Goose follicular granule cell lipid droplet separation and multi-dimensional quality evaluation system and method

Abstract

The invention relates to the technical field of biological information analysis, in particular to a goose follicular granule cell lipid droplet separation and multi-dimensional quality evaluation system and method, and the system comprises a lipid droplet image acquisition module, a lipid droplet separation module, a Raman spectrum detection module, a chemical component analysis module and a quality comprehensive evaluation module. According to the method, a multi-stage data acquisition system based on space identification and dynamic separation is constructed, so that the form of the lipid droplet at a single cell scale is accurately captured and positioned, and the target lipid droplet is screened and transferred by utilizing a fluid control strategy on the premise of not damaging the structure; high-purity molecular information is obtained by combining the field enhancement effect of spectral signals and lattice parameter tuning, the proportion difference of fatty acid and triglyceride is deconstructed through signal characteristics, the quantitative corresponding relation between morphological parameters and chemical compositions is established, multi-dimensional comprehensive characterization of lipid droplets is achieved, and the precision and result consistency of lipid metabolism state analysis are improved.

Description

Technical Field

[0001] This invention relates to the field of bioinformatics analysis technology, and in particular to a system and method for lipid droplet separation and multidimensional quality assessment of goose follicle granulosa cells. Background Technology

[0002] The field of bioinformatics analysis technology encompasses a comprehensive technological system for the acquisition, processing, and analysis of biomolecular information using computer science, mathematics, and statistics. Its core components include the interpretation of genome sequencing data, the integrated analysis of transcriptomic and proteomic information, and the construction and computational modeling of metabolic pathways and molecular interaction networks. By combining algorithms with experimental data, it enables the digital characterization and multi-dimensional feature analysis of biological processes, providing crucial support for modern life science research and precision medicine.

[0003] The traditional goose follicle granulosa cell lipid droplet separation and multidimensional quality assessment system refers to an experimental system for extracting and quantifying lipid droplet components in granulosa cells of poultry ovaries. It uses density gradient centrifugation to physically separate lipid droplets, and then combines fluorescence microscopy and biochemical colorimetry to evaluate the morphology, size and content of lipid droplets. At the same time, it uses Western blotting or mass spectrometry to detect the expression level of lipid droplet-related proteins, so as to achieve a comprehensive characterization of the lipid metabolism status of granulosa cells.

[0004] Current technologies for studying lipid droplets in goose follicle granulosa cells rely on centrifugation and staining to obtain information. The separation process easily damages the lipid droplet structure, causing deviations in morphology and chemical characteristics. The results can only reflect the average state of the overall sample and cannot reveal the true distribution and differences of individual lipid droplets within the cell. Fluorescence imaging and colorimetric detection are affected by signal overlap and photobleaching, resulting in insufficient data stability. Protein detection lacks a direct correlation with lipid droplet properties, making it difficult to form a unified indicator system between morphology, composition, and function. Consequently, lipid droplet quality assessment lacks spatial resolution and molecular specificity, affecting the accurate determination of lipid metabolism status. Summary of the Invention

[0005] To address the shortcomings of existing technologies in goose follicle granulosa cell lipid droplet research, which rely on centrifugation and staining to obtain information, the separation process easily damages the lipid droplet structure, causing deviations in morphology and chemical characteristics. The results only reflect the average state of the overall sample and cannot reveal the true distribution and differences of individual lipid droplets within the cell. Furthermore, fluorescence imaging and colorimetric detection are affected by signal overlap and photobleaching, resulting in insufficient data stability. The lack of a direct correlation between protein detection and lipid droplet properties makes it difficult to establish a unified indicator system between morphology, composition, and function, leading to a lack of spatial resolution and molecular specificity in lipid droplet quality assessment and affecting the accurate determination of lipid metabolic status. This invention provides a system and method for lipid droplet separation and multi-dimensional quality assessment in goose follicle granulosa cells. The technical solution is as follows: On the one hand, a system for lipid droplet separation and multi-dimensional quality assessment of goose follicle granulosa cells is provided, which includes: The lipid droplet image acquisition module acquires images of goose follicle granulosa cell samples through a microscope, performs edge detection on the lipid droplet region based on image processing algorithms, calculates the area value and roundness coefficient, and marks the spatial location coordinates to generate a lipid droplet morphology localization dataset, which is then transmitted to the lipid droplet separation module. The lipid droplet separation module calls the lipid droplet morphology localization dataset, and uses a microfluidic device to adjust the flow rate of the lipid droplets and control the interception valve according to the spatial coordinates to separate the target lipid droplets, generate a lipid droplet separation sample information set, and transmit it to the Raman spectroscopy detection module. The Raman spectroscopy detection module calls the lipid droplet separation sample information set, carries the lipid droplet sample through a nano-depression array, adjusts the depression parameters to optimize the electric field gradient, excites Raman scattering and collects the signal, analyzes the peak position wave value and peak intensity value, generates a lipid droplet spectral feature set, and transmits it to the chemical composition analysis module. The chemical composition analysis module calls the lipid droplet spectral feature set, uses the plasmon resonance peak tuning model to dynamically adjust the lattice spacing according to the peak wave value, identifies the content ratio of triglycerides and fatty acid components, generates a lipid droplet multidimensional feature dataset, and transmits it to the quality comprehensive evaluation module.

[0006] As a further aspect of the present invention, the lipid droplet morphology localization dataset includes lipid droplet area parameters, lipid droplet roundness coefficients, and lipid droplet spatial location information; the lipid droplet separation sample information set includes target lipid droplet number, fluid flow velocity parameters, and valve control status data; the lipid droplet spectral feature set includes Raman peak position wave value, Raman peak intensity value, and spectral distribution characteristics; and the lipid droplet multidimensional feature dataset includes triglyceride content ratio, fatty acid composition ratio, and lattice spacing parameters.

[0007] As a further aspect of the present invention, the lipid droplet image acquisition module includes: The microscopic image acquisition submodule acquires images of goose follicle granulosa cell samples through a microscope, adjusts the light intensity and focal length, detects the pixel signal intensity distribution and performs grayscale equalization processing, standardizes and compresses the pixel matrix, numbers and caches image frames, and generates standardized image matrix data. The lipid droplet edge detection submodule, based on the standardized image matrix data, uses an image processing algorithm to scan the pixel grayscale gradient and compare it with the gradient threshold, filters out regions with brightness abrupt changes greater than the threshold and records their coordinates, determines the connectivity of pixel points, aggregates and corrects the boundaries, and generates a lipid droplet boundary coordinate set. The gradient threshold is dynamically set based on the average gray-level difference of the sample images; The morphological parameter calculation submodule calculates the number of pixels in the closed region and converts it to area based on the lipid droplet boundary coordinate set, calculates the perimeter of the contour path and the roundness coefficient, extracts the geometric center coordinates and integrates the parameters to generate a lipid droplet morphological localization dataset.

