A digital cloning accurate counting and identification system and method

Abstract

This invention relates to the field of microbial counting technology, and more particularly to a digital clone accurate counting and identification system and method. The method includes the following steps: injecting the bacterial solution to be detected into a preset microchamber array and constructing several partitioned units; placing the microchamber array in a preset constant-temperature incubation environment for parallel cultivation and acquiring chamber image sets for each partitioned unit; segmenting the chamber image sets for each partitioned unit and extracting the corresponding image contours; identifying positive chambers based on the image contours to determine the predicted loading number of single-clone growth chambers; evaluating the relative growth capacity of each positive chamber based on the predicted loading number and the growth kinetic parameter set of each single-clone growth chamber, and identifying the target dominant clone type corresponding to each positive chamber. This application not only improves the screening efficiency and accuracy of probiotics and other microbial research and development, but also significantly enhances the stability, functionality, and safety of microbial products.

Description

Technical Field

[0001] This invention relates to the field of microbial counting technology, and in particular to a digital clone accurate counting and identification system and method. Background Technology

[0002] In the research and development and production of probiotics and other microorganisms, accurate measurement of microbial quantity and activity is fundamental to achieving product stability, functionality, and safety. However, traditional microbial counting methods have significant limitations. For example, the plate count method relies on the formation of visible colonies through culture, a process that typically takes 24–72 hours and is susceptible to environmental microbial contamination, operator variations, and fluctuations in culture conditions, resulting in low measurement accuracy and poor repeatability. Furthermore, this method struggles to capture clonal heterogeneity information at the single-cell level, failing to identify a few dominant strains or potentially highly active clones, thus posing significant limitations for high-throughput screening and evolutionary experiments.

[0003] In microbial evolution screening and directed evolution experiments, existing methods lack digital tools to accurately track the number of individual clones and identify dominant phenotypic types. Especially under high-throughput conditions, a large number of clones exhibit high heterogeneity, and traditional manual statistical and observational methods cannot achieve real-time monitoring and accurate analysis, thus limiting the quantitative study of the evolutionary process and making it difficult to provide a reliable basis for subsequent strain optimization and quality control. Summary of the Invention

[0004] Therefore, it is necessary for the present invention to provide a digital clone accurate counting and identification system and method to solve at least one of the above-mentioned technical problems.

[0005] To achieve the above objectives, a method for accurate counting and identification of digital clones includes the following steps:

[0006] Step S1: Inject the serially diluted bacterial solution to be tested into the preset microchamber array, and seal the injected microchamber array to construct several partitioned units;

[0007] Step S2: Place the microchamber array in a preset constant temperature culture environment for parallel culture, and perform full-field scanning imaging of each partition unit at preset time intervals to obtain the chamber image set of each partition unit;

[0008] Step S3: Perform image segmentation on the chamber image set of each partition unit, and extract the image contour corresponding to each partition unit based on the image segmentation results;

[0009] Step S4: Extract the growth kinetic parameter set based on the image contour to identify positive chambers, and classify the positive chambers into monoclonal growth chambers and polyclonal growth chambers according to the growth of each parameter in the growth kinetic parameter set; remove polyclonal growth chambers, and determine the predicted loading number of monoclonal growth chambers.

[0010] Step S5: Based on the predicted loading number and the growth kinetic parameter set of each monoclonal growth chamber, deduce the actual number of single bacteria in each monoclonal growth chamber to determine the number of single bacteria to be added in each monoclonal chamber, and perform amplification culture to make the number of bacteria in the microbial community consistent and achieve normalization.

[0011] This application achieves high-throughput, parallel culture and monitoring of microbial samples through the standardized construction of microchamber arrays and automated full-field scanning imaging. Compared to the traditional plate counting method, which is time-consuming and susceptible to operational variations and environmental contamination, this application can complete continuous observation of a large number of independent samples in a short time, greatly improving detection efficiency and reducing errors introduced by human operation at the source, significantly improving the repeatability and reliability of counting results. Secondly, based on image segmentation and contour extraction technology, combined with dynamic analysis of growth kinetic parameter sets, this application achieves a fundamental leap from static morphological recognition to dynamic growth behavior determination. It no longer relies solely on a single final state result, but accurately extracts core indicators such as maximum growth rate and incubation period by fitting curves of morphological parameters such as cell area, perimeter, and equivalent diameter over time. This innovation makes the determination of positive chambers no longer a subjective, experience-based observation, but a data-driven, quantitative, and precise identification, effectively eliminating stray noise and non-target structural interference, ensuring high sensitivity and specificity of the identification results. This application can clearly distinguish between monoclonal and polyclonal growth chambers based on the characteristics of growth kinetic parameters, and accurately eliminate polyclonal interference samples. Simultaneously, by statistically analyzing the proportion of negative chambers and performing reverse iterative solutions, the predicted loading number of monoclonal chambers can be accurately derived, achieving precise digital analysis of microbial clonal populations and filling the gap in traditional techniques for single-cell level heterogeneity analysis. By deriving the actual number of single bacteria and performing amplification culture, precise and standardized data support is provided for subsequent strain optimization and quality control. It ensures the consistency of initial cell numbers in all analyzed samples, providing a rigorous scientific basis and comparability for subsequent evolutionary experiments and screening studies, thereby comprehensively guaranteeing the stability, functionality, and safety of microbial product research and development and production processes.

[0012] Optionally, this application also provides a digital clone precise counting and identification system for performing the digital clone precise counting and identification method as described above, the digital clone precise counting and identification system comprising:

[0013] The sample injection module is used to inject the serially diluted bacterial solution to be tested into a preset microchamber array and to seal the injected microchamber array to construct several partitioned units.

[0014] The scanning imaging module is used to place the microchamber array in a preset constant temperature culture environment for parallel culture, and to perform full-field scanning imaging of each partition unit at preset time intervals to obtain the chamber image set of each partition unit.

[0015] The image processing module is used to segment the chamber image set of each partition unit and extract the image contour corresponding to each partition unit based on the image segmentation results.

[0016] The predicted loading number determination module is used to extract a set of growth kinetic parameters based on image contours to identify positive chambers, and classify positive chambers into monoclonal growth chambers and polyclonal growth chambers according to the growth of each parameter in the set of growth kinetic parameters; polyclonal growth chambers are eliminated, and the predicted loading number of monoclonal growth chambers is determined at the same time.

[0017] The microbial community normalization module is used to deduce the actual number of single bacteria in each monoclonal growth chamber based on the predicted loading number and the growth kinetic parameter set of each monoclonal growth chamber, so as to determine the number of single bacteria to be added in each monoclonal chamber, and perform amplification culture to make the number of bacteria in the microbial community consistent and achieve normalization.

