Mycobacterium tuberculosis gene rapid screening service system oriented to health physical examination center

Abstract

The invention discloses a mycobacterium tuberculosis gene rapid screening service system oriented to a health physical examination center, and relates to the technical field of health services, and the mycobacterium tuberculosis gene rapid screening service system is characterized in that a sample is subjected to automatic uniform mixing and transfer to complete standardized pretreatment; the mechanical arm accurately loads the preprocessed sample to each independent channel of the micro-fluidic chip; bacterial lysis, nucleic acid adsorption, washing and elution are automatically completed in the sealed chip; the channels of the chip are subjected to thermal cycle amplification in parallel and generate target specific fluorescence signals; the system collects fluorescence intensity data of each channel in real time and automatically completes filtering and normalization processing; the trained deep learning model automatically analyzes a fluorescence curve and classifies results; and the system automatically generates an encrypted report and distributes the encrypted report to a subject and a health management center. By integrating closed parallel processing of the multi-channel micro-fluidic chip and fluorescence curve automatic analysis based on deep learning, automation and intelligentization of the whole process from sample loading to report generation can be realized.

Description

Technical Field

[0001] This invention belongs to the field of health service technology, and in particular relates to a rapid tuberculosis gene screening service system for health check-up centers. Background Technology

[0002] Tuberculosis genes refer to specific nucleic acid sequences (including DNA and RNA) present in the genome of Mycobacterium tuberculosis. These sequences carry genetic information related to the survival, proliferation, and pathogenicity of the pathogen. Among them, some highly conserved and specific gene fragments, such as the IS6110 insertion sequence, are often used as marker targets for molecular diagnosis due to their stable copy number and species specificity within the Mycobacterium tuberculosis complex. By detecting and identifying these characteristic gene fragments, it is possible to confirm whether a sample contains the genetic material of Mycobacterium tuberculosis, thus providing a direct basis for etiological diagnosis.

[0003] In applications involving annual centralized health checkups for large enterprises, traditional tuberculosis gene screening methods suffer from a mismatch between the high throughput and rapid turnaround required for large-scale population screening and the fragmented and highly manual nature of existing molecular detection technologies. Current technologies typically rely on sending samples in batches to a central laboratory, involving opening the plates, manual separation, nucleic acid extraction in multi-well plates, and then setting up PCR reactions. This series of open operations becomes a bottleneck in terms of speed when sample volumes surge, and the frequent manual intervention and plate opening introduce risks of sample confusion and cross-contamination. Furthermore, the interpretation of surged fluorescence amplification curves often relies on experienced technicians to review each result individually and set thresholds. This process is time-consuming in large-scale screening and subject to subjective judgment differences, becoming a key factor restricting the timeliness and consistency of reports. Therefore, the following solutions are proposed to address these issues. Summary of the Invention

[0004] The purpose of this invention is to provide a rapid tuberculosis gene screening service system for health check-up centers. By integrating multi-channel microfluidic chip closed parallel processing and deep learning-based automatic fluorescence curve analysis, it can achieve full automation and intelligence from sample loading to report generation. This solves the problems of limited detection throughput, easy contamination of the process, and difficulty in ensuring result consistency caused by reliance on multiple manual operations and manual result interpretation in the prior art.

[0005] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution:

[0006] This invention is a rapid tuberculosis gene screening service system for health check-up centers, comprising:

[0007] The sample collection and identification module is used to receive and uniquely identify biological samples from the test subject.

[0008] An integrated microfluidic chip has multiple independent and sealed detection channels for automated nucleic acid extraction, purification, and amplification reactions of input samples;

[0009] A temperature control and amplification module, coupled to the microfluidic chip, is used to provide precise temperature cycling control for the nucleic acid amplification reaction within the chip.

[0010] A fluorescence detection module is used to acquire fluorescence signals from each detection channel in real time during the amplification reaction.

[0011] The data analysis module integrates a pre-trained deep learning model to receive and analyze the fluorescence signal, and output classification results corresponding to the presence or absence of Mycobacterium tuberculosis target genes.

