[0012] A control environment with and without moso bamboo leaf volatile oil is set up in a solid culture medium or a biofilm culture device, different doses of moso bamboo leaf volatile oil are added to the culture medium to form a treatment gradient, and bacteria are inoculated and grown on the culture medium surface under the set culture conditions to obtain culture sample batches;

[0013] Based on the batch of culture samples, a multispectral imaging device is used to sequentially select multiple preset visible light to near-infrared bands, a unified imaging parameter is set for each sample, and each culture sample is regionally scanned one by one, and the light signal response intensity of each scanning point in each band is synchronously recorded to generate an original spectral signal set;

[0014] Based on the original spectral signal set, the light signal response intensity values ​​recorded at each scanning point in different bands are converted into the brightness values ​​of the corresponding pixels in the image matrix, and the brightness image matrix of each band is organized according to the band information. The multi-band image data of all culture samples are integrated to obtain multi-channel image data.

[0015] Preferably, the steps for obtaining the morphological damage index are as follows:

[0016] Based on the multi-channel image data, feature reference points are selected in each channel image, and pixel positions between each multispectral channel are calibrated and aligned using a geometric transformation method to complete image registration. Then, based on pre-acquired standard reference object image data and device inherent signal data, pixel-by-pixel illumination and sensor response correction is performed on each channel image to obtain a registration-corrected image set;

[0017] Based on the registered and corrected image set, the boundaries of the colony or biofilm are segmented as the region of interest according to the pixel brightness distribution characteristics and the regional edge gradient characteristics. For the image data of each region of interest, the multi-directional grayscale co-occurrence matrix is ​​calculated and the contrast, energy and homogeneity are extracted. At the same time, the image is subjected to multi-scale and multi-directional filtering transformation processing and the amplitude and phase characteristics of the filtering response are extracted to obtain a multi-dimensional morphological feature set.

[0018] Preferably, the step of obtaining the morphological damage index also includes: based on the multidimensional morphological feature set, reducing the dimensionality of the extracted multiple morphological feature value data, selecting principal components or projection combinations that represent variation information or have distinguishing capabilities, and evaluating and combining the screened feature components according to their contribution to obtain the morphological damage index.

[0019] Preferably, the steps of obtaining the differential volatile organic compound list are as follows:

[0020] By setting up different concentration gradients of bamboo leaf volatile oil treatment groups and a control group without volatile oil, the target strain suspension was cultured under set environmental conditions to a predetermined time point. Before the end of the culture cycle, a certain volume of gas sample was extracted and collected from the top closed space of each culture system to obtain a headspace gas concentration sample;

[0021] Based on the headspace gas concentration sample, a solid phase microextraction probe is used to adsorb and enrich volatile organic compounds in the headspace gas concentration sample under target conditions, and then the adsorbed probe is placed in the inlet of a gas chromatography system for thermal desorption, so that the volatile organic compounds enter the chromatographic column with the carrier gas for separation, and are then detected by a mass spectrometer and the retention time and mass spectrum information of each component are recorded to generate volatile compound spectrum data;

[0022] Based on the volatile compound spectrum data, the chromatographic peaks of each sample are identified, integrated, and baseline corrected. The detected chromatographic peaks are identified by referring to the standard spectral library information. The signals of all identified compounds in each sample data are normalized. The difference in the mean of the normalized signals of each identified compound between the treatment group and the control group is compared. Compounds with changed signal responses are screened to obtain a differential volatile organic compound list.

[0023] Preferably, the steps for obtaining the metabolic network topology change evaluation results are specifically as follows:

[0024] Based on the differential volatile organic compound list, the total number of differential volatile organic compounds in the list is counted, and the relative change amplitude of the signal response of each differential volatile organic compound between the moso bamboo leaf volatile oil treatment group and the control group is quantified one by one to obtain a quantitative statistical set of differential substances;

[0025] Based on the quantitative statistical set of differential substances and the differential volatile organic compound list, each volatile organic compound in the differential volatile organic compound list is associated with the corresponding biochemical reaction pathway in the known metabolic pathway database. The change in the amount of all associated differential volatile organic compounds on the key metabolic pathway of the target is evaluated and calculated according to their importance in the pathway to obtain a metabolic perturbation index.

[0026] Preferably, the step of obtaining the metabolic network topology change assessment result further includes:

[0027] Based on the normalized signal data of each compound in the differential volatile organic compound list in different samples, the correlation strength between any two differential volatile organic compound signals is analyzed, and compound pairs with correlation are screened. The compounds are regarded as nodes and the correlations as connecting edges. A volatile organic compound correlation network is constructed, and the connectivity and clustering of the volatile organic compound correlation network are analyzed to obtain the metabolic network topology change assessment results.