[0008] As a further aspect of the present invention, the lipid droplet separation module includes: The lipid droplet localization and analysis submodule obtains the spatial coordinates of the lipid droplet morphology localization dataset, analyzes the lipid droplet coordinates and calculates the path distribution in the microfluidic channel, calculates the fluid velocity gradient in the channel based on the morphological parameters and performs position index mapping to generate a lipid droplet spatial distribution matrix. The flow rate control submodule calls the lipid droplet spatial distribution matrix, calculates the flow rate correction coefficient based on the channel segment resistance value and pressure difference data, compares the flow rate threshold to determine the flow rate deviation and makes proportional adjustments, maps the correction coefficient to the lipid droplet diameter parameter, and generates a lipid droplet flow rate control parameter set. The velocity threshold is determined based on the ratio of average velocity to local velocity fluctuation rate; The valve-controlled separation submodule, based on the set of lipid droplet flow rate control parameters, calls the flow rate correction coefficient and channel positioning index data to calculate the opening time of the interception valve and the pressure adjustment range, determines the residence time of lipid droplets within the valve-controlled interval, and filters the target morphology lipid droplet data to generate a lipid droplet separation sample information set.

[0009] As a further aspect of the present invention, the Raman spectroscopy detection module includes: The lipid droplet sample carrying submodule acquires the lipid droplet separation sample information set, injects the sample into the surface of the nano-depression array, calculates the surface tension distribution based on the array geometric parameters, judges the sample distribution state, adjusts the array unit spacing, records the positioning coordinates and tension deviation, and generates a lipid droplet spatial distribution parameter set. The electric field gradient control submodule, based on the lipid droplet spatial distribution parameter set, calls the array surface electrode driving signal, calculates the electric field intensity range, adjusts the indentation depth and electrode bias direction, calculates the mapping function relationship between the electric field change rate and the indentation depth, and obtains the indentation electric field gradient parameter set. The spectral signal analysis submodule applies excitation light to the lipid droplet sample according to the set of concave electric field gradient parameters, collects scattering signals, calculates the peak wavenumber and peak intensity range, filters scattering points according to the ratio of wavenumber to intensity, and generates a lipid droplet spectral feature set.

[0010] As a further aspect of the present invention, the chemical composition analysis module includes: The spectral signal receiving submodule acquires the multi-channel reflectance spectral signal in the lipid droplet spectral feature set, calculates the light intensity stability based on the amplitude change of the band signal, filters out noise signals and performs interval equalization processing, calculates the average absorption intensity value based on the absorption rate distribution, and generates a spectral intensity distribution parameter set. The peak position dynamic adjustment submodule calls the spectral intensity distribution parameter set, calculates the center wavenumber offset difference based on the plasmonic resonance peak tuning model, adjusts the lattice spacing to correct the resonance peak position according to the offset difference, records the wavenumber difference sequence, and generates a set of resonance peak position adjustment coefficients. The component ratio calculation submodule extracts the absorption intensity of triglycerides and fatty acids based on the resonance peak position adjustment coefficient set, calculates the intensity ratio and normalizes it, and performs weighted calculation on the absorption intensity of the bands through weighting factors to generate a lipid droplet multidimensional feature dataset.

[0011] As a further aspect of the present invention, the comprehensive quality assessment module calls the lipid droplet multidimensional feature dataset, compares the roundness coefficient with the preset morphology standard threshold to calculate the morphology score, compares the fatty acid component content ratio with the preset component standard range to calculate the component score, and calculates the comprehensive quality score by weighting, classifies the lipid droplet quality level, and outputs the lipid droplet quality level identifier. The lipid droplet quality rating includes a morphological score, a compositional score, and a comprehensive quality score.

[0012] As a further aspect of the present invention, the comprehensive quality assessment module includes: The feature data comparison submodule acquires the roundness coefficient and fatty acid component content ratio data in the lipid droplet multidimensional feature dataset, calculates the deviation value according to the morphological standard threshold and the component standard range, integrates the roundness and component deviation results into a numerical vector, and generates a feature comparison difference vector. The morphological scoring calculation submodule calls the feature comparison difference vector, performs weighted calculation based on the weight ratio set for the roundness deviation term and fatty acid deviation amount, normalizes the deviation components, and generates a lipid droplet weighted scoring matrix. The quality grade classification submodule calculates the score difference by calling the boundary values ​​of the quality grade benchmark interval based on the lipid droplet weighted scoring matrix, determines the corresponding interval position and maps the grade number, and generates a lipid droplet quality grade identifier.

[0013] As a further aspect of the present invention, the morphological standard threshold is determined based on the statistical distribution law of lipid droplet roundness coefficient in multiple batches of samples, and in combination with the results of quantitative analysis of microscopic images and morphological comparison experiments. The component standard range is determined by statistical regression analysis based on the content ratio characteristics of fatty acid components in different sample groups, combined with the chemical component detection standards and the mean range of laboratory measurements. The boundary values ​​of the quality grade benchmark interval are determined based on the statistical distribution results of the lipid droplet weighted scoring matrix.

[0014] On the other hand, the method for lipid droplet separation and multidimensional quality assessment of goose follicle granulosa cells, which is based on the aforementioned system for lipid droplet separation and multidimensional quality assessment of goose follicle granulosa cells, includes the following steps: S1: Obtain images of goose follicle granulosa cell samples using a microscope, perform edge detection on lipid droplet regions based on image processing algorithms, calculate area values ​​and roundness coefficients, and mark spatial location coordinates to generate a lipid droplet morphology localization dataset; S2: Call the lipid droplet morphology localization dataset, and use a microfluidic device to adjust the flow rate of the lipid droplets and control the interception valve according to the spatial coordinates to separate the target lipid droplets and generate a lipid droplet separation sample information set; S3: Call the lipid droplet separation sample information set, carry the lipid droplet sample through a nano-depression array, adjust the depression parameters to optimize the electric field gradient, excite Raman scattering and collect the signal, analyze the peak wave value and peak intensity value, and generate a lipid droplet spectral feature set; S4: Call the lipid droplet spectral feature set, use the plasmonic resonance peak tuning model to dynamically adjust the lattice spacing according to the peak wave value, identify the content ratio of triglycerides and fatty acid components, and generate a lipid droplet multidimensional feature dataset; S5: Call the lipid droplet multidimensional feature dataset, compare the roundness coefficient with the preset morphology standard threshold to calculate the morphology score, compare the fatty acid component content ratio with the preset component standard range to calculate the component score, and calculate the weighted comprehensive quality score to classify the lipid droplet quality level and output the lipid droplet quality level identifier.