[0012] The system also includes a report generation and distribution module, which automatically generates screening reports based on the classification results and distributes the reports to the relevant examinees and the health check-up center information system.

[0013] Furthermore, the sample collection and identification module includes a sampling container with a radio frequency identification tag, and a read / write device for reading and writing the tag information, which is associated with the examinee's unique identifier in the health check-up center database.

[0014] Furthermore, the integrated microfluidic chip has a plate structure and includes multiple parallel channels corresponding to the number of standard multiwell plates. Each channel sequentially integrates a sample inlet, a lysis unit, a nucleic acid purification unit based on magnetic particles, an amplification reaction chamber, and an optical detection window. Each unit is connected through a microfluidic channel and the fluid flow is controlled by a built-in microvalve.

[0015] Furthermore, the deep learning model integrated in the data analysis module is a convolutional neural network model, which is configured to receive preprocessed sequence data of fluorescence intensity changing over time, process it through a network structure containing convolutional layers and fully connected layers, and finally output a classification result containing the probabilities of positive, negative and invalid states.

[0016] Furthermore, it also includes an automated sample loading device, which includes a robotic arm and a vision positioning system for precisely transferring samples from the sample acquisition and identification module to a predetermined volume and injecting them into the sample inlet of a designated detection channel of the microfluidic chip.

[0017] Furthermore, the lysis unit, nucleic acid purification unit, and connected fluid channels of the integrated microfluidic chip are pre-stored or can be sequentially injected with lysis reagents, binding reagents, washing reagents, and elution reagents. The purification unit controls the movement of magnetic particles through an external magnetic field generator to complete the separation and purification of nucleic acids within a sealed channel.

[0018] Furthermore, the fluorescence detection module includes a light source and a photoelectric sensor corresponding to each detection channel, used to excite and collect fluorescence signals at specific stages of each temperature cycle of the amplification reaction, and convert analog signals into digital signal sequences.

[0019] Furthermore, the data analysis module also includes a signal preprocessing unit, which performs smoothing filtering, background fluorescence subtraction, and signal normalization on the digital signal sequence acquired by the fluorescence detection module to generate standard fluorescence curve data that conforms to the input format of the deep learning model.

[0020] Furthermore, the report generation and distribution module is configured to: determine the final screening conclusion based on the classification results and preset decision rules, automatically retrieve the examinee's information from the database, generate a structured report with a digital signature, and send the report to the examinee's email address and the confidential information platform of the health check-up center through an encrypted communication link.

[0021] A rapid tuberculosis gene screening method for health check-up centers, the method being implemented based on any of the above-described systems, comprising the following steps:

[0022] Samples are collected using labeled containers, and the container labels are associated with the subject's information;

[0023] Samples are automatically loaded into an independent detection channel integrated with a microfluidic chip;

[0024] Within the sealed channel of the chip, the purification steps of sample lysis, nucleic acid binding to magnetic particles, impurity washing, and nucleic acid elution are automatically and sequentially completed.

[0025] In the amplification reaction chamber of the chip, polymerase chain reaction targeting Mycobacterium tuberculosis specific target genes is completed, and fluorescence signals are monitored in real time;

[0026] The acquired real-time fluorescence signals are preprocessed to generate standardized amplification curve data;

[0027] The standardized amplification curve data is input into a pre-trained deep learning model to obtain the classification probability distribution;

[0028] Based on the comparison between the classification probability distribution and the preset threshold, the screening result is automatically determined to be positive, negative, or invalid.

[0029] Based on the results of the automatic assessment, an electronic screening report is generated and automatically distributed to the relevant examinees and health check-up centers.

[0030] The present invention has the following beneficial effects:

[0031] 1. This invention achieves simultaneous processing of multiple samples through the design and automated control of a multi-channel microfluidic chip. This technical approach enables sample lysis, nucleic acid purification, and amplification to be completed continuously in a microscale space, reducing the transfer and waiting time between processing steps. It helps to increase the number of samples that can be tested per unit time, adapting to the need to process large-scale samples in a short time. The entire process is executed sequentially under the drive of a preset program, reducing fluctuations caused by differences in manual operation.