[0028] Compared with the prior art, the advantages and positive effects of the present invention are:

[0029] The present invention sets up a bamboo leaf essential oil treatment and control environment in a solid culture medium or biofilm culture device, and uses a multispectral imaging device to capture multi-channel image data from visible light to near-infrared bands, thereby capturing the microscopic morphological changes of the colony or biofilm. This surpasses the limitations of traditional methods that rely solely on naked eye observation or single-band imaging, and can reveal the differences in physiological states such as bacterial density, pigment production, and metabolite accumulation reflected by different spectral information. The multi-channel image data is further subjected to image registration, radiation correction, and region of interest extraction. The texture features are extracted using a grayscale co-occurrence matrix, and the response information is extracted using a multi-scale and multi-directional filtering transform. The colony edge sharpness, internal uniformity, surface roughness, and biofilm density and structural complexity are quantified to construct a morphological damage index. This provides a visual and quantifiable assessment of the "structural" destructive effect of bamboo leaf essential oil, and supplements the deficiency of traditional live bacteria counting methods that cannot reflect morphological changes. At the same time, by culturing the target strains and collecting their headspace volatile organic compounds, qualitative and quantitative analysis was performed to obtain a list of differential volatile organic compounds of bacteria under the action of bamboo leaf volatile oil. This approach of indirectly evaluating the antibacterial effect at the metabolite level can reflect the impact of the drug on the overall metabolic network of bacteria. Subsequently, based on the differential volatile organic compound list, the total number of differences was counted, the sum of the content change multiples was calculated, and it was mapped to the metabolic pathway for weighted summation to obtain the metabolic perturbation index. At the same time, the association network between the differential VOCs was constructed and the changes in their topological parameters were analyzed to obtain the evaluation results of the metabolic network topology changes. This not only provides indirect evidence of the antibacterial effect from the metabolomics level, but also helps to identify potential targets and biomarkers. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 Schematic diagram of the steps of the present invention. DETAILED DESCRIPTION

[0031] 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.

[0032] See also Figure 1 The present invention provides a technical solution, a multi-dimensional evaluation method for the antibacterial effect of volatile oil from moso bamboo leaves, comprising the following steps:

[0033] By setting up control environments with and without moso bamboo leaf volatile oil in a solid culture medium or biofilm culture device, controlling the culture temperature and time parameters, the bacteria are allowed to grow and form colonies or biofilms, and using a multispectral imaging device to select visible light and near-infrared bands, each culture sample is scanned one by one, and the reflectivity or transmittance value of each pixel in different bands is recorded to obtain multi-channel image data;

[0034] Based on multi-channel image data, a fiducial point alignment method is used to complete image registration between multispectral channels. Radiation correction is performed based on a standard white plate and dark current data. The colony or biofilm area is segmented as a region of interest. The gray-level co-occurrence matrix of each channel image is calculated for the region of interest, and contrast, energy, and homogeneity values ​​are extracted. Multi-scale and multi-directional filtering transformations are also performed on the image to extract response amplitude and phase information. The colony edge sharpness, internal uniformity, and surface roughness are quantified to obtain a morphological damage index.

[0035] By setting up different concentration gradients of bamboo leaf volatile oil treatment groups and a control group without volatile oil, the target strain was cultured in a constant temperature shaking incubator, gas samples were regularly extracted from the top space of the culture bottle, and volatile organic compounds were adsorbed by a solid phase microextraction probe at a set temperature and time. The probe was inserted into the gas chromatography inlet for thermal desorption. After separation on the chromatographic column, the retention time and mass spectrum of each component were recorded by the mass spectrometer detector. The obtained chromatographic peaks were integrated and baseline corrected, and the compounds were identified with reference to the standard spectral library. The data of each sample were normalized by total ion current, and the mean peak area of ​​each compound in the treatment group and the control group were compared. The compounds with changed peak area were screened to obtain a list of differential volatile organic compounds.

[0036] Based on the list of differential volatile organic compounds, the total number of volatile organic compounds in the list was counted, the change multiples of the content of each differential volatile organic compound between the treatment group and the control group were calculated and the absolute value sum was obtained. The disease was mapped to the known metabolic pathway, and the change amount of the associated volatile organic compounds on the specific pathway was weighted and summed to obtain the metabolic perturbation index. Based on the compound concentration data in the list of differential volatile organic compounds, the correlation coefficient between each compound was calculated, and the threshold was set to screen the associated compound pairs to construct a network. The number of connections, clustering coefficient and module division of each compound node in the network were analyzed to obtain the evaluation results of the metabolic network topology changes.

[0037] The specific steps for acquiring multi-channel image data are as follows:

[0038] A control environment with and without moso bamboo leaf volatile oil is set up in a solid culture medium or a biofilm culture device, different doses of moso bamboo leaf volatile oil are added to the culture medium to form a treatment gradient, and bacteria are inoculated and grown on the culture medium surface under the set culture conditions to obtain culture sample batches;

[0039] Based on the batch of culture samples, a multispectral imaging device is used to sequentially select multiple preset visible light to near-infrared bands. Uniform imaging parameters are set for each sample. Each culture sample is scanned one by one, and the light signal response intensity of each scanning point in each band is synchronously recorded to generate a set of original spectral signals.

[0040] Based on the original spectral signal set, the light signal response intensity values ​​recorded at each scanning point in different bands are converted into the brightness values ​​of the corresponding pixels in the image matrix. The brightness image matrix of each band is organized according to the band information, and the multi-band image data of all culture samples are integrated to obtain multi-channel image data.

