A method and system for detecting a new pollutant of ecological environment

Abstract

The present application relates to the field of environmental monitoring and biosensing technology, and particularly relates to a detection method and system for new pollutants in ecological environment, comprising the following steps: S100: culturing specific effector cells in a biosensing channel, and collecting trans-epithelial electrical resistance (TEER) signals of the effector cells in real time; S200: monitoring a change rate of the TEER signals, and generating an enrichment triggering instruction when the change rate exceeds a preset threshold; S300: in response to the enrichment triggering instruction, introducing a fluid to be tested, which generates the change rate, into a surface-enhanced Raman scattering (SERS) detection cavity, and starting an active enrichment unit to in-situ concentrate the fluid to be tested, so that the volume of the fluid to be tested is contracted to a detection area of a SERS substrate; the present application uses impedance response of effector cells as a preliminary screening "sentinel", realizes targeted capture of only abnormal samples with acute toxicity, avoids ineffective detection of a large number of non-toxic water samples, and significantly improves monitoring efficiency.

Description

Technical Field

[0001] This invention relates to the field of environmental monitoring and biosensing technology, specifically to a method and system for detecting new pollutants in the ecological environment. Background Technology

[0002] In recent years, with the acceleration of industrialization, "new pollutants" such as antibiotics, endocrine disruptors, and perfluorinated compounds have been frequently detected in aquatic environments. These pollutants are characterized by high biotoxicity, high environmental persistence, and extremely low concentrations (typically in the ng / L to μg / L range). Traditional physicochemical detection methods, such as gas chromatography-mass spectrometry (GC-MS) or liquid chromatography-mass spectrometry (LC-MS), while accurate in qualitative and quantitative analysis, rely on large laboratory equipment, involve cumbersome sample pretreatment, and have long detection cycles, making it difficult to meet the needs for on-site, real-time, and rapid early warning.

[0003] Limitations of Existing Biosensing Technologies: To achieve rapid early warning, cell-based biosensors have emerged. In particular, sensors based on transepithelial electrical impedance (TEER) technology can reflect the overall acute toxicity of water by monitoring the integrity of a cell monolayer. However, existing cell impedance sensors suffer from a significant lack of specificity—they can only tell users "the water is toxic," but cannot identify "what substance is causing the toxicity." Furthermore, the single impedance signal is easily interfered with by non-specific factors such as temperature and flow rate fluctuations, leading to a high false alarm rate.

[0004] On the other hand, surface-enhanced Raman scattering (SERS) technology, due to its "fingerprint recognition" capability, is considered a powerful tool for identifying new pollutants. However, in practical applications, SERS detection faces two major physical bottlenecks: first, "diffusion limitation": at extremely low concentrations, the probability of target molecules diffusing and adsorbing onto the hot spot region of the SERS substrate is extremely low, resulting in weak signals or even undetectable signals; second, the "coffee ring effect": conventional droplet drying and enrichment methods lead to uneven solute deposition, mainly accumulating at the droplet edges, resulting in poor reproducibility of spectral signals and difficulty in quantitative analysis.

[0005] While current technologies have attempted to combine biosensing with chemical analysis, most are simply physical stacking processes lacking intelligent triggering and linkage mechanisms. For example, systems often indiscriminately perform complex SERS detection on all samples, resulting in a large number of non-toxic samples consuming valuable detection resources and data processing power. Furthermore, at the data analysis level, existing retrieval algorithms are mostly based on single spectral matching, which is highly susceptible to background interference when dealing with complex environmental water matrices, leading to erroneous matching results. There is a lack of multimodal identification methods that utilize biotoxicological characteristics to assist in chemical qualitative analysis.

[0006] Therefore, there is an urgent need to develop an integrated detection system that can use biotoxicity response as a primary screening trigger and combine efficient physical enrichment technology with multimodal AI recognition algorithms to solve the technical problem that existing technologies cannot balance sensitivity, specificity and real-time performance. Summary of the Invention

[0007] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a method for detecting new pollutants in the ecological environment, comprising the following steps: S100: culturing specific effector cells in a biosensor channel and acquiring transepithelial electrical impedance signals of the effector cells in real time; S200: monitoring the rate of change of the TEER signal, and generating an enrichment trigger command when the rate of change exceeds a preset threshold; S300: responding to the enrichment trigger command, introducing the test fluid that generates the rate of change into the SERS detection chamber, and activating an active enrichment unit to concentrate the test fluid in situ, shrinking its volume to the detection area of ​​the SERS substrate; S400: acquiring the SERS characteristic spectral data of the concentrated test fluid; S500: inputting the rate of change feature that triggers the enrichment trigger command and the SERS characteristic spectral data together into a dual-modal recognition model; the model limits the database range to be searched based on the rate of change feature, matches the SERS characteristic spectral data within this range, and outputs the detection result.