[0032] 2. The fully enclosed microfluidic processing of this invention strictly confines samples and reagents within independent microchannels; this physical isolation method limits the possibility of unintended contact between different samples and between samples and the external environment; it can provide a controlled fluid environment for the detection process, which helps to maintain the stability of reaction conditions in each step; from sample loading to final signal acquisition, no open operation is required, thus supporting the consistency and repeatability of detection results.

[0033] 3. This invention combines real-time fluorescence signal acquisition with a data-driven analysis model, enhancing the ability to analyze complex detection signals. By learning the characteristics of a large amount of historical detection data, the model can perform pattern recognition and classification of fluorescence curves generated during amplification. This method helps to distinguish between specific signals caused by the target nucleic acid and background signals or non-specific amplification signals caused by other factors, thereby supporting the judgment of the presence of the target in complex sample backgrounds and improving the reliability of detection.

[0034] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0035] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is a schematic diagram of the structure of a rapid tuberculosis gene screening service system for health check-up centers according to the present invention. Detailed Implementation

[0037] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0038] Please see Figure 1 As shown, this invention is a rapid tuberculosis gene screening service system for health check-up centers, comprising:

[0039] The sample collection and identification module is used to receive and uniquely identify biological samples from the test subject.

[0040] An integrated microfluidic chip with multiple independent and sealed detection channels is used for automated nucleic acid extraction, purification, and amplification reactions of input samples.

[0041] The temperature control and amplification module, coupled to the microfluidic chip, is used to provide precise temperature cycling control for the nucleic acid amplification reaction within the chip.

[0042] The fluorescence detection module is used to acquire the fluorescence signal of each detection channel in real time during the amplification reaction;

[0043] The data analysis module integrates a pre-trained deep learning model to receive and analyze fluorescence signals, and output classification results corresponding to the presence or absence of Mycobacterium tuberculosis target genes.

[0044] It also includes a report generation and distribution module, which automatically generates screening reports based on the classification results and distributes the reports to the relevant examinees and the health check-up center information system.

[0045] The sample collection and labeling module includes a sampling container with a radio frequency identification tag and a reader / writer for reading and writing tag information, which is associated with the examinee's unique identifier in the health check-up center's database.

[0046] The integrated microfluidic chip has a plate-like structure and contains multiple parallel channels corresponding to the number of standard multiwell plates. Each channel sequentially integrates a sample inlet, a lysis unit, a nucleic acid purification unit based on magnetic particles, an amplification reaction chamber, and an optical detection window. Each unit is connected by a microfluidic channel and the fluid flow is controlled by a built-in microvalve.

[0047] The deep learning model integrated in the data analysis module is a convolutional neural network model, which is configured to receive preprocessed fluorescence intensity sequence data changing over time, process it through a network structure containing convolutional layers and fully connected layers, and finally output a classification result containing the probabilities of positive, negative and invalid states.

[0048] It also includes an automated sample loading device, which includes a robotic arm and a vision positioning system, for precisely transferring samples from the sample acquisition and labeling module to a predetermined volume and injecting them into the sample inlet of a designated detection channel of the microfluidic chip.

[0049] The lysis unit, nucleic acid purification unit, and connected fluid pipeline of the integrated microfluidic chip are pre-stored or can be sequentially injected with lysis reagents, binding reagents, washing reagents, and elution reagents. The purification unit controls the movement of magnetic particles through an external magnetic field generator to complete the separation and purification of nucleic acids in a sealed channel.

[0050] The fluorescence detection module includes a light source and a photoelectric sensor for each detection channel, used to excite and collect fluorescence signals at specific stages of each temperature cycle of the amplification reaction, and convert analog signals into digital signal sequences.

[0051] The data analysis module also includes a signal preprocessing unit, which performs smoothing filtering, background fluorescence subtraction, and signal normalization on the digital signal sequence acquired by the fluorescence detection module to generate standard fluorescence curve data that conforms to the input format of the deep learning model.

