Fluorescent live cell long-term focal plane drift compensation method, system and apparatus

Abstract

The application provides a fluorescent live cell long-term focal plane drift compensation method, system and device, the method comprising: periodically sampling the multi-focal plane of the hole edge in the bright field, determining the hole edge observation focal plane based on the sharpness evaluation, and calculating the hole edge observation drift amount relative to the current effective reference; reading the incubator temperature rise, calculating the temperature drift prediction drift amount according to the pre-calibrated temperature drift coefficient; weighting and fusing the two-way drift amount according to the hole edge detection quality, obtaining the drift compensation amount through sliding average and superimposing it on the subsequent fluorescence collection focal plane; refreshing the current effective reference and resetting the compensation amount according to the predetermined period or deviation threshold. The application realizes the automatic maintenance of the focal plane for the long-term multi-hole site fluorescence monitoring in the incubator under the premise of not frequently using the fluorescence focusing, and is suitable for the continuous live cell imaging task for several hours to several days.

Description

Technical Field

[0001] This invention relates to the field of live-cell fluorescence imaging technology, and in particular to a method, system, and device for compensating for long-term focal plane drift in live-cell fluorescence imaging. Background Technology

[0002] In the field of live-cell fluorescence microscopy, long-term monitoring of cells within multi-well plates (e.g., for several hours to several days) typically requires an incubator with temperature and humidity control. During this continuous monitoring, the microscope stage, multi-well plate, and imaging container undergo micrometer-level thermal expansion and mechanical deformation due to minute changes in temperature and humidity in the working environment. This causes the initially aligned fluorescence imaging focal plane to gradually shift axially, resulting in defocused and blurred live-cell fluorescence images. Furthermore, due to differences in the physical and mechanical structure and heat conduction paths of each well in the multi-well plate, the degree of focal plane drift varies across different wells, making it impossible to correct for using a single global compensation value.

[0003] To address the issue of maintaining the focal plane in long-term imaging, existing technologies typically employ two main approaches. The first is the fluorescence autofocus continuous scanning method. Its principle is to activate the fluorescence excitation optical path before each formal imaging acquisition, driving the stage to perform continuous multi-layer scanning along the Z-axis, acquiring a series of fluorescent cell images at different axial positions. The optimal focal layer is then found by calculating the maximum value of the focus score curve for the entire image series. The second is the infrared hardware focusing method. Its principle relies on an independent near-infrared paraxial optical path added to the microscope system. Near-infrared laser light is obliquely incident on the reflective surface at the interface between the container bottom and the air. A position-sensitive detector captures the spatial displacement of the reflected light spot to achieve hardware focusing of the optical focal plane.

[0004] Therefore, there is an urgent need for a method, system, and device for long-term focal plane drift compensation of fluorescent live cells to solve the problems existing in the prior art. Summary of the Invention

[0005] This invention provides a method, system, and device for long-term focal plane drift compensation of fluorescent live cells. It addresses the problems of existing focal plane maintenance schemes in long-term multi-well live cell fluorescence monitoring scenarios in incubators, such as frequent fluorescence excitation leading to high phototoxicity, difficulty in deploying additional hardware optical paths, inability to independently compensate for temperature drift deformation per well, and lack of focal plane error accumulation suppression mechanism to adapt to continuous operation for several days.

[0006] The core technology of this invention is to determine the focal plane of the aperture observation by acquiring aperture edge images in bright field mode with periodic intervals, and to perform dual-source weighted fusion and moving average by combining incubator temperature drift prediction. The drift compensation amount is output according to the aperture position and superimposed on the fluorescence acquisition focal plane. At the same time, the reference is periodically refreshed according to a predetermined strategy to suppress long-term cumulative error.

[0007] In a first aspect, the present invention provides a method for compensating for long-term focal plane drift in fluorescent live cells, the method comprising the following steps:

[0008] During the inspection period, multi-focal surface sampling is performed on the aperture edge in bright field mode. The aperture edge observation focal plane is determined based on sharpness evaluation, and the aperture edge observation drift relative to the current effective reference is calculated. Read the temperature change information of the incubator and calculate the predicted drift amount based on the pre-calibrated temperature drift coefficient; The quality of borehole edge detection is comprehensively evaluated based on multiple observable indicators. Based on the evaluation results, the observed drift amount and the predicted drift amount are assigned fusion weights. The two are weighted and merged according to the fusion weights and then the drift compensation amount is obtained by moving average. The drift compensation amount is then superimposed on the focal plane of subsequent fluorescence acquisition. The current valid baseline is updated according to the predetermined refresh strategy, and the drift compensation amount is reset according to the updated baseline.