[0018] The present invention relates to a digital clone precise counting and identification system. This system can implement any of the digital clone precise counting and identification methods of the present invention. It serves as a medium for the operation and signal transmission between various modules to complete the digital clone precise counting and identification method. The modules within the system cooperate with each other, thereby improving the screening efficiency and accuracy of probiotics and other microorganisms research and development. It can also be used for the precise quantification and quality control of tolerant strains in industrial production, significantly improving the stability, functionality and safety of microbial products. Attached Figure Description

[0019] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0020] Figure 1 This is a schematic diagram of the steps of the digital clone accurate counting and identification method of the present invention;

[0021] Figure 2 This is a chamber image of the partition unit in an embodiment of the present invention;

[0022] Figure 3 This is a block diagram of the digital clone precision counting and identification system in an embodiment of the present invention;

[0023] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0024] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0025] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0026] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0027] To achieve the above objectives, please refer to Figures 1 to 3 This invention provides a method for accurate counting and identification of digital clones, the method comprising the following steps:

[0028] Step S1: Inject the serially diluted bacterial solution to be tested into the preset microchamber array, and seal the injected microchamber array to construct several partitioned units;

[0029] In this embodiment, a serially diluted bacterial solution is injected into a microchamber array using a micro-injection device. The microchamber array consists of 1000 independent and regularly arranged microchambers, each with a volume of 0.5 μL, matched to the dilution concentration of the bacterial solution to be tested to ensure a single-bacterial loading probability close to 0.8. After injection, the openings of the microchamber array are sealed using a bio-inert transparent polymer to facilitate optical imaging. Simultaneously, the overall sealing status is detected using a liquid connectivity sensor to detect the cross-connection of liquids between microchambers. Microchambers meeting the independent sealing conditions are defined as partitioned units, while microchambers not meeting the conditions are grouped into a single partitioned unit based on their connectivity.

[0030] Step S2: Place the microchamber array in a preset constant temperature culture environment for parallel culture, and perform full-field scanning imaging of each partition unit at preset time intervals to obtain the chamber image set of each partition unit;

[0031] In a further embodiment, the microchamber array is placed in a constant-temperature incubator, with the temperature set at 37±0.2℃ and the humidity maintained above 95%. The array is scanned at preset time intervals (e.g., every 30 minutes) along a pre-calculated full-coverage scanning path using a two-dimensional displacement mechanism of the imaging platform. The two-dimensional displacement mechanism has a movement accuracy of 5. The system captures initial and final images at each field of view position, and controls the exposure time (range 50-150ms) and light intensity (range 100-300lx) based on the grayscale distribution of the initial image. Then, it performs image registration and fusion processing on the initial and final images to generate a set of partitioned unit chamber images, and establishes a mapping relationship between the field of view position and the chamber index.

[0032] Step S3: Perform image segmentation on the chamber image set of each partition unit, and extract the image contour corresponding to each partition unit based on the image segmentation results;

[0033] In a further embodiment, image segmentation processing is performed on the image sets of each partition unit chamber using a U-Net model based on a convolutional neural network. The input is a chamber image (resolution 2048×2048 pixels), and the output is a binary contour map. Noise is removed through edge detection and morphological filtering, and the contour pixel set of each partition unit is extracted. The contour position coordinates, area, and morphological indices are recorded as contour parameters.

[0034] Notably, a U-Net convolutional neural network model was constructed for accurate segmentation of the chamber images. This model comprises an encoder and a decoder. The encoder consists of four convolutional layers, each containing two 3×3 convolutional kernels, a ReLU activation function, and a 2×2 max-pooling operation to extract multi-scale features. The decoder is symmetrically constructed, fusing encoder features through upsampling and skip connections to achieve high-resolution image restoration. The input image is a 2048×2048 pixel micro-chamber image, and the output is a binary contour map, where the contour pixel values ​​represent colony locations. The model training data comes from a pre-labeled set of positive colony images, with each pixel labeled as either a colony or background. During training, a cross-entropy loss function and an Adam optimizer were used, with a learning rate of 0.001, 500 iterations, and a batch size of 8. After training, the contour map was overlaid with the time-series images. Morphological parameters such as contour area, roundness, and boundary roughness were used to extract the location and morphological features of each colony, thereby identifying positive colonies and constructing a set of growth kinetic parameters.

[0035] Step S4: Extract the growth kinetic parameter set based on the image contour to identify positive chambers, and classify the positive chambers into monoclonal growth chambers and polyclonal growth chambers according to the growth of each parameter in the growth kinetic parameter set; remove polyclonal growth chambers, and determine the predicted loading number of monoclonal growth chambers.

[0036] In a further embodiment, based on the chamber image corresponding to the partition unit, the grayscale contour in the image is extracted, and the grayscale value, contour area, roundness, and aspect ratio are used as constraint parameters for screening. By setting minimum and maximum area thresholds and roundness ranges, stray noise and non-target structures are effectively eliminated, thereby obtaining candidate positive chambers and establishing their spatial index. Subsequently, for the candidate chambers, contour tracking is performed on continuously acquired images at preset time intervals during the culture cycle to extract morphological parameters such as the area, perimeter, and equivalent diameter of the bacterial cell region, and a morphological time series data matrix is ​​constructed in chronological order. At the same time, the pixels within the contour region are statistically analyzed in conjunction with grayscale feature analysis, and the average grayscale value and grayscale discretization are used to obtain the data. The method determines the potential bacterial cell region and performs binarization segmentation using an adaptive threshold method. Further, it combines area and morphological constraints to screen for real bacterial cells, achieving precise morphological identification. Based on this, the change in bacterial cell area over time is used as a growth characterization metric. Curve fitting is performed to obtain growth kinetic parameters such as maximum growth rate, latency period, and final cell size. Candidate chambers are comprehensively analyzed based on a pre-defined multi-parameter joint judgment rule. A chamber is classified as positive if it meets the constraints in terms of growth rate, latency period, and final size; otherwise, it is classified as negative. This transitions the judgment from static morphological features to dynamic growth behavior. Furthermore, based on the extracted growth kinetic parameters... The positive chambers were classified into clonal types based on the continuity and uniformity of the cell area growth curve, the uniformity of the maximum growth rate, the dispersion of the latency period, and the distribution characteristics of the final cell size. If the growth kinetic parameters showed typical characteristics of single-cell proliferation with increasing culture time, the growth rate had no obvious step abrupt changes, and the latency period was concentrated within a certain range, then the positive chamber was classified as a monoclonal growth chamber. If the growth kinetic parameters showed characteristics of superimposed proliferation of multiple cell populations, the growth rate had multiple discrete mutation nodes, or the latency period showed a multi-level differentiation state, then the positive chamber was classified as a polyclonal growth chamber. After the classification was completed, the polyclonal growth chamber was removed from the positive chamber. The data from each chamber was centrally removed, retaining only the monoclonal growth chambers as valid samples for subsequent analysis. Furthermore, within the monoclonal growth chambers, a single-bacterial random loading hypothesis was introduced, modeling the cell loading process within the chamber as a discrete random event. A Poisson distribution model with the average loading number as a parameter was constructed. By performing multiple sets of random sampling on the loading numbers to form a probability matrix, the statistical distribution characteristics of whether the chamber was empty or loaded with a single cell were characterized. Based on this, statistical analysis was performed on the entire microchamber array to obtain the actual proportion of negative chambers. This proportion was then used as a constraint input into the Poisson distribution model. Through iterative solving, the theoretical distribution was matched with the actual statistical results, thereby inversely determining the predicted loading number for each monoclonal growth chamber.