[0052] The report generation and distribution module is configured to: determine the final screening conclusion based on the classification results and preset decision rules, automatically retrieve the examinee's information from the database, generate a structured report with a digital signature, and send the report to the examinee's email address and the confidential information platform of the health check-up center through an encrypted communication link.

[0053] A rapid tuberculosis gene screening method for health check-up centers, the method being implemented based on any of the above-mentioned systems, includes the following steps:

[0054] Samples are collected using labeled containers, and the container labels are associated with the subject's information;

[0055] Samples are automatically loaded into an independent detection channel integrated with a microfluidic chip;

[0056] Within the sealed channel of the chip, the purification steps of sample lysis, nucleic acid binding to magnetic particles, impurity washing, and nucleic acid elution are automatically and sequentially completed.

[0057] In the amplification reaction chamber of the chip, polymerase chain reaction targeting Mycobacterium tuberculosis specific target genes is completed, and fluorescence signals are monitored in real time;

[0058] The acquired real-time fluorescence signals are preprocessed to generate standardized amplification curve data;

[0059] Standardized amplification curve data are input into a pre-trained deep learning model to obtain the classification probability distribution;

[0060] Based on the comparison between the classification probability distribution and the preset threshold, the screening result is automatically determined to be positive, negative or invalid;

[0061] Based on the results of the automatic assessment, an electronic screening report is generated and automatically distributed to the relevant examinees and health check-up centers.

[0062] One specific application of this embodiment is:

[0063] Step S1: Sample Collection and Standardization Preprocessing

[0064] At the health check-up center, examinees provide samples using disposable sputum samplers (5mL capacity, containing 1mL of nucleic acid preservation solution with 0.1% sodium azide and RNase inhibitor). The sampler is equipped with an RFID tag on top, which is automatically encoded by an RFID reader during collection. The code is associated with the examinee's unique identifier (such as employee ID) in the health check-up center's database. After collection, the sampler is immediately placed on an automated conveyor belt system (conveyor belt speed 0.5m / s, carrier spacing 10cm) and transported to the pre-processing station.

[0065] The pretreatment station includes a vortex mixer and a precision liquid processing unit; first, the sampler is picked up by a robotic arm and placed into the vortex mixer, where it is mixed at 2500 rpm for 30 seconds to ensure that the sputum and the preservation solution are evenly mixed.

[0066] Subsequently, the liquid processing unit used disposable pipette tips (200 μL capacity, ±1 μL accuracy) to aspirate 200 μL of the mixed sample and transfer it to standardized sample tubes (2 mL tubes, pre-filled with 800 μL of lysis buffer containing 0.5% Triton X-100 and 50 mM Tris-HCl, pH 8.0); the standardized sample tubes also had RFID tags, which were linked to the original sampler tag via a reader.

[0067] After transfer, the sample tubes are sealed and placed on a temperature-controlled rack (temperature maintained at 4°C) to await the next loading step.

[0068] Step S2: Load the sample into the microfluidic chip

[0069] The microfluidic chip is a customized polymer chip (material: polymethyl methacrylate, size: standard 96-well plate format, 128 mm long, 86 mm wide, and 5 mm thick), containing 96 independent channels, each with the same structure: sample inlet (1 mm in diameter), lysis chamber (50 μL volume), purification chamber (100 μL volume), reaction chamber (25 μL volume), and optical detection window (2 mm in diameter); the chip surface is coated with a hydrophobic film to prevent liquid evaporation;

[0070] The loading process is performed by an automated robotic arm (accuracy ±0.1mm). The robotic arm is equipped with a vision system (camera resolution 1280×720, frame rate 30fps) to identify the position of the standardized sample tube and the coordinates of the chip inlet. The robotic arm uses a disposable pipette tip (tip capacity 10μL, accuracy ±0.5μL) to draw 10μL of pre-treated sample from the sample tube, moves it to the corresponding channel inlet of the chip, and injects it at a flow rate of 0.5μL / s. After injection, the chip inlet closes with a self-sealing membrane (material: polydimethylsiloxane) to prevent leakage. After loading, the chip is placed on a dedicated carrier (the carrier integrates a temperature sensor and an airflow interface). Each carrier holds one chip, and the carrier is associated with the sample RFID information via a barcode.