[0041] Specifically, by setting a control environment with and without moso bamboo leaf volatile oil in a solid culture medium or a biofilm culture device, the specific operation is to first prepare a series of concentrations of moso bamboo leaf volatile oil. For example, if preliminary experiments show that the minimum inhibitory concentration (MIC) of moso bamboo leaf volatile oil is about 0.5% (v / v), then a concentration gradient can be set, such as 0% (control group, without volatile oil), 0.1%, 0.25%, 0.5%, 0.75%, 1.0% (v / v). These concentrations of volatile oil are mixed with an appropriate amount of Tween 80 (for example, a final concentration of After mixing, add the mixture to a sterilized liquid culture medium (such as tryptone soy broth, TSB) or solid culture medium (such as tryptone soy agar, TSA) cooled to about 45 to 50 degrees Celsius and mix well. For the control group, add equal amounts of Tween 80 and culture medium. Then, prepare a standardized bacterial suspension of the target strain. For example, culture the strain in TSB to the logarithmic growth phase and adjust the turbidity of the bacterial suspension with sterile saline or phosphate buffered saline (PBS) to reach 0.5 McFarland standard, which roughly corresponds to 1.5×10 8 CFU / mL bacterial concentration, then, the standardized bacterial suspension is inoculated into a culture medium containing different concentrations of volatile oils at a specific ratio (for example, 1:100). For solid culture medium, 100 microliters of bacterial suspension can be taken for surface coating to ensure uniform inoculation. Subsequently, all cultures of the treatment group and the control group (such as culture dishes or culture plates) are placed in a constant temperature incubator, and the culture temperature is set to the optimal growth temperature of the specific bacteria, such as Escherichia coli or Staphylococcus aureus is usually 37°C. The culture time is set according to the experimental requirements. Generally, colony formation is observed for 18 to 24 hours, and biofilm formation requires 24 to 72 hours. During the culture period, ensure that the environmental humidity is appropriate, for example, by placing a water tray at the bottom of the incubator to maintain a relative humidity of about 85%. When the culture reaches the predetermined time point, the complete experimental group containing all different treatment concentrations and controls is regarded as a batch to obtain a culture sample batch.

[0042] Based on the culture sample batch, the multispectral imaging equipment should be preheated and calibrated during implementation. Multiple preset visible to near-infrared bands should cover the spectral region that can reflect the growth status and potential damage of bacteria. For example, blue light (such as 450nm±10nm, used to observe pigment changes or fluorescent markers), green light (such as 550nm±10nm, general imaging), and red light (such as 660nm±10nm, reflecting colony thickness or certain metabolites) should be selected for the visible light part. Bands such as 750nm±10nm, 850nm±10nm, and 950nm±10nm should be selected for the near-infrared part. These bands are more sensitive to water content, cell structure integrity, etc. For each sample in each culture sample batch (including different concentration treatment groups and control groups), uniform imaging parameters are set, including light source intensity (for example, set to 80% of the maximum output to ensure sufficient signal-to-noise ratio). , integration time or exposure time (for example, determined according to preliminary experiments to ensure that the brightest band and the most concentrated colony area are not saturated, such as 50 milliseconds), camera gain (for example, set to a medium level, such as 12 dB), and object distance and focal length (remain fixed to ensure consistent spatial resolution). Then, a scanning device of a multispectral imaging device is used, such as a push-broom or area array camera with a filter wheel or a tunable filter, to perform regional scanning on each culture sample (well of a culture dish or microplate) one by one. The scanning method can be to move the sample on an XY platform or to move the scanning probe to ensure that the entire colony or biofilm growth area is covered. During the scanning process, the device synchronously records the light signal response intensity of each scanning point (corresponding to a pixel or a superpixel in the future image) in each selected band. This intensity is usually expressed as the raw digital quantity (DN value) output by the detector. The set of these raw digital quantities constitutes the raw spectral signal set.

[0043] Based on the original spectral signal set, the light signal response intensity values ​​of each scanning point recorded in different bands are processed. First, the original digital quantity (DN value) recorded by each scanning point in a specific band is converted into the brightness value of the corresponding pixel in the image matrix through linear or nonlinear mapping (usually linear mapping, such as directly using the DN value or scaling it to the range of 0-255 or 0-65535). For example, if the DN value range is 0-4095 (12-bit ADC), this value can be used directly, or it can be linearly stretched to an 8-bit (0-255) or 16-bit (0-65535) grayscale range to form a single-channel grayscale image under this band. For all preset bands of each sample, This conversion is performed on all samples to obtain a series of single-channel brightness image matrices corresponding to different bands. Then, these brightness image matrices of different bands obtained for the same culture sample are organized and stacked according to the band information (such as wavelength from small to large or in a specific order) to form a multi-band image data cube, where the x and y dimensions of the image represent the spatial position, and the z dimension represents the different spectral bands. Finally, the multi-band image data collected from all culture samples (including all samples of different treatment gradients and control groups) are integrated, that is, the multi-band image data cubes of each sample are brought together to form a data set containing the imaging information of all samples in the entire experimental batch, and multi-channel image data are obtained.

[0044] The specific steps for obtaining the morphological damage index are as follows:

[0045] Based on multi-channel image data, feature reference points are selected in each channel image, and the pixel positions between each multispectral channel are calibrated and aligned using geometric transformation to complete image registration. Then, based on the pre-acquired standard reference object image data and the device's inherent signal data, the illumination and sensor response corrections are performed pixel by pixel on each channel image to obtain a registered and corrected image set.

[0046] Based on the registered and corrected image set, the boundaries of the colony or biofilm are segmented as regions of interest according to the pixel brightness distribution characteristics and regional edge gradient characteristics. For the image data of each region of interest, the multi-directional gray-level co-occurrence matrix is ​​calculated and the contrast, energy and homogeneity are extracted. At the same time, the image is subjected to multi-scale and multi-directional filtering transformation and the amplitude and phase characteristics of the filter response are extracted to obtain a multi-dimensional morphological feature set.

[0047] Based on the multidimensional morphological feature set, the dimensionality reduction of the extracted multiple morphological feature value data is performed, and the principal components or projection combinations representing variation information or with distinguishing ability are selected. The screened feature components are evaluated and combined according to their contribution to obtain the morphological damage index.