[0009] Furthermore, the steps of sampling multiple focal planes at the aperture edge in bright-field mode and determining the aperture edge observation focal plane based on sharpness evaluation include: Using the current compensated fluorescence acquisition focal plane as the central reference point, multiple brightfield images are acquired at equal intervals within the preset focal plane range; Sharpness evaluation is performed on the region of interest at the aperture edge of the brightfield image to obtain the sharpness value corresponding to each focal plane; The focal plane corresponding to the maximum sharpness is taken as the coarse peak position. When the coarse peak position is not at both ends of the sampling range, the sharpness values ​​of the coarse peak position and its adjacent sampling points are fitted with a quadratic curve. The focal plane corresponding to the vertex of the fitted curve is taken as the focal plane of the aperture edge observation in this period. When the coarse peak position is at both ends of the sampling range, the sampling range is reduced and the step size is decreased to supplement the sampling, and the aperture edge observation focal plane is refitted to obtain the focal plane.

[0010] Furthermore, the region of interest at the edge of the hole is located by a geometric edge detection algorithm to locate the annulus on the hole wall, and the annulus with a radius equal to a preset proportion of the hole radius is taken as the region of interest at the edge of the hole; the sharpness evaluation is to calculate the Laplacian variance of the grayscale image in the region of interest at the edge of the hole.

[0011] Furthermore, the step of comprehensively evaluating the hole edge detection quality based on multiple observable indicators, and assigning fusion weights to the observed hole edge drift and the predicted temperature drift based on the evaluation results, includes: The quality level of the borehole edge inspection is determined by comprehensively evaluating at least one of the following: positioning deviation of the borehole wall annulus, peak position of the sharpness curve, peak contrast, and recent drift consistency. The fusion weights of the observed drift and the predicted drift are assigned according to the quality level of the borehole edge detection, with the weight of a single route limited to a preset range. The observed drift at the borehole edge and the predicted drift at temperature are weighted and summed according to the assigned weights to obtain the recommended compensation value for this period.

[0012] Furthermore, the pre-calibrated temperature drift coefficient is calibrated in segments according to temperature ranges, with different temperature ranges corresponding to different temperature drift coefficient values; and the temperature drift coefficient is corrected according to the spatial position of the hole in the orifice plate, wherein the temperature drift coefficient of the edge hole is greater than that of the center hole.

[0013] Furthermore, the moving average includes: obtaining the compensation suggestion value for the most recent N inspection cycles of this hole position, calculating the arithmetic mean of the most recent N compensation suggestion values, and using the resulting average value as the current drift compensation amount for this hole position, where N is an integer between 3 and 7.

[0014] Further, the steps of updating the current valid benchmark according to a predetermined refresh strategy and resetting the drift compensation amount include: Record the starting reference focal plane of the task and the current valid reference focal plane that participates in the drift calculation independently for each aperture position; When the timer period arrives, the sharpest focal plane is confirmed by sampling the edge of the bright field aperture, the current effective reference is updated to that focal plane, and the drift compensation is reset to zero or the residual is checked. When the absolute value of the drift compensation exceeds the preset upper limit threshold, the continuous fluorescence clarity is lower than the threshold, or the aperture edge and temperature drift estimate continuously exceed the tolerance, the current effective benchmark will be forcibly refreshed. For well sites that have not been monitored for a long time, the sampling inspection benchmark is refreshed according to a preset ratio.

[0015] Furthermore, it also includes a low-dose fluorescence verification step: Fluorescence verification is triggered when the deviation between the observed drift at the aperture edge and the predicted drift at temperature exceeds the first threshold, the change in the compensation recommendation value for the current period relative to the previous period exceeds the second threshold, or the fluorescence clarity for a consecutive preset number of frames is lower than the preset proportion of the historical average. Multiple verification images are acquired at equal intervals within a preset range near the focal plane corresponding to the current drift compensation amount, using an excitation dose lower than that used in the formal fluorescence acquisition. The drift compensation amount for this well is corrected based on the sharpness fitting peak position of the verification image in the cell region.

[0016] In a second aspect, the present invention provides a long-term focal plane drift compensation system for fluorescent live cells, comprising: The bright-field imaging unit is configured to acquire multiple images of the hole edge in a bright-field manner within a preset focal plane range during the inspection cycle. Temperature reading unit, configured to acquire temperature change information of the incubator; Temperature drift calibration storage unit, configured to store pre-calibrated temperature drift coefficient; The aperture edge analysis unit is configured to determine the aperture edge observation focal plane based on the sharpness evaluation function and compare it with the current effective benchmark to obtain the aperture edge observation drift. The drift fusion unit is configured to comprehensively evaluate the quality of aperture edge detection based on multiple observable indicators, and assign fusion weights to the aperture edge observed drift and temperature drift predicted drift based on the evaluation results. The two are weighted and merged according to the fusion weights and then the drift compensation is obtained by moving average. The drift compensation is then superimposed on the focal plane of subsequent fluorescence acquisition. The reference refresh unit is configured to update the current valid reference according to a predetermined refresh strategy, and reset the drift compensation amount according to the updated reference. The borehole position compensation recording unit is configured to maintain reference information and compensation history according to the borehole position.

[0017] Thirdly, the present invention provides an electronic device including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the above-described method for long-term focal plane drift compensation of fluorescent live cells.