[0037] Step S5: Based on the predicted loading number and the growth kinetic parameter set of each monoclonal growth chamber, deduce the actual number of single bacteria in each monoclonal growth chamber to determine the number of single bacteria to be added in each monoclonal chamber, and perform amplification culture to make the number of bacteria in the microbial community consistent and achieve normalization.

[0038] In a further embodiment, the standard growth parameters of the target bacterial species are obtained by calling the bacterial growth characteristic database, and the time point of the positive signal appearance in each monoclonal growth chamber is determined according to the change characteristics of the exponential growth phase in each monoclonal growth chamber. Then, the fluorescence signal intensity of the corresponding chamber is collected at the time point, and the fluorescence signal intensity is converted into the corresponding bacterial number by combining the fluorescence calibration relationship in the bacterial growth characteristic database. After obtaining the bacterial number at the positive signal time, the initial cell number is calculated backward by combining the exponential growth law, and the predicted loading number from the statistical results of the loading process is introduced as a prior constraint to perform probability screening and correction on the initial number to obtain the range of the initial single cell number. On this basis, a bacterial growth trajectory simulation sequence is further constructed by Monte Carlo simulation. The simulation results are statistically analyzed by introducing growth perturbation, and the structural similarity of the simulated bacterial number distribution is compared with the actual chamber image sequence to determine the actual single bacterial number in each monoclonal chamber. Finally, the actual single bacterial number in the entire array chamber is statistically analyzed and compared with the target loading number to obtain the distribution of the number of single bacteria to be filled in each monoclonal chamber. Next, based on the coordinate index information of each positive chamber in the microchamber array, the corresponding target microchamber to be filled is located in the microchamber array, and the target microchamber to be filled is opened at a fixed point. For the culture in each target microchamber to be filled, the culture is transferred to a preset independent culture carrier according to the corresponding number of single bacteria to be filled, and the amplification culture time and the corresponding number of bacteria to be filled are determined according to the exponential growth of the growth kinetic parameters of the culture. Amplification culture is carried out in the independent culture carrier to obtain the number of bacteria that meet the filling requirements. Then, the amplified culture is used to fill the bacterial population in the original microchamber array, thereby realizing the normalization of the number of bacteria in each chamber of the microchamber array.

[0039] Optionally, the amplification culture performed in step S5 includes:

[0040] Based on the coordinate index information of each positive chamber, the corresponding target micro-chamber in the micro-chamber array is located, so as to open the target micro-chamber at a fixed point;

[0041] In this embodiment, the target microchamber to be filled in the microchamber array is located based on the coordinate index information of each positive chamber. The coordinate index information includes row number, column number, and depth information to ensure accurate spatial positioning of the microchamber array. Based on the positioning results, a command is sent to the microchamber opening and closing control device to precisely open the microvalve or microfluidic channel of the target microchamber to be filled, thereby establishing a controlled pathway between the culture inside the chamber and the external operating environment.

[0042] Based on the number of single bacteria to be added to each target microchamber, the cultures in each target microchamber are transferred to a pre-set independent culture carrier. The amplification culture time and the corresponding number of bacteria to be added are determined based on the exponential growth of the growth kinetic parameter set corresponding to the culture, so as to carry out amplification culture in the independent culture carrier.

[0043] In a further embodiment, based on the number of single bacteria to be filled in each microchamber, the culture within the microchamber is transferred to a pre-defined independent culture carrier using a microfluidic manipulation device. The independent culture carrier can be a single culture well, a microplate, or a microfluidic amplification chamber, ensuring complete isolation of the culture in each microchamber from other chambers. Based on the set of growth kinetic parameters corresponding to the transferred culture, including maximum growth rate, latency, and exponential growth curve, the required amplification culture time and the corresponding number of filler bacteria are calculated, and the culture device is started for timed culture. During the amplification culture, the growth of the culture can be monitored in real-time using signal detection. Culture is stopped once the pre-defined number of bacteria is reached, and an amplification culture record is generated. This operation ensures that the required number of bacteria to fill the gap is obtained in the independent carrier, guaranteeing the fairness of subsequent cultures.

[0044] It is worth noting that the culture time required to amplify the culture to the target number of filler cells in an independent culture vector can be calculated using the following formula: ;in, The initial number of bacteria to be transferred to an independent culture vector. This corresponds to the maximum growth rate. To replenish the target number of bacterial cells. The required amplification and culture time is set. The control system sets the amplification and culture time based on the calculation results and starts the culture device of the independent culture carrier for timed culture. At the same time, real-time signal detection (such as fluorescence intensity or optical density) can be optionally equipped to dynamically confirm the culture progress and ensure that the number of bacteria reaches the predetermined replenishment target.

[0045] It is worth noting that during the amplification and culture process, the culture status is continuously monitored, and the culture is automatically stopped when the target quantity is reached or the preset culture time is completed to avoid overgrowth that could cause the bacterial count to exceed the replenishment requirements. Subsequently, the amplified culture is used to replenish the bacterial population in the corresponding microchambers of the microchamber array, achieving normalization of the bacterial population in the microchambers to be replenished. Meanwhile, for microchambers that do not require replenishment, their microvalves or microfluidic channels are kept closed, allowing the culture to maintain its original state without transfer or additional amplification culture. Microchambers that do not require replenishment maintain culture stability and isolation throughout the entire operation, ensuring no cross-contamination or bacterial interference between the microchambers in the microchamber array.

[0046] Optionally, step S1 includes:

[0047] The bacterial solution to be tested is injected into the micro-chamber array through a preset micro-injection device; wherein the micro-chamber array includes several independent and regularly arranged micro-chambers, and the volume of each micro-chamber is matched with the dilution concentration of the bacterial solution to be tested;

[0048] In this embodiment, a micro-injection device (injection accuracy 0.1) is used. Injection speed 0.5–2 (Adjustable) The serially diluted bacterial solution is injected into the microchamber array. The microchamber array consists of 500 independent microchambers, each with a volume of 0.8... The microchambers are arranged in a regular 10×50 matrix, with physical diaphragms between them to prevent cross-flow of liquid. During injection, a high-speed camera system monitors the liquid level in real time to ensure uniform filling of each microchamber and prevent bubble formation.

[0049] After each microchamber in the microchamber array is filled, the open end of the microchamber array is sealed.

[0050] In a further embodiment, to seal the microcavities, after injection, a flexible thin-film cover and a pressing device are used to seal the opening ends of the microcavity array, with the sealing pressure controlled at [pressure value missing]. This ensures that the liquid inside the microchamber does not leak and prevents cross-contamination. After sealing, infrared imaging is used to detect the temperature uniformity at the top of the microchamber, verifying the integrity of the seal and the stability of the liquid, ensuring that the subsequent culture environment is controllable.

[0051] The overall closure status of the micro-cavities after injection is detected to determine the cross-connection status.