[0015] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: By constructing a multi-level data acquisition system based on spatial recognition and dynamic separation, the morphology of lipid droplets at the single-cell scale is accurately captured and located. A fluid control strategy is used to complete the screening and transfer of target lipid droplets without destroying the structure. High-purity molecular information is obtained by combining the field enhancement effect of spectral signals and lattice parameter tuning. The difference in the ratio of fatty acids and triglycerides is deconstructed through signal characteristics, and a quantitative correspondence between morphological parameters and chemical composition is established, achieving multi-dimensional comprehensive characterization of lipid droplets and improving the accuracy and consistency of lipid metabolism state analysis. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the accompanying drawings without creative effort.

[0017] Figure 1 This is a system schematic diagram of the present invention; Figure 2 This is a schematic diagram of the system framework of the present invention; Figure 3 This is a flowchart of the lipid droplet image acquisition module in this invention; Figure 4 This is a flowchart of the lipid droplet separation module in this invention; Figure 5 This is a flowchart of the Raman spectroscopy detection module in this invention; Figure 6 This is a flowchart of the chemical composition analysis module in this invention; Figure 7 This is a flowchart of the comprehensive quality assessment module in this invention; Figure 8 This is a flowchart of the method of the present invention. Detailed Implementation

[0018] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0019] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0020] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0021] In this embodiment of the invention, sometimes the subscript such as W1 is written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0022] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0023] This invention provides a system for separating lipid droplets from goose follicle granulosa cells and conducting multi-dimensional quality assessment, such as... Figure 1-2 The diagram shown illustrates a system for separating lipid droplets from goose follicle granulosa cells and performing multidimensional quality assessment. This system includes: The lipid droplet image acquisition module acquires images of goose follicle granulosa cell samples through a microscope, performs edge detection on the lipid droplet region based on image processing algorithms, calculates the area value and roundness coefficient, and marks the spatial location coordinates to generate a lipid droplet morphology localization dataset, which is then transmitted to the lipid droplet separation module. The lipid droplet separation module calls the lipid droplet morphology localization dataset, and uses a microfluidic device to adjust the flow rate of the lipid droplets and control the interception valve according to the spatial coordinates to separate the target lipid droplets, generate a lipid droplet separation sample information set, and transmit it to the Raman spectroscopy detection module. The Raman spectroscopy detection module calls the lipid droplet separation sample information set, carries the lipid droplet sample through a nano-depression array, adjusts the depression parameters to optimize the electric field gradient, excites Raman scattering and collects the signal, analyzes the peak wave value and peak intensity value, generates the lipid droplet spectral feature set, and transmits it to the chemical composition analysis module. The chemical composition analysis module calls the lipid droplet spectral feature set, uses the plasmonic resonance peak tuning model to dynamically adjust the lattice spacing according to the peak wave value, identifies the content ratio of triglycerides and fatty acid components, generates a lipid droplet multidimensional feature dataset, and transfers it to the quality comprehensive evaluation module. The comprehensive quality assessment module calls the lipid droplet multidimensional feature dataset, compares the roundness coefficient with the preset morphology standard threshold to calculate the morphology score, compares the fatty acid component content ratio with the preset component standard range to calculate the component score, and calculates the comprehensive quality score by weighting, classifies the lipid droplet quality level, and outputs the lipid droplet quality level identifier. The lipid droplet morphology localization dataset includes lipid droplet area parameters, lipid droplet roundness coefficients, and lipid droplet spatial location information. The lipid droplet separation sample information dataset includes target lipid droplet numbers, fluid flow velocity parameters, and valve control status data. The lipid droplet spectral feature dataset includes Raman peak position values, Raman peak intensity values, and spectral distribution characteristics. The lipid droplet multidimensional feature dataset includes triglyceride content ratio, fatty acid composition ratio, and lattice spacing parameters. The lipid droplet quality grade identifier includes morphology score, composition score, and comprehensive quality score.

[0024] Specifically, such as Figure 2 , 3 As shown, the lipid droplet image acquisition module includes: The microscopic image acquisition submodule acquires images of goose follicle granulosa cell samples through a microscope, adjusts the light intensity and focal length, detects the pixel signal intensity distribution and performs grayscale equalization processing, standardizes and compresses the pixel matrix, numbers and caches image frames, and generates standardized image matrix data. First, activate the high-sensitivity charge-coupled device camera mounted on the inverted microscope and set the light source brightness parameter to [value missing]. Lumens were used to ensure sufficient and uniform illumination of the goose follicle granulosa cell sample within the field of view, and then the stepper motor was controlled to... The step-driven platform in Move up and down along the axis for continuous data acquisition Images with different focal planes were analyzed, and their sharpness was evaluated by calculating the Laplacian variance of each image. The image with the largest variance was selected as the sharpest sample frame. Then, the photoelectric signal response value of each pixel in the sample frame was read, and the analog signal was converted into... to The system iterates through the range of digital grayscale values ​​to construct an initial grayscale matrix. For each pixel, the frequency of gray levels is counted and a gray-level histogram is plotted. If the histogram peaks are detected to be concentrated in... to For the low grayscale range, a grayscale linear stretching operation is performed to stretch the original grayscale values. Mapped to new grayscale value For example, the interval Mapping extended to This enhances the contrast between the cytoplasm and lipid droplets. Subsequently, the pixel matrix is ​​standardized, and the arithmetic mean of the current image pixel grayscale values ​​is calculated. and standard deviation For example, in a containing The average gray level is measured within a local region of 1 pixel. Standard deviation For the grayscale value in this area target pixels The system converts it into a standard score. The calculation process is as follows Divide by get This value eliminates the absolute value deviation caused by uneven lighting. Next, dynamic range compression is performed, using logarithmic transformation logic to map the standard score with a wider value range to... to Standard byte space, set compression factor for The calculation formula is reflected as Multiply by the natural logarithm ,like for The transformed value is Rounded down to the nearest integer The system performs bit-depth compression of single-point data, and finally uses the current system timestamp. and sample number Generate a unique image frame index The processed matrix data is written to a cache queue to generate standardized image matrix data.