[0018] The main contributions and innovations of this invention are as follows: 1. This invention uses bright field detection instead of fluorescence for focal plane detection in routine monitoring cycles, and only uses low-dose fluorescence for verification under abnormal conditions. Fluorescence excitation is not required in more than 90% of the inspection cycles, which greatly reduces the phototoxic exposure of live cells in long-term monitoring and helps maintain the stability of cell activity and physiological state.

[0019] 2. This invention maintains an independent drift compensation record for each borehole location and combines borehole edge observation and temperature drift prediction information for fusion estimation. It can provide personalized compensation for the differential deformation caused by different spatial positions of each borehole location, overcoming the problem of insufficient accuracy of uniform compensation across the entire plate.

[0020] 3. This invention adopts a dual-reference architecture, which periodically resets or zeros the drift compensation amount through a periodic reference refresh mechanism. Combined with moving average filtering, it suppresses single random errors and effectively avoids the unidirectional accumulation of drift compensation amount during several days of continuous operation, so that the focal plane can be maintained for several days or even more than a week.

[0021] Details of one or more embodiments of the present invention are set forth in the following drawings and description, so that other features, objects and advantages of the invention will be more readily understood. Attached Figure Description

[0022] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings: Figure 1This is a flowchart of a method for compensating for long-term focal plane drift of fluorescent live cells according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the architecture of a fluorescent live cell long-term focal plane drift compensation system according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0023] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with one or more embodiments of this specification. Rather, they are merely examples of apparatuses and methods consistent with some aspects of one or more embodiments of this specification as detailed in the appended claims.

[0024] It should be noted that the steps of the corresponding methods are not necessarily performed in the order shown and described in this specification in other embodiments. In some other embodiments, the methods may include more or fewer steps than described in this specification. Furthermore, a single step described in this specification may be broken down into multiple steps in other embodiments; and multiple steps described in this specification may be combined into a single step in other embodiments.

[0025] In the field of live-cell fluorescence microscopy, long-term continuous monitoring of cells within multi-well plates typically requires an incubator with temperature and humidity control. During continuous operation for hours to days (up to 7 days), the stage and well plate slowly deform with temperature changes, causing a micrometer-level shift in the original fluorescence alignment focal plane. To avoid the high phototoxicity and photobleaching induced by frequent high-intensity fluorescence scanning in the Z-axis to find the optimal focal plane, this invention employs non-invasive, multi-level sampling and focusing at the well edges using a low-dose bright field during detection intervals. Simultaneously, it combines temperature drift prediction with changes in the incubator's ambient temperature. Based on the well edge detection quality, the two estimates are weighted, fused, and time-series smoothed to output the immediate drift compensation for that well, which is then superimposed on the focal plane of subsequent fluorescence acquisition. The compensation is reset through a periodically or forcedly triggered effective benchmark recalibration mechanism, thereby suppressing the cumulative error during long-term continuous operation.

[0026] The following describes in detail and comprehensively the specific implementation methods of the present invention in conjunction with several specific embodiments and application scenarios.

[0027] Example 1 This invention proposes a method for long-term focal plane drift compensation in fluorescence live cell imaging. Specifically, this embodiment uses a live cell microscopy imaging system equipped with a 20× dry objective lens and a sealed incubator as the application scenario to perform long-term monitoring of a typical 96-well glass substrate containing culture medium. The system automatically triggers a standard focusing and compensation procedure without fluorescence excitation every 2 hours, after every 20 fluorescence acquisitions or when switching to a new well position, and when more than 3 hours have passed since the last bright-field examination. Figure 1 As shown, the specific steps include: Step 1: Non-invasive multilayer sampling at the edge of the bright-field aperture and quadratic curve extreme value fitting Since the central area of ​​the multi-well plate contains live cells to be monitored, in order not to interfere with cell growth and reduce light dose, the system uses the extremely low dose bright field illumination that microscopes are used for routine edge band sampling. The bright field exposure time is set to 5% to 10% of the formal fluorescence acquisition exposure time.

[0028] First, using Hough circle detection or edge gradient detection algorithms in computer vision, the circular rings on the container pore wall are identified and located on low-magnification bright-field images. The system automatically extracts a ring band with a radius of 0.85 to 0.95 times the pore radius as the region of interest at the pore edge, with the ring band width set to 8 to 12 pixels. This region is located at the interface between the pore wall and the liquid surface and does not include the central cell distribution area.

[0029] Subsequently, using the planned focal plane position (already superimposed with the historical time-series drift compensation of this aperture) recorded by the current reading of the stage's Z-axis encoder as the central reference point, five non-invasive bright-field images were automatically acquired at equal intervals within a span of ±4µm in both the negative and positive Z-axis directions with a fixed axial step size of 2µm. The focal plane axial coordinates corresponding to these five bright-field images are, in order: center point -4µm, center point -2µm, center point, center point +2µm, and center point +4µm. This ±4µm sampling range covers the normal thermal drift within a single cycle and leaves a margin.