[0071] Step S3: Nucleic acid extraction and purification within the microfluidic chip

[0072] The chip carrier is inserted into the nucleic acid extraction module, which is connected to an external fluid controller and a pneumatic pump (pressure range 0-100 kPa, accuracy ±1 kPa); the microvalves and channels inside the chip are pre-filled with reagents: lysis buffer (containing 0.1 M NaOH and 1% SDS), magnetic silica particle suspension (particle diameter 1 μm, surface modified with silanol groups), washing buffer (70% ethanol, volume concentration), and elution buffer (10 mM Tris-HCl, pH 8.5).

[0073] The extraction process is automated: First, the lysis chamber is heated to 56°C (controlled by an integrated heater with an accuracy of ±0.5°C) and maintained for 5 minutes. Simultaneously, a pneumatic pump applies 20 kPa pressure to circulate and mix the sample within the lysis chamber, causing the tuberculosis cells to lyse and release nucleic acids. Second, a valve is switched to inject a magnetic particle suspension (5 μL) into the lysis chamber. After standing at room temperature for 2 minutes, the nucleic acids are adsorbed onto the particle surface via silica gel. Subsequently, an external electromagnet (magnetic field strength 0.5T) is activated, attracting the particles to the purification chamber. Washing buffer flows through the purification chamber twice at a flow rate of 10 μL / s (100 μL each time) to remove impurities. Finally, the particles are transferred to the reaction chamber, where elution buffer is applied at 65°C at a flow rate of 5 μL / s (20 μL). The nucleic acids dissolve in the elution buffer, completing the purification process. The entire process is conducted within a sealed channel, preventing cross-contamination.

[0074] Step S4: Multichannel parallel PCR amplification

[0075] The chip carrier was transferred to the PCR amplification module (the module integrates 96 independent thermal cycling units, each unit corresponding to one channel reaction chamber, with a temperature control range of 4-99°C and an accuracy of ±0.2°C); the reaction chamber was pre-filled with lyophilized PCR reaction mixture (each chamber contains: 0.2 μM forward primer (sequence: 5'-CCT GCG AGC GTA GGC GTC GG-3'), 0.2 μM reverse primer (sequence: 5'-CTCGTC CAG CGC CGC TTC GG-3'), 0.1 μM TaqMan probe (reporter group FAM, quencher group BHQ1), 0.2 mM dNTPs, 1 U DNA polymerase, The freeze-dried mixture was prepared by vacuum drying prior to chip loading.

[0076] The amplification program was software-controlled: initial denaturation at 94°C for 2 minutes; followed by 40 cycles, each cycle consisting of denaturation at 94°C for 15 seconds, annealing at 60°C for 30 seconds, extension at 72°C for 30 seconds; and a final extension at 72°C for 2 minutes; during thermal cycling, each channel was independently temperature-controlled at a rate of 5°C / s; during the annealing step, the TaqMan probe hybridized with the Mycobacterium tuberculosis IS6110 gene target, and polymerase hydrolyzed the probe to release a fluorescent signal.

[0077] Step S5: Real-time fluorescence detection and signal preprocessing

[0078] The PCR amplification module integrates a miniature fluorescence detection system: each reaction chamber is equipped with an LED light source (wavelength 485nm, power 10mW) and a photodiode detector (detection wavelength 535nm, bandwidth ±10nm); at the end of the annealing step of each cycle, the detector collects the fluorescence intensity, converts it into a digital signal through an analog-to-digital converter (ADC, resolution 16-bit), and stores it as time series data;

[0079] Signal preprocessing was performed in real time on an embedded processor (ARM Cortex-A72, 1.8 GHz); for each channel, the raw fluorescence sequence was denoted as... ,in, For the cycle number ( );

[0080] First, apply a moving average filter to smooth the data: Boundary values ​​are filled with nearest neighbors;

[0081] Secondly, calculate the baseline fluorescence: take the average value of cycles 3 to 15 as the background fluorescence. The formula is ;

[0082] Finally, normalized fluorescence values: In the formula, As a reference fluorescence value, the maximum average fluorescence value of the positive control was measured using a calibration plate and set to 1000 units; after normalization, the fluorescence curve for each sample was obtained. , as subsequent input.

