Intelligent auxiliary system for single-tube detection of food-borne pathogenic bacteria

Abstract

According to the intelligent auxiliary system for single-tube detection of the food-borne pathogenic bacteria, the One-pot-RPA-CRISPR / Cas12a reaction system is combined with the embedded artificial intelligence detection device, so that high-sensitivity and high-specificity detection and automatic grading judgment of the food-borne pathogenic bacteria are realized; through wireless interaction between the portable detection equipment and the intelligent terminal, the portability and field applicability of the equipment are realized; by introducing a YOLOv8n model and an OpenCV image processing algorithm, automatic identification and quantitative analysis of fluorescence signals are realized, subjective errors of manual interpretation are avoided, and the detection efficiency and consistency are improved; the system has good expandability, can meet the detection requirements of various pathogens or biomarkers by replacing primers and crRNA, is suitable for various scenes such as food pollution monitoring, primary medical diagnosis, family health management and the like, and has wide social benefits and application prospects.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of detection, in particular to an intelligent auxiliary system for detecting foodborne pathogenic bacteria in a single tube. BACKGROUND

[0002] Diseases caused by foodborne pathogenic bacteria are a global public health concern, especially for children, pregnant women and immunodeficient populations. According to the World Health Organization, hundreds of millions of people suffer from foodborne diseases every year due to the consumption of contaminated food, and children under the age of five bear nearly 40% of the foodborne disease burden. In remote areas with limited medical resources, the lack of rapid and accurate pathogen detection methods further exacerbates health risks and economic burdens.

[0003] Currently, the detection of pathogenic microorganisms in food mainly relies on nucleic acid detection technology, among which quantitative polymerase chain reaction (qPCR) is considered the "gold standard" in a laboratory environment. Although qPCR technology has high sensitivity and stability, it relies on a complex, bulky and expensive PCR instrument that requires professional operation in a standardized laboratory, which severely limits its application in grassroots, family or on-site environments. In addition, the qPCR detection period is long, usually taking several hours from sample processing to result output, which cannot meet the demand for rapid response to "sample in-result out" in public health emergencies.

[0004] To overcome the limitations of qPCR, isothermal nucleic acid amplification technology (INAAT) has gradually developed, such as recombinase polymerase amplification (RPA), which can achieve rapid DNA amplification at a constant temperature of about 37°C, with the advantages of fast reaction speed, simple primer design, and low equipment requirements. However, RPA technology itself has problems such as strong non-specific amplification, easy aerosol pollution, and difficulty in quantifying results. In recent years, the CRISPR system has been introduced into the field of molecular diagnosis due to its high specific recognition ability, and combining RPA with CRISPR / Cas12a can construct a sensitive, specific, fast and on-site deployable nucleic acid detection platform.

[0005] With the growing demand for home detection and point-of-care testing (POCT), portable detection devices have gradually emerged. However, existing POCT devices are easily disturbed by external factors such as temperature and humidity when used in non-standardized environments such as bedside, home or outdoor, and rely on manual interpretation of results (such as using ImageJ image analysis software) during operation, which has large subjective errors, low efficiency and difficulty in large-scale processing. In recent years, artificial intelligence technology has shown its superiority over human eyes and traditional algorithms in image recognition and classification, and its integration into POCT devices is expected to significantly improve the automation and reliability of detection results.

[0006] Although some studies have attempted to combine CRISPR technology with portable devices, most systems still have the following problems: first, the whole process from sample processing to result output is not integrated; second, there is a lack of intelligent analysis and automatic classification ability for fluorescence signals; third, the volume, power consumption or cost of the device is still insufficient to support extensive field applications; fourth, there is a lack of a data interaction platform that seamlessly interfaces with mobile terminals. Therefore, there is an urgent need in the art to develop a foodborne pathogen rapid detection system that integrates miniaturized devices, high-sensitivity detection methods, artificial intelligence assisted analysis, and mobile terminal interaction to achieve truly "sample-in and result-out" intelligent diagnosis. SUMMARY

[0007] The purpose of the present application is to provide an intelligent auxiliary system for single-tube detection of foodborne pathogenic bacteria to solve the problems existing in the prior art.

[0008] To achieve the above purpose, the present application provides the following solutions: The present application provides an intelligent auxiliary system for single-tube detection of foodborne pathogenic bacteria, comprising: a portable detection device, the portable detection device comprising a microcontroller unit, an optical detection unit, a temperature control unit, a power management unit, a Wi-Fi communication unit and an AI data processing unit; a mobile terminal application program, the mobile terminal application program being wirelessly connected to the portable detection device, for realizing parameter setting, data receiving, result display and user interaction; a cloud server, the cloud server being used for storing detection data, model parameters and user information.

