A method for blood drug concentration analysis based on mass spectrometry detection

Abstract

The application relates to the technical field of blood drug concentration detection, and discloses a blood drug concentration analysis method based on mass spectrometric detection, which comprises the following steps: adopting microfluidic technology in combination with solid-phase extraction to pretreat a blood sample to obtain a concentrated solution, so as to remove interfering components and concentrate target drug molecules; according to the physicochemical properties of the target drug molecules, the concentrated solution is selected to flow through a biosensor or is directly introduced into a chromatographic unit, and the target drug molecules in the liquid phase are converted into gas-phase ions when flowing through the biosensor; chromatography-mass spectrometry is adopted to detect the gas-phase ions or the concentrated solution, so as to obtain mass spectrum data; a machine learning model is used to analyze the spectrum data, to identify the drug types and calculate the concentration of the target drug molecules, and to generate an analysis report. The method is suitable for target drugs with different physicochemical properties, simplifies the detection process, and improves the applicability and accuracy of detection.

Description

Technical Field

[0001] This application relates to the field of blood drug concentration detection technology, specifically to a blood drug concentration analysis method based on mass spectrometry detection. Background Technology

[0002] In the field of blood drug concentration detection, mass spectrometry is widely used due to its advantages of high sensitivity and high specificity. However, in actual detection, due to the differences in the physicochemical properties of target drugs, the process design of traditional detection methods lacks specificity and is difficult to flexibly adapt to the detection needs of different types of drugs. At the same time, the connection between the pretreatment stage and the subsequent detection stage is not smooth enough, and the removal effect of interfering components and the concentration efficiency of target drugs during the pretreatment process are limited, which affects the accuracy of the detection results and the overall analysis efficiency, and needs further improvement. Summary of the Invention

[0003] To solve, or at least partially solve, the above-mentioned technical problems, this application provides a blood drug concentration analysis method based on mass spectrometry detection, comprising the following steps: S1. Pre-treat the blood sample using microfluidic technology combined with solid-phase extraction to obtain a concentrate, remove interfering components from the blood sample and concentrate the target drug molecules; S2. Based on the physicochemical properties of the target drug molecule, select whether to allow the concentrate to flow through the biosensor or to directly introduce the concentrate into the chromatographic unit; if it flows through the biosensor, the target drug molecule in the liquid phase is converted into gaseous ions by the biosensor. S3. The gaseous ions converted by the biosensor or the concentrated solution directly introduced into the chromatographic unit are detected by chromatography-mass spectrometry to obtain mass spectrometry data. S4. Utilize a machine learning model to analyze the mass spectrometry data, identify the drug type, calculate the concentration of the target drug molecule in the blood, and generate an analysis report.

[0004] Optionally, in step S1, the microfluidic technology is implemented using a microfluidic chip; The solid-phase extraction utilizes the specific adsorption structure on the microfluidic chip to retain the interfering components and adsorb the target drug molecules. The interfering components are proteins and lipids.

[0005] Optionally, in step S2, the biosensor is a paper-based biosensor, which captures the target drug molecule through specific recognition sites modified on its surface and then converts the target drug molecule into gaseous ions.

[0006] Optionally, in step S2, the paper-based biosensor converts the captured target drug molecules into gaseous ions via electrospraying or thermally assisted evaporation.

[0007] Optionally, in step S3, the chromatography-mass spectrometry technique is gas chromatography-mass spectrometry or liquid chromatography-mass spectrometry. The corresponding coupling technique is selected according to the physicochemical properties of the target drug molecule to separate and detect the concentrate or the gas phase ions.

[0008] Optionally, when using the gas chromatography-mass spectrometry (GC-MS) technique, the concentrated liquid is allowed to flow through the biosensor, and the gaseous ions converted by the biosensor are carried by a carrier gas into the chromatographic column of the gas chromatography unit. After the drug components corresponding to the gaseous ions are separated by the chromatographic column, the separated drug components are delivered to the ionization chamber of the mass spectrometry unit for further ionization and mass-to-charge ratio analysis.

[0009] Optionally, when using the liquid chromatography-mass spectrometry (LC-MS) technique, the concentrate is directly introduced into the liquid chromatography unit. The concentrate is then delivered to the chromatographic column of the liquid chromatography unit via the high-pressure delivery system of the liquid chromatography unit. After separating the drug component corresponding to the target drug molecule from the interfering substances in the concentrate, the separated drug component is delivered to the composite ionization source interface. After ionization treatment, it is delivered to the mass spectrometry unit via a capillary tube for mass-to-charge ratio analysis.

[0010] Optionally, the composite ionization source interface includes an electrospray ionization region and an atmospheric pressure chemical ionization region. Depending on the polarity and molecular weight of the target drug molecule, ionization can be performed in the electrospray ionization region using high voltage or in the atmospheric pressure chemical ionization region using a coronal discharge needle and heating-assisted ionization.

[0011] Optionally, in step S4, when the machine learning model analyzes the mass spectrometry data, it first performs noise reduction processing on the mass spectrometry data, then extracts feature peaks, compares the feature peaks with a standard drug spectrum database, and then identifies the drug type.