[0025] The lipid droplet edge detection submodule, based on standardized image matrix data, uses image processing algorithms to scan pixel grayscale gradients and compares them with gradient thresholds. It filters out regions with brightness abrupt changes greater than the threshold and records their coordinates. It judges the connectivity of pixel points, aggregates and corrects the boundaries, and generates a lipid droplet boundary coordinate set. The gradient threshold is dynamically set based on the average gray-level difference of the sample images; First, define a The sliding window traverses the normalized image matrix, calculating the relationship between the center pixel and the surrounding pixels. The gray-level difference of neighboring pixels is calculated separately in the horizontal direction. and vertical direction The gradient components, for example, the gray level of the center pixel is The horizontal left pixel is The right side is Then the horizontal gradient Approximately The vertically upward pixel is Below is Vertical gradient Approximately Comprehensive gradient magnitude Calculated as The square of and The arithmetic square root of the sum of the squares of , i.e. The system then selects a background region in the image where no cells are distributed, and calculates the value of that region. Average gradient value of each pixel and gradient standard deviation For example, measured , Set sensitivity coefficient for Dynamically calculate gradient threshold Its value is Plus Multiply ,Right now This threshold serves as the benchmark for determining whether a region is a lipid droplet edge. The system calculates the gradient magnitude of each pixel across the entire image. With threshold Perform a comparison one by one, if a certain pixel point gradient Greater than If the point is determined to be a candidate point in the region of sudden brightness change, its coordinates are recorded. If gradient Less than If a point is not found, it is marked as a background point. The specific judgment data is shown in Table 1. After the filtering is completed, the system performs a connectivity judgment, checking the area around each candidate edge point. Do there exist other candidate points in each direction? If a certain point... exist If there are no connection points in the neighborhood, it is considered an isolated noise point and removed; otherwise, it is retained. Finally, the broken edges are aggregated and corrected. When the distance between the endpoints of two edge segments is less than... When there are 1 pixel, linear interpolation is used to automatically fill in the missing pixels, such as endpoints. With endpoints The system automatically fills in the pixels between them. Generate the lipid droplet boundary coordinate set using the closed contour.

[0026] Table 1: Pixel Gradient Screening Judgment Table for Lipid Droplet Edges

[0027] As shown in Table 1, the gradient magnitude of different coordinate points is binarized and filtered according to the dynamically set threshold. Only when the local gradient is significantly higher than the threshold calculated by the background noise is the point retained as an edge candidate point.

[0028] The morphological parameter calculation submodule calculates the number of pixels in the closed region and converts it into area based on the lipid droplet boundary coordinate set, calculates the perimeter and roundness coefficient of the contour path, extracts the geometric center coordinates and integrates the parameters to generate a lipid droplet morphological localization dataset. First, identify the region enclosed by each closed boundary, and then use scanline fill logic to count the total number of pixels within the closed region. For example, for the number The lipid droplets were statistically analyzed to obtain the number of internal pixels as follows: Based on microscope calibration parameters, the actual physical length represented by a single pixel is known. for The physical area of ​​a single pixel for Multiply equal The system will calculate the total number of pixels. Multiply by unit area The actual cross-sectional area of ​​the lipid droplet was calculated. for Next, calculate the perimeter of the contour path, traverse the boundary coordinate list, and accumulate the Euclidean distance between adjacent boundary points. Horizontal or vertical adjacency is denoted as... Unit length, diagonally adjacent is denoted as Unit length, assuming cumulative pixel distance is One unit, multiplied by physical length To obtain the actual perimeter for Then the roundness coefficient is extracted. The calculation formula is set as follows Multiply by pi Multiply by the area Divide by perimeter The square of , substituted into the numerical value, is calculated as Divide by ,Right now Divide by The result is approximately This value is used to quantify the degree to which lipid droplets deviate from a perfect circle, while the system also assigns the x-coordinate of the boundary points. and ordinate Calculate the arithmetic mean separately, for example, for the boundary point set. The sum of the coordinates is The number of points is Then the x-coordinate of the geometric center for Similarly, the ordinate can be obtained. for The area calculated above ,perimeter Roundness and center coordinates The lipid droplets are structured and encapsulated according to their numbers to generate a lipid droplet morphology localization dataset.

[0029] Specifically, such as Figure 2 , 4 As shown, the lipid droplet separation module includes: The lipid droplet localization and parsing submodule obtains the spatial coordinates of the lipid droplet morphology localization dataset, parses the lipid droplet coordinates and calculates the path distribution in the microfluidic channel, calculates the fluid velocity gradient in the channel based on the morphological parameters and performs position index mapping to generate the lipid droplet spatial distribution matrix. First, extract the data with the number [number] from the dataset. lipid droplets in pixel center coordinates at time Call the microscope calibration parameters, i.e., the physical length of a single pixel. , pixel coordinates Multiply Physical conversion coordinate Similarly, convert physics coordinate ,get Physical location at any moment Then in Time (e.g.) Get New pixel coordinates , converted to Physical location at any moment System records and Form path points and calculate instantaneous velocity in direction Its value is Divide by ,Right now The system then uses morphological parameters With position Query the preset flow field model of the microfluidic channel and extract... Theoretical fluid velocity corresponding to the position The velocity slip difference was calculated. And this slip difference is related to the lipid droplet area. Perform associative storage, and finally execute location index mapping to define a Logical grid matrix To characterize The channel area, each grid cell Size is droplets of fat The coordinates are divided by Rounded down, we get and System access Unit, update the lipid droplet count within that unit. (For example from) Become ) and average area (For example from) Updated to ), generating the lipid droplet spatial distribution matrix.

[0030] The flow rate control submodule calls the lipid droplet spatial distribution matrix, calculates the flow rate correction coefficient based on the channel segment resistance value and pressure difference data, compares the flow rate threshold to determine the flow rate deviation and makes proportional adjustments, maps the correction coefficient to the lipid droplet diameter parameter, and generates a lipid droplet flow rate control parameter set. The velocity threshold is determined based on the ratio of average velocity to local velocity fluctuation rate; From the matrix Extract the average flow velocity of the current region from the cell. and average area At the same time, acquire the channel segment. Preset fluid resistance and the currently applied pressure difference Set the target flow rate for this area. System measurement within the area velocity values ​​of consecutive sampling points Calculate its standard deviation Then, the local velocity fluctuation rate can be obtained. Its value is the standard deviation. Divide by average flow velocity ,get The system is based on the average flow velocity With volatility The ratio relationship determines the flow velocity threshold. Set sensitivity coefficient Flow rate threshold The calculation process is as follows Multiply Multiply by ,Right now The threshold This represents the acceptable range of flow velocity deviation; the system calculates the current flow velocity deviation. Its value is the target flow rate. Subtract average flow velocity ,get Compare flow rate thresholds Determine the flow velocity deviation. Greater than If a significant deviation is detected, a proportional adjustment is performed, and a flow rate correction factor is calculated. Its value is Divide by ,Right now The coefficient Will be used for adjustment Finally, the correction coefficient will be... Mapping the lipid droplet diameter to the corresponding parameter, the lipid droplet diameter... From average area Conversion, System storage mapping relationship (Seg_6, D=3.09) (K=1.333), generate a set of lipid droplet flow rate control parameters.