[0008] Further, in step S100: an interdigitated electrode array is integrated at the bottom of the biosensing channel, the surface of the interdigitated electrode array is modified with an extracellular matrix protein layer, and the specific effector cells adhere to and grow on the extracellular matrix protein layer to form a dense monolayer structure; the real-time acquisition of the transepithelial electrical impedance signal of the effector cells specifically includes: applying a perturbation AC voltage with a frequency range of 10Hz to 100kHz to the interdigitated electrode array using an impedance analyzer, measuring the current response flowing through the monolayer of the specific effector cells, and extracting the TEER signal by calculating the impedance modulus and phase angle.

[0009] Further, in step S200: monitoring the rate of change of the TEER signal specifically includes: performing sliding window smoothing filtering on the real-time acquired TEER signal sequence to eliminate high-frequency noise interference, and then calculating the first derivative of the smoothed TEER signal with respect to time as the rate of change; the preset threshold is set based on: statistically analyzing the standard deviation of the TEER signal baseline under conditions where no test fluid or blank buffer solution is introduced, and setting the preset threshold to 3 to 5 times the standard deviation.

[0010] Further, in step S200: the generation of enrichment trigger instruction specifically includes: recording the start time and end time when the rate of change exceeds the preset threshold, calculating the fluid transmission delay time based on the fluid velocity and pipe volume between the biosensing channel and the SERS detection chamber in the microfluidic chip, and generating an enrichment trigger instruction containing time window parameters in combination with the start time and end time.

[0011] Further, in step S300: the active enrichment unit includes a concentric segmented driving electrode disposed at the bottom of the SERS detection cavity, a dielectric layer covering the driving electrode, and an uppermost hydrophobic layer, wherein the SERS substrate is located at the center of the concentric segmented driving electrode.

[0012] Further, in step S300: the in-situ concentration of the test fluid by activating the active enrichment unit specifically includes: applying an AC driving voltage to the concentric segmented driving electrode, adjusting the wettability of the test fluid on the hydrophobic layer surface by utilizing the dielectric wetting effect; by controlling the voltage timing, making the SERS substrate region hydrophilic and the peripheral region of the SERS substrate hydrophobic, thereby driving the droplet edge of the test fluid to contract towards the center, pinning the solute in the fluid to the surface of the SERS substrate.

[0013] Further, in step S400: the SERS substrate adopts a nanoporous noble metal composite structure with a three-dimensional hotspot distribution, including a porous anodic alumina template and gold or silver nanoparticle clusters deposited in the pores of the template; the acquisition of SERS characteristic spectral data specifically includes: controlling the excitation beam of the Raman spectrometer to focus on the geometric center of the active enrichment unit, selecting 785 The near-infrared band was used as the excitation wavelength for spectral scanning; and the acquired raw spectral data were processed by removing cosmic ray interference, polynomial fitting baseline correction, and maximum value normalization to obtain SERS characteristic spectral data with enhanced signal-to-noise ratio.

[0014] Further, in step S500: the dual-modal recognition model pre-constructs an association database, which establishes a mapping relationship between known chemical pollutants and their corresponding SERS standard spectra and cytotoxic mechanism tags; the model limits the scope of the database to be searched based on the rate of change feature, specifically including: extracting the time response curve feature vector of the rate of change at the trigger time, inputting it into the first classifier to identify the toxicological mechanism category of the pollutant, and filtering out a subset of candidate chemicals with the same toxicological mechanism tag from the association database according to the identified toxicological mechanism category.

[0015] Further, in step S500: matching the SERS characteristic spectral data within the range specifically includes: converting the SERS characteristic spectral data into a spectral vector to be measured. The standard spectra of each chemical in the candidate chemical subset are converted into reference spectral vectors. The matching score between the two can be calculated using the following cosine similarity formula: ; in, Indicates that the measured spectral vector is at the th... Normalized intensity values ​​at each Raman shift point The reference spectral vector is represented as being in the th... Normalized intensity values ​​at each Raman shift point This indicates the number of spectral sampling points; the system outputs the chemical with the highest calculated Similarity value as the target detection result.