[0083] Step S6: Deep learning-assisted data analysis and result classification

[0084] The preprocessed fluorescence curve data was transmitted to a server via Ethernet (server configuration: GPU NVIDIA Tesla V100, 32GB memory); the deep learning model was built based on a convolutional neural network (CNN), and the model input was the standardized fluorescence curve. The length is fixed at 40; the data is first normalized to zero mean and unit variance: In the formula, The mean of the curve, Standard deviation;

[0085] The model architecture is as follows:

[0086] Input layer: 40×1 dimension;

[0087] Convolutional layer 1: 32 filters, size 3, stride 1, using ReLU activation function, output dimension 38×32;

[0088] Max pooling layer 1: pooling size 2, stride 2, output dimension 19×32;

[0089] Convolutional layer 2: 64 filters, size 3, stride 1, ReLU activation, output dimension 17×64;

[0090] Flattening layer: Flattens the output into a 1088-dimensional vector;

[0091] Fully connected layer 1: 128 neurons, ReLU activated;

[0092] Output layer: 3 neurons, using the softmax activation function, output probability distribution. , representing the probabilities of positive, negative, and invalid, respectively;

[0093] The model was pre-trained on a historical dataset (containing 10,000 fluorescence curves labeled as positive, negative, or invalid, with invalid indicating amplification failure or abnormal signal). Training used the cross-entropy loss function and the Adam optimizer (learning rate 0.001). During deployment, the model performed forward propagation on new sample curves to calculate probabilities. The decision rule was set as follows: if... If, then it is classified as positive; if If the result is positive, the test is classified as negative; otherwise, it is classified as invalid and needs to be retested. All calculations are automated and require no human intervention.

[0094] Step S7: Automatic Interpretation of Results and Generation of Reports

[0095] The classification results are linked to the sample RFID identifier. Examinee information (such as name, age, and department) is retrieved by querying the health check-up center database. The system automatically generates a structured report: the report template is in XML format, and the fields to be filled include the test date, time, result (positive / negative / invalid), and confidence score (i.e., ...). or The report includes the results (values), and reference notes (e.g., a positive result recommends further chest X-ray examination); the report is converted to a PDF file via a TLS encryption protocol and includes a digital signature;

[0096] Report distribution is automated via an email system: examinees receive their personal reports (email addresses are extracted from the database), and the reports are simultaneously uploaded to the health check-up center's information management system (based on cloud storage and compliant with HIPAA security standards); if the result is positive, the system triggers an alarm module, sending a real-time notification (via SMS or internal messaging system) to designated medical personnel and marking the examinee's file for follow-up; all process data (including fluorescence curves, classification probabilities, and timestamps) are stored in an encrypted database and retained for at least 5 years for auditing and statistical analysis.

[0097] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0098] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A rapid tuberculosis gene screening service system for health check-up centers, characterized in that, The application relates to a system for tuberculosis screening, comprising: a sample collection and identification module for receiving and uniquely identifying biological samples from subjects; an integrated microfluidic chip having a plurality of independent and sealed detection channels for automated nucleic acid extraction, purification and amplification of input samples; a temperature control and amplification module coupled with the microfluidic chip for providing precise temperature cycling control for nucleic acid amplification reactions in the chip; a fluorescence detection module for real-time acquisition of fluorescence signals from each of the detection channels during the amplification reactions; a data analysis module having a pre-trained deep learning model integrated therein for receiving and analyzing the fluorescence signals to output a classification result corresponding to the presence or absence of target genes of Mycobacterium tuberculosis; and a report generation and distribution module for automatically generating a screening report based on the classification result and distributing the report to the corresponding subjects and health examination center information system.