[0030] After acquiring the five bright-field images at different axial levels, the image grayscale matrix is ​​calculated within the defined region of interest at the edge of the aperture. The statistical variance of the second derivative response is used as the sharpness score of the focal plane. Based on the Laplace operator The characteristic of being sensitive to high-frequency details in the contour is calculated using the following formula:

[0031] Statistical variance The higher the value, the richer the high-frequency details and the clearer the focal plane at the interface between the hole edge and the liquid surface in the current sampling level. The system compares the sharpness scores of the five sampling levels. The level with the largest extremum is extracted as the candidate focal plane for the coarse peak. When the coarse peak is not located at either end of the sampling range, the system extracts the score and axial position of the coarse peak and one level to its left and right (a total of three discrete data points), and constructs a quadratic parabolic function model:

[0032] in , , These are the coefficients for fitting the quadratic curve. The coefficients of the quadratic term of this fitted function are obtained by solving for them. coefficient of the first term and constant term Directly solve for the theoretical axial coordinates of the vertex of the fitted parabola:

[0033] The system will calculate the extreme value coordinates The aperture edge observation focal plane for this cycle is confirmed. Finally, the aperture edge observation drift relative to the effective reference focal plane is calculated by subtracting the currently effective reference focal plane, which is independently maintained at the bottom layer of the aperture location, from this aperture edge observation focal plane. :

[0034] in This is the current effective reference aperture focal plane.

[0035] Step 2: Segmented temperature drift mapping model modeling and environmental temperature drift feedforward estimation Due to differences in thermal expansion and conduction rates among the components of the imaging system during incubator heating and environmental fluctuations, the focal plane displacement exhibits a non-linear mapping relationship with temperature changes. This embodiment pre-calibrates and integrates a piecewise temperature drift mapping model into the imaging system.

[0036] During calibration, a typical 96-well glass plate containing culture medium was placed on the microscope stage, and the temperature of the incubator was recorded using its built-in temperature reading unit. The temperature was slowly increased in 0.5°C increments from the normal incubation temperature (e.g., 37.0°C) to the stable upper limit (e.g., 39.0°C). After each temperature increment and holding for 30 minutes, a Z-axis displacement scan was performed on the representative wells, and the Z-axis encoder reading corresponding to the highest resolution was recorded. Using the initial reading of 37.0°C as the reference zero point, the Z-axis displacement relative to the reference for subsequent temperature increments was calculated and divided by the temperature increase. The basic coefficient of axial drift per unit temperature for each segment interval was obtained through piecewise fitting. In this embodiment, with a 20× dry objective lens and a 96-well plate configuration, the basic coefficient of unit temperature for the segment intervals and the central core representative well (e.g., well B5) is as follows: In the range of 37.0 to 37.5°C, the coefficient... It is 1.8µm / ℃; in the range of 37.5 to 38.5℃, the coefficient is... The coefficient is 2.5µm / ℃; when the temperature is above 38.5℃, the coefficient... It is 3.2µm / ℃.

[0037] Meanwhile, due to the more efficient convection of hot air at the edges of the perforated plate, the deformation of the edge holes is significantly greater than that of the center holes. The system is configured with spatial modification parameters. : The space modification number for the center representative hole position (such as hole B5) is set to 1.0; The space modification number of the outer edge representing the hole position (such as hole A1) is set to 1.2 to 1.3 times that of the center hole (in this embodiment, 1.28 is selected by default).

[0038] During the formal long-term monitoring cycle, the temperature reading unit reads the cumulative temperature change of the incubator from the previous monitoring time to the present in real time. The system automatically determines the temperature range the current absolute temperature falls into and retrieves the corresponding base coefficient. ( The values ​​1, 2, or 3 correspond to the three temperature intervals and the spatial repair system number of the target hole position. The predicted temperature drift amount calculated using feedforward extrapolation is obtained. :

[0039] Step 3: Adaptive weighted fusion of detection quality and temporal smoothing mean filtering To suppress single-sense sensing errors caused by obstructions from liquid surface burrs, tiny bubbles, or external airflow disturbances within the field of view, the system calculates... and Perform dynamic weighted fusion and time-series smoothing mean filtering.

[0040] The system comprehensively evaluates the quality of borehole edge inspection during this inspection cycle based on the following indicators: 1. Hole wall annular positioning stability: The magnitude of the residual between the Hough circle fitting edge and the actual acquired hole edge contour; 2. Centered extreme value shape: Fitted extreme value coordinates Is it within the middle range of the normal ±4µm range? 3. Extreme Contrast: Sharpness score of the sharpest and blurriest layers among the 5 bright-field sampled images. The magnitude of the ratio; 4. Sequence consistency: Whether the jump difference between the currently calculated aperture drift and the previous adjacent period is stable; Based on the aforementioned comprehensive evaluation indicators, the system categorizes the quality of this borehole edge inspection into three levels: "Good," "Average," and "Poor." Within a preset upper and lower limit range of 30% to 80% for the weighting coefficient of a single route, the system allocates weights to both parties. When the hole edge detection quality is "good", the hole edge observation weights are assigned. Temperature drift prediction weight

[0041] When the hole edge detection quality is "average", the two contributions are balanced and allocated. ,

[0042] When the hole edge inspection quality is "poor", the allocation is... ,

[0043] The system calculates the recommended value for a single inspection in this cycle by weighting and summing the values ​​according to the assigned weights. :