[0012] The method provided in this application has the following beneficial effects: The blood drug concentration analysis method of this application utilizes microfluidic technology combined with solid-phase extraction to pretreat blood samples. This pretreatment effectively removes interfering components such as proteins and lipids from the blood while efficiently concentrating the target drug molecules. This provides analytes with higher purity and more suitable concentrations for subsequent detection, reducing the impact of interfering components on the detection results. Different processing paths are selected based on the physicochemical properties of the target drug molecules. If the sample flows through a paper-based biosensor, its surface-modified specific recognition sites can accurately capture the target drug molecules, which are then efficiently converted into gaseous ions via electrospray ionization or thermally assisted evaporation, meeting the detection requirements of gas chromatography-mass spectrometry (GC-MS). If the sample is directly introduced into a chromatographic unit, it matches the workflow characteristics of liquid chromatography-mass spectrometry (LC-MS). These two paths flexibly adapt to drugs with different physicochemical properties, expanding the applicability of the method. Detection using either GC-MS or LC-MS allows for effective separation of the target drug's components, further eliminating residual interference and ensuring the specificity and accuracy of the detection process. When using machine learning models to analyze mass spectrometry data, noise reduction is performed before extracting characteristic peaks. By comparing with a standard drug spectrum database, drug types can be accurately identified and concentrations calculated, reducing errors from manual analysis. The overall method improves flexibility and applicability while ensuring detection accuracy, and can stably generate reliable analysis reports, meeting the practical needs of clinical blood drug concentration testing. Attached Figure Description

[0013] Figure 1 A schematic flowchart of a blood drug concentration analysis method based on mass spectrometry detection is provided for an embodiment of this application; Figure 2 A schematic diagram of a gas chromatography-mass spectrometry (GC-MS) module provided in this application embodiment; Figure 3 A schematic diagram of a gas chromatography-mass spectrometry (GC-MS) module provided in this application embodiment; Figure 4 This is a schematic diagram of a mass spectrometry pattern provided in an embodiment of this application.

[0014] Figure reference numerals: A1, Gas chromatography unit; A2, Mass spectrometry unit; 201, Chromatographic column; 202, Ionization chamber; 203, Heating jacket; 204, Insulation layer; 205, Analysis chamber; 301, HPLC injector; 302, Nebulizer; 303, Electrospray ionization zone; 304, Atmospheric pressure chemical ionization zone; 305, Capillary; 306, Drying gas system; 307, Coronal discharge needle. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this application clearer, specific embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely for explaining this application and not for limiting it. It should also be noted that, for ease of description, only the parts relevant to this application are shown in the drawings, not all of them. Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe operations (or steps) as sequential processes, many of these operations can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but may also have additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0016] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0017] See Figures 1 to 4 This application provides a method for analyzing blood drug concentrations based on mass spectrometry, comprising the following steps: S1. Pretreatment of blood samples using microfluidic technology combined with solid-phase extraction to obtain a concentrate, removing interfering components from the blood samples and concentrating the target drug molecules; S2. Based on the physicochemical properties of the target drug molecule, choose to either allow the concentrate to flow through the biosensor or directly introduce the concentrate into the chromatographic unit; if it flows through the biosensor, the target drug molecule in the liquid phase is converted into gaseous ions by the biosensor. S3. The gaseous ions converted by the biosensor or the concentrated solution directly introduced into the chromatographic unit are detected by chromatography-mass spectrometry to obtain mass spectrometry data. S4. Use machine learning models to analyze mass spectrometry data, identify drug types, calculate the concentration of target drug molecules in the blood, and generate an analysis report.

[0018] Specifically, blood samples contain a variety of components that may affect the test results. It is necessary to first remove interference and concentrate the target drug through pretreatment, then select an appropriate subsequent processing path according to the characteristics of the drug itself, and finally obtain the analysis results through detection and analysis, forming a complete analysis process.

[0019] In practice, the blood sample to be tested is first acquired and pretreated using microfluidic technology combined with solid-phase extraction. During pretreatment, microfluidic technology precisely controls the flow and reaction process of the blood sample, while solid-phase extraction uses specific adsorption materials to retain interfering components in the blood sample and selectively adsorbs target drug molecules. After this treatment, a concentrated solution is obtained, which not only removes interfering components that affect detection but also increases the relative content of target drug molecules in the concentrated solution, providing a more suitable analyte for subsequent detection.

[0020] Subsequently, the appropriate processing path is selected based on the physicochemical properties of the target drug molecule. If the characteristics of the target drug molecule are more suitable for conversion into gaseous ions before detection, the pretreated concentrate is passed through the biosensor. Upon contact with the concentrate, the biosensor converts the liquid-phase target drug molecule into gaseous ions. If the characteristics of the target drug molecule are more suitable for direct chromatographic separation, the pretreated concentrate is directly introduced into the chromatographic unit. The selection of these two paths can adapt to different types of target drug molecules, avoiding the limitations of a single processing method for the detection of some drugs.

[0021] Next, the gaseous ions obtained by the biosensor, or the concentrated solution directly introduced into the chromatographic unit, are detected using chromatography-mass spectrometry (GC-MS). GC-MS is an analytical technique that combines the separation capabilities of chromatography with the detection advantages of mass spectrometry, enabling the separation and qualitative and quantitative analysis of target components. GC-MS first separates the components in the mixture using the chromatographic component, then detects the separated target drug-related components using the mass spectrometric component, ultimately obtaining mass spectrometric data that reflects the drug's components and concentration information. This process further eliminates residual interference, ensuring the specificity of the detection.

[0022] Finally, a machine learning model is used to analyze the acquired mass spectrometry data. The machine learning model identifies and processes the relevant information in the mass spectrometry data, first distinguishing the features related to the target drug, then comparing it with known drug information to accurately identify the drug type. At the same time, based on information such as signal intensity in the mass spectrometry data, the concentration of the target drug molecule in the blood is calculated, and finally an analysis report containing key information such as drug type and concentration is generated.

[0023] Through the above process, this method can effectively remove interfering components in blood samples, accurately capture and detect target drug molecules, and is adaptable to drugs with different physicochemical properties, thus improving the applicability of the analysis. Utilizing machine learning models to interpret mass spectrometry data reduces the subjectivity and error of human analysis, improving the accuracy and consistency of analytical results. The overall process is stable and reliable, meeting the practical application needs of blood drug concentration analysis.