[0048] Specifically, based on multi-channel image data, feature reference points are first selected in each channel image. These reference points can be marking points on the edge of the culture dish, fixed defects in the culture medium, or reference objects with clear geometric shapes placed in the imaging field of view (such as tiny checkerboard corners). If there are no obvious physical reference points, the characteristics of the image itself can also be used, such as clear edge points of the colony or significant internal texture points. These same-name points are manually or automatically identified in multiple channels, and then a geometric transformation method is used, such as selecting an affine transformation or a polynomial transformation model, and fitting the coordinate relationship of these same-name reference points by the least squares method. The transformation parameters are calculated and applied to calibrate and align the pixel positions between each multispectral channel to complete the image registration, ensuring that the same spatial position in images of different bands can accurately correspond. Then, radiation correction is performed based on the pre-acquired standard reference image data and the inherent signal data of the device. The standard reference image data usually refers to a standard white plate (such as a barium sulfate pressed plate or a commercial white plate) photographed under the same imaging conditions. A whiteboard with a known and close to 100% Lambertian reflectance within the target band) image and a dark field image (taken with the lens cap on or in a dark environment to record the device's own dark current noise and fixed pattern noise) are used to perform pixel-by-pixel illumination and sensor response correction on each channel image. The specific calculation method can be the relative reflectance R pixel =(DN sample,pixel -DN dark,pixel ) / (DN white,pixel -DN dark,pixel ), where DN sample,pixel is the original brightness value of the corresponding pixel of the sample image, DN dark,pixel is the brightness value of the corresponding pixel in the dark field image, DN white,pixel is the brightness value of the corresponding pixel in the whiteboard image. After this step, the registration and correction image set is obtained.

[0049] Based on the registered and corrected image set, the boundary of the colony or biofilm is first segmented as the region of interest (ROI) according to the pixel brightness distribution characteristics and the regional edge gradient characteristics. For example, a threshold-based segmentation method such as the Otsu method can be used to automatically determine the global threshold, or for images with unclear contrast, an edge-based segmentation method such as the Canny edge detection algorithm or the Sobel operator can be used to calculate the gradient amplitude, and then the edge is closed to form a complete region through methods such as region growing or level set. The setting of the segmentation threshold, for example, for the Otsu method, it will automatically find a threshold to maximize the inter-class variance; for the gradient-based method, a lower limit of the gradient amplitude can be set, such as the position where the gradient amplitude is greater than 1.5 times the average value of the gradient amplitude of the entire image plus one standard deviation is considered to be an edge candidate point. For the image data extracted from each region of interest, the multi-directional gray-level co-occurrence matrix (GLCM) is calculated, for example, in the four directions of 0°, 45°, 90°, and 135° with a pixel spacing of 1 (or other appropriate distances such as 3 or 5 pixels are selected according to the scale of the colony texture) to construct GLCM, and texture features are extracted from these GLCMs, including contrast (measurement of the intensity of local grayscale changes), energy (also known as angular second moment, measurement of image texture uniformity), and homogeneity (also known as inverse disparity, measurement of local texture similarity). For each feature, the values ​​calculated in the four directions can be averaged or the maximum value can be taken as the representative value of the feature. At the same time, the image is subjected to multi-scale and multi-directional filtering transformation processing, such as using a Gabor filter bank, which contains Gabor filters with different center frequencies (scales) and different directions. The ROI image is convolved with each filter in the Gabor filter bank to extract the amplitude and phase characteristics of the filter response, where the amplitude response represents the intensity of the texture at a specific scale and direction, and the phase response is related to the symmetry and structure of the texture. All extracted GLCM texture features and Gabor filter features are combined to obtain a multi-dimensional morphological feature set.

[0050] Based on the multidimensional morphological feature set, the dimensionality reduction processing of the extracted multiple morphological feature value data is first performed, such as using the principal component analysis (PCA) method to calculate the feature covariance matrix and perform eigenvalue decomposition on it, and select the principal components whose cumulative contribution rate reaches the preset threshold (for example, select the principal components that can explain more than 95% of the total variance of the original data), or use supervised dimensionality reduction methods such as linear discriminant analysis (LDA). If different treatment groups are known (such as control group, low concentration group, high concentration group), the projection direction combination that can maximize the inter-class divergence while minimizing the intra-class divergence is selected. These principal components or projection combinations represent the main variation information in the original feature space or feature subcomponents with the ability to distinguish different treatment effects. Then, the selected characteristic components, i.e., the selected principal component scores or LDA projection values, are evaluated and combined according to their contribution to distinguishing different treatment groups or reflecting the degree of injury. The contribution can be evaluated by analyzing the statistical significance of each principal component in distinguishing the control group from the treatment group (such as analyzing the mean difference through t-test or ANOVA) or its feature importance score in the classification model (such as support vector machine or decision tree). For example, if the characteristic values ​​(or proportion of explained variance) of the first three principal components PC1, PC2, and PC3 are λ1, λ2, and λ3 respectively, and their F values ​​for distinguishing the treatment group from the control group are F1, F2, and F3 respectively, then a weighting scheme can be set, and the weight w i Proportional to λ i ×F i And normalized, for example, if λ1×F1=10, λ2×F2=5, λ3×F3=2, and the total is 17, then the weights are w1=10 / 17, w2=5 / 17, w3=2 / 17, and the morphological damage index MDI can be calculated as in is the principal component score after standardization (e.g., subtracting the mean of the control group and dividing by the standard deviation of the control group) to obtain the morphological damage index.