[0052] In a further embodiment, for overall closure detection, a microfluidic resistance measurement system is used to scan and detect each column of the microchamber array. By measuring the liquid connectivity conductivity between the microchambers, it is determined whether cross-connectivity exists. The measurement results are input into the control software to generate a two-dimensional connectivity matrix, with each matrix unit corresponding to one microchamber. The conductivity threshold is set to [value missing]. , used to determine independence.

[0053] Based on the overall closed-state detection results, the area is divided into several partition units.

[0054] In a further embodiment, the microchambers are divided into partition units based on the overall closed-state matrix. If the overall closed-state matrix is ​​higher than the conductivity threshold, each microchamber is considered as a partition unit; if the overall closed-state matrix is ​​lower than the conductivity threshold, microchambers exhibiting cross-connectivity are grouped together as a partition unit. After partitioning, a partition unit index table is generated, recording the microchamber number, location coordinates, and volume information of each partition unit, providing a data foundation for subsequent positive chamber identification and growth kinetic analysis.

[0055] Optionally, injecting the bacterial solution to be tested into the micro-chamber array via a preset micro-injection device includes controlling the injection flow rate and injection time of the bacterial solution to be tested, wherein the methods for setting the injection flow rate and injection time include:

[0056] The dilution concentration of the bacterial solution to be tested is detected, and the loading probability of a single cell in the microchamber is estimated by combining the volume of the microchamber. The target single cell distribution density is determined based on the loading probability.

[0057] In this embodiment, a hemocytometer combined with a microscopic imaging system is used to count cells in the bacterial suspension to be tested. The number of cells per unit volume is counted under multiple fields of view to obtain the bacterial suspension dilution concentration. For example, a cell concentration of approximately [missing value] is detected. Given that the volume of a single microchamber is approximately 0.8 nL, the theoretical average number of cells loaded per unit volume is calculated by converting the cell concentration per unit volume to the chamber volume. And substitute this average loading number into the Poisson distribution relationship. The probability of 0 cells in a single chamber can be obtained. 1 cell probability The probability of finding two or more cells is approximately 0.26. Therefore, the target single-cell loading interval is determined to be... The corresponding single-cell occupancy probability is approximately 34%–37%, and the multi-cell occupancy probability is no more than 30%. Based on this, the bacterial concentration is controlled at approximately [missing value]. scope.

[0058] The injection concentration range of the bacterial solution to be tested is set based on the target single cell distribution density, and the appropriate injection flow rate is determined according to the channel structure of the microchamber array and the total volume of the microchambers.

[0059] In a further embodiment, the bacterial concentration is controlled at approximately [value missing] based on the target single-cell distribution density. Within a certain range, to maintain a stable interval for single-cell loading probability; subsequently, the structural parameters of the microchamber array, such as the microchannel width, are statistically analyzed. Depth is Effective channel length is And calculate the cross-sectional area of ​​the microchannel to be approximately Meanwhile, the microchamber array contains 400 chambers, each with a volume of [missing information]. The total volume of the chamber is approximately And combined with the internal volume of the microchannel approximately The total filling volume of the array is approximately Based on the cross-sectional area of ​​the microchannel and the stable laminar flow conditions, the average flow velocity within the channel is controlled to be approximately... Within this range, the appropriate injection flow rate is calculated to be approximately [value missing]. And set the stepper drive pump of the injection device to approximately This allows the bacterial solution to be stably transported in the microchannels and evenly distributed into each microchamber.

[0060] The injection time required for the bacterial solution to complete the array filling is calculated based on the volume of each microchamber in the microchamber array, the total volume of the array, and the set injection flow rate.

[0061] In a further embodiment, the microchamber array comprises approximately 400 microchambers, each with a volume of approximately [missing information]. The total theoretical filling volume of the array is approximately Meanwhile, considering that the microchannel volume is approximately Therefore, approximately [amount] needs to be injected overall. Bacterial solution. At a set injection flow rate of... Under these conditions, the time required for the bacterial solution to complete array filling can be calculated to be approximately Towards the end of the filling process, the flow rate was gradually reduced to approximately [value missing] by controlling the injection device. This allows the bacterial solution to enter the microchamber stably and reduces the generation of air bubbles, thereby achieving uniform filling of the microchamber array and providing stable conditions for subsequent single-cell culture and imaging detection.

[0062] Optionally, based on the overall closure status detection results, several partition units may be divided, including:

[0063] When the overall closure status test results meet the preset independent closure conditions, each microchamber is treated as a partition unit;

[0064] In this embodiment, if the overall closed state matrix is ​​less than the conductivity threshold, it is determined that there is no liquid communication between the microchambers, and each microchamber is managed as an independent partition unit. Subsequently, a two-dimensional coordinate index table is established according to the array arrangement order, for example, (i,j) represents the microchamber in the i-th row and j-th column, and each microchamber is individually marked as a partition unit, thereby forming a single-chamber-single-partition structural mapping relationship, providing a stable spatial index for subsequent image extraction and colony identification.

[0065] In another embodiment, a closed-state detection image of the micro-chamber array can be acquired using a microscopic imaging device, and the closed region at the boundary of each micro-chamber can be identified based on a grayscale threshold segmentation method. When the grayscale value of the boundary between adjacent micro-chambers stabilizes within a preset threshold range (e.g., grayscale value greater than 180 and continuous width greater than...), the closed region is determined. When the boundary is not connected to the liquid, it is determined that there is no liquid connection at that boundary.

[0066] When the overall closed state detection result does not meet the preset independent closed conditions, the liquid communication state between each microchamber is detected, and the combined area of ​​interconnected microchambers is taken as a partition unit according to the liquid communication state.

[0067] In another embodiment, if there is a case in the overall closed state matrix where the conductivity value is greater than the conductivity threshold, it is determined that there is liquid communication between the corresponding microchambers. Subsequently, a set of microchamber connectivity relationships is constructed based on the connectivity relationships in the conductivity matrix where the conductivity value is greater than the threshold, and the interconnected microchambers are combined to form a combined region, which is then managed as a partition unit by numbering.

[0068] In another embodiment, the continuity of liquid grayscale in the boundary regions of each microchamber is extracted from the closed-state detection image. By calculating the grayscale gradient and the continuity of liquid texture in the boundary regions of adjacent chambers, liquid connectivity is determined when the grayscale change in the boundary region is less than a preset threshold (e.g., grayscale gradient less than 10) and the width of the continuous region is greater than 8 μm. Subsequently, a microchamber connectivity matrix is ​​constructed, where 0 and 1 represent whether the chambers are connected. Region merging is performed based on the connectivity relationships in the matrix, combining interconnected microchambers into a combined region. For example, 3 to 5 connected chambers are divided into the same partition unit, thereby ensuring that each partition unit contains actual connected space while maintaining the culture independence between different partition units.

[0069] Optionally, the full-field scanning imaging in step S2 includes:

[0070] The full-coverage scanning path of the micro-cavity array is determined based on its physical size and arrangement.