[0024] In some implementations, in S1, microfluidics is implemented using a microfluidic chip; Solid-phase extraction uses specific adsorption structures on microfluidic chips to retain interfering components and adsorb target drug molecules. The interfering components are proteins and lipids.

[0025] Microfluidic chips are miniature devices that integrate microchannels, reaction regions, and other structures. Their channel sizes are adapted to the processing needs of small blood samples, enabling precise control of parameters such as sample flow rate and reaction time, making the implementation of microfluidic technology more stable and operable. Specific adsorption structures are specially designed functional regions on microfluidic chips. Their surfaces are modified with materials or groups that have affinity for target drug molecules and can interact with specific interfering components, enabling the selective separation of different components.

[0026] During pretreatment, after the blood sample is introduced into the microfluidic chip, it flows orderly within the chip's tiny channels. When the blood sample flows through the specific adsorption structure, the structure retains proteins and lipids in the blood sample through physical adsorption or chemical binding. These two components are common interfering components in blood that significantly affect drug concentration detection; their molecular structures have strong interactions with the modification materials of the specific adsorption structure, thus enabling efficient retention. Simultaneously, the specific adsorption structure has a specific affinity for the target drug molecules, selectively adsorbing the target drug molecules while retaining interfering components, preventing the loss of the target drug.

[0027] This design allows microfluidics to precisely manipulate samples using the miniaturized structure of microfluidic chips, while solid-phase extraction relies on the specific adsorption structures on the chip to achieve targeted component separation. The combination of the two is more tightly integrated and efficient. Targeted retention of proteins and lipids makes the removal of interfering components more precise, reduces the influence of irrelevant components on the adsorption process, further improves the purity of the final concentrate, and ensures a more stable relative content of target drug molecules in the concentrate, laying a more reliable foundation for subsequent transformation, detection, and analysis steps.

[0028] In some embodiments, in S2, the biosensor is a paper-based biosensor. After capturing the target drug molecule through specific recognition sites modified on its surface, the paper-based biosensor converts the target drug molecule into gaseous ions.

[0029] In practical applications, the concentrate is subjected to broad-spectrum preliminary screening to obtain a list of candidate drugs, and then the downstream fine analysis path is adaptively selected according to the physicochemical properties of each candidate drug. a) Broad-spectrum initial screening: Take 1–2 µL of concentrate and perform a full scan of rapid gradient LC-MS / MS (m / z 50–1000, cycle 2 min) to obtain a high-resolution first-order spectrum; compare in real time with a pre-set general screening model to output the candidate drugs with the probability Top-K (K≤5) and their confidence level; b) Path Decision: The system has a built-in drug-path mapping table that automatically assigns the optimal analytical path based on the volatility, polarity, and thermal stability of candidate drugs. If all candidate drugs are vaporizable and thermally stable, the paper-based biosensor-GC-MS channel will be used; if at least one candidate drug is difficult to vaporize or thermally unstable, the direct LC-MS channel will be used. If the candidate list contains two categories, then dual channels are opened in parallel to detect the same sample separately, and the results are finally merged. c) The paper-based biosensor only operates when the GC-MS pathway is triggered. Its surface recognition sites employ a broad-spectrum-specific combination strategy: first, common sedative drugs such as benzodiazepines and barbiturates are captured using a molecularly imprinted polymer (MIP) layer, and then antidepressant / antiepileptic drugs are captured using aptamer bands, ensuring that all candidates in the initial screening list can be captured; after capture, GC-MS is performed immediately in S3 for precise quantification. d) The spectra generated during the initial screening stage are synchronously written into a temporary database for continuous updating of the general screening model, enabling self-learning.

[0030] Optionally, the specific recognition sites of the surface modification adopt three complementary schemes, which are optimally combined according to the differences in molecular weight, polarity and functional groups of the target drug.

[0031] In one embodiment, the paper substrate is selected as Whatmanchromatography 1Chr (pore size 11μm, thickness 0.18mm), and a high hydroxyl density surface is obtained through the following pretreatment to provide anchor points for subsequent grafting: Oxygen plasma 100W, 5min, introduced –OH density ≈ 1.2 × 10⁻⁶ 15 cm -2 ; 2wt% 3-aminopropyltriethoxysilane (APTES) / 95% ethanol-5% H2O, pH 4.5, 60°C for 2h, yield –NH2 terminator; 0.5% glutaraldehyde PBS at room temperature for 1 hour generates a –CHO intermediate layer, which can be used to covalently couple proteins or aptamers.

[0032] Optionally, the MIP modification process for the small molecule antiepileptic drug carbamazepine (CBZ, 236Da) is as follows: Template molecule: CBZ 0.2 mmol; functional monomer: methacrylic acid (MAA) 0.8 mmol; crosslinking agent: ethylene glycol dimethacrylate (EGDMA) 4 mmol; initiator: AIBN 1 wt%; solvent: acetonitrile 10 mL.

[0033] Prepolymerization: Degas by ultrasonication for 10 min, then stand at 0°C for 2 h to form a pre-hydrogen bond complex.

[0034] Graft polymerization: The paper substrate was placed in the above solution and protected with nitrogen at 60°C for 12 hours to form a 50–80 nm thick MIP layer on the surface (measured by AFM).

[0035] Template elution: Soxhlet extraction with methanol-acetic acid (9:1, v / v) for 8 h until CBZ was undetectable by HPLC; cavity size ≈ 1.2 × 10⁻⁶. 12 sitescm -2 .

[0036] Blocking: 1% BSAPBS at room temperature for 30 min to block nonspecific adsorption.