[0044] To suppress mechanical vibration or accidental electrical signal noise, the system selects the nearest borehole location in the timing buffer queue. The suggested values ​​for a single inspection in a consecutive period are calculated using the arithmetic mean sliding calculation. This embodiment selects a time window length. (in other customized configurations) (The value can be an integer between 3 and 7), and the calculated output is the instantaneous drift compensation amount used for drive axial compensation. :

[0045] Step 4: Planning focal plane overlay and state database maintenance The instantaneous drift compensation amount after smoothing is calculated. Subsequently, the micro-focusing control unit directly superimposes it as an axial correction value onto the planned focal plane for the subsequent formal acquisition task. In the middle. The microscope stage drive module automatically drives the stage to move to the target physical coordinates:

[0046] Subsequently, the microscopic imaging system was positioned at the target's physical coordinates. The fluorescence excitation source and camera are then activated to acquire fluorescence images of the live cell region. Subsequently, the system writes complete well status data back to the underlying persistent well record table. Table 1 Structure of Hole Location Historical Compensation and Baseline Status Data

[0047] Example 2 This embodiment specifically discloses the corresponding adaptive encrypted sampling steps and the abnormal gating low-dose fluorescence verification mechanism. In automated monitoring, when encountering drastic environmental changes or sudden changes such as evaporation disturbances, the conventional bright-field sampling span may not be able to encompass the true focal plane. This embodiment provides closed-loop fault-tolerant and bias-correcting capabilities under complex operating conditions.

[0048] I. Adaptive Encryption Sampling Steps During routine inspections, the system acquires 5 brightfield images within a range of ±4 µm (fixed step size 2 µm) from the central reference point to calculate the sharpness score. If the system identifies that the highest value falls into the two outermost endpoints of the scanning range (i.e., the maximum value appears at -4 µm or +4 µm), the system determines that the true optimal physical focal plane has exceeded the normal search range or is in the extreme value region of the range slope.

[0049] At this point, the system does not accept extrapolated coordinates, but immediately or in the next adjacent inspection cycle triggers an adaptive encrypted sampling process for this specific aperture location: the system narrows the focusing range from ±4 µm to a fine span of ±2 µm from the current center reference point; simultaneously, it reduces the Z-axis sampling step size from the conventional 2 µm to 1 µm. The physical locations corresponding to the five bright-field images acquired at this time are, in order: -2 µm, -1 µm, 0 µm, +1 µm, and +2 µm from the center reference point. Based on the finely encrypted sampling data, the system re-executes the quadratic parabolic fitting operation to obtain new extreme point positions, updating them to the aperture edge observation focal plane for this cycle. After completing this encrypted sampling, the system automatically reverts to the conventional ±4 µm, 2 µm step size low-power mode in subsequent inspection cycles.

[0050] II. Low-dose fluorescence re-verification steps triggered by multidimensional gating In routine cycles, over 90% of focal plane tracking is handled solely by bright-field aperture edge sampling and temperature drift closure, thus avoiding continuous fluorescence irradiation. However, when abnormal deviations occur, the system sets the following three quantization gating conditions; meeting any one of them triggers fluorescence closure verification: 1. Estimated difference exceeds the limit: The absolute value of the difference between the observed drift at the borehole edge and the predicted drift due to temperature drift in the current detection cycle exceeds the limit, i.e.:

[0051] (The differential threshold is relaxed to 2.5µm during the initial 24 hours of monitoring.) 2. Excessive Compensation Sudden Change: The difference between the currently calculated instantaneous drift compensation amount and the change in the previous inspection cycle is greater than the threshold, i.e.:

[0052] in This is the instantaneous drift compensation amount for this period. This is the instantaneous drift compensation amount for the previous cycle.

[0053] 3. Significantly reduced fluorescence clarity: In the officially acquired live cell fluorescence images, the Laplacian variance clarity calculated in the central cell ROI region was lower than 70% of the average level of the most recent 10 consecutive wells for two consecutive frames (relaxed to 65% in the initial monitoring phase).

[0054] When any of the above conditions are triggered, the system initiates a closed-loop fluorescence verification process. Extremely low light dose parameters are used during verification: the duty cycle of the fluorescence excitation source is limited to 15% to 25% of that used in the formal fluorescence image acquisition; the exposure time is shortened to 20% to 30% of the formal acquisition exposure time. The verification sampling range is narrowed to a span of ±2µm near the current instantaneous target focal plane, and five low-exposure fluorescence images are quickly captured at equal intervals of 1µm.

[0055] The system extracts the central viable cell region of interest (ROI) from these five verification images, calculates the second derivative Laplacian variance in parallel, and performs parabolic fitting on the discrete scores to determine the axial absolute position of the physical layer with the clearest image of the cell. Based on this clearest layer value, the system directly performs absolute overlay and numerical reset on the originally calculated instantaneous drift compensation, completing the correction operation.