[0051] The specific steps for obtaining the differential volatile organic compound list are as follows:

[0052] By setting up different concentration gradients of bamboo leaf volatile oil treatment groups and a control group without volatile oil, the target strain suspension was cultured under set environmental conditions to a predetermined time point. Before the end of the culture cycle, a certain volume of gas sample was extracted and collected from the top closed space of each culture system to obtain a headspace gas concentration sample;

[0053] Based on the headspace gas concentration sample, a solid phase microextraction probe is used to adsorb and enrich the volatile organic compounds in the headspace gas concentration sample under target conditions. The adsorbed probe is then placed in the inlet of the gas chromatography system for thermal desorption. The volatile organic compounds enter the chromatographic column with the carrier gas and are separated. After that, the mass spectrometer detects and records the retention time and mass spectrum information of each component to generate volatile compound spectrum data.

[0054] Based on the volatile compound spectrum data, the chromatographic peaks of each sample were identified, integrated and baseline corrected. The compound structures of the detected chromatographic peaks were identified with reference to the standard spectral library information. The signals of all identified compounds in each sample data were normalized. The differences in the mean values ​​of the normalized signals of each identified compound in the treatment group and the control group were compared. Compounds with changed signal responses were screened to obtain a list of differential volatile organic compounds.

[0055] Specifically, by setting up a bamboo leaf volatile oil treatment group with different concentration gradients and a control group without volatile oil, for example, based on the minimum inhibitory concentration (MIC value, for example, 0.2% v / v) of bamboo leaf volatile oil against the target strain (such as Staphylococcus aureus) determined in the preliminary experiment, a bamboo leaf volatile oil concentration gradient including 0% (control group, containing only an equal amount of solvent such as 0.1% Tween 80 as an emulsifier), 0.05% (0.25xMIC), 0.1% (0.5xMIC), 0.2% (1xMIC), and 0.4% (2xMIC) v / v is established, and these concentrations of volatile oil are added to a liquid culture medium (such as Luria-Bertani broth) that has been sterilized and cooled to about 45°C, and mixed thoroughly. Subsequently, the target strain suspension cultured to the logarithmic growth phase (adjusted to an OD600 of about 0.1, equivalent to about 10 7 A 1% inoculum of 5% CFU / mL (CFU / mL) is added to the culture flasks of the above-mentioned concentration treatment groups and the control group (for example, using headspace flasks, such as 20 mL size, filled with 10 mL of culture medium, ensuring sufficient headspace). All culture flasks are cultured under set environmental conditions, such as in a constant temperature shaker at 37°C and 180 rpm, and the target strain suspension is cultured to a predetermined time point. The predetermined time point is determined based on the bacterial growth curve, for example, the time point when the bacteria reaches the late logarithmic phase but has not yet entered the stationary phase, such as 12 hours of culture. Approximately 10 minutes before the end of the culture period (e.g., 12 hours), the culture system is ensured to be airtight (e.g., the headspace flask is sealed with an aluminum cap with a polytetrafluoroethylene / silicone septum). A certain volume of gas sample, for example, 5 mL of headspace gas, is slowly drawn and collected from the sealed space at the top of each culture system using a clean, airtight syringe (e.g., 10 mL size) preheated to, for example, 50°C to obtain a headspace gas concentrate sample.

[0056] Based on the headspace gas concentration sample, a solid phase microextraction (SPME) probe is immediately used for adsorption enrichment of volatile organic compounds (VOCs). An SPME fiber head with good adsorption performance for the metabolites of the target strain is selected, such as a 50 / 30 μm coated fiber. The SPME probe is inserted into the septum of the headspace bottle containing the headspace gas concentration sample to expose the fiber head to the headspace. Adsorption is performed under target conditions. The target conditions are set as follows: the sample bottle is maintained in a water bath for 30 minutes at, for example, 60°C, while magnetic stirring is performed (for example, 250 rpm) to promote the release of VOCs from the liquid phase to the gas phase and adsorption by the fiber. After the adsorption enrichment is completed, the adsorbed SPME probe is immediately pulled out of the sample bottle and quickly inserted into the inlet of the gas chromatography (GC) system for thermal decomposition. The inlet temperature is set to, for example, 250°C, and the decomposition time is, for example, 3 minutes. The sample is injected in the non-split mode for 1 minute and then switched to the split mode (split ratio, for example, 20 : 1), volatile organic compounds are made to enter the chromatographic column with high-purity helium carrier gas (flow rate is for example 1.0 mL / min) and be separated, the chromatographic column is selected for example HP-5MS capillary column (30 m×0.25 mm×0.25 μm), the column temperature program is set to: initial temperature 40 ℃ is kept for 2 minutes, then the temperature is increased to 150 ℃ at a rate of 5 ℃ / min, then the temperature is increased to 280 ℃ at a rate of 10 ℃ / min, and is kept at 280 ℃ for 5 minutes, each component after separation enters mass spectrometry (MS) detector in sequence, the mass spectrometry detector adopts electron bombardment ion source (EI source), electron energy is 70 eV, ion source temperature is 230 ℃, quadrupole temperature is 150 ℃, transmission line temperature is 280 ℃, carry out full scan mode detection, mass scan range is m / z 35-550, the system automatically records the retention time (RetentionTime, RT) when each component effluents and the corresponding mass spectrum information, generates volatile compound spectrogram data.