[0031] The valve-controlled separation submodule, based on the lipid droplet flow rate control parameter set, calls the flow rate correction coefficient and channel positioning index data, calculates the opening time of the interception valve and the pressure adjustment range, determines the residence time of lipid droplets in the valve-controlled interval and filters the target morphology lipid droplet data, and generates a lipid droplet separation sample information set. Call the mapping relationship (Seg_6, D=3.09) (K=1.333), obtain the flow velocity correction coefficient. and the corresponding channel positioning index and obtain lipid droplets Current location With target flow rate The inquiry revealed Corresponding shut-off valve lie in First, calculate the pressure adjustment range. retrieve current pressure The calculated range is ,Right now Apply this pressure increment to This increases the lipid droplet velocity to Then, the opening time of the shut-off valve and the time required for the grease droplets to reach the valve were calculated. Distance Divide by speed ,get , setting system reaction delay Calculate the valve trigger time Next, calculate the valve opening time. lipid droplets Time required for complete passage through the valve for Divide by ,get ,set up safety factor ,calculate The system determines the valve control range (length). Duration of lipid droplet residence time Its value is Divide by ,get This value Less than The lipid droplets were determined to be able to pass through smoothly, and finally the lipid droplet data were screened according to the target morphology parameters shown in Table 2.

[0032] Table 2: Lipid Droplet Morphology Screening Criteria and Separation Decision Table

[0033] As shown in Table 2, the system retrieves... Morphological parameters and Perform comparison The judgment is true, and If the condition is true, the filtering decision is "withdraw", and the system... Triggered at any time Start, continue ,Will Import the collection channel and record its information. Generate a lipid droplet separation sample information set.

[0034] Specifically, such as Figure 2 , 5 As shown, the Raman spectroscopy detection module includes: The lipid droplet sample carrying submodule acquires the lipid droplet separation sample information set, injects the sample into the surface of the nano-depression array, calculates the surface tension distribution based on the array geometric parameters, judges the sample distribution state, adjusts the array unit spacing, records the positioning coordinates and tension deviation, and generates the lipid droplet spatial distribution parameter set. First, target lipid droplets are extracted from the lipid droplet separation sample information set. morphological information and The sample was then injected into the surface of the nano-recessed array chip, and the surface energy of the array substrate material was set. for Surface tension of the culture medium in which lipid droplets of goose follicle granulosa cells are located for According to the radius of the depression and the depth of the depression The geometric parameters were used to calculate the radius of curvature at the contact point between the lipid droplet and the depression boundary based on the Young-Laplace equation. For example, setting the contact angle for radius of curvature The calculation process is as follows Divide by ,Right now Next, the surface tension distribution is calculated. Its value is Multiply Divide by Substituting the values, we get This tension is the capillary force that drives the lipid droplets into the depression. The system determines the sample distribution state and sets the critical tension required for the lipid droplets to be completely embedded in the depression. for The calculated tension With critical tension By comparison, we found The system determines that the lipid droplets can be stably captured within the depression, and then adjusts the array cell spacing, setting the initial cell center spacing. for and set the target spacing. Based on lipid droplet diameter Confirm, calculate The value is ,Right now , initial spacing Distance from target By comparing the results, the spacing adjustment amount is calculated. The system sends displacement commands to the drive piezoelectric ceramic actuator. To increase the array spacing and achieve sparser arrangement of lipid droplets, a high-precision vision system was then used to record the positioning coordinates of the lipid droplets within the depressions. ,For example And calculate the tension deviation. Its value is minus ,Right now This deviation value This reflects the current surplus of capture capacity, and finally the positioning coordinates are determined. and tension deviation The parameters are stored in a structured manner to generate a set of lipid droplet spatial distribution parameters.

[0035] The electric field gradient control submodule, based on the lipid droplet spatial distribution parameter set, calls the array surface electrode driving signal, calculates the electric field intensity range, adjusts the indentation depth and electrode bias direction, calculates the mapping function relationship between the electric field change rate and the indentation depth, and obtains the indentation electric field gradient parameter set. First, lipid droplets are extracted from the parameter set. positioning coordinates and the depth of the depression The array surface electrode driving signal is invoked to set the initial driving voltage of the microelectrode. for Electrode spacing for Calculate the electric field strength Its value is Divide by ,Right now Define the electric field strength range and set the electric field strength Divided into low electric field regions , medium electric field area High electric field region According to the initial strength Determining that it is in the medium electric field region, the system adjusts the indentation depth and sets the adjustment amount for the indentation depth. Based on lipid droplet diameter With initial depth Determine the differences and calculate the adjustment amount. for If the adjustment amount is greater than Then, a depth increase operation will be performed, adjusting the depth to... Then, the electrode bias direction was adjusted, and the initial bias angle was set. for Based on the roundness coefficient of lipid droplets ,like (i.e., non-circular), then the offset angle will be... Adjusted to By applying a gradient force in the asymmetric direction, the rate of change of the electric field, i.e., the electric field gradient, is calculated. Its value is the electric field strength. Change Divide by depth change Set the electrode voltage to The final electric field strength Calculate the rate of change of electric field for Divide by ,get Finally, the system calculates the mapping function relationship between the rate of change of electric field and the depth of the depression. Set coefficients The value is Substitute depth ,calculate: ; This calculated value Compared with measured values By comparing and confirming the mapping relationship, the set of concave electric field gradient parameters is obtained.

[0036] The spectral signal analysis submodule applies excitation light to the lipid droplet sample based on the concave electric field gradient parameter set, collects the scattering signal, calculates the peak wavenumber and peak intensity value range, filters scattering points based on the wavenumber and intensity ratio, and generates a lipid droplet spectral feature set. Extract the depression depth from the parameter set and electric field gradient For lipid droplets Excitation light is applied to the sample, and the excitation light wavelength is set to... Power is Duration The system uses a spectrometer to continuously collect the scattered light from lipid droplets, recording wavenumbers in the range of... to Raman scattering signals between, for example in Signal strength measured at [location] Counting, in Signal strength measured at [location] Counting, in Signal strength measured at [location] The system counts the peak wavenumber and determines it by finding local maxima. Corresponding wave number Assuming the primary peak position, calculate the peak intensity range, and define the high range of peak intensity values ​​as... Count, the middle interval is Count, lower interval is Count, then peak intensity If the count indicates the location is in the high range, the system filters scattering points based on the wavenumber-intensity ratio and sets a wavenumber ratio reference value. for Set a reference value for the strength ratio for The system iterates through the collection points and calculates adjacent scattering points. and wavenumber ratio Strength ratio ,like and When a feature point is determined to be a valid feature point, for example and wavenumber ratio Greater than Strength ratio Greater than The scattering point is The selected effective feature points are then integrated based on their wavenumber, intensity, and ratio values ​​to generate a lipid droplet spectral feature set.