[0056] Example 3 In automated live cell monitoring tasks lasting up to 7 days (168 hours), avoiding the infinite accumulation of integral displacement caused by slow deformation of the well plate and material stress is a core technical challenge in the industry. This embodiment specifically discloses a dual-layer reference decoupling architecture and a periodic recalibration mechanism.

[0057] In the underlying motion control mapping logic, the system independently maintains two reference values ​​with completely different physical meanings and without coupling for each monitored hole position on the entire board: 1. Task starting reference focal plane (z_baseline_task): Recorded only after the first successful fluorescence focusing is completed on the first day of the experiment. This value is completely locked and remains unchanged during the subsequent 7-day operational lifecycle, and is only used for overall task sinking log output or trend verification. It is absolutely not involved in any drift displacement subtraction calculations during any inspection cycle.

[0058] 2. Current effective reference focal plane (z_baseline_active): This is the formula by which the system closed-loop control actually performs subtraction to calculate the aperture edge observation drift (i.e., the formula). The instant reference origin.

[0059] The system employs a periodic benchmark recalibration strategy for the current effective benchmark: Every 24 hours of the default monitoring cycle (e.g., during low-frequency periods like early morning), the system initiates a benchmark recalibration process for the current target well location. The system re-determines the current, clearest focal plane position using multi-layer sampling at the well edge in bright field, supplemented by ultra-low-dose fluorescence imaging. The confirmed axial coordinate values ​​are updated to the current effective benchmark focal plane field (z_baseline_active) for that well location. Simultaneously, the accumulated instantaneous drift compensation (drift_offset) calculated over the past 24 hours for that well location is either zeroed or reset to a small residual value (e.g., 0.1µm) after verification and correction. The sliding time-series queue is also cleared. In the subsequent 24-hour cycle, new observational drift values ​​are calculated relative to this updated effective benchmark, thus breaking the long-span error accumulation divergence path relative to the original starting point.

[0060] In addition to the 24-hour timed refresh, the system is equipped with a forced benchmark recalibration trigger gating: if the absolute value of the instantaneous drift compensation calculated for the target hole position in the daily weekly cycle exceeds the extreme value safety boundary (i.e., If the bright field aperture edge and temperature drift calculation data deviate from the set threshold twice consecutively, the system will immediately trigger the effective reference recalibration process and reset the historical compensation amount to zero.

[0061] In addition, for tasks where only a small number of wells in a perforated plate are monitored at high frequency, in order to prevent irreversible deformation of static, unmonitored wells that have been in an unvisited state for a long time due to prolonged constant temperature, the system's internal clock automatically performs a pre-set ratio (e.g., randomly selecting 10% of the wells) sampling benchmark recalibration process on the static, unvisited wells every 48 hours. This ensures that the Z-axis spatial geometric coordinate mapping field of the entire perforated plate remains consistent throughout the 7-day task cycle, guaranteeing the focusing accuracy of the wells that are sampled or called up on an ad hoc basis.

[0062] Example 4 like Figure 2 As shown, this embodiment discloses in detail the implementation method corresponding to the long-term focal plane drift replenishment system for fluorescent live cells.

[0063] The fluorescence live-cell long-term focal plane drift compensation system is integrated into the industrial control system architecture of the live-cell imaging system, and includes the following basic units: The bright-field imaging unit consists of a microscope bright-field light source, a well stage, objectives, a Z-axis drive mechanism with a micrometer-level stepper motor, and a digital camera focused on the imaging optical path. It is configured to perform multi-level tomographic bright-field acquisition of the region of interest at the edge of the target well during the fluorescence cycle acquisition intervals.

[0064] Temperature reading unit: It consists of at least one temperature sensor and an analog-to-digital converter chip distributed inside the sealed incubator and near the edge of the stage. It can continuously read and report the cumulative sequence of ambient temperature fluctuations to the core platform in real time.

[0065] Temperature drift calibration storage unit: It is carried by an embedded non-volatile storage medium and internally persistently stores entries of segmented temperature drift basic coefficient table and spatial modification coefficient mapping table for adapting to the geometric topology of the perforated plate.

[0066] The hole edge analysis unit is carried by a processing kernel with computing power. It performs geometric contour edge detection, Laplacian second-order variance calculation, and quadratic parabolic numerical fitting solution to generate the hole edge observation drift.

[0067] Drift fusion unit and reference refresh unit: integrate floating-point calculation logic. They read the evaluation metrics converged from the detection channel and dynamically allocate weighted fusion coefficients. They maintain a timing sliding window for each target hole position in its memory buffer and output real-time axial positioning correction instructions. When the refresh trigger gate is reached, they invoke the laser and electronic shutter to verify and confirm the updated valid origin of the target hole.

[0068] Hole position compensation record unit: Based on the system relational database and the underlying shared memory architecture, an independent index is created for each hole position, and all benchmark fields in Table 1 are dynamically maintained to provide parameter retrieval guidance when scanning multiple holes.

[0069] Scenario Example 1: 48-hour routine continuous monitoring Initial conditions: Select the core well position C5 in the microplate to be measured, and accurately record the absolute coordinates of the reference focal plane at the starting point of the task. After 12 hours of stable operation, a routine open-field check is triggered.