[0057] Based on the volatile matter spectrum data, the total ion current chromatogram (TIC) of each sample is first processed using the chemical workstation software (such as Agilent Mass Hunter Qualitative Analysis) supporting the gas chromatography-mass spectrometry instrument, and the chromatographic peaks are automatically or manually identified. The signal-to-noise ratio threshold is set to 5 for peak detection, the identified chromatographic peaks are integrated, the peak area or peak height is calculated, and a baseline correction is performed, such as using the automatic baseline correction algorithm provided by the instrument software. Subsequently, the mass spectrum of each chromatographic peak is compared with the standard mass spectral library, and the matching threshold is set to 800 (full score 1000). The experimental retention time is then compared with the retention index (RI, calculated by running the normal alkane standard) reported in the standard substance or literature. (If the RI deviation is less than 20 units), the detected chromatographic peaks are subjected to compound structure identification, and all identified compound signals in each sample data are normalized (usually using peak area). The normalization method can be total peak area normalization, that is, the peak area of ​​each compound is divided by the total peak area of ​​all identified compounds in the sample, and then multiplied by a constant (such as 10000). The difference in the mean of the normalized signal of each identified compound in the treatment group (each concentration of bamboo leaf volatile oil) and the control group without volatile oil is compared, and the compounds with significant changes in signal response are screened. The judgment standard of significant change is set as follows: the mean of the normalized signal of each compound between the treatment group and the control group is compared by two-sample t-test or one-way analysis of variance (ANOVA), and the p value is less than the preset significance level (for example, p threshold =0.05), and the absolute value of the fold change (FoldChange, FC) of the average signal intensity of the compound between the treatment group and the control group is greater than or equal to a preset threshold (e.g., |FC| threshold =1.5), where FC is calculated as the average signal of the treatment group / the average signal of the control group. For example, if the average normalized signal of a compound in the control group is 100 and the average normalized signal in a treatment group is 160, then FC=1.6. If p < 0.05 and 1.6 ≥ 1.5, the compound is determined to have a changed signal response, and a differential volatile organic compound list is obtained.

[0058] The specific steps for obtaining the metabolic network topology change evaluation results are as follows:

[0059] Based on the differential volatile organic compound list, the total number of differential volatile organic compounds in the list was counted, and the relative change amplitude of the signal response of each differential volatile organic compound between the bamboo leaf volatile oil treatment group and the control group was quantified one by one to obtain the quantitative statistical set of differential substances;

[0060] Based on the quantitative statistical set of differential substances and the differential volatile organic compound list, each volatile organic compound in the differential volatile organic compound list is associated with the corresponding biochemical reaction pathway in the known metabolic pathway database. The changes in all associated differential volatile organic compounds on the key metabolic pathway of the target are evaluated and calculated based on their importance in the pathway to obtain the metabolic perturbation index;

[0061] Based on the normalized signal data of each compound in the differential volatile organic compound list in different samples, the correlation strength between any two differential volatile organic compound signals was analyzed, and the compound pairs with correlation were screened. The compounds were regarded as nodes and the correlation as connecting edges to construct a volatile organic compound association network. The connectivity and clustering of the volatile organic compound association network were analyzed to obtain the evaluation results of metabolic network topology changes.

[0062] Specifically, based on the differential volatile organic compound list, the total number of differential volatile organic compounds contained in the list is first directly counted and recorded. Then, for each differential volatile organic compound in the list, the relative change amplitude of the normalized signal response between the bamboo leaf volatile oil treatment group and the volatile oil-free control group is quantified one by one. This relative change amplitude can be expressed by various indicators, for example, the fold change (FC) of each differential volatile organic compound at each treatment concentration is calculated, and the calculation formula is: MeanNormalizedSignal compound_i,treatment_concentration_j is the average normalized signal value of compound i under the treatment of volatile oil concentration j, MeanNormalizedSignal compound_i,control The average normalized signal value of compound i in the control group is recorded, and the corresponding statistically significant p-value (for example, from a t-test or ANOVA analysis result) is recorded. The logarithmic-transformed fold change, such as log2(FC), can also be calculated. This transformation helps to make the up-regulated and down-regulated changes more symmetrical in value and stabilize the variance. The name or unique identifier of each differential volatile organic compound, its FC value (or log2(FC) value) in each treatment group relative to the control group, and the corresponding p-value are organized into a structured data table. This collection of all differential volatile organic compounds and their quantitative change information is the differential substance quantitative statistical set.

[0063] Based on the quantitative statistical set of differential substances and the differential volatile organic compound list, each volatile organic compound in the differential volatile organic compound list is first matched with a known public metabolic pathway database (such as the KEGGPATHWAY database and the MetaCyc database) using its accurate compound name, CAS number or KEGGID identifier to identify the biochemical reaction pathways in which these differential volatile organic compounds participate. Next, the key metabolic pathways of the target are screened or defined. The selection of these key pathways can be based on the research purpose, such as selecting pathways related to bacterial core metabolism (such as glycolysis, tricarboxylic acid cycle, amino acid metabolism). ), cell wall synthesis, virulence factor production and other related pathways, or by performing pathway enrichment analysis on differential volatile organic compounds (for example, using tools such as MetaboAnalyst, based on hypergeometric test or Fisher's exact test, pathways with p-values ​​less than 0.05 are considered significantly enriched) to determine the significantly affected pathways, for each selected target key metabolic pathway, all associated differential volatile organic compounds are extracted from the quantitative statistical set of differential substances (for example, using log2 (FC) values), and weighted according to the importance of the volatile organic compound in the pathway, and the importance weight (w i ) can be set based on its position in the pathway (for example, the weight of the starting substrate of the pathway is set to 0.8, the weight of the intermediate product regulated by the key enzyme is set to 1.2, and the weight of the terminal product of the pathway is set to 1.0), degree centrality (if the compound is the intersection of multiple reactions, the weight is higher, such as 1.5), or its known contribution to the overall flow of the pathway. For example, if the log2(FC) of a key intermediate A in a pathway P is 1.5, its weight w A According to the central position is set to 1.2, the log2(FC) of the other end product B is -0.8, and its weight w B If set to 1.0, the metabolic disturbance index (MDI) is calculated by summing these weighted changes, specifically MDI pathway =∑ i∈pathway w i ×|log2(FC i )|, where FC i It is the multiple change of the i-th differential volatile organic compound on the pathway. The absolute value is used to reflect the degree of disturbance. The metabolic disturbance index of the specific pathway is calculated. If there are multiple key pathways, their metabolic disturbance indices are calculated separately.