[0037] Affinity verification: Kd = 6.3 × 10⁻⁶ was obtained by surface plasmon resonance (SPR). -9 M; cross-reactivity with the structural analogue oxcarbazepine is 1.8%.

[0038] Paper-based biosensors are functional devices based on paper materials. The paper substrate has excellent hydrophilicity and permeability, allowing concentrated solutions to penetrate and fully contact the material. Furthermore, the material is lightweight and cost-effective, making it suitable for micro-volume processing of blood samples. The surface is modified to form multiple specific recognition sites. These recognition sites are designed based on the structural characteristics of common target drug molecules. The surface-modified active groups can specifically bind to the target drug molecules, exhibiting affinity only for them and avoiding binding to other components.

[0039] During the S2 processing, when the path through the biosensor is selected based on the physicochemical properties of the target drug molecule, the pretreated concentrate is dropwise added to the sample introduction area of ​​the paper-based biosensor. The concentrate slowly diffuses under capillary action on the paper substrate, flowing through the functional region modified with specific recognition sites. At this point, the recognition sites rapidly capture the target drug molecule flowing through the region, fixing it to the paper surface, while other residual impurities in the concentrate continue to diffuse, thus achieving secondary enrichment and purification of the target drug molecule. After capture, the paper-based biosensor initiates a conversion process, transforming the target drug molecule, originally in a liquid phase, into gaseous ions, providing a suitable analytical form for subsequent separation and detection by gas chromatography-mass spectrometry (GC-MS).

[0040] The application of this paper-based biosensor, with its precise capture capability at specific recognition sites, further improves the separation accuracy of target drug molecules and reduces the interference of residual impurities on subsequent transformation and detection. At the same time, the characteristics of the paper substrate make the operation simpler, and the transformation process can be seamlessly connected with the capture process, ensuring that the target drug molecules are not easily lost during the transformation process, providing a reliable guarantee for obtaining accurate mass spectrometry data in the future.

[0041] Additionally, a dual recognition layer can be constructed on the surface of the specific recognition sites of paper-based biosensors: the first layer consists of broad-spectrum affinity groups that initially bind to target drug molecules and structurally similar compounds; the second layer is a spatial structure matching unit that allows only target drug molecules to remain through spatial fit, while structurally similar interfering compounds are repelled due to spatial mismatch. Simultaneously, an ultrathin anti-fouling coating can be applied to the paper surface. This coating does not affect the binding of target molecules to the recognition sites but reduces the non-specific adsorption of large molecules such as proteins and lipids in the blood. This design allows the capture process to retain targeting while eliminating cross-interference through dual screening. The anti-fouling coating further reduces background interference, resulting in higher purity of the gaseous ions in subsequent conversions, providing more accurate analytes for the detection stage.

[0042] In some embodiments, in S2, the paper-based biosensor converts the captured target drug molecules into gaseous ions via electrospraying or thermally assisted evaporation.

[0043] Electrospraying is a technique that uses high voltage to form charged droplets from a liquid. It is suitable for target drug molecules with strong polarity and relatively stable molecular structures. Thermal assisted evaporation, on the other hand, uses gentle heating to provide energy and causes liquid molecules to transform into a gaseous state. It is more suitable for target drug molecules with better heat resistance. The two methods can be flexibly selected according to actual analytical needs.

[0044] Once a paper-based biosensor captures a target drug molecule through a specific recognition site, it initiates a corresponding conversion process based on the physicochemical properties of the drug molecule. If an electrospray method is used, the sample introduction area of ​​the paper-based biosensor is connected to a high-voltage generator, creating a strong electric field in the area where the target drug molecule is attached. The paper surface containing the captured drug molecule generates tiny droplets due to the electric field. These droplets are continuously atomized and broken down under the force of the electric field, ultimately freeing the target drug molecule from its liquid phase and converting it into charged ions.

[0045] If a heat-assisted evaporation method is used, the functional area of ​​the paper-based biosensor will activate the built-in heating module to release uniform and stable heat. The heat is transferred through the paper substrate to the attachment site of the target drug molecules, providing them with sufficient energy without damaging the drug molecule structure, so that the liquid-phase drug molecules gradually evaporate into a gaseous state, and then form gaseous ions.

[0046] Both of these conversion methods are closely integrated with the capture function of paper-based biosensors. During the conversion process, the integrity of the target drug molecules can be preserved to the greatest extent. The generated gaseous ions are highly pure and stable, and can be directly introduced into the separation and detection stage of subsequent gas chromatography-mass spectrometry, effectively reducing the error in the conversion process and further ensuring the accuracy of the overall analytical results.

[0047] Alternatively, a miniature ion pre-regulation module can be integrated into the ion outlet of the paper-based biosensor. This module incorporates a weak electric field generating unit and an inert gas auxiliary channel, without altering the original conversion process, and only functions during ion output. When gaseous ions generated by electrospray or thermal assisted evaporation leave the paper-based biosensor, they first enter the pre-regulation module: the weak electric field generating unit produces a uniform low-voltage electric field, causing charged ions to align orderly along the electric field direction, breaking up any potential ion aggregates; simultaneously, a trace amount of clean inert gas is introduced into the inert gas auxiliary channel, mixing with the gaseous ions to further maintain their dispersion and prevent re-aggregation. The regulated gaseous ions have a uniform charge distribution and are individually dispersed, allowing for more stable transport with the carrier gas. This results in sharper peaks when passing through the chromatographic column, higher ion capture efficiency during mass spectrometry detection, and effectively reduced detection errors caused by uneven ion states.