[0009] Preferably, the optical detection unit comprises: a blue LED light source for exciting fluorescence signals; a macro camera for capturing fluorescence images of the reaction tube; a high-pass filter disposed in front of the camera for filtering excitation light and passing fluorescence signals.

[0010] Preferably, the temperature control unit comprises: a ceramic heating sheet for providing a constant temperature reaction environment; a temperature sensor for real-time monitoring of reaction temperature; a PID algorithm temperature controller for dynamically adjusting heating power to control temperature fluctuations within ±0.7℃.

[0011] Preferably, the AI ​​data processing unit uses a Rockchip RK3588 SoC chip and embeds a trained YOLOv8n model and an OpenCV image processing program; the YOLOv8n model is used to automatically locate and select the reaction tube region in the fluorescence image; the OpenCV program is used to quantify and analyze the fluorescence brightness within the selected region and classify the risk level of the detection results according to a preset threshold.

[0012] Preferably, the hazard classification includes: Risk-free: 0 CFU / mL; Low risk: 1×10 0 –1×10 2 CFU / mL; Medium risk: 1×10 2 –1×10 5 CFU / mL; High risk: >1×10 5 CFU / mL.

[0013] Preferably, the mobile terminal application includes: A Wi-Fi connectivity module is used to establish wireless communication with the portable testing device. The parameter setting module is used to set the reaction temperature and time; The real-time data module is used to display fluorescence curves and temperature changes; The results output module is used to display the test results and the risk level.

[0014] This invention also provides a detection method for an intelligent auxiliary system for single-tube detection of foodborne pathogens, comprising the following steps: S1. Sample pretreatment, extracting DNA from the sample to be tested; S2. Reaction system preparation: Add DNA to a One-pot-RPA-CRISPR / Cas12a reaction tube; S3. Start the device: Place the reaction tube into the portable detection device and start the detection via the mobile terminal application; S4. Results Output: The device automatically completes fluorescence acquisition, AI analysis, and result grading, and displays the test results through a mobile terminal application.

[0015] Preferably, the RPA reaction and CRISPR / Cas12a detection are performed sequentially in the One-pot-RPA-CRISPR / Cas12a reaction tube.

[0016] Preferably, the One-pot-RPA-CRISPR / Cas12a reaction tube is designed with specific primers and crRNA for the Listeria monocytogenes hly gene.

[0017] Preferably, the reaction conditions of the One-pot-RPA-CRISPR / Cas12a reaction tube are as follows: the temperature is 43℃, the Mg 2+ The concentration is 75mM, the RNP concentration is 100nM, the gRNA:Cas12a ratio is 3:1, the ssDNA reporter concentration is 100μM, and the reaction time is 8min.

[0018] The present application has the following beneficial technical effects compared with the prior art: The intelligent auxiliary system for single-tube detection of foodborne pathogens provided by the present application combines the One-pot-RPA-CRISPR / Cas12a reaction system with an embedded artificial intelligence module, realizes high sensitivity and high specificity detection and automatic classification of foodborne pathogens, realizes the portability and on-site applicability of the device through wireless interaction between the portable detection device and the intelligent terminal, realizes automatic recognition and quantitative analysis of the fluorescence signal through the introduction of the YOLOv8n model and the OpenCV image processing algorithm, avoids subjective errors of manual interpretation, improves detection efficiency and consistency, and has good scalability, can adapt to the detection needs of multiple pathogens or biomarkers by replacing primers and crRNA, and is suitable for food contamination monitoring, primary medical diagnosis, family health management and other scenes, has wide social benefits and application prospects. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0020] Figure 1 The intelligent auxiliary system for single-tube detection of foodborne pathogens provided by the present application provides a technical line diagram; Figure 2 The intelligent auxiliary system for single-tube detection of foodborne pathogens provided by the present application provides an exploded view of the portable detection device; Figure 3 The intelligent auxiliary system for single-tube detection of foodborne pathogens provided by the present application provides a mobile terminal application architecture diagram; Figure 4 The detection method provided by the present application is compared with the qPCR gold standard; Figure 5 The results of the triple-blind test of the present application are shown in Table 1. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work are within the protection scope of the present application.

[0022] The purpose of the present application is to provide an intelligent auxiliary system for single-tube detection of foodborne pathogenic bacteria. The system integrates miniaturized hardware devices, high-sensitivity biochemical detection methods, embedded artificial intelligence analysis, and mobile application, and realizes rapid, accurate, and automated detection of foodborne pathogenic bacteria on site. The system is particularly suitable for primary medical care, home self-detection, or food safety monitoring, and can effectively solve the problems of traditional detection methods, such as large equipment, complex operation, and dependence on professional personnel.