[0071] In this embodiment, the scanning range is determined based on the physical dimensions and array arrangement of the micro-cavity array. For example, the overall size of the micro-cavity array is 20mm × 20mm, the array is arranged in a 100 × 100 pattern, and the spacing between each micro-cavity is 200. The center-to-center spacing of each microchamber is 200. The imaging platform uses a 4× magnification objective lens, corresponding to a single imaging field of view of approximately 2mm × 2mm. The number of scans to cover is calculated based on the array size and field of view size: In the X direction, [the required area is...]. Ten fields of view are needed in the Y direction, so the array is divided into 10×10 scanning field of view units. To ensure stable image stitching, an overlap of approximately 0.1 mm is reserved between adjacent fields of view, and the field of view step distance is recalculated to 1.9 mm. A scanning coordinate table is then established with the top left corner of the array as the starting coordinate (0,0), and a complete scanning path is generated in a serpentine pattern of "first row from left to right, second row from right to left," thus forming a full-coverage scanning path table covering the entire microcavity array. The positions of each scanning field of view are recorded as a two-dimensional coordinate sequence for subsequent scanning control.

[0072] When the preset time interval is reached, the microcavity array is scanned according to the full-coverage scanning path through the two-dimensional displacement mechanism of the preset imaging platform;

[0073] In a further embodiment, a fixed time interval is set to trigger the scanning operation under a constant temperature incubation environment, for example, a full array scan is performed every 10 minutes. When the preset time interval is reached, the two-dimensional displacement mechanism of the imaging platform moves sequentially to each field of view position according to a pre-generated scanning path. The two-dimensional displacement mechanism includes linear drive structures for the X and Y axes, and its minimum step resolution is [missing information]. Movement speed set to The system moves point by point according to the coordinate sequence in the scanning path, stopping at each field of view position for about 0.3 seconds to complete image acquisition, thereby gradually acquiring a multi-field image sequence covering the entire microcavity array and forming the array scanning image data corresponding to the current time point.

[0074] The images corresponding to each field of view in the stitched scanning results are combined, and an index mapping relationship is established for the image stitching results to obtain the chamber image set of each partition unit.

[0075] In a further embodiment, after completing the scanning of the entire field of view, the images of each field of view are stitched together based on the field of view coordinate information recorded in the scanning path. First, the spatial offset between images is calculated based on a preset 5% overlap area between adjacent fields of view, and the overlapping areas are aligned using an image registration method. Then, all field of view images are stitched together according to the array arrangement order to form a complete array image. At the same time, a spatial index table is established based on the aforementioned partition unit division results. For example, (i,j,k) represents the image data corresponding to the partition unit in the i-th row and j-th column at the k-th scan time. Images from multiple time points belonging to the same partition unit are grouped in chronological order to obtain the chamber image set corresponding to each partition unit.

[0076] Of particular importance, scanning the microchamber array according to the full-coverage scanning path also includes:

[0077] When the two-dimensional displacement mechanism reaches the preset field of view position, it performs focal length correction based on the detected current focal plane state, thereby capturing the initial image of the current field of view position;

[0078] In this embodiment, after the two-dimensional displacement mechanism moves to the specified field-of-view coordinates in the scanning path, the current focal plane state is first detected by the autofocus structure of the imaging platform. The autofocus structure includes a fine-tuning lifting platform that moves along the Z-axis and a focal length evaluation unit. The minimum adjustment step of the lifting platform is... The system is in The focal length is gradually adjusted within the specified focal length range, and a low-resolution preview image is acquired at each focal length position. The optimal focal position is determined by calculating the gradient sharpness value of the image. When the sharpness value reaches its maximum value, this position is determined as the optimal focal plane for the current field of view, and an initial image with a resolution of 2048×2048 pixels is acquired at this focal plane position as a reference image for subsequent imaging parameter adjustments.

[0079] Based on the grayscale distribution of the initial image, the exposure time and light intensity of the imaging platform are controlled, and the last image of the current field of view is captured at the preset stable time point after the imaging parameters are controlled.

[0080] In a further embodiment, statistical analysis is performed on the grayscale histogram of the initial image. Specifically, the 2048×2048 pixel image is divided into 16×16 sub-regions, and the average grayscale value and grayscale distribution range of each sub-region are calculated. The average grayscale intensity of the entire image is also statistically analyzed. When the average grayscale value is lower than a preset grayscale threshold of 120, the light intensity is increased by 10% to 15% by controlling the driving current of the LED ring light source; when the average grayscale value is higher than 180, the light intensity is reduced accordingly. Simultaneously, the exposure time of the CMOS sensor is adjusted according to the grayscale distribution range, for example, by gradually adjusting the exposure time within the range of 5ms to 20ms. After completing the exposure and light intensity control, the system waits for approximately 0.2s of optical stabilization time, and then acquires another frame as the last image at the current field of view position.

[0081] Perform image registration and fusion processing on the initial image and the last image at the current field of view position, and use the fusion processing result as the chamber image at that field of view position.

[0082] In a further embodiment, after obtaining the initial and final images, spatial alignment is performed on the two images. First, an image registration method based on feature point matching is used to extract corner features from the two images and establish corresponding matching relationships. Based on the matching results, a two-dimensional translation offset is calculated to correct the spatial position of the final image. After registration, pixel-level fusion processing is performed on the two images. Specifically, a weighted average of the grayscale values ​​at corresponding pixel positions is calculated, with the initial image weight set to 0.4 and the final image weight set to 0.6, to improve the image signal-to-noise ratio and reduce the impact of illumination fluctuations. Finally, a stable fused image is obtained, and this fusion result is recorded as the chamber image at the current field of view for subsequent colony identification and growth analysis.

[0083] Optionally, identifying a positive chamber in step S4 includes:

[0084] Candidate positive chambers are selected based on the contour area and morphological parameters of the image contour corresponding to each partition unit.

[0085] In this embodiment, the area, roundness, and aspect ratio of each grayscale contour region are used as judgment parameters. The area is obtained by counting the number of grayscale pixels within the contour (the number of pixels exceeding the grayscale value threshold of 120), and the roundness is calculated based on the relationship between the grayscale contour area and its perimeter. To avoid noise interference, a minimum contour area threshold of 40 pixels and a maximum contour area threshold of 5000 pixels are set, while the roundness threshold is limited to the range of 0.55 to 1.20. When both the area and roundness of a contour region fall within the above threshold range, the corresponding microcavity is marked as a candidate positive cavity, and the center coordinates and contour boundary information of the contour are recorded to establish a spatial index for the candidate cavities.

[0086] Based on the image contours at each time point, bacterial cell morphology parameters were extracted, and corresponding morphological time series data were constructed.

[0087] In a further embodiment, after obtaining candidate positive chambers, the image contours corresponding to each time point during the culture process are continuously extracted. Specifically, according to a preset imaging time interval (e.g., images are acquired every 10 minutes), the area, perimeter, and equivalent diameter of the bacterial cell contour within the same chamber are extracted at 12 consecutive time points. These parameters are then arranged in chronological order to construct a morphological time series data matrix, where the matrix rows represent the time point sequence and the columns represent different morphological parameters. For example, a T×3 data matrix is ​​formed (T being the number of time points). This time series data can reflect the expansion and changes of the bacterial cells during the culture process, providing basic data for subsequent growth curve fitting.