[0048] In some embodiments, in S3, the chromatography-mass spectrometry technique is gas chromatography-mass spectrometry or liquid chromatography-mass spectrometry. The appropriate coupling technique is selected according to the physicochemical properties of the target drug molecule to separate and detect the concentrate or gas phase ions.

[0049] Gas chromatography-mass spectrometry (GC-MS) is an analytical technique that combines the separation capabilities of gas chromatography with the detection capabilities of mass spectrometry. Its core feature is that it relies on a carrier gas to propel the sample components to separate within the chromatographic column, making it more suitable for target drug molecules with a certain degree of volatility and good thermal stability. Liquid chromatography-mass spectrometry (LC-MS), on the other hand, combines the separation function of high-performance liquid chromatography (HPLC) with the detection function of mass spectrometry. It uses a high-pressure liquid delivery system to propel the sample to flow and separate within the chromatographic column, making it more suitable for target drug molecules with strong polarity, poor volatility, or insufficient thermal stability.

[0050] In actual testing, it is necessary to first determine the physicochemical properties of the target drug molecule, including its polarity, volatility, and thermal stability, and then select the appropriate coupling technology accordingly. If the target drug molecule has suitable volatility and is not easily structurally damaged during heating or carrier gas delivery, i.e., it has good thermal stability, then gas chromatography-mass spectrometry (GC-MS) is selected. This technology can efficiently carry sample components with a carrier gas to achieve separation, and then obtain accurate data through mass spectrometry. If the target drug molecule is highly polar and difficult to separate effectively by carrier gas, or has poor thermal stability and is prone to structural changes after heating, then liquid chromatography-mass spectrometry (LC-MS) is selected. Its high-pressure delivery system can promote sample separation under gentle conditions, avoiding damage to the drug molecule structure.

[0051] This approach, which selects coupling technologies specifically based on the physicochemical properties of the target drug molecules, allows the separation and detection process to better align with the characteristics of the drug itself. It avoids problems such as poor separation results, damage to drug molecule structure, or weak detection signals caused by improper technology selection. This ensures that whether the concentrate is directly introduced or the gaseous ions are converted by a biosensor, efficient separation and accurate detection can be achieved through appropriate coupling technologies, providing high-quality mass spectrometry data for subsequent analysis steps.

[0052] In some embodiments, when gas chromatography-mass spectrometry is used, the concentrate is passed through a biosensor, and the gaseous ions converted by the biosensor are carried by a carrier gas into the chromatographic column of the gas chromatography unit. After the drug components corresponding to the gaseous ions are separated by the chromatographic column, the separated drug components are delivered to the ionization chamber of the mass spectrometry unit for further ionization and mass-to-charge ratio analysis.

[0053] The carrier gas is a chemically stable gas that does not react with drug components. Its core function is to provide stable transport power for gaseous ions, ensuring the orderly flow of sample components within the detection system. The chromatographic column of the gas chromatography unit is the core component for component separation. It is filled with a specific stationary phase material. Different substances have different adsorption and desorption strengths with the stationary phase, which is the key to achieving separation. The ionization chamber of the mass spectrometry unit is a functional area that provides a further ionization environment for substances, allowing the separated drug components to form more stable charged ions, which facilitates subsequent detection.

[0054] When gas chromatography-mass spectrometry (GC-MS) is selected based on the physicochemical properties of the target drug molecule, the processing path for the concentrated liquid to flow through the biosensor must be simultaneously chosen. This path selection should be compatible with the operating characteristics of GC-MS, providing a morphologically suitable analyte for the technique. After the biosensor converts the target drug molecule in the liquid phase into gaseous ions, the carrier gas is continuously and steadily introduced into the system. The gaseous ions are uniformly entrained by the carrier gas and enter the chromatographic column of the gas chromatography unit along the preset channel.

[0055] During the flow of gaseous ions within the chromatographic column, different drug components interact with the stationary phase. Due to differences in molecular structure, polarity, and other properties, the adsorption strength and desorption rate of each component differ, resulting in different elution orders along the column. This leads to the complete separation of drug components from residual trace interfering substances. The separated individual drug components are then propelled by the carrier gas to the ionization chamber of the mass spectrometry unit. Under the specific environment of the ionization chamber, the drug components undergo further ionization, forming stable charged ions. These charged ions then enter the detection region of the mass spectrometry unit. By detecting the mass-to-charge ratio of the ions, detection data that accurately reflects the characteristics of the drug components is obtained, providing a reliable basis for subsequent analysis steps.

[0056] Throughout the process, the conversion function of the biosensor matches the separation and detection requirements of gas chromatography-mass spectrometry (GC-MS), the stable transport of the carrier gas ensures that the sample components are not lost or degraded, the separation function of the chromatographic column further eliminates interference, and the secondary ionization in the ionization chamber improves the stability of ions, thus effectively ensuring the specificity and accuracy of the detection results.

[0057] like Figure 2 This is a schematic diagram of a gas chromatography-mass spectrometry (GC-MS) module provided in an embodiment of this application. Figure 2 As shown, the system consists of a gas chromatography (GC) unit A1 and a mass spectrometry (MS) unit A2 connected via a dedicated interface. Its core components include a chromatographic column 201, an ionization chamber 202, a heating mantle 203, a heat insulation layer 204, and an analysis chamber 205. The end of the chromatographic column 201 must extend 1-2 mm into the ionization chamber to ensure efficient entry of the components into the ionization region. The GC unit A1 uses a carrier gas to carry volatile drug components ionized by a biosensor, achieving efficient separation through the chromatographic column 201 based on the distribution differences between the compound and the stationary phase. The separated components are then transported to the MS unit A2 via the interface. The heating mantle 203 and the heat insulation layer 204 maintain the high-temperature environment of the ionization chamber 202, preventing drug component condensation and ensuring efficient gas phase transport. The MS unit A2 further ionizes the components entering the ionization chamber 202 using methods such as electron bombardment, and then performs mass-to-charge ratio analysis in the analysis chamber 205 to achieve qualitative and quantitative drug analysis.