[0037] Specifically, such as Figure 2 , 6 As shown, the chemical composition analysis module includes: The spectral signal receiving submodule acquires multi-channel reflectance spectral signals from the lipid droplet spectral feature set, calculates light intensity stability based on the amplitude change of the band signal, filters out noise signals and performs interval equalization processing, calculates the average absorption intensity value based on the absorption rate distribution, and generates a spectral intensity distribution parameter set. Start the high-sensitivity spectrometer and set the integration time to... ,exist to Set within the spectral range One discrete acquisition channel, continuous acquisition Frame the original spectral data and construct the original signal matrix. The system performs light intensity stability calculations for each channel. Extract its in Sequence of light intensity values ​​in a frame Calculate the arithmetic mean of the sequence. with standard deviation For example, targeting The passage at that location was measured. The light intensity value collected this time is Count and calculate the average value. Count, standard deviation Counting, based on the formula Calculate the light intensity stability and substitute the values ​​to obtain Set stability screening threshold for ,Will and The comparison is performed to determine if the signal is valid and retain it; otherwise, if the stability of a certain channel is lower than that of the signal, the signal is retained. This channel is then marked as high-noise and set to zero. Subsequently, interval equalization is performed using a moving average filtering algorithm, with the window size set to [value missing]. Each channel smooths the retained signal to eliminate random spikes, then calculates the absorption rate distribution and calls the reference whiteboard light intensity stored in the system. Counting, using formulas Converting reflected light intensity into absorbance, for ,calculate Finally, the average absorption intensity value was calculated, and the characteristic band of triglycerides was selected. Accumulate within this band The absorbance values ​​of each channel are averaged, for example, the sum is... , divided by Obtain the average absorption intensity The specific spectral data processing results are shown in Table 3, which generates a set of spectral intensity distribution parameters.

[0038] Table 3: Lipid Droplet Spectral Signal Quality Assessment and Absorbance Conversion Table

[0039] As shown in Table 3, quality gating is performed on spectral signals of different bands based on the calculated stability index, and only high-stability signals are retained for subsequent absorbance conversion, thus ensuring the accuracy of the characteristic parameters.

[0040] The peak position dynamic adjustment submodule calls the spectral intensity distribution parameter set, calculates the center wavenumber offset difference based on the plasmonic resonance peak tuning model, adjusts the lattice spacing to correct the resonance peak position according to the offset difference, records the wavenumber difference sequence, and generates a set of resonance peak position adjustment coefficients. First, identify the actual spectral peak positions in the parameter set, for example, measure the current triglyceride level. The peak of symmetrical stretching vibration is located at The standard theoretical resonance peak position of the system is set as Calculation of center wavenumber shift difference based on plasmonic resonance peak tuning model The model defines the offset. Measured wavenumber Subtract standard wavenumber Substituting the numerical values, we obtain This positive offset indicates a mismatch between the localized surface plasmon resonance (LSPR) mode of the current nano-recessed array and the excitation light. The system adjusts the lattice spacing based on the offset difference and invokes the response function of the piezoelectric ceramic actuator. ,in Initial lattice spacing , The tuning coefficient is set to [value]. (The negative sign indicates that the wavenumber redshift requires a reduction in the lattice spacing to achieve a blueshift resonance mode), calculate the target lattice spacing correction. The system output voltage command drives the array substrate to shrink, adjusting the lattice spacing to... To correct the position of the resonance peak to Nearby, to enhance signal coupling efficiency, the wavenumber difference sequence is then recorded. The system performs this every [period]. Record the residual deviation after one adjustment, for example, the sequence is Finally, the final adjustment coefficient will be... (Normalized gain) and final bias Integrate and generate a set of resonance peak position adjustment coefficients.

[0041] The component ratio calculation submodule extracts the absorption intensity of triglycerides and fatty acids based on the resonance peak position adjustment coefficient set, calculates the intensity ratio and normalizes it, and generates a lipid droplet multidimensional feature dataset by weighting the absorption intensity of the bands through weighting factors. First, confirm that the resonance peaks have been calibrated. Then, extract the absorption intensities of the triglyceride (TG) and fatty acid (FA) bands and select... absorbance at Represents triglyceride content, selected absorbance at Represents the content of free fatty acids and terminal methyl groups, and calculates the intensity ratio. Its calculation formula is Divide by ,Right now This ratio reflects the relative abundance of esterified lipids and free lipids within the lipid droplet. After normalization, a standard ratio range for this type of cell lipid droplet was set. Using the linear mapping formula Calculate the normalized fraction, substitute it into the equation, and you will get... This value Map the ratio to The intervals are used to facilitate subsequent cluster analysis. Then, the absorption intensity of the bands is weighted by weighting factors to construct multidimensional eigenvalues. The formula is defined as follows: The advantage of this formula lies in the introduction of weighting factors. Different evaluation weights were assigned to the absolute amounts of triglycerides, fatty acids, and their relative proportions, respectively, thus integrating the total amount of lipid droplets and component structure information into a single feature value. Substitute into the calculation The system will use this feature value Compared with the preset high-quality lipid droplet reference range The comparison showed that... If the lipid droplet sample falls within this range, it indicates that the sample exhibits typical characteristics of maturity and a balanced composition. The final system will then... The data is packaged to generate a lipid droplet multidimensional feature dataset.

[0042] Specifically, such as Figure 2 , 7 As shown, the comprehensive quality assessment module includes: The feature data comparison submodule obtains the roundness coefficient and fatty acid component content ratio data from the lipid droplet multidimensional feature dataset, calculates the deviation value according to the morphological standard threshold and the component standard range, integrates the roundness and component deviation results into a numerical vector, and generates a feature comparison difference vector. Extract from dataset Roundness coefficient of lipid droplets and the normalized component ratios calculated above. The deviation value is calculated based on the morphological standard threshold and the composition standard range. in accordance with The sphericity statistical distribution of high-viability goose follicle lipid droplet samples from batches, taking its first value. percentile And combined with quantitative analysis of microscopic images Based on the morphological comparison results of the above roundness samples, a comprehensive determination was made. for Ingredient Standard Range in accordance with The proportion of fatty acid components in lipid droplet samples determined by chemical composition analysis (e.g., high performance liquid chromatography). Based on the distribution characteristics, statistical regression analysis determined that the mean range corresponding to its high survival rate was [missing value]. The system first calculates the roundness deviation. The calculation process involves morphological standard thresholds. Subtract the measured roundness coefficient ,Right now This value The difference between the representative morphology and the ideal standard is then used to calculate the component deviation. Determine the proportion of the measured components Is it within the standard range? Inside, Greater than or equal to and less than or equal to It was determined to be within the acceptable range, therefore the component deviation was negligible. Calculated as ,like Below (For example The deviation is... ,like Higher than (For example The deviation is... Finally, the roundness deviation With component deviation Integrate into numerical vectors to generate feature comparison difference vectors. .