[0016] A detection system for new environmental pollutants, and a method for detecting new environmental pollutants according to any one of claims 1-9, characterized in that the system comprises: A biosensing module, configured in the upstream channel of a microfluidic chip, includes a specific effector cell culture interface and an impedance acquisition unit, used to acquire the transepithelial electrical impedance signal of the effector cells in real time as the fluid to be tested flows through it. The signal monitoring and triggering module is electrically connected to the biosensing module and is used to calculate the rate of change of the TEER signal in real time. When the absolute value of the rate of change exceeds a preset threshold, the current fluid to be tested is determined to be an abnormal sample and an enrichment trigger command is generated. The cascaded enrichment execution module is located in the downstream channel of the microfluidic chip and is communicatively connected to the signal monitoring and triggering module. In response to the enrichment triggering command, the cascaded enrichment execution module introduces the abnormal sample into the detection cavity and starts the active enrichment unit to perform in-situ volume concentration of the abnormal sample, so that it accumulates in the surface-enhanced Raman scattering substrate region. The spectral acquisition module, optically coupled to the SERS substrate region, is used to perform laser excitation and scanning on the concentrated anomalous sample to acquire SERS characteristic spectral data. The intelligent identification and tracing module is connected to the signal monitoring and triggering module and the spectral acquisition module, respectively, and is used to receive the rate of change feature and the SERS feature spectral data that trigger the enrichment trigger command, and input them into the dual-modal identification model; the dual-modal identification model is configured to use the rate of change feature to limit the database search range, and match the SERS feature spectral data within the range to output the detection result.

[0017] Beneficial effects Compared with known public technologies, the technical solution provided by this invention has the following beneficial effects: This invention innovatively constructs a hierarchical joint detection system combining biological response triggering, physical active enrichment, and chemical spectral fingerprinting. Its beneficial effects are as follows: First, by utilizing the impedance response of effector cells as a primary screening "sentinel," it achieves targeted capture of only anomalous samples with acute toxicity, avoiding ineffective detection of a large number of non-toxic water samples and significantly improving monitoring efficiency. Second, through in-situ microfluidic concentration achieved by dielectric wetting (EWOD) driving and contact wire pinning technology, it effectively overcomes the "diffusion limitation" and uneven "coffee ring effect" existing in traditional SERS detection, increasing the enrichment factor of trace pollutants to over 100 times, greatly enhancing detection sensitivity. Finally, the unique dual-modal recognition model utilizes the characteristics of biological toxicology mechanisms to "reduce the dimensionality" of massive chemical spectral databases, solving the problem of single spectral matching being easily affected by background interference and having a high false alarm rate in complex water matrices, thus achieving accurate source tracing of new pollutants with high biological relevance, high sensitivity, and high specificity. Attached Figure Description

[0018] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a flowchart of the signal processing and triggering logic for step S200 of the present invention. Detailed Implementation

[0019] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0020] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but includes other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0021] The present invention will now be described in further detail with reference to the accompanying drawings: Example:

[0022] As shown in the figure, a method for detecting new pollutants in the ecological environment includes the following steps: S100: Specific effector cells are cultured in a biosensing channel, and transepithelial electrical impedance signals of the effector cells are acquired in real time; The specific implementation process is as follows: S101: First, the underlying substrate of the microfluidic chip is prepared. A standard glass slide is selected as the substrate material, and a thickness of 20 mm is deposited on the surface of the glass substrate using conventional photolithography and lift-off processes. Chromium adhesion layer with a thickness of 100 A gold (Au) conductive layer is used to pattern an interdigitated electrode array. The interdigitated electrode array is designed to contain 20 pairs of finger electrodes, with a finger width and a finger spacing of 20. The effective sensing area is approximately 2 mm². Subsequently, a microfluidic channel layer was fabricated using polydimethylsiloxane (PDMS) via soft lithography, with a channel height designed to be 100 mm. Width is 500 The bonding surfaces of the glass substrate and the PDMS layer are treated with oxygen plasma to irreversibly bond the microfluidic channel layer to the glass substrate containing the interdigital electrode array. This ensures that the interdigital electrode array is fully exposed within the main flow path of the microfluidic channel, and the electrode pins are led out to the external interface of the chip via wires.

[0023] S102: The surface of the interdigitated electrode array within the channel is biocompatiblely modified to promote cell adhesion and growth. A concentration of 50% is injected into the assembled biosensor channel via a microfluidic pump. A solution of fibronectin or type I collagen was prepared, and the chip was incubated in a 37°C incubator for 1 to 2 hours to allow extracellular matrix proteins to bind tightly to the gold electrode surface through physical adsorption. After incubation, sterile phosphate buffer was used to... The flow rate was used to flush the channel for 5 minutes to remove unadsorbed protein residues and form a uniform extracellular matrix modification layer.