[0064] Based on the normalized signal data of each compound in the differential volatile organic compound list in different samples (including all treatment groups and control groups), the correlation coefficient between the normalized signal intensities of any two compounds in the differential volatile organic compound list in all samples is first calculated, for example, using the Pearson correlation coefficient (PCC). If the data does not conform to a normal distribution, the Spearman rank correlation coefficient is used. The calculated correlation coefficients of all compound pairs and their corresponding p-values ​​are screened, and a correlation threshold is set to determine compound pairs with significant correlation. The threshold setting standard is: the absolute value of the Pearson correlation coefficient (|r|) is greater than a preset value (for example, |r| threshold = 0.7, this value is based on experience or literature, and is selected to capture strong correlations and avoid overly dense networks, such as selecting a threshold where the absolute value of all correlation coefficients is in the top 5%), and the corresponding p-value is less than the statistical significance level (for example, p threshold =0.05, multiple testing correction can be considered, such as when the q value after FDR correction is less than 0.05). Compound pairs that meet these conditions are considered to be associated. Then, each compound in the differential volatile organic compound list is regarded as a node in the network, and the association between the compound pairs with significant association screened out is regarded as the edge connecting the two nodes. The weight of the edge can be set to the absolute value of the correlation coefficient. In this way, a volatile organic compound association network is constructed. Networks for the control group and each treatment group can be constructed separately, or an overall network containing all samples can be constructed and the node attributes of different groups can be compared in subsequent analysis. Then, the topological parameters of the constructed volatile organic compound association network are analyzed using network analysis software. It includes calculating the number of connections (degree) of each compound node in the network, that is, the number of edges directly connected to the node, analyzing the network's clustering coefficient (Clustering Coefficient), which measures the tightness of the connection between node neighbors, such as calculating the global average clustering coefficient, and dividing the network into modules, such as using the Louvain algorithm or the FastGreedy algorithm to identify tightly connected compound subsets (modules) in the network, comparing the changes in these network topology parameters (such as average degree, average clustering coefficient, number and composition of modules) under different treatment conditions (for example, the control group and the bamboo leaf volatile oil treatment group), and obtaining the evaluation results of metabolic network topology changes.

[0065] 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 multi-dimensional evaluation method for the antibacterial effect of bamboo leaf volatile oil, characterized in that: The following steps are involved: By setting up control environments with and without moso bamboo leaf volatile oil in a solid culture medium or biofilm culture device, controlling the culture temperature and time parameters, the bacteria are allowed to grow and form colonies or biofilms, and using a multispectral imaging device to select visible light and near-infrared bands, each culture sample is scanned one by one, and the reflectivity or transmittance value of each pixel in different bands is recorded to obtain multi-channel image data; Based on the multi-channel image data, a reference point alignment method is used to complete image registration between multispectral channels. Radiation correction is performed based on a standard white plate and dark current data. The colony or biofilm area is segmented as the region of interest. The grayscale co-occurrence matrix of each channel image is calculated for the region of interest, and the contrast, energy and homogeneity values ​​are extracted. At the same time, a multi-scale and multi-directional filtering transformation is performed on the image to extract the response amplitude and phase information. The colony edge sharpness, internal uniformity and surface roughness are quantified to obtain a morphological damage index.

2. The multidimensional evaluation method for the antibacterial effect of bamboo leaf volatile oil according to claim 1, wherein The method further comprises: By setting up different concentration gradients of bamboo leaf volatile oil treatment groups and a control group without volatile oil, the target strain was cultured in a constant temperature shaking incubator, gas samples were regularly extracted from the top space of the culture bottle, and volatile organic compounds were adsorbed by a solid phase microextraction probe at a set temperature and time. The probe was inserted into the gas chromatography inlet for thermal desorption. After separation on the chromatographic column, the retention time and mass spectrum of each component were recorded by the mass spectrometer detector. The obtained chromatographic peaks were integrated and baseline corrected, and the compounds were identified with reference to the standard spectral library. The data of each sample were normalized by total ion current, and the mean peak area of ​​each compound in the treatment group and the control group were compared. The compounds with changed peak area were screened to obtain a list of differential volatile organic compounds. Based on the differential volatile organic compound list, the total number of volatile organic compounds in the list is counted, the change multiple of the content of each differential volatile organic compound between the treatment group and the control group is calculated and the absolute value sum is obtained, the disease is mapped to a known metabolic pathway, and the change amount of the associated volatile organic compounds on the specific pathway is weighted and summed to obtain a metabolic perturbation index. Based on the compound concentration data in the differential volatile organic compound list, the correlation coefficient between each compound is calculated, and a threshold is set to screen the associated compound pairs to construct a network. The number of connections, clustering coefficient and module division of each compound node in the network are analyzed to obtain the metabolic network topology change assessment results.