[0088] Figure 2 This is an image of the chamber of a partitioned unit in an embodiment of the present invention; identifying bacterial cells based on this chamber image includes extracting the pixel grayscale values ​​within the image outline to form a grayscale matrix; subsequently calculating the average grayscale value of the grayscale matrix. and grayscale standard deviation To characterize the brightness distribution features of the contour region; in this embodiment, if and If the region is identified as a candidate region where bacteria may exist, then the region is considered a candidate region. A threshold of 120 corresponds to dark-colored bacterial pixels. A threshold of 15 is used to exclude regions with uniform background noise. Next, the grayscale matrix of the candidate regions is binarized, and the Otsu thresholding method is used to automatically calculate the binarization threshold T, generating a binary image. ;in Represents bacterial cell pixels. Background information; finally, statistics. The area and shape features of consecutive pixel blocks, and compared with a preset area threshold (e.g. ) and roundness threshold (e.g. The system performs a matching operation. If the conditions are met, the contour region is determined to be a real bacterial cell.

[0089] The growth curves of the bacterial cells are determined based on morphological time series data to obtain a set of bacterial cell growth kinetic parameters.

[0090] In a further embodiment, the data on the change of cell area over time is used as a growth index, and a Logistic growth function is used for fitting, wherein the curve parameters include the maximum growth rate. Incubation period and final bacterial cell area The system calculates the above parameters using a least-squares fitting method, thereby forming a set of bacterial growth kinetic parameters.

[0091] If the set of bacterial growth kinetic parameters of any candidate positive chamber meets the preset positive determination rule, then the candidate positive chamber is determined to be a positive chamber.

[0092] If any parameter in the set of bacterial growth kinetic parameters of a candidate positive chamber does not meet the positive determination rule, then the candidate positive chamber is determined to be a negative chamber.

[0093] In a further embodiment, a comprehensive determination is performed on candidate positive chambers. Specifically, the maximum growth rate is considered. Incubation period and final bacterial cell area In this embodiment, the maximum growth rate is used as the criterion for judgment. The threshold is set to Incubation period The threshold is set to Final bacterial cell area The threshold is set to If the maximum growth rate of the bacterial cells corresponding to any partition unit... And the incubation period And the final bacterial cell area If the microchamber corresponding to the partition unit is determined to be a positive chamber; if the maximum growth rate of the bacterial cells corresponding to any partition unit is... or incubation period or final bacterial cell area If so, the microchamber corresponding to the partition unit is determined to be a negative chamber.

[0094] Optionally, determining the predicted loading number of monoclonal growth chambers in step S4 includes:

[0095] In each monoclonal growth chamber, the loading number of a single bacterium was randomly assumed, and a monoclonal Poisson distribution model was established.

[0096] In this embodiment, the loading number of a single bacterium in the monoclonal growth chamber is randomly assumed. Specifically, within the partition unit corresponding to each monoclonal growth chamber, the discrete value range of the loading number of a single bacterium is set to 0-5 cells, and a set of loading number sample sequences is generated according to a random sampling method, for example, 1000 sets of loading number samples are generated between 0 and 5 using a uniform random function. Subsequently, the loading number... A single-clonal Poisson distribution model is established for discrete variables, where the model uses the average loading number. Statistical parameters, and passed through the Poisson probability function. Calculate the probability value when the loading number is 0 or 1, and form a probability matrix structure consisting of the loading number vector {k0,k1} and the corresponding probability vector {P0,P1}, where P0 represents the probability that the chamber is empty and P1 represents the probability that a single cell is loaded into the chamber. This probability matrix is ​​used to describe the statistical distribution characteristics of random loading of a single bacterium in a monoclonal chamber.

[0097] The total number of monoclonal growth chambers and negative chambers in the microchamber array is counted, and the proportion of negative chambers in the total number is input into the monoclonal Poisson distribution model to inversely determine the predicted loading number corresponding to each monoclonal chamber.

[0098] In a further embodiment, a statistical count is performed on the entire microchamber array to record the total number of monoclonal growth chambers and negative chambers. For example, in an array containing 4096 partitioned units, the proportion of negative chambers is statistically determined. The actual proportion of negative chambers is then input into a monoclonal Poisson distribution model, based on the Poisson distribution... The relationship between the loading numbers of the monoclonal chamber was solved through numerical iteration. The iteration step size was set to 0.01 during the calculation process, and the iteration was stopped when the error was less than 0.001, so as to obtain the predicted number of monoclonal chambers loaded that matched the proportion of negative chambers.

[0099] Optionally, determining the number of missing bacteria to be added in each monoclonal compartment in step S5 includes:

[0100] The time point at which the positive signal appears in each monoclonal chamber is determined based on the exponential growth of the growth kinetic parameter set of each monoclonal chamber.

[0101] In this embodiment, a pre-defined microbial growth characteristic database is invoked to determine the standard growth parameters of the target microbial species. This database is indexed by the species number, and each record contains the maximum growth rate, latency period, and typical time range of the exponential growth phase, stored in a parameter table structure. After obtaining the growth kinetic parameter set of the monoclonal chamber, this parameter set is matched with the parameters of each microbial species in the database to determine the target microbial species, and the exponential growth phase is identified in a sampling sequence with a time resolution of 2 minutes. When the growth rate change rate is less than 5% over three consecutive sampling periods, this time period is determined as a stable exponential growth interval, and the start time of this interval is taken as the time point when a positive signal appears.

[0102] It is worth noting that the construction process of the bacterial growth characteristic database includes selecting the target bacterial species to be studied, continuously monitoring the bacterial growth process under standard culture conditions, and collecting basic growth data such as bacterial count, fluorescence signal intensity, and colony area at different time points. Subsequently, the collected data is processed into a time series, and key growth parameters of the bacteria are extracted using curve fitting methods, including parameters such as latency, maximum growth rate, duration of the exponential growth phase, and upper limit of bacterial count in the stationary phase. After obtaining various growth parameters, the parameters of different bacterial species are structured and organized to establish database record units indexed by the bacterial species number. Each record unit includes basic information about the bacterial species, a growth kinetic parameter table, and a calibration relationship matrix between fluorescence signal and bacterial count. The growth kinetic parameter table stores the latency, maximum growth rate, and characteristic parameters of typical growth curves, while the calibration relationship matrix stores the mapping relationship between fluorescence signal intensity ranges and corresponding bacterial count ranges.

[0103] The fluorescence signal intensity values ​​at the time points when the positive signal appeared were collected, and the number of bacteria at the time points when the positive signal appeared in each monoclonal chamber was determined based on the fluorescence signal intensity values.

[0104] In a further embodiment, after determining the time point where the positive signal appears, the fluorescence image of the corresponding time frame is retrieved from the fluorescence image sequence according to the time index. The image is then segmented based on pre-calibrated chamber spatial coordinates to extract the image region corresponding to each monoclonal chamber. The average pixel grayscale value is calculated as the fluorescence signal intensity value within each chamber region, and a 3×3 pixel window mean filter is applied to this region to reduce imaging noise interference. Subsequently, a fluorescence calibration sub-table from the bacterial growth characteristic database is called. This sub-table forms a calibration matrix with bacterial quantity ranges and fluorescence intensity ranges. Through table lookup and linear interpolation, the fluorescence intensity value is mapped to the corresponding bacterial quantity, thereby obtaining the bacterial quantity data for each monoclonal chamber at the positive time point.