[0024] S103: Human colon adenocarcinoma cells or human bronchial epithelial cells were selected as specific effector cells, digested and resuspended in complete culture medium, and prepared to a cell density of approximately [missing information]. The cell suspension was administered using a microinfusion pump. The microfluidic chip was injected at a low flow rate into the modified biosensor channel. Once the channel was filled with cell suspension, flow was stopped, and the chip was placed in a 37°C, 5% CO2 incubator for 4 hours to allow for initial cell adhesion. Subsequently, the microfluidic perfusion system was activated at a flow rate of 2... Fresh culture medium is continuously introduced at a constant flow rate to simulate the shear force environment of blood or body fluids in the body. The cells are continuously and dynamically cultured for 3 to 5 days until cells in the channel are observed to be interconnected under an optical microscope, forming a dense and complete cell monolayer.

[0025] S104: Electrochemical impedance spectroscopy (EIS) scanning of cell growth status was performed using an interdigitated electrode array integrated at the bottom of the channel to verify the compactness of the cell monolayer. The electrode interface of the chip was connected to a high-precision impedance analyzer, and frequency sweep tests were performed in the frequency range of 100Hz to 100kHz. When the impedance amplitude at the low-frequency end (e.g., 500Hz) reached a stable plateau and the phase angle deviated significantly from the blank substrate data, it indicated that the tight junctions between cells had been fully established. At this point, the cell monolayer possessed the functional conditions to serve as a biosensing interface and could be used for subsequent toxicity monitoring.

[0026] S105: The system enters real-time monitoring mode (i.e., baseline acquisition phase). Maintain the test fluid (initially a pure buffer solution or standard water sample) at... A constant flow rate flows through the biosensing channel. An impedance analyzer applies an effective value of 10 to the interdigital electrode array. The system measures the perturbation AC voltage and locks the detection frequency at 10kHz (this frequency is the characteristic frequency point with the optimal signal-to-noise ratio). The system continuously reads and records the current response flowing through the cell monolayer at a sampling rate of 1 second / sample, calculates the real-time transepithelial impedance modulus value according to Ohm's law, and generates a time-impedance response baseline as a reference benchmark for determining the acute toxicity of new pollutants in subsequent steps.

[0027] S200: Monitor the rate of change of the TEER signal, and generate an enrichment trigger command when the rate of change exceeds a preset threshold; In this embodiment, the process of monitoring the TEER signal change rate, determining anomalies, and generating trigger commands in step S200 is specifically implemented through the following sub-steps S201 to S205: The specific implementation process is as follows: S201: The processor first preprocesses the raw TEER signal sequence acquired in real time in step S100 to eliminate high-frequency random noise caused by microfluidic pump pulsation or external electromagnetic interference. This embodiment uses a sliding window averaging method to smooth the raw impedance data. The size of the sliding window is set to... For the current sampling time Smoothed impedance value The calculation formula is: ; in, This represents the original impedance modulus value. This step preserves the low-frequency trend characteristics of cell impedance changes while significantly improving the signal-to-noise ratio.

[0028] S202: Based on smoothed impedance data, the system calculates the first derivative of impedance with time in real time, i.e., the rate of change of impedance, to quantify the cell's acute toxic response to the flowing fluid. Since the sampling is discrete, the backward difference method is used to approximate the calculation of the current moment. rate of change of impedance The calculation formula is: ; in, The rate of change is the time interval between two adjacent samples. It can keenly reflect the instantaneous degree of damage to the integrity of the cell layer.

[0029] S202: Determine the acute toxicity threshold used to determine positive samples Before the system is formally tested, a non-toxic blank culture medium is introduced for baseline calibration, and the rate of change of baseline impedance is continuously collected for a period of time (e.g., 10 minutes). The standard deviation of the rate of change of impedance during this baseline period is calculated. and according to Statistical criteria set thresholds, and the calculation formula is as follows: ; in, This represents the total number of sampling points during the baseline calibration phase. For the baseline phase Rate of change at each point The mean of the baseline rate of change. This is the sensitivity coefficient. It should be noted that, in this embodiment, it is preferably set to... This ensures a confidence level of over 99.7% and effectively avoids false positives.