[0070] Focusing and Prediction: Positive and Negative at the Reference Point Within the range Five bright-field images were acquired using a step distance method. The focal plane for aperture observation was obtained through calculation and fitting. The drift amount observed at the edge of the hole was obtained. The temperature reading unit reports the temperature difference. , in Middle interval, coefficient Central regional spatial coefficient Calculate the predicted drift amount based on temperature drift. .

[0071] Weighted Smoothing: A good fit, moderate extreme values, and no jumps in the periodicity indicate "good" detection quality. The weight of the pore edge is then assigned accordingly. Temperature drift Calculate the single suggested value. The instantaneous drift compensation amount is obtained by pushing the data into the time-series queue for the last 5 iterations and calculating its arithmetic mean. .

[0072] Operational performance: The Z-axis module will drive the positioning to... Formal fluorescence imaging was performed. During 48 hours of continuous fully automated operation, all acquired fluorescence images were clear, and the low-dose fluorescence anomaly verification procedure was not triggered.

[0073] Scenario Example 2: Continuous Monitoring for 7 Days (168 Hours) and Periodic Recalibration Experimental conditions: Live cell monitoring was conducted for 7 days (168 hours) for monitoring well C5. The system initiated the "benchmark recalibration process" at 24, 48, 72, 96, 120 and 144 hours.

[0074] Procedure: Precision aperture bright-field measurement is performed at the daily recalibration time, supplemented by ultra-low dose fluorescence rapid imaging for verification. Assume the absolute physical location of the clearest focal plane confirmed at the 24th hour is... The system immediately updates the currently valid baseline field z_baseline_active to the new value. Set the accumulated instantaneous compensation amount drift_offset to zero. And clear the moving average buffer.

[0075] Operational performance: The daily 2-hour cycle still executes two-way weighted calculations, with all drift calculations subtracted from the latest valid origin established that day. After 168 hours of continuous operation, this cell detection channel eliminated the error accumulation path caused by microplate deformation and maintained a normal focal plane range for 7 consecutive days.

[0076] Scenario Example 3: Low-dose fluorescence correction triggered by mutational fusion deviation under abnormal operating conditions Operating condition input: Select the peripheral edge orifice E8, which is subject to localized liquid level disturbance. During a certain inspection cycle, the calculated orifice edge observation drift jumps to... The temperature drift prediction derived from the concurrent environmental temperature drift model is: .

[0077] Gating trigger: The absolute value of the estimation difference between the two parallel paths reaches... It exceeded the upper limit of the safety threshold. The system kernel determined that the normal bright-field view had been misled by micro-interference from the liquid surface.

[0078] Low-light correction: The main controller immediately terminates the sliding operation and initiates the verification logic: reducing the light source's duty cycle to the normal operating condition. The exposure time was adjusted to the normal parameters. Above and below the current preset target focus Within the range Five live cell images are captured quickly with a stride distance.

[0079] Performance: Based on the extracted central cell population ROI, the Laplacian variance is calculated and extreme value fitting is performed to directly determine that the current true focus is the offset. Location. The system immediately invokes this. The absolute value replaces the previously calculated instantaneous compensation amount, forcibly converging the closed-loop correction to the effective focus, and writing back the correction log. After the liquid surface disturbance stabilizes in the next cycle, it automatically resets to the normal full-field loop.

[0080] Scenario Example 4: Typical Temperature Drift Segmented Calibration Mapping Parameter calibration: A transparent 96-hole glass base plate was selected as the calibration carrier, and the representative central hole position B5 and the representative edge hole position A1 were designated.

[0081] Operating procedure: Set the incubator to the normal temperature. The steps rise to the extreme value The temperature rise span is set to Then let it stand for 30 minutes to maintain its temperature. After each temperature setting stabilizes, scan to obtain the focal plane reading with maximum sharpness.

[0082] Operation archiving: with The gear position reading is used as a reference zero point to calculate the displacement and step ratio. The corresponding piecewise coefficients are then fitted to hole position B5 as follows: , , The unit temperature coefficient of the edge hole position A1 is approximately higher than that of the center hole B5. Establish spatial repair system data The aforementioned mapping table is written to non-volatile memory to ensure that during automated operation, the system can automatically match differentiated temperature drift prediction parameters for internal and edge holes according to the specific physical hole topology location called.

[0083] The above description is only a specific preferred embodiment of the present invention disclosed under the conditions of a 20× microscope objective, a typical microplate, and a closed incubator for live cells. Well positions A1, B5, C5, E8, etc., follow the international standard naming convention for 96-well plates—the letters AH represent row numbers, and the numbers 1-12 represent column numbers. This is common knowledge and can be understood by those skilled in the art without further explanation, therefore it will not be elaborated upon. Those skilled in the art, without departing from the essential inventive concept of the present invention, may modify some parameter settings (such as changing the bright-field axial sampling span) based on the basic principles disclosed in this specification. or , length of moving average processing Take as to Any equivalent substitution or transformation made by other integers in the specification (or by replacing the carrier with other porous flux specifications) shall also fall within the scope of protection set forth in the claims of this invention.