3. The multidimensional evaluation method for the antibacterial effect of bamboo leaf volatile oil according to claim 1, wherein The steps of acquiring the multi-channel image data are specifically as follows: A control environment with and without moso bamboo leaf volatile oil is set up in a solid culture medium or a biofilm culture device, different doses of moso bamboo leaf volatile oil are added to the culture medium to form a treatment gradient, and bacteria are inoculated and grown on the culture medium surface under the set culture conditions to obtain culture sample batches; Based on the batch of culture samples, a multispectral imaging device is used to sequentially select multiple preset visible light to near-infrared bands, a unified imaging parameter is set for each sample, and each culture sample is regionally scanned one by one, and the light signal response intensity of each scanning point in each band is synchronously recorded to generate an original spectral signal set; Based on the original spectral signal set, the light signal response intensity values ​​recorded at each scanning point in different bands are converted into the brightness values ​​of the corresponding pixels in the image matrix, and the brightness image matrix of each band is organized according to the band information. The multi-band image data of all culture samples are integrated to obtain multi-channel image data.

4. The multidimensional evaluation method for the antibacterial effect of bamboo leaf volatile oil according to claim 1, wherein The steps for obtaining the morphological damage index are specifically as follows: Based on the multi-channel image data, feature reference points are selected in each channel image, and pixel positions between each multispectral channel are calibrated and aligned using a geometric transformation method to complete image registration. Then, based on pre-acquired standard reference object image data and device inherent signal data, pixel-by-pixel illumination and sensor response correction is performed on each channel image to obtain a registration-corrected image set; Based on the registered and corrected image set, the boundaries of the colony or biofilm are segmented as the region of interest according to the pixel brightness distribution characteristics and the regional edge gradient characteristics. For the image data of each region of interest, the multi-directional grayscale co-occurrence matrix is ​​calculated and the contrast, energy and homogeneity are extracted. At the same time, the image is subjected to multi-scale and multi-directional filtering transformation processing and the amplitude and phase characteristics of the filtering response are extracted to obtain a multi-dimensional morphological feature set.

5. The multidimensional evaluation method for the antibacterial effect of bamboo leaf volatile oil according to claim 4, wherein The step of obtaining the morphological damage index also includes: based on the multidimensional morphological feature set, reducing the dimensionality of the extracted multiple morphological feature value data, selecting principal components or projection combinations that represent variation information or have distinguishing capabilities, and evaluating and combining the screened feature components according to their contribution to obtain the morphological damage index.

6. The multidimensional evaluation method for the antibacterial effect of the bamboo leaf volatile oil according to claim 2, wherein The steps for obtaining the differential volatile organic compound list are specifically as follows: By setting up different concentration gradients of bamboo leaf volatile oil treatment groups and a control group without volatile oil, the target strain suspension was cultured under set environmental conditions to a predetermined time point. Before the end of the culture cycle, a certain volume of gas sample was extracted and collected from the top closed space of each culture system to obtain a headspace gas concentration sample; Based on the headspace gas concentration sample, a solid phase microextraction probe is used to adsorb and enrich volatile organic compounds in the headspace gas concentration sample under target conditions, and then the adsorbed probe is placed in the inlet of a gas chromatography system for thermal desorption, so that the volatile organic compounds enter the chromatographic column with the carrier gas for separation, and are then detected by a mass spectrometer and the retention time and mass spectrum information of each component are recorded to generate volatile compound spectrum data; Based on the volatile compound spectrum data, the chromatographic peaks of each sample are identified, integrated, and baseline corrected. The detected chromatographic peaks are identified by referring to the standard spectral library information. The signals of all identified compounds in each sample data are normalized. The difference in the mean of the normalized signals of each identified compound between the treatment group and the control group is compared. Compounds with changed signal responses are screened to obtain a differential volatile organic compound list.

7. The multidimensional evaluation method for the antibacterial effect of the bamboo leaf volatile oil according to claim 2, wherein The steps for obtaining the metabolic network topology change evaluation results are specifically as follows: Based on the differential volatile organic compound list, the total number of differential volatile organic compounds in the list is counted, and the relative change amplitude of the signal response of each differential volatile organic compound between the moso bamboo leaf volatile oil treatment group and the control group is quantified one by one to obtain a quantitative statistical set of differential substances; Based on the quantitative statistical set of differential substances and the differential volatile organic compound list, each volatile organic compound in the differential volatile organic compound list is associated with the corresponding biochemical reaction pathway in the known metabolic pathway database. The change in the amount of all associated differential volatile organic compounds on the key metabolic pathway of the target is evaluated and calculated according to their importance in the pathway to obtain a metabolic perturbation index.

8. The multidimensional evaluation method for the antibacterial effect of bamboo leaf volatile oil according to claim 7, wherein The step of obtaining the metabolic network topology change evaluation result further includes: Based on the normalized signal data of each compound in the differential volatile organic compound list in different samples, the correlation strength between any two differential volatile organic compound signals is analyzed, and compound pairs with correlation are screened. The compounds are regarded as nodes and the correlations as connecting edges. A volatile organic compound correlation network is constructed, and the connectivity and clustering of the volatile organic compound correlation network are analyzed to obtain the metabolic network topology change assessment results.

Images