[0030] S204: Execute real-time decision logic. The processor will calculate the absolute value of the real-time impedance change rate obtained in S202. The threshold set in S203 A comparison is performed. When the conditions are met... And this state persists beyond the judgment time window. For example, if three consecutive sampling points are detected, the system determines that the fluid currently flowing through the biosensor channel contains an acutely toxic contaminant, and marks it as a "positive sample." The system records the start time of this triggering event. And generate a Level 1 alarm sign.

[0031] S205: Generates an enriched trigger command containing precise timing control parameters. Due to the physical distance between the biosensing channel (upstream) and the SERS detection chamber (downstream), the fluid transport delay time must be calculated. This ensures that the positive sample is captured precisely when the downstream valve opens. Based on the microfluidic chip's geometric parameters and fluid dynamics equations, the calculation formula is as follows: ; in, The dead volume of the flow channel between the two modules. These are the length, width, and height of the connecting channel, respectively. The current microfluidic volumetric flow rate is set. The processor will calculate Write trigger commands to control the downstream micro-valve at a specific time. The flow path is switched on time, and the electrode activation duration of the SERS detection chamber is set to a time window that matches the passage time of the positive sample, thereby achieving precise physical capture of trace amounts of toxic fluid.

[0032] S300: In response to the enrichment trigger command, the test fluid that generates the rate of change is introduced into the SERS detection chamber, and the active enrichment unit is activated to concentrate the test fluid in situ, so that its volume shrinks to the detection area of ​​the SERS substrate. The specific implementation process is as follows: S301: The central controller monitors the enriched trigger commands generated by S200 in real time. When the system time... Reaching the preset start time in the command Upon activation, the controller immediately sends a high-level drive signal to the shunt microvalve located downstream of the biosensing channel. This microvalve (preferably a piezoelectrically actuated or pneumatically actuated microvalve) completes a state switch within 10 ms, changing the flow path from the waste liquid outlet to the SERS detection chamber inlet. This open state is maintained for a preset time window (e.g., 10 seconds), capturing a positive sample droplet of approximately 20 μL into the SERS detection chamber. Subsequently, the valve resets, closing the SERS detection chamber to form an independent reaction microenvironment.

[0033] S302: After the positive sample droplet enters the SERS detection chamber, it naturally covers the electrowetting drive array at the bottom. The drive array consists of a series of concentric circular ring electrodes, with the SERS enhancement substrate (e.g., 200mm in diameter) at the center. (Gold nanostar array region). In the initial state without applied voltage, the hydrophobic layer covering the electrode surface (e.g., Teflon AF1600, approximately 50 nm thick) allows the droplet to maintain a large contact angle. The droplets are in a loose, spread-out state.

[0034] S303: Start the Active Electrowetting (EWOD) driver. The controller applies an AC drive voltage with a frequency of 1kHz to the concentric electrodes. At this point, the contact angle of the droplet at the solid-liquid interface changes, and its change follows the Lippmann-Young equation. To ensure that the droplet can be effectively driven and to prevent dielectric layer breakdown, the formula for setting the optimal driving voltage is as follows: ; in, The contact angle after applying voltage. The initial contact angle, The vacuum permittivity, For dielectric layers (e.g., SU-8 photoresist, thickness) The relative permittivity of ) This refers to the liquid-gas interfacial tension.

[0035] It should be noted that the driving voltage is calculated and set in this embodiment. This significantly reduces the contact angle. This generates an unbalanced capillary force at the edge of the droplet, driving the droplet to move towards the central region where the electric field strength is highest.

[0036] S304: Perform centripetal contraction and pinning operations. The controller employs a timing-based control strategy from the outside in, sequentially activating the outermost to the innermost concentric electrodes. As the outer electrodes are de-energized and become hydrophobic, and the inner electrodes are energized and become hydrophilic, the droplet is "squeezed" layer by layer towards the SERS substrate region at the center. Finally, only the electrode below the SERS substrate at the center remains active, using an electric field to forcibly pin the droplet's three-phase contact line to the geometric boundary of the SERS substrate, preventing the droplet from drifting during subsequent processing.

[0037] S305: Initiate auxiliary thermal evaporation concentration. While maintaining the pinned state, activate the micro-thin-film heater integrated on the bottom of the chip to maintain the substrate temperature at 50°C (below the temperature that destroys biomolecules). Because the contact lines are pinned by the electric field, solvent evaporation causes a sharp decrease in droplet volume. The solute (new pollutant molecules) cannot evaporate with the solvent and is instead continuously transported and accumulated in the SERS hotspot region at the center under the "reverse coffee ring effect" of the microfluidic. After approximately 60 seconds of concentration, the droplet volume decreases from the initial 20... Shrink to 0.2 It achieves a physical enrichment coefficient of approximately 100 times, thereby increasing the surface concentration of trace pollutants to above the SERS detection limit.