[0058] In some implementations, when using liquid chromatography-mass spectrometry (LC-MS), the concentrate is directly introduced into the liquid chromatography unit. The concentrate is then delivered to the chromatographic column of the liquid chromatography unit via the high-pressure delivery system of the liquid chromatography unit. After separating the drug component corresponding to the target drug molecule from the interfering substances in the concentrate, the separated drug component is delivered to the composite ionization source interface. After ionization treatment, it is delivered to the mass spectrometry unit via a capillary tube for mass-to-charge ratio analysis.

[0059] A liquid chromatography unit is a device specifically designed for separating components of liquid samples. Its core components include a high-pressure delivery system and a chromatographic column. The high-pressure delivery system has a stable pressure output capability, enabling it to propel the concentrate through the chromatographic column at a uniform flow rate. The chromatographic column is filled with a specific stationary phase material, and the difference in the interaction strength between different substances and the stationary phase is the core principle for achieving separation. The composite ionization source interface is a component that connects the liquid chromatography unit and the mass spectrometry unit, converting the separated liquid drug components into gaseous ions. The capillary is a thin tubular component with good sealing and transmission performance, used for the precise delivery of ionized ions.

[0060] When liquid chromatography-mass spectrometry (LC-MS) is selected based on the physicochemical properties of the target drug molecule, the pretreated concentrate is directly introduced into the LC unit. This approach is compatible with the working characteristics of the LC technique, eliminating the need for pre-conversion to gaseous ions and allowing direct processing of the liquid concentrate. The high-pressure delivery system of the LC unit is activated, providing a stable pressure that propels the concentrate into the chromatographic column at a uniform and controllable flow rate.

[0061] During the flow of the concentrate inside the chromatographic column, the drug components corresponding to the target drug molecules and residual interfering substances interact with the stationary phase inside the column through adsorption, desorption, and other processes. Due to the differences in molecular structure, polarity, and other physicochemical properties between the drug components and interfering substances, their interaction strengths with the stationary phase differ, resulting in different migration velocities within the chromatographic column. Ultimately, they elute from the column in a specific order, achieving complete separation of the drug components and interfering substances.

[0062] The separated individual drug components are then transported to the composite ionization source interface. Under the specific environmental conditions of the interface, the liquid drug components undergo ionization, transforming into stable gaseous ions. These gaseous ions are then transported through a capillary tube, which prevents ion loss or contamination during transport, ensuring that the ions are delivered to the mass spectrometry unit in a stable state. Once in the mass spectrometry unit, the mass-to-charge ratio of the gaseous ions is detected to obtain accurate detection data reflecting the characteristics of the drug components, providing high-quality basic data for subsequent analysis steps.

[0063] Throughout the process, the stable delivery of the high-pressure infusion system ensured the consistency of the separation process, the targeted separation of the chromatographic column further eliminated interfering components, the composite ionization source interface enabled the efficient conversion of liquid components into gaseous ions, and the precise transmission of the capillary reduced ion loss. The coordination of each link allowed the advantages of liquid chromatography-mass spectrometry to be fully utilized, ensuring the stability of the detection process and the accuracy of the detection results.

[0064] like Figure 3 This is a schematic diagram of a gas chromatography-mass spectrometry (GC-MS) module provided in an embodiment of this application. Figure 3 As shown, a high-performance liquid chromatography (HPLC) unit serves as the core separation mechanism, connected to the mass spectrometry (MS) unit via a combined ionization source interface. Its core components include an HPLC injector 301, an nebulizer 302, an electrospray ionization (ESI) zone 303, an atmospheric pressure chemical ionization (APCI) zone 304, a capillary tube 305, a drying gas system 306, and a coronal discharge needle 307. The HPLC unit delivers pretreated blood samples to the chromatographic column via a high-pressure infusion system, achieving efficient separation of drug components from interfering substances. The separated components enter the MS unit through an interface that integrates both electrospray ionization and atmospheric pressure chemical ionization, allowing for flexible switching based on drug characteristics. After the liquid sample enters through the HPLC injector 301, it is nebulized into tiny droplets by the nebulizer 302. Ionization is achieved in the ESI region 303 by high voltage (suitable for polar, large molecule drugs), or in the APCI region 304 by coronal discharge needle 307 with heating assistance (suitable for non-polar, small molecule drugs). The drying gas system 306 accelerates the evaporation of droplets and promotes the formation of gaseous ions. Finally, the ions are efficiently transported to the MS unit for detection through the capillary 305.

[0065] Continue reading Figure 3 In some embodiments, the composite ionization source interface includes an electrospray ionization region 303 and an atmospheric pressure chemical ionization region 304. Depending on the polarity and molecular weight of the target drug molecule, ionization is performed in the electrospray ionization region 303 by high voltage or in the atmospheric pressure chemical ionization region 304 by a coronal discharge needle and heating-assisted ionization.

[0066] It should be noted that the composite ionization source interface adopts a dual-zone parallel zero-dead-volume switching architecture, which makes real-time ionization mode decisions for the effluent during continuous liquid chromatography separation, ensuring continuous detection of multiple components without data loss.

[0067] The electrospray ionization region 303 is a functional area in the composite ionization source interface that uses high voltage to achieve ionization. Its core is to use an electric field to make the liquid drug components form charged droplets, which are then converted into gaseous ions. The atmospheric pressure chemical ionization region 304 is a region that relies on the coronal discharge needle 307 and heating assistance to achieve ionization. The coronal discharge needle 307 can generate the charge required for ionization, and the heating module can provide mild energy to assist the liquid components in converting into a gaseous state.