[0023] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0024] Embodiment 1: As shown in Figure 1 , the core of the intelligent auxiliary system for single-tube detection of foodborne pathogenic bacteria provided in the present embodiment is the cooperative work of the portable detection device, the mobile terminal application program, and the cloud server. The portable detection device is responsible for completing the constant temperature reaction, fluorescence signal excitation and collection, image data processing and result analysis; the mobile terminal application program provides a user interface for device connection, parameter setting, real-time data monitoring and result visualization; the cloud server is used for storing historical detection data, algorithm model updating and user management, supporting data backtracking and system optimization.

[0025] Specifically, in terms of hardware design, as shown in Figure 2 , the portable detection device is mainly composed of five modules: an optical detection unit, a temperature control unit, a power management unit, a Wi-Fi communication unit, and an AI data processing unit. The device shell is made of high-strength engineering plastic, with a size of 14cm x 7.5cm x 11cm and a weight of about 180 grams, which is convenient for single-handed holding and carrying. The internal layout of the device is compact, and each module is integrated and controlled efficiently through a microcontroller unit.

[0026] The optical detection unit includes a 1W blue LED light source with an emission wavelength of 470nm, used to excite the fluorescence signal in the reaction tube; a macro camera is used to acquire fluorescence images; and a high-pass filter is set in front of the camera with a center wavelength of 600nm and a bandwidth of ±20nm, which can effectively filter out the excitation light and allow 500-550nm green fluorescence to pass through, ensuring clear capture of the fluorescence signal.

[0027] The temperature control unit employs two sets of high-precision ceramic heating elements, evenly distributed around the reaction vessel, with a total power of 6W. Combined with a temperature sensor and a PID algorithm temperature controller, it achieves precise control of the reaction temperature, keeping temperature fluctuations within ±0.7℃, meeting the 43℃ constant temperature environment requirement of the One-pot-RPA-CRISPR / Cas12a reaction system. A dual ball-bearing DC fan with a speed of 3000 RPM is integrated on the right side of the heating module for rapid cooling after the reaction, preventing thermal inertia from interfering with subsequent operations.

[0028] The power management unit uses a 3-cell lithium battery pack with a total capacity of 3400mAh. It is equipped with a balanced protection circuit and supports overcharge, over-discharge, overcurrent and short circuit protection. The module integrates a PD140W bidirectional fast charging power bank chip, which is compatible with multiple fast charging protocols such as PD3.1 and QC4.0, and displays the remaining power in real time through a 4-level LED indicator.

[0029] The Wi-Fi communication unit is integrated into the RK3588 core board, supports Wi-Fi 6E, and can operate on both 2.4GHz and 5GHz dual-band, with a maximum transmission rate of 2.4Gbps. The device establishes a long connection with the mobile terminal application via TCP / IP protocol, with a transmission interval configurable from 100ms to 10s. The average network latency is less than 50ms (within a distance of 10 meters). Image data is transmitted via USB 3.0 interface with a bandwidth of 5Gbps and a single frame image transmission latency of less than 10ms.

[0030] Furthermore, in terms of artificial intelligence processing, such as Figure 3 As shown, the AI ​​data processing unit is based on the Rockchip RK3588 SoC, which embeds a trained YOLOv8n model and OpenCV image processing program. YOLOv8n is the model with the fewest parameters and fastest speed in the YOLOv8 series, designed specifically for edge devices, with 3.2M parameters and 6.2MB weights (FP16 format). The model training process includes steps such as sample preparation, dataset construction, training, and optimization. First, DNA (NC, 1×10⁻⁶) extracted from different bacterial cultures was prepared. 0 –1×10 2 CFU / mL, 1×10 2 –1×105 CFU / mL, >1 x 105 0 -1 x 105 5 CFU / mL) sample tubes as training sample set, totally 7000 tubes were made, 1750 tubes for each bacteria concentration, each DNA gradient in the sample set was randomly divided into three groups: training set (70%), validation set (20%) and test set (10%), respectively composed of 4900, 1400 and 700 tubes. Image acquisition was carried out in the dark room of the experimental device, using 470 nm light irradiation, two samples of each DNA gradient generated an image, the number of training, validation and test data sets was 2450, 700 and 350 images respectively. Subsequently, the reaction tube area in the image was manually labeled using the labelling tool, and the original image of the validation data set was retained for accuracy comparison.