[0038] S400: Acquire the SERS characteristic spectral data of the concentrated fluid to be tested; Following step S300 above, the process of excitation scanning and acquisition of characteristic spectral data of the enriched positive sample described in step S400 of this embodiment is specifically implemented through the following sub-steps S401 to S405: The specific implementation process is as follows: S401: Place the entire microfluidic chip on the stage of a confocal micro Raman spectrometer. Adjust the three-axis displacement platform of the microscope and focus through the transparent cover plate (e.g., glass or PDMS) on top of the microfluidic chip using a 50x long focal length objective lens. Observe using a white light imaging system to precisely locate the area where the droplets were pinned and concentrated in step S300, and focus the laser spot at the densest cluster of gold nanoparticles at the geometric center of the SERS substrate, which is the core region of the electromagnetic field-enhanced "hot spot".

[0039] S402: Configure the excitation parameters of the spectrometer. Considering that new pollutants (mostly small organic molecules) are prone to strong fluorescence background interference under visible light excitation, this embodiment selects a wavelength of 785 nm. A near-infrared diode laser was used as the excitation source. To prevent the trace organic molecules enriched on the substrate from photodegrading or carbonizing due to high-energy laser light, the power of the laser reaching the sample was set to 10. Up to 20 Setting the single-scan spectral integration time to 5 to 10 seconds and the cumulative number of scans to 3 can initially reduce random thermal noise through hardware averaging.

[0040] S403: Perform spectral scanning and raw data acquisition. Start the spectrometer. Raman scattering signals were collected within the wavenumber range of the fingerprint region. The charge-coupled device within the spectrometer converted the optical signal into an electrical signal, generating the raw spectral data. ,in The Raman shift system temporarily stores the raw data in a cache. At this point, the raw spectrum typically includes superimposed cosmic ray spikes, the sample's own fluorescence background envelope, and instrument baseline drift.

[0041] S404: Performs a standardized preprocessing algorithm on the raw spectral data. First, a medium-range filtering algorithm is applied to identify and remove abnormal high-intensity peaks with a width less than 3 pixels. Second, for 785... For the residual fluorescence background after excitation, an adaptive iterative reweighted penalized least squares method is used for baseline fitting. The processor calculates a smooth background baseline through multiple iterations. The corrected net Raman spectrum is obtained by subtracting it from the original spectrum. The calculation formula is: This effectively eliminates the interference of non-Raman scattered light on the intensity of characteristic peaks.

[0042] S405: The enhancement effect of the spectral signal is evaluated and normalized. To quantify the detection performance of this system under the synergistic effect of physical enrichment and substrate enhancement, the processor calculates the surface-enhanced Raman scattering enhancement factor based on the characteristic peaks. The formula for calculating the EF value is as follows: ; in, The measured SERS characteristic peak intensity, The intensity of the conventional Raman peak is the same material measured on an unreinforced substrate (such as ordinary glass). This represents the number of molecules being measured within the laser focal spot volume. This is for the routine measurement of the number of molecules within a volume.

[0043] Calculations show that the EF value in this embodiment reaches [value missing]. That concludes the above. Finally, to adapt to the input requirements of subsequent AI models, [further details are needed]. Perform Min-Max normalization to map all spectral intensities to The interval is used to generate the final SERS characteristic spectral data vector.

[0044] S500: The rate of change feature that triggers the enrichment trigger instruction and the SERS feature spectral data are jointly input into the dual-modal recognition model; the model limits the database range to be searched based on the rate of change feature, and matches the SERS feature spectral data within this range to output the detection result.

[0045] In this embodiment, the process of using a dual-modal recognition model for database limitation and spectral matching in step S500 is specifically implemented through the following sub-steps S501 to S505: The specific implementation process is as follows: S501: Construct a bimodal relational reference database. During the system initialization phase, a standard dataset containing various known novel pollutants (e.g., antibiotics, pesticides, endocrine disruptors, etc.) is pre-established. This database adopts a relational structure, and each entry record contains: 1) Chemical identification information (name, molecular formula, CAS number); (2) Standard SERS spectral vector (fingerprint spectrum measured under the same experimental conditions as S400). (3) Cell toxicology mechanism labels (e.g., “cell membrane perforation type”, “oxidative stress type”, “mitochondrial damage type”, etc.). The toxicology mechanism labels are determined by cluster analysis based on the impedance response curve characteristics (e.g., slope of decrease, response delay time, maximum decrease amplitude) caused by the chemical to specific effector cells (e.g., Caco-2).