[0068] The polarity and molecular weight of the target drug molecule are key characteristics that affect the ionization effect: drug molecules with strong polarity and relatively small molecular weight have polar groups in their molecular structure that are more likely to form charged particles in an electric field, making them suitable for electrospray ionization; while drug molecules with weak polarity and large molecular weight are difficult to charge quickly in an electric field on their own, and require the energy provided by the charge provided by the coronal discharge needle 307 and the energy provided by heating assistance to complete the ionization conversion more efficiently and are less likely to suffer molecular structure damage.

[0069] In the actual ionization process, when the separated drug components are delivered to the composite ionization source interface, the appropriate ionization method is first determined based on the polarity and molecular weight of the target drug molecules. If the drug molecules are highly polar and have a small molecular weight, the electrospray ionization zone 303 is activated. By applying a high voltage to this zone, the liquid drug components entering the zone are atomized into tiny charged droplets under the action of a strong electric field. The droplets gradually evaporate in subsequent processes, eventually transforming into stable gaseous ions.

[0070] If the drug molecules are weakly polar and have a large molecular weight, then atmospheric pressure chemical ionization region 304 is selected. At this time, the coronal discharge needle 307 will generate an electric charge, and the heating module will be activated to provide gentle heat. Under the dual action, the liquid drug components will first gradually evaporate into a gaseous state, and then combine with the charge generated by the coronal discharge needle 307 to form gaseous ions. The whole process can avoid damage to the drug molecule structure due to excessive energy.

[0071] The integrated design of two ionization methods and a composite ionization source interface allows the ionization process to be flexibly adjusted according to the drug characteristics. This ensures efficient ionization of highly polar, small molecular weight drug molecules, while also enabling stable conversion of weakly polar, large molecular weight drug molecules. The generated gaseous ions are highly pure and stable, and can be smoothly transported to the mass spectrometry unit through the capillary, providing a reliable ion source for subsequent mass-to-charge ratio detection and further ensuring the accuracy of the detection results.

[0072] In some implementations, in S4, when the machine learning model analyzes the mass spectrometry data, it first performs noise reduction on the mass spectrometry data, then extracts the characteristic peaks, compares the characteristic peaks with a standard drug spectrum database, and then identifies the drug type.

[0073] Figure 4This is a schematic diagram of a mass spectrometry pattern provided in an embodiment of this application. (See also...) Figure 4 , Figure 4 The diagram only schematically illustrates the positions of characteristic peaks for functional groups such as -C=O (carbonyl) and -CH2- (methylene). The horizontal axis represents the mass-to-charge ratio, and the vertical axis represents the ionic intensity. Units and specific values ​​for these axes are not provided here. In addition to the effective signal corresponding to the target drug molecule, mass spectrometry data may also contain background interference signals generated during detection. These interference signals can affect the identification of effective information, therefore noise reduction processing is necessary. Characteristic peaks are signal peaks in mass spectra that reflect the unique physicochemical properties of drug molecules. The positions and shapes of characteristic peaks differ among different drug molecules, serving as a key basis for drug identification. The standard drug spectrum database stores a large amount of standard characteristic peak information for known drugs; the information contained therein has been verified and can be used as a reliable reference for comparison.

[0074] During the analysis process, the machine learning model first performs noise reduction on the acquired mass spectrometry data. This process specifically identifies and filters out background interference signals in the spectrum. These interference signals do not originate from the target drug molecules but may be generated by factors such as the detection environment and instrument noise. Filtering makes the outline of the effective signal in the spectrum clearer and avoids interference signals affecting subsequent steps.

[0075] After noise reduction, the machine learning model extracts feature peaks. Using pre-defined recognition rules, the model filters out characteristic peaks representing the properties of the target drug molecule from the cleared mass spectra, eliminating meaningless minor fluctuations in the spectra to ensure that the extracted feature peaks accurately reflect the core information of the target drug molecule.

[0076] Subsequently, the machine learning model comprehensively compares the extracted feature peaks with information in a standard drug spectrum database. During the comparison process, key features such as the position and relative intensity of each feature peak are matched one by one. If the extracted feature peaks highly match the standard feature peaks of a known drug in the database, the type of the target drug can be determined.

[0077] Optionally, the machine learning model described in S4 employs a lightweight residual one-dimensional convolution and attention architecture: Mass spectra are one-dimensional signals, eliminating the need for two-dimensional convolution and reducing the number of parameters; residual connections can alleviate gradient vanishing and are suitable for long sequences with more than 16,000 sampling points. A Channel-Attention module is introduced to automatically weight low-abundance feature peaks in the 200-800 m / z range, thereby improving the recall rate for low-concentration drugs (<5 ng / mL). The specific network structure is shown in Table 1.

[0078] Table 1 ResNet1D-Attention Network Configuration (Input: 16384×1 spectrogram)

[0079] In one implementation, the model training data came from 62,000 clinical blood samples accumulated from three central laboratories between 2018 and 2023, covering 87 common psychotropic and neurological drugs (such as valproic acid, carbamazepine, clonazepam, etc.). Each sample was analyzed using reference methods (LC-MS / MS or isotope dilution GC-MS) to obtain the gold standard concentration, and two senior mass spectrometry engineers independently labeled the drug type; inconsistent samples were arbitrated by a third party, ultimately forming a ternary set of spectrum-drug-concentration. To ensure diversity, stratified sampling was employed. Ages 0–100 years, with each 10-year age range accounting for approximately 10%; Matrix: Normal serum, hemolysis, lipemia, and jaundice each account for 70%, 10%, 10%, and 10%, respectively; Concentration distribution: uniformly cover 0.2×LOQ–2×ULOQ (LOQ=0.1ngmL) according to log10. -1 ).