[0031] During the training process, first use Pytorch2.0.0 to train YOLOv8n for the first time, the model mAP50 is 0.872, and the actual test confidence is 0.78-0.85; in order to improve the accuracy of the bounding box, use the model generated by the first training to generate pseudo-labels, build a data set of about 20000 pictures of reagent sample tubes of different concentrations, and carry out the second strengthening training, the model mAP50 is improved to 0.995, and the confidence is improved to 0.89-0.95. After training, the model is converted to.rkm format and embedded in the device.

[0032] The OpenCV image processing program is used to quantify the brightness of the fluorescence region bounded by YOLOv8n, which uses the inRange function for threshold segmentation. This parameter can accurately distinguish the fluorescence intensity of each concentration, and according to the fluorescence value range, the results are divided into four risk levels: no risk (0 CFU / mL), low risk (1 x 105 0 -1 x 105 2 CFU / mL), medium risk (1 x 105 2 -1 x 105 5 CFU / mL) and high risk (>1 x 105 5 CFU / mL), this grading is based on the dose-response relationship of Listeria infection in the WHO related report, and is verified in the actual sample test.

[0033] Further, the mobile terminal application program is developed based on Android Studio, written in Java language, and compatible with Android and above operating systems. The application program mainly includes four functional modules: the Wi-Fi connection module is used to search and connect device hotspots, and the device needs to be connected manually for the first use, and the subsequent system automatically identifies and connects; the parameter setting module allows users to freely set the reaction temperature and duration to meet the needs of different detection scenarios; the real-time data module dynamically displays the fluorescence intensity and temperature changes in the form of a curve, and the application program collects the fluorescence value of the test tube every 30 seconds until two complete fluorescence curves are drawn; the result output module displays the photographed fluorescence image and the AI interpreted risk level. The application program establishes real-time data communication with the device through the Socket protocol, realizes remote control and result feedback, and the overall detection process does not require manual intervention, realizing the full closed-loop automation from parameter setting, data collection to result output.

[0034] Example 2: On the detection method, the application adopts One-pot-RPA-CRISPR / Cas12a technology, and takes Listeria monocytogenes hly gene as the target (GenBank ID: LC259879.1). First, 12 groups of RPA primers are designed for the target sequence, and the F1-R3 primer combination is selected as the best primer pair through experiments, which has the most significant difference in the positive and negative amplification curves. The crRNA sequence is designed according to the target region, containing a PAM sequence (TTTV) and a 20-nucleotide specific recognition region. After multi-parameter optimization of the reaction system, the temperature is 43℃, the Mg 2+ concentration is 75mM, the RNP complex concentration is 100nM, the gRNA to Cas12a ratio is 3:1, the ssDNA-FQ reporter concentration is 100μM, and the reaction time is 8 minutes. Under this condition, the system can excite the maximum fluorescence and the difference between the negative and positive is the largest. The detection limit of this system can reach 1×10 1 CFU / mL, and has no cross-reaction with other common foodborne pathogens (such as Escherichia coli O157:H7, Salmonella, etc.), indicating its high specificity. In addition, through continuous fluorescence monitoring of the sample for 7 days, it is found that the fluorescence signal remains stable, proving that the method has good stability.

[0035] To verify the performance of the above system, the applicant conducted a triple-blind comparison experiment. As shown in Figure 4 and 5 , three researchers were responsible for sample preparation, qPCR verification and platform detection respectively. The first researcher prepared 30 simulated overnight refrigerator samples, including 15 negative samples and 15 positive samples (covering low, medium and high infection levels); the second researcher verified the samples by qPCR detection, and the R 2The result is 0.998, which is completely consistent with the result reported by the first researcher; the third researcher uses the platform to detect the same batch of samples, and the results show that the sensitivity and specificity of the platform are both 100%, indicating that the detection accuracy is comparable to the laboratory gold standard method. In addition, the applicability test of the device on different brands of mobile phones (such as Honor, OPPO, Huawei) shows that the mobile terminal application and the device are connected stably and function normally, proving that the system has good universality.

[0036] In summary, the present application constructs a complete, efficient and reliable foodborne pathogen on-site detection system through the deep integration of hardware miniaturization, detection method optimization and artificial intelligence algorithm, which provides strong technical support for food safety monitoring and public health protection. The system not only realizes the rapid detection mode of "sample in-result out", but also reduces human error through intelligent analysis, significantly improves the detection efficiency and consistency.

[0037] The technical features of the above embodiments can be combined arbitrarily, and to make the description concise, not all possible combinations of the technical features in the above embodiments are described, however, as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present application.

[0038] It should be noted that the components mentioned in the above embodiments are all general standard components or components known to those skilled in the art, and their structure and principle are known to those skilled in the art through technical manuals or through conventional experimental methods.