[0105] It is worth noting that the fluorescence images corresponding to the time frames are continuously acquired by the fluorescence imaging system of the microchamber array. Specifically, during the cultivation and monitoring process, a fluorescence camera set on the microscopic imaging platform takes periodic pictures at fixed time intervals, such as 2 minutes per acquisition cycle, to perform full-field imaging of the microchamber array, and stores the acquired images in chronological order as an image sequence file. When the system identifies the time point of the positive signal based on the growth kinetic parameters, it reads the corresponding frame image according to the time index of that time point in the image sequence. For example, when the sampling sequence number is t, the system retrieves the t-th frame fluorescence image from the image cache or storage array, and performs region segmentation of the image using pre-calibrated chamber position coordinates to obtain the fluorescence image region corresponding to each positive chamber, which is used for subsequent fluorescence intensity extraction and cell count calculation.

[0106] Based on the number of bacteria at the time point of positive signal appearance and the corresponding exponential growth, the initial number of cells in each monoclonal chamber is inferred, and the corresponding predicted loading number is used as a prior range to constrain the initial number of cells for probability constraint correction, so as to obtain the range of initial single cell number in each monoclonal chamber.

[0107] In a further embodiment, after obtaining the number of bacteria at the positive signal moment, the exponential growth process is discretized and back-calculated based on the maximum growth rate and latency parameters in the bacterial growth characteristic database. This back-calculates the number of bacteria at the positive signal moment back to the initial moment, obtaining an estimate of the initial cell number. Subsequently, a predicted loading number is introduced as a priori constraint, reflecting the expected range of cell loading in the monoclonal chamber. A probability distribution interval is constructed centered on this predicted loading number, and the back-calculated initial cell number is probabilistically filtered. A cumulative probability threshold of 0.95 is used to determine the range of initial single-cell numbers that satisfy the prior distribution constraint, thereby outputting the initial single-cell number interval for each monoclonal chamber.

[0108] Based on the initial single-cell number range and corresponding exponential growth of each positive chamber, the bacterial growth trajectory of each positive chamber is simulated and the bacterial number distribution is statistically analyzed. The results are then compared with the chamber image set of each partition unit to determine the actual single bacterial number of each monoclonal growth chamber.

[0109] In a further embodiment, a cell growth trajectory simulation framework is constructed based on growth parameters from a bacterial strain growth characteristic database. This framework generates a growth matrix with time series as the horizontal axis and cell number as the vertical axis, and discretizes the exponential growth relationship into a growth sequence with a fixed time step, for example, setting the time step to 2 minutes and the total simulation duration to 240 minutes, thus forming a time series matrix of length 120. Subsequently, a random perturbation coefficient is introduced into each time step to reflect growth fluctuations in the actual culture environment. The perturbation coefficient is randomly sampled according to a normal distribution with a mean of 0 and a standard deviation of 0.05, and applied to the maximum growth rate parameter in the bacterial strain growth characteristic database. Based on this perturbation mechanism, a Monte Carlo simulation of the cell growth process is performed, for example, 500 independent random simulations are executed. Each simulation generates a complete cell growth trajectory, resulting in a growth trajectory matrix of 500×120 dimensions. To further improve the stability of the simulation results, a local growth slope can be calculated for each trajectory, and trajectory samples with abnormally deviating growth rates by ±15% can be removed, retaining only the set of trajectories that meet the growth characteristic constraints. The selected growth trajectories were then converted into grayscale matrices, establishing a linear mapping between bacterial count and image grayscale values. Structural similarity was then calculated between these matrices and the corresponding chamber image sequences, using a structural similarity index matrix with a matching threshold of 0.9. When the similarity of a simulated trajectory exceeded this threshold, it was determined that the bacterial count distribution corresponding to that trajectory most closely resembled the actual growth pattern of the chamber, thus determining the actual number of single bacteria in each monoclonal growth chamber.

[0110] The actual number of single bacteria in each monoclonal growth chamber was counted to determine the current number of single bacteria to be added in each positive chamber.

[0111] In a further embodiment, after determining the actual number of single bacteria corresponding to each positive chamber, statistical analysis is performed on the entire microchamber array. Specifically, the actual number of single bacteria in each positive chamber is recorded in a statistical matrix and compared with the target single-cell loading quantity. For example, when the target single-cell loading quantity is set to 1 cell, chambers with an actual single-bacterial quantity less than 1 are marked, and the difference is calculated as the quantity to be filled. A list of filling requirements can be generated according to the partition unit order, and a local average correction can be performed on 10 adjacent chambers to reduce local random errors, thereby obtaining a stable distribution result of the number of single bacteria to be filled.

[0112] Optionally, this application also provides a digital clone precise counting and identification system 100 for performing the digital clone precise counting and identification method as described above, the digital clone precise counting and identification system comprising:

[0113] The sample injection module 101 is used to inject the serially diluted bacterial solution to be tested into a preset microchamber array and to seal the injected microchamber array to construct several partitioned units.

[0114] The scanning imaging module 102 is used to place the microchamber array in a preset constant temperature culture environment for parallel culture, and to perform full-field scanning imaging on each partition unit at preset time intervals to obtain the chamber image set of each partition unit.

[0115] The image processing module 103 is used to perform image segmentation on the chamber image set of each partition unit, and extract the image contour corresponding to each partition unit based on the image segmentation result.

[0116] The prediction loading number determination module 104 is used to extract a set of growth kinetic parameters based on the image contour to identify positive chambers, and to divide the positive chambers into monoclonal growth chambers and polyclonal growth chambers according to the growth of each parameter in the set of growth kinetic parameters; to eliminate polyclonal growth chambers, and at the same time determine the prediction loading number of monoclonal growth chambers.

[0117] The microbial community normalization module 105 is used to deduce the actual number of single bacteria in each single-clone growth chamber based on the predicted loading number and the growth kinetic parameter set of each single-clone growth chamber, so as to determine the number of single bacteria to be added in each single-clone chamber, and perform amplification culture to make the number of bacteria in the microbial community consistent and achieve normalization.