[0080] In one implementation, performance verification uses a five-center external test set, completely independent of the training set, with n=4800. Key metrics are shown in Table 2.

[0081] Table 2 External validation results (4800 cases, 87 drugs)

[0082] The entire analysis process, by first reducing noise and then extracting characteristic peaks, effectively eliminates irrelevant interference, ensuring the accuracy and reliability of the information used for comparison. Standard drug spectrum databases provide authoritative references, making the comparison results more convincing. The automated processing of the machine learning model avoids the subjective errors of manual identification, making the drug identification process more stable and the results more accurate, laying a solid foundation for subsequent calculations of target drug molecule concentrations and the generation of analysis reports.

[0083] This application also provides a blood drug concentration analysis system based on mass spectrometry detection, including: The pretreatment module is used to pretreat blood samples using microfluidic technology combined with solid-phase extraction to obtain a concentrate, remove interfering components from the blood samples and concentrate the target drug molecules; The selection module is used to select whether to allow the concentrate to flow through the biosensor or directly introduce it into the chromatography unit based on the physicochemical properties of the target drug molecule; if it flows through the biosensor, the target drug molecule in the liquid phase is converted into gaseous ions by the biosensor. The detection module is used to detect gaseous ions converted by the biosensor or concentrated solutions directly introduced into the chromatographic unit using chromatography-mass spectrometry (GC-MS) to obtain mass spectrometry data. The analysis module is used to analyze mass spectrometry data using machine learning models, identify drug types, calculate the concentration of target drug molecules in the blood, and generate analysis reports.

[0084] The system embodiments provided in this application have the same or at least partially the same technical features as the above method embodiments, and therefore can achieve the same or at least partially the same technical effects, which will not be repeated here.

[0085] The above description is merely a preferred embodiment and the technical principles employed in this application. This application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions that can be made by those skilled in the art will not depart from the scope of protection of this application. Therefore, although this application has been described in detail through the above embodiments, this application is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of this application.

Claims

1. A method for analyzing blood drug concentrations based on mass spectrometry, characterized in that, Includes the following steps: S1. Pre-treat the blood sample using microfluidic technology combined with solid-phase extraction to obtain a concentrate, remove interfering components from the blood sample and concentrate the target drug molecules; S2. Based on the physicochemical properties of the target drug molecule, select whether to allow the concentrate to flow through the biosensor or to directly introduce the concentrate into the chromatographic unit; if it flows through the biosensor, the target drug molecule in the liquid phase is converted into gaseous ions by the biosensor. S3. The gaseous ions converted by the biosensor or the concentrated solution directly introduced into the chromatographic unit are detected by chromatography-mass spectrometry to obtain mass spectrometry data. S4. Utilize a machine learning model to analyze the mass spectrometry data, identify the drug type, calculate the concentration of the target drug molecule in the blood, and generate an analysis report.

2. The method according to claim 1, characterized in that, In step S1, the microfluidic technology is implemented through a microfluidic chip; The solid-phase extraction utilizes the specific adsorption structure on the microfluidic chip to retain the interfering components and adsorb the target drug molecules. The interfering components are proteins and lipids.

3. The method according to claim 1 or 2, characterized in that, In step S2, the biosensor is a paper-based biosensor. After capturing the target drug molecule through specific recognition sites modified on its surface, the paper-based biosensor converts the target drug molecule into the gaseous ions.

4. The method according to claim 3, characterized in that, In step S2, the paper-based biosensor converts the captured target drug molecules into gaseous ions through electrospraying or thermally assisted evaporation.

5. The method according to claim 1, characterized in that, In step S3, the chromatography-mass spectrometry technique is either gas chromatography-mass spectrometry or liquid chromatography-mass spectrometry. The corresponding coupling technique is selected according to the physicochemical properties of the target drug molecule to separate and detect the concentrate or the gas phase ions.

6. The method according to claim 5, characterized in that, When the gas chromatography-mass spectrometry (GC-MS) technique is used, the concentrated liquid is selected to flow through the biosensor, and the gas phase ions converted by the biosensor are carried by the carrier gas into the chromatographic column of the gas chromatography unit. After the drug components corresponding to the gas phase ions are separated by the chromatographic column, the separated drug components are delivered to the ionization chamber of the mass spectrometry unit for further ionization and mass-to-charge ratio analysis.

7. The method according to claim 5, characterized in that, When using the liquid chromatography-mass spectrometry (LC-MS) technique, the concentrate is directly introduced into the liquid chromatography unit. The concentrate is then delivered to the chromatographic column of the liquid chromatography unit via the high-pressure delivery system of the liquid chromatography unit. After separating the drug component corresponding to the target drug molecule from the interfering substances in the concentrate, the separated drug component is delivered to the composite ionization source interface. After ionization treatment, it is delivered to the mass spectrometry unit via a capillary tube for mass-to-charge ratio analysis.

8. The method according to claim 7, characterized in that, The composite ionization source interface includes an electrospray ionization region and an atmospheric pressure chemical ionization region. Depending on the polarity and molecular weight of the target drug molecule, ionization is performed in the electrospray ionization region using high voltage or in the atmospheric pressure chemical ionization region using a coronal discharge needle and heating-assisted ionization.

9. The method according to claim 1, characterized in that, In step S4, when the machine learning model analyzes the mass spectrometry data, it first performs noise reduction on the mass spectrometry data, then extracts feature peaks, compares the feature peaks with a standard drug spectrum database, and then identifies the drug type.

Images