2. The Mycobacterium tuberculosis gene rapid screening service system for health checkup center according to claim 1, characterized in that, The sample collection and identification module comprises a sampling container with a radio frequency identification tag and a read-write device for reading and writing the tag information, which is associated with a unique identifier of the subject in the health examination center database.

3. The Mycobacterium tuberculosis gene rapid screening service system for health checkup center according to claim 1, characterized in that, The integrated microfluidic chip has a plate structure and contains a plurality of parallel channels corresponding to a standard multi-well plate, each channel being integrated with a sample inlet, a lysis unit, a nucleic acid purification unit based on magnetic particles, an amplification reaction chamber and an optical detection window in sequence, and each unit is connected by a microfluidic pipeline and controlled by a built-in microvalve.

4. The Mycobacterium tuberculosis gene rapid screening service system for health examination center according to claim 1, characterized in that, The deep learning model integrated in the data analysis module is a convolutional neural network model configured to receive preprocessed fluorescence intensity sequence data over time, process the data through a network structure containing convolutional layers and fully connected layers, and finally output a classification result containing positive, negative and invalid state probabilities.

5. The Mycobacterium tuberculosis gene rapid screening service system for health checkup center according to claim 1, characterized in that, The system further comprises an automated sample loading device containing a mechanical arm and a visual positioning system for accurately transferring and injecting samples from the sample collection and identification module into the sample inlet of the designated detection channel of the microfluidic chip with a predetermined volume.

6. The Mycobacterium tuberculosis gene rapid screening service system for health checkup center according to claim 3, characterized in that, The lysis unit, nucleic acid purification unit and connected fluid pipeline of the integrated microfluidic chip prestore or can sequentially inject lysis reagents, binding reagents, washing reagents and elution reagents, and the purification unit controls the movement of magnetic particles through an external magnetic field generating device to complete the separation and purification of nucleic acids in the sealed channel.

7. The health checkup center-oriented Mycobacterium tuberculosis gene rapid screening service system according to claim 1, characterized in that, The fluorescence detection module comprises a light source and a photoelectric sensor corresponding to each detection channel for exciting and collecting fluorescence signals at specific stages of each temperature cycle of the amplification reaction and converting analog signals into digital signal sequences.

8. The health checkup center-oriented Mycobacterium tuberculosis gene rapid screening service system according to claim 4, characterized in that, The data analysis module further comprises a signal preprocessing unit for performing smoothing filtering, background fluorescence subtraction and signal normalization processing on the digital signal sequences collected by the fluorescence detection module to generate standard fluorescence curve data conforming to the input format of the deep learning model.

9. The health checkup center-oriented Mycobacterium tuberculosis gene rapid screening service system according to claim 1, characterized in that, The report generation and distribution module is configured to determine a final screening conclusion according to the classification result and a preset decision rule, automatically call the information of the examinee from a database, generate a structured report with a digital signature, and send the report to an email of the examinee and a confidential information platform of the health check center respectively through an encrypted communication link.

10. A method for rapid screening of Mycobacterium tuberculosis gene for health check-up center, characterized in that, The method is implemented based on the system according to any one of claims 1 to 9, and comprises the following steps: Collecting samples through containers with labels and associating the container labels with the information of the examinees; Automatically loading the samples into independent detection channels of integrated microfluidic chips; In the sealed channels of the chips, automatically and sequentially completing the purification steps of sample lysis, nucleic acid binding to magnetic particles, impurity washing and nucleic acid elution; In the amplification reaction chambers of the chips, completing polymerase chain reactions for specific target genes of Mycobacterium tuberculosis and monitoring fluorescence signals in real time; Pretreating the collected real-time fluorescence signals to form standardized amplification curve data; Inputting the standardized amplification curve data into a pre-trained deep learning model to obtain a classification probability distribution; Based on the comparison between the classification probability distribution and a preset threshold, automatically determining the screening result as positive, negative or invalid; According to the automatically determined result, generating an electronic screening report and automatically distributing the report to the corresponding examinee and the health check center.

Images