[0043] The morphology scoring calculation submodule calls the feature comparison difference vector, performs weighted calculation based on the weight ratio set by the roundness deviation term and fatty acid deviation amount, normalizes the deviation components, and generates a lipid droplet weighted scoring matrix. extract Difference vectors of lipid droplets Based on the roundness deviation item Deviation from fatty acid Weighted calculations are performed by setting weight ratios. The basis for setting these weight ratios is that, in lipid droplet quality assessment, the regularity of morphology (roundness) is a significant indicator of subsequent developmental potential. The deviation of the component ratio accounts for Therefore, a weight is set for the roundness deviation. The weight of fatty acid bias Next, the deviation components are normalized, and the roundness deviation is... Normalized reference maximum deviation Its value is a threshold. Subtract the theoretical minimum value ,Right now Calculate the normalized roundness deviation for Divide by ,get , fatty acid deviation Normalized reference maximum deviation Its value is Outside the scope Maximum deviation, i.e. Calculate the normalized fatty acid bias for Divide by ,get The system then calculates the weighted sum of deviations. The calculation process is as follows Substitute the value as ,get Finally, the lipid droplet weighted score was calculated. The calculation process is as follows Subtract the total weighted deviation ,Right now The rating will be as The quality evaluation values ​​of lipid droplets are stored in a matrix to generate a weighted scoring matrix of lipid droplets. .

[0044] The quality grade classification submodule calculates the score difference based on the lipid droplet weighted scoring matrix, calls the boundary values ​​of the quality grade benchmark interval, determines the corresponding interval position and maps the grade number, and generates a lipid droplet quality grade identifier. Call Weighted score of lipid droplets And call the quality grade benchmark interval boundary value, the quality grade benchmark interval boundary value according to The statistical distribution of the weighted scoring matrix of historical lipid droplet samples was determined, and the scoring was... from arrive Divide the scope and set For level (excellent), For level (good), For level (middle), For level (Difference), the specific boundary values ​​are shown in Table 4.

[0045] Table 4: Criteria for Classifying Lipid Droplet Quality Grades

[0046] As shown in Table 4, system call levels boundary values ,grade boundary values ,grade boundary values Calculate the score difference and then... and Compare, Less than or equal to Continue and Compare, Greater than Determine the position of the corresponding interval. And map the level number, the interval corresponding to the level The system will The quality grade of lipid droplets is determined as follows: (Good), and this identifier Store and generate lipid droplet quality grade identifiers. .

[0047] Please see Figure 8 The method for lipid droplet separation and multidimensional quality assessment of goose follicle granulosa cells is based on the above-mentioned system for lipid droplet separation and multidimensional quality assessment of goose follicle granulosa cells, and includes the following steps: S1: Obtain images of goose follicle granulosa cell samples using a microscope, perform edge detection on lipid droplet regions based on image processing algorithms, calculate area values ​​and roundness coefficients, and mark spatial location coordinates to generate a lipid droplet morphology localization dataset; S2: Call the lipid droplet morphology localization dataset, and use a microfluidic device to adjust the flow rate of the lipid droplets and control the interception valve according to the spatial coordinates to separate the target lipid droplets and generate a lipid droplet separation sample information set; S3: Call the lipid droplet separation sample information set, carry the lipid droplet sample through a nano-depression array, adjust the depression parameters to optimize the electric field gradient, excite Raman scattering and collect the signal, analyze the peak wave value and peak intensity value, and generate a lipid droplet spectral feature set; S4: Call the lipid droplet spectral feature set, use the plasmonic resonance peak tuning model to dynamically adjust the lattice spacing according to the peak wave value, identify the content ratio of triglycerides and fatty acid components, and generate a lipid droplet multidimensional feature dataset; S5: Call the lipid droplet multidimensional feature dataset, compare the roundness coefficient with the preset morphology standard threshold to calculate the morphology score, compare the fatty acid component content ratio with the preset component standard range to calculate the component score, and calculate the weighted comprehensive quality score to classify the lipid droplet quality level and output the lipid droplet quality level identifier.

[0048] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A goose follicular granulosa cell lipid droplet isolation and multi-dimensional quality assessment system, characterized in that, The system comprises: The lipid droplet image acquisition module acquires images of goose granulosa cell samples through a microscope, performs edge detection on lipid droplet regions based on an image processing algorithm, calculates area values, roundness coefficients, and spatial position coordinates, generates lipid droplet morphology positioning data sets, and transmits them to the lipid droplet separation module; The lipid droplet separation module calls the lipid droplet morphology positioning data sets, adjusts the flow rate of lipid droplets and controls the interception valve based on the spatial position coordinates through a microfluidic device, separates target lipid droplets, generates lipid droplet separation sample information sets, and transmits them to the Raman spectrum detection module; The Raman spectrum detection module calls the lipid droplet separation sample information sets, carries lipid droplet samples through a nano-recess array, adjusts recess parameters to optimize the electric field gradient, excites Raman scattering and collects signals, analyzes peak wavenumber values and peak intensity values, generates lipid droplet spectrum feature sets, and transmits them to the chemical component analysis module; The chemical component analysis module calls the lipid droplet spectrum feature sets, dynamically adjusts the lattice spacing based on the peak wavenumber values using the plasmonic resonance peak tuning model, identifies the content ratio of triglycerides and fatty acid components, generates lipid droplet multi-dimensional feature data sets, and transmits them to the quality comprehensive evaluation module.

2. The goose follicular granulosa cell lipid droplet isolation and multi-dimensional quality assessment system of claim 1, wherein, The lipid droplet morphology positioning data sets include lipid droplet area parameters, lipid droplet roundness coefficients, and lipid droplet spatial position information, the lipid droplet separation sample information sets include target lipid droplet numbers, fluid flow rate parameters, and valve control state data, the lipid droplet spectrum feature sets include Raman peak wavenumber values, Raman peak intensity values, and spectrum distribution features, and the lipid droplet multi-dimensional feature data sets include triglyceride content ratios, fatty acid composition ratios, and lattice spacing parameters.

3. The goose follicular granulosa cell lipid droplet isolation and multi-dimensional quality assessment system of claim 1, wherein, The lipid droplet image acquisition module comprises: The microscopic image acquisition sub-module acquires images of goose granulosa cell samples through a microscope, adjusts light intensity and focal length, detects pixel signal intensity distribution and performs gray scale equalization processing, normalizes and dynamically compresses pixel matrices, numbers and buffers image frames, and generates standardized image matrix data; The lipid droplet edge detection sub-module, based on the standardized image matrix data, uses an image processing algorithm to scan pixel gray scale gradients and compare them with gradient thresholds, filters out brightness mutation regions greater than the threshold and records the coordinates, judges pixel point connectivity, aggregates and corrects the boundaries, and generates lipid droplet boundary coordinate sets; The morphology parameter calculation sub-module calculates the number of pixels in the closed region and converts the area based on the lipid droplet boundary coordinate sets, calculates the contour path circumference and roundness coefficient, extracts the geometric center coordinates and integrates the parameters, and generates the lipid droplet morphology positioning data sets.