[0046] S502: Perform feature vectorization extraction on the bioimpedance signal that triggered the alarm in step S200. The processor then traces back to the trigger time. The impedance change data before and after the trigger is used to extract a complete time response window (e.g., data within 5 minutes after triggering).

[0047] Key morphological features of the response curve were extracted, including: the maximum rate of descent. Time required to reach maximum inhibition rate and impedance recovery rate The above features are combined to construct a multidimensional biological feature vector. .

[0048] S503: Perform a first-level coarse classification based on biometric features to classify the biometric feature vector. The input is a pre-trained support vector machine or random forest classifier. Based on the impedance response pattern of the input, the classifier predicts the toxicological mechanism category of the current pollutant. According to this prediction, the system sends a query command to the associated reference database, retrieving only chemical entries with the same "cell membrane perforation" tag, thus generating a significantly reduced subset of candidate chemicals. This step, by introducing prior biological knowledge, compresses the originally massive full-database search space to 10% or even less, significantly improving the efficiency and accuracy of subsequent matching.

[0049] S504: Perform second-level fine-grained matching based on spectral features. The processor reads the SERS feature spectral data of the sample to be tested, which was output in step S400, and defines it as the vector to be tested. Simultaneously, the candidate chemical subsets generated in S503 are traversed, and the standard spectrum of each chemical in the subset is defined as a reference vector. The processor calculates one by one. With each Similarity scores between them.

[0050] In this embodiment, to overcome the influence of absolute intensity fluctuations in the spectrum, S505 focuses on matching peak positions with relative intensities, and uses cosine similarity as the core discrimination algorithm. The calculation formula is as follows: ; in, Indicates that the measured spectral vector is at the th... Normalized intensity values ​​at each Raman shift point The reference spectral vector is represented as being in the th... Normalized intensity values ​​at each Raman shift point This indicates the number of band points sampled in the spectrum.

[0051] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting new pollutants in the ecological environment, characterized in that, Includes the following steps: S100: Specific effector cells are cultured in a biosensor channel, and transepithelial electrical impedance signals of the effector cells are acquired in real time; S200: Monitor the rate of change of the TEER signal, and generate an enrichment trigger command when the rate of change exceeds a preset threshold; S300: In response to the enrichment trigger command, the test fluid that generates the rate of change is introduced into the SERS detection chamber, and the active enrichment unit is activated to concentrate the test fluid in situ, so that its volume shrinks to the detection area of ​​the SERS substrate. S400: Acquire the SERS characteristic spectral data of the concentrated fluid to be tested; S500: The rate of change feature that triggers the enrichment trigger instruction and the SERS feature spectral data are jointly input into the dual-modal recognition model; the model limits the database range to be searched based on the rate of change feature, and matches the SERS feature spectral data within this range to output the detection result.

2. The method for detecting new environmental pollutants according to claim 1, characterized in that, In step S100: an interdigitated electrode array is integrated at the bottom of the biosensing channel, and an extracellular matrix protein layer is modified on the surface of the interdigitated electrode array. The specific effector cells adhere to and grow on the extracellular matrix protein layer to form a dense monolayer structure. The real-time acquisition of the transepithelial electrical impedance signal of the effector cells specifically includes: applying a perturbation AC voltage with a frequency range of 10Hz to 100kHz to the interdigital electrode array using an impedance analyzer, measuring the current response flowing through the monolayer of the specific effector cells, and extracting the TEER signal by calculating the impedance modulus and phase angle.

3. The method for detecting new environmental pollutants according to claim 2, characterized in that, In step S200: monitoring the rate of change of the TEER signal specifically includes: performing sliding window smoothing filtering on the real-time acquired TEER signal sequence to eliminate high-frequency noise interference, and then calculating the first derivative of the smoothed TEER signal with respect to time as the rate of change; The preset threshold is set based on the following: when no test fluid or blank buffer solution is introduced, the standard deviation of the TEER signal baseline is calculated, and the preset threshold is set to 3 to 5 times the standard deviation.

4. The method for detecting new environmental pollutants according to claim 3, characterized in that, In step S200: The specific steps of generating the enrichment trigger instruction include: recording the start and end times when the rate of change exceeds the preset threshold; calculating the fluid transmission delay time based on the fluid velocity and pipe volume between the biosensing channel and the SERS detection chamber in the microfluidic chip; and generating an enrichment trigger instruction containing time window parameters in combination with the start and end times.