[0039] The present application applies specific examples to explain the principles and implementation methods of the present application, and the above examples are only used to help understand the method of the present application and its core idea; at the same time, for those skilled in the art, according to the idea of the present application, the specific implementation method and application range will be changed. In summary, the content of the specification should not be understood as a limitation of the present application.

Claims

1. An intelligent assistant system for single tube detection of foodborne pathogen, characterized in that, The application relates to a portable detection device, a mobile terminal application and a cloud server. The portable detection device comprises a microcontroller unit, an optical detection unit, a temperature control unit, a power management unit, a Wi-Fi communication unit and an AI data processing unit. The mobile terminal application is wirelessly connected with the portable detection device and is used for realizing parameter setting, data receiving, result display and user interaction. The cloud server is used for storing detection data, model parameters and user information.

2. The intelligent assistant system for single tube detection of food borne pathogen as claimed in claim 1 wherein, The optical detection unit comprises a blue light LED light source, a macro camera and a high-pass optical filter. The blue light LED light source is used for exciting a fluorescent signal. The macro camera is used for collecting a fluorescent image of a reaction tube. The high-pass optical filter is arranged in front of the camera and is used for filtering excitation light and passing fluorescent signals.

3. The intelligent assistant system for single tube detection of food borne pathogen as claimed in claim 1 wherein, The temperature control unit comprises a ceramic heating sheet, a temperature sensor and a PID algorithm temperature controller. The ceramic heating sheet is used for providing a constant-temperature reaction environment. The temperature sensor is used for monitoring a reaction temperature in real time. The PID algorithm temperature controller is used for dynamically adjusting heating power and controlling temperature fluctuation within a range of + / -0.7 DEG C.

4. The intelligent assistant system for single tube detection of food borne pathogen as claimed in claim 1 wherein, The AI data processing unit adopts a Rockchip RK3588 SoC chip and is embedded with a trained YOLOv8n model and an OpenCV image processing program.

5. The intelligent assistant system for single tube detection of food borne pathogen as claimed in claim 4 wherein, The YOLOv8n model is used for automatically positioning and framing a reaction tube region in a fluorescent image. The OpenCV program is used for quantitatively analyzing fluorescent brightness in the framed region and classifying a detection result according to a preset threshold value. Low risk: 1 x 10 0 - 1 x 10 2 CFU / mL; Medium risk: 1 x 10 2 - 1 x 10 5 CFU / mL; High risk: >1 x 10 5 CFU / mL.

6. The intelligent assistant system for single tube detection of food borne pathogen as claimed in claim 1 wherein, The risk classification comprises the following: No risk: 0 CFU / mL. The mobile terminal application comprises a Wi-Fi connection module, a parameter setting module, a real-time data module and a result output module. The Wi-Fi connection module is used for establishing wireless communication with the portable detection device. The parameter setting module is used for setting a reaction temperature and time.

7. The detection method for the intelligent assistant system for single tube detection of foodborne pathogenic bacteria according to any one of claims 16, characterized in that, The real-time data module is used for displaying a fluorescent curve and temperature change. The result output module is used for displaying a detection result and a risk level. The application comprises the following steps: S1. Sample pretreatment, extracting DNA of a sample to be measured; S2. Reaction system configuration, adding the DNA into a One-pot-RPA-CRISPR / Cas12a reaction tube; 8. The detection method for the intelligent assistant system for single tube detection of foodborne pathogenic bacteria according to claim 7, characterized in that, S3. Device starting, placing the reaction tube into the portable detection device and starting detection through the mobile terminal application; 9. The detection method for the intelligent assistant system for single tube detection of foodborne pathogenic bacteria according to claim 7, characterized in that, S4. Result output, automatically completing fluorescent collection, AI analysis and result classification by the device and displaying a detection result through the mobile terminal application. The RPA reaction and the CRISPR / Cas12a detection in the One-pot-RPA-CRISPR / Cas12a reaction tube are sequentially performed. The One-pot-RPA-CRISPR / Cas12a reaction tube is designed with specific primers and crRNA for the hly gene of Listeria monocytogenes.

10. The detection method for the intelligent assistant system for single tube detection of foodborne pathogenic bacteria according to claim 7, characterized in that, The reaction conditions of the One-pot-RPA-CRISPR / Cas12a reaction tube are as follows: the temperature is 43°C, the Mg 2+ The concentration is 75mM, the RNP concentration is 100nM, the gRNA:Cas12a ratio is 3:1, the ssDNA reporter concentration is 100μM, and the reaction time is 8min.

Images