4. The goose follicular granulosa cell lipid droplet isolation and multi-dimensional quality assessment system of claim 3, wherein, The lipid droplet separation module comprises: The lipid droplet positioning analysis sub-module acquires the spatial position coordinates in the lipid droplet morphology positioning data sets, analyzes the lipid droplet coordinates and calculates the path distribution in the microfluidic channel, calculates the fluid velocity gradient in the channel based on the morphology parameters and performs position index mapping, and generates a lipid droplet spatial distribution matrix; The flow rate regulation sub-module calls the lipid droplet spatial distribution matrix, calculates the flow rate correction coefficient based on the channel segment resistance value and pressure difference data, compares the flow rate threshold to determine the flow rate deviation and performs proportional adjustment, maps the correction coefficient and the lipid droplet diameter parameter, and generates lipid droplet flow rate control parameter sets; The valve control separation sub-module calls flow rate correction coefficients and channel positioning index data according to the set of lipid droplet flow rate control parameters, calculates interception valve opening time and pressure adjustment amplitude, judges lipid droplet residence time in the valve control interval, and filters target morphological lipid droplet data to generate lipid droplet separation sample information set.

5. The goose follicular granulosa cell lipid droplet isolation and multi-dimensional quality assessment system of claim 4, wherein, The Raman spectrum detection module comprises: The lipid droplet sample bearing sub-module obtains the lipid droplet separation sample information set, injects samples into the surface of the nanometer concave array, calculates surface tension distribution according to array geometric parameters, judges sample distribution state, adjusts array unit spacing, records positioning coordinates and tension deviation, and generates lipid droplet spatial distribution parameter set; The electric field gradient regulation sub-module calls array surface electrode driving signals based on the lipid droplet spatial distribution parameter set, calculates electric field intensity interval, adjusts concave depth and electrode bias direction, calculates mapping function relationship between electric field change rate and concave depth, and obtains concave electric field gradient parameter group; The spectrum signal analysis sub-module applies excitation light to lipid droplet samples according to the concave electric field gradient parameter group, collects scattering signals, calculates peak wave number and peak intensity value interval, filters scattering points according to wave number and intensity proportion relationship, and generates lipid droplet spectrum feature set.

6. The goose follicular granulosa cell lipid droplet isolation and multi-dimensional quality assessment system of claim 5, wherein, The chemical component analysis module comprises: The spectrum signal receiving sub-module obtains multi-channel reflection spectrum signals in the lipid droplet spectrum feature set, calculates light intensity stability according to waveband signal amplitude variation, filters out noise signals and performs interval equalization processing, calculates average absorption intensity value according to absorption rate distribution, and generates spectrum intensity distribution parameter set; The peak dynamic adjustment sub-module calls the spectrum intensity distribution parameter set, calculates center wave number deviation based on plasmon resonance peak tuning model, adjusts resonance peak position by correcting lattice spacing according to deviation, records wave number difference sequence, and generates resonance peak position adjustment coefficient set; The component proportion calculation sub-module extracts triglyceride and fatty acid waveband absorption intensity based on the resonance peak position adjustment coefficient set, calculates intensity ratio and performs normalization processing, performs weighted operation on waveband absorption intensity through weight factor, and generates lipid droplet multi-dimensional feature data set.

7. The goose follicular granulosa cell lipid droplet isolation and multi-dimensional quality assessment system of claim 1, wherein, The quality comprehensive evaluation module calls the lipid droplet multi-dimensional feature data set, compares roundness coefficient with preset morphological standard threshold to calculate morphological score value, compares fatty acid component content proportion with preset component standard range to calculate component score value, and weightedly calculates quality comprehensive score to divide lipid droplet quality grade and output lipid droplet quality grade identifier; The lipid droplet quality grade identifier comprises morphological score value, component score value and comprehensive quality score.

8. The goose follicular granulosa cell lipid droplet isolation and multi-dimensional quality assessment system of claim 7, wherein, The quality comprehensive evaluation module comprises: The feature data comparison sub-module obtains roundness coefficient and fatty acid component content proportion data in the lipid droplet multi-dimensional feature data set, calculates deviation value according to morphological standard threshold and component standard range, integrates roundness and component deviation results into numerical vector, and generates feature comparison difference vector; The morphological score calculation sub-module calls the feature comparison difference vector, performs weighted operation according to roundness deviation term and fatty acid deviation amount weight proportion, performs normalization processing on deviation components, and generates lipid droplet weighted score matrix; The quality grade classification sub-module calculates a score difference according to a quality grade benchmark interval boundary value based on the lipid droplet weighted score matrix, judges a corresponding interval position and maps a grade number, and generates a lipid droplet quality grade identifier.

9. The goose follicular granulosa cell lipid droplet isolation and multi-dimensional quality assessment system of claim 8, wherein, The morphological standard threshold is determined based on the statistical distribution of the lipid droplet roundness coefficient in multiple batches of samples, and combined with the results of quantitative analysis of microscopic images and morphological comparison experiments; The component standard range is determined based on the content proportion characteristics of fatty acid components in different sample populations, combined with chemical component detection standards and statistical regression analysis of laboratory measurement mean intervals; The quality grade benchmark interval boundary value is determined based on the statistical distribution results of the lipid droplet weighted score matrix.

10. A method for isolation and multi-dimensional quality assessment of lipid droplets from goose follicular granulosa cells, characterized in that, The goose follicular granulosa cell lipid droplet separation and multi-dimensional quality evaluation system according to any one of claims 1-9 is executed, comprising the following steps: S1: Obtain a goose follicular granulosa cell sample image through a microscope, perform edge detection on the lipid droplet region based on an image processing algorithm, calculate the area value, roundness coefficient, and mark the spatial position coordinates, and generate a lipid droplet morphological positioning data set; S2: Call the lipid droplet morphological positioning data set, adjust the flow rate and control the interception valve of the lipid droplet based on the spatial position coordinates through a microfluidic device, separate the target lipid droplet, and generate a lipid droplet separation sample information set; S3: Call the lipid droplet separation sample information set, load the lipid droplet sample on a nanometer concave array, adjust the concave parameters to optimize the electric field gradient, excite Raman scattering and collect signals, analyze the peak wavenumber value and peak intensity value, and generate a lipid droplet spectral feature set; S4: Call the lipid droplet spectral feature set, dynamically adjust the lattice spacing according to the peak wavenumber value using the plasmon resonance peak tuning model, identify the content proportion of triglycerides and fatty acid components, and generate a lipid droplet multi-dimensional feature data set; S5: Call the lipid droplet multi-dimensional feature data set, compare the roundness coefficient with the preset morphological standard threshold to calculate a morphological score value, compare the content proportion of fatty acid components with the preset component standard range to calculate a component score value, and weightedly calculate a quality comprehensive score to classify the lipid droplet quality grade, and output the lipid droplet quality grade identifier.

Images