5. The method for detecting new environmental pollutants according to claim 4, characterized in that, In step S300: the active enrichment unit includes a concentric segmented driving electrode disposed at the bottom of the SERS detection cavity, a dielectric layer covering the driving electrode, and an uppermost hydrophobic layer, and the SERS substrate is located at the center of the concentric segmented driving electrode.

6. The method for detecting new environmental pollutants according to claim 5, characterized in that, In step S300: The in-situ concentration of the test fluid by activating the active enrichment unit specifically includes: applying an AC driving voltage to the concentric segmented driving electrode to adjust the wettability of the test fluid on the hydrophobic layer surface using the dielectric wetting effect; and controlling the voltage timing to make the SERS substrate region hydrophilic and the peripheral region of the SERS substrate hydrophobic, thereby driving the droplet edge of the test fluid to contract towards the center, pinning the solute in the fluid to the SERS substrate surface.

7. The method for detecting new environmental pollutants according to claim 6, characterized in that, In step S400: The SERS substrate adopts a nanoporous noble metal composite structure with a three-dimensional hot spot distribution, including a porous anodic aluminum oxide template and gold or silver nanoparticle clusters deposited in the pores of the template; The acquisition of SERS characteristic spectral data specifically includes: controlling the excitation beam of the Raman spectrometer to focus on the geometric center of the active enrichment unit, and selecting a 785... The near-infrared band was used as the excitation wavelength for spectral scanning. The acquired raw spectral data were processed to remove cosmic ray interference, perform polynomial fitting baseline correction, and normalize the maximum value to obtain SERS characteristic spectral data with enhanced signal-to-noise ratio.

8. The method for detecting new environmental pollutants according to claim 7, characterized in that, In step S500: The dual-modal recognition model has a pre-built relational database that maps known chemical pollutants to their corresponding SERS standard spectra and cytotoxic mechanism tags. The model, based on the rate of change feature, limits the scope of the database to be searched, specifically by: extracting the time response curve feature vector of the rate of change at the trigger time, inputting it into a first classifier to identify the toxicological mechanism category of the pollutant, and filtering out a subset of candidate chemicals with the same toxicological mechanism label from the association database according to the identified toxicological mechanism category.

9. The method for detecting new environmental pollutants according to claim 8, characterized in that, In step S500: The specific steps of matching the SERS characteristic spectral data within the range include: converting the SERS characteristic spectral data into a spectral vector to be measured. The standard spectra of each chemical in the candidate chemical subset are converted into reference spectral vectors. The matching score between the two can be calculated using the following cosine similarity formula: ； in, Indicates that the measured spectral vector is at the th... Normalized intensity values ​​at each Raman shift point The reference spectral vector is represented as being in the th... Normalized intensity values ​​at each Raman shift point This indicates the number of spectral sampling points; the system outputs the chemical with the highest calculated Similarity value as the target detection result.

10. A detection system for new environmental pollutants, comprising a detection method for new environmental pollutants according to any one of claims 1-9, characterized in that the system... include: A biosensing module, configured in the upstream channel of a microfluidic chip, includes a specific effector cell culture interface and an impedance acquisition unit, used to acquire the transepithelial electrical impedance signal of the effector cells in real time as the fluid to be tested flows through it. The signal monitoring and triggering module is electrically connected to the biosensing module and is used to calculate the rate of change of the TEER signal in real time. When the absolute value of the rate of change exceeds a preset threshold, the current fluid to be tested is determined to be an abnormal sample and an enrichment trigger command is generated. The cascaded enrichment execution module is located in the downstream channel of the microfluidic chip and is communicatively connected to the signal monitoring and triggering module. In response to the enrichment triggering command, the cascaded enrichment execution module introduces the abnormal sample into the detection cavity and starts the active enrichment unit to perform in-situ volume concentration of the abnormal sample, so that it accumulates in the surface-enhanced Raman scattering substrate region. The spectral acquisition module, optically coupled to the SERS substrate region, is used to perform laser excitation and scanning on the concentrated anomalous sample to acquire SERS characteristic spectral data. The intelligent identification and tracing module is connected to the signal monitoring and triggering module and the spectral acquisition module, respectively, and is used to receive the rate of change feature and the SERS feature spectral data that trigger the enrichment trigger command, and input them into the dual-modal identification model; the dual-modal identification model is configured to use the rate of change feature to limit the database search range, and match the SERS feature spectral data within the range to output the detection result.

Images