[0118] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0119] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for accurate counting and identification of digital clones, characterized in that, Includes the following steps: Step S1: Inject the serially diluted bacterial solution to be tested into the preset microchamber array, and seal the injected microchamber array to construct several partitioned units; Step S2: Place the microchamber array in a preset constant temperature culture environment for parallel culture, and perform full-field scanning imaging of each partition unit at preset time intervals to obtain the chamber image set of each partition unit; Step S3: Perform image segmentation on the chamber image set of each partition unit, and extract the image contour corresponding to each partition unit based on the image segmentation results; Step S4: Extract the growth kinetic parameter set based on the image contour to identify positive chambers, and classify the positive chambers into monoclonal growth chambers and polyclonal growth chambers according to the growth of each parameter in the growth kinetic parameter set; remove polyclonal growth chambers, and determine the predicted loading number of monoclonal growth chambers. Step S5: Based on the predicted loading number and the growth kinetic parameter set of each monoclonal growth chamber, deduce the actual number of single bacteria in each monoclonal growth chamber to determine the number of missing bacteria in each monoclonal chamber, and perform amplification culture to ensure that the number of bacteria in the bacterial population is consistent, thus achieving normalization; the determination of the number of missing bacteria in each monoclonal chamber in step S5 includes: The time point at which the positive signal appears in each monoclonal chamber is determined based on the exponential growth of the growth kinetic parameter set of each monoclonal chamber. The fluorescence signal intensity values ​​at the time points when the positive signal appeared were collected, and the number of bacteria at the time points when the positive signal appeared in each monoclonal chamber was determined based on the fluorescence signal intensity values. Based on the number of bacteria at the time point of positive signal appearance and the corresponding exponential growth, the initial number of cells in each monoclonal chamber is inferred, and the corresponding predicted loading number is used as a prior range to constrain the initial number of cells for probability constraint correction, so as to obtain the range of initial single cell number in each monoclonal chamber. Based on the initial single-cell number range and corresponding exponential growth of each monoclonal chamber, the bacterial growth trajectory of each monoclonal chamber is simulated and the bacterial number distribution is statistically analyzed. The results are then compared with the chamber image set of each partition unit to determine the actual number of single bacteria in each monoclonal growth chamber. The actual number of single bacteria in each monoclonal growth chamber is counted to determine the current number of single bacteria to be added to each monoclonal chamber.

2. The method for accurate counting and identification of digital clones according to claim 1, characterized in that, Step S5, which involves amplification culture, includes: Based on the coordinate index information of each positive chamber, the corresponding target micro-chamber in the micro-chamber array is located, so as to open the target micro-chamber at a fixed point; Based on the number of single bacteria to be added to each target microchamber, the cultures in each target microchamber are transferred to a pre-set independent culture carrier. The amplification culture time and the corresponding number of bacteria to be added are determined based on the exponential growth of the growth kinetic parameter set corresponding to the culture, so as to carry out amplification culture in the independent culture carrier.

3. The method for accurate counting and identification of digital clones according to claim 1, characterized in that, Step S1 includes: The bacterial solution to be tested is injected into the micro-chamber array through a preset micro-injection device; wherein the micro-chamber array includes several independent and regularly arranged micro-chambers, and the volume of each micro-chamber is matched with the dilution concentration of the bacterial solution to be tested; After each microchamber in the microchamber array is filled, the open end of the microchamber array is sealed. The overall closure status of the micro-cavities after injection is detected to determine the cross-connection status. Based on the overall closed-state detection results, the area is divided into several partition units.

4. The method for accurate counting and identification of digital clones according to claim 3, characterized in that, Injecting the bacterial solution to be tested into the micro-chamber array through a preset micro-injection device includes controlling the injection flow rate and injection time of the bacterial solution to be tested, wherein the methods for setting the injection flow rate and injection time include: The dilution concentration of the bacterial solution to be tested is detected, and the loading probability of a single cell in the microchamber is estimated by combining the volume of the microchamber. The target single cell distribution density is determined based on the loading probability. The injection concentration range of the bacterial solution to be tested is set based on the target single cell distribution density, and the appropriate injection flow rate is determined according to the channel structure of the microchamber array and the total volume of the microchambers. The injection time required for the bacterial solution to complete the array filling is calculated based on the volume of each microchamber in the microchamber array, the total volume of the array, and the set injection flow rate.

5. The method for accurate counting and identification of digital clones according to claim 3, characterized in that, Based on the overall closure status detection results, the area is divided into several partition units, including: When the overall closure status test results meet the preset independent closure conditions, each microchamber is treated as a partition unit; When the overall closed state detection result does not meet the preset independent closed conditions, the liquid communication state between each microchamber is detected, and the combined area of ​​interconnected microchambers is taken as a partition unit according to the liquid communication state.

6. The method for accurate counting and identification of digital clones according to claim 1, characterized in that, Step S2, which involves full-field scanning imaging, includes: The full-coverage scanning path of the micro-cavity array is determined based on its physical size and arrangement. When the preset time interval is reached, the microcavity array is scanned according to the full-coverage scanning path through the two-dimensional displacement mechanism of the preset imaging platform; The images corresponding to each field of view in the stitched scanning results are combined, and an index mapping relationship is established for the image stitching results to obtain the chamber image set of each partition unit.

7. The method for accurate counting and identification of digital clones according to claim 1, characterized in that, Step S4, which identifies positive chambers, includes: Candidate positive chambers are selected based on the contour area and morphological parameters of the image contour corresponding to each partition unit. Based on the image contours at each time point, bacterial cell morphology parameters were extracted, and corresponding morphological time series data were constructed. The growth curves of the bacterial cells are determined based on morphological time series data to obtain a set of bacterial cell growth kinetic parameters. If the set of bacterial growth kinetic parameters of any candidate positive chamber meets the preset positive determination rule, then the candidate positive chamber is determined to be a positive chamber. If any parameter in the set of bacterial growth kinetic parameters of a candidate positive chamber does not meet the positive determination rule, then the candidate positive chamber is determined to be a negative chamber.

8. The method for accurate counting and identification of digital clones according to claim 1, characterized in that, Step S4, which determines the predicted loading number of monoclonal growth chambers, includes: In each monoclonal growth chamber, the loading number of a single bacterium was randomly assumed, and a monoclonal Poisson distribution model was established. The total number of monoclonal growth chambers and negative chambers in the microchamber array is counted, and the proportion of negative chambers in the total number is input into the monoclonal Poisson distribution model to inversely determine the predicted loading number corresponding to each monoclonal chamber.

9. A digital clone accurate counting and identification system, characterized in that, For performing the precise counting and identification method for digital clones as described in claim 1, the precise counting and identification system for digital clones includes: The sample injection module is used to inject the serially diluted bacterial solution to be tested into a preset microchamber array and to seal the injected microchamber array to construct several partitioned units. The scanning imaging module is used to place the microchamber array in a preset constant temperature culture environment for parallel culture, and to perform full-field scanning imaging of each partition unit at preset time intervals to obtain the chamber image set of each partition unit. The image processing module is used to segment the chamber image set of each partition unit and extract the image contour corresponding to each partition unit based on the image segmentation results. The predicted loading number determination module is used to extract a set of growth kinetic parameters based on image contours to identify positive chambers, and classify positive chambers into monoclonal growth chambers and polyclonal growth chambers according to the growth of each parameter in the set of growth kinetic parameters; polyclonal growth chambers are eliminated, and the predicted loading number of monoclonal growth chambers is determined at the same time. The microbial community normalization module is used to deduce the actual number of single bacteria in each monoclonal growth chamber based on the predicted loading number and the growth kinetic parameter set of each monoclonal growth chamber, so as to determine the number of single bacteria to be added in each monoclonal chamber, and perform amplification culture to make the number of bacteria in the microbial community consistent and achieve normalization.

Images