[0084] Example 5 This embodiment also provides an electronic device, see reference. Figure 3 It includes a memory 402 and a processor 401, the memory 402 storing a computer program and the processor 401 being configured to run the computer program to perform the steps in any of the above method embodiments.

[0085] Specifically, the processor 401 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of the present invention.

[0086] The memory 402 may include a mass storage device for data or instructions. For example, and not limitingly, the memory 402 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk drive, a magneto-optical disk drive, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 402 may include removable or non-removable (or fixed) media. Where appropriate, the memory 402 may be internal or external to a data processing device. In a particular embodiment, the memory 402 is non-volatile memory. In a particular embodiment, the memory 402 includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable read-only memory (PROM), an erasable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), an electrically alterable read-only memory (EAROM), or flash memory, or a combination of two or more of these. Where appropriate, the RAM can be Static Random-Access Memory (SRAM) or Dynamic Random-Access Memory (DRAM). DRAM can be Fast Page Mode Dynamic Random Access Memory (FPMDRAM), Extended Data Out Dynamic Random Access Memory (EDODRAM), Synchronous Dynamic Random-Access Memory (SDRAM), etc.

[0087] The memory 402 can be used to store or cache various data files that need to be processed and / or communicated, as well as possible computer program instructions executed by the processor 401.

[0088] The processor 401 reads and executes computer program instructions stored in the memory 402 to implement any of the long-term focal plane drift compensation methods for fluorescent live cells in the above embodiments.

[0089] Optionally, the electronic device may further include a transmission device 403 and an input / output device 404, wherein the transmission device 403 is connected to the processor 401 and the input / output device 404 is connected to the processor 401.

[0090] The transmission device 403 can be used to receive or send data via a network. Specific examples of the network described above may include wired or wireless networks provided by the communication provider of the electronic device. In one example, the transmission device includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 403 may be a Radio Frequency (RF) module used for wireless communication with the Internet.

[0091] Input / output device 404 is used for inputting or outputting information. It can be a speaker, microphone, monitor, or keyboard.

[0092] Example 6 This embodiment also provides a readable storage medium storing a computer program, the computer program including program code for controlling a process to execute the process, the process including the long-term focal plane drift compensation method for fluorescent live cells according to Embodiment 1.

[0093] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.

[0094] Generally, various embodiments can be implemented in hardware or dedicated circuitry, software, logic, or any combination thereof. Some aspects of the invention can be implemented in hardware, while others can be implemented by firmware or software executed by a controller, microprocessor, or other computing device, but the invention is not limited thereto. Although various aspects of the invention may be shown and described as block diagrams, flowcharts, or using some other graphical representation, it should be understood that, by way of non-limiting example, these blocks, apparatuses, systems, techniques, or methods described herein can be implemented in hardware, software, firmware, dedicated circuitry or logic, general-purpose hardware or controllers or other computing devices, or some combination thereof.

[0095] Embodiments of the present invention can be implemented by computer software, which may be executable by a data processor of a mobile device, such as a processor entity, or by hardware, or by a combination of software and hardware. Computer software or programs (also referred to as program products) including software routines, applets, and / or macros can be stored in any device-readable data storage medium, and they include program instructions for performing specific tasks. The computer program product may include one or more computer-executable components configured to perform the embodiments when the program is run. The one or more computer-executable components may be at least one piece of software code or a portion thereof. Additionally, it should be noted in this respect that, as Figure 1 Any box in the logical flow can represent a program step, or interconnected logic circuits, boxes and functions, or a combination of program steps and logic circuits, boxes and functions. Software can be stored on physical media such as memory chips or blocks of storage implemented within a processor, magnetic media such as hard disks or floppy disks, and optical media such as DVDs and their data variants, CDs, etc. The physical medium is a non-transient medium.

[0096] Those skilled in the art should understand that the technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments have been described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0097] The above embodiments are merely illustrative of several implementations of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined by the appended claims.

Claims

1. A method for compensating for long-term focal plane drift in fluorescent live cells, characterized in that, Includes the following steps: During the inspection cycle, multi-focal surface sampling is performed on the aperture edge in bright field mode. The aperture edge observation focal plane is determined based on sharpness evaluation, and the aperture edge observation drift amount relative to the current effective reference is calculated. The temperature change information of the incubator is read, and the predicted temperature drift is calculated according to the pre-calibrated temperature drift coefficient. The temperature drift coefficient is calibrated in segments according to temperature ranges, and different temperature ranges correspond to different temperature drift coefficient values. The temperature drift coefficient is corrected according to the spatial position of the well in the well plate, wherein the temperature drift coefficient of the edge well is greater than that of the center well. The borehole edge detection quality level is determined by comprehensively evaluating at least one of the following: borehole wall annular positioning deviation, sharpness curve peak position, peak contrast, and recent drift consistency. Based on the borehole edge detection quality level, the observed drift amount and the predicted drift amount are assigned a fusion weight, where the weight of a single route is limited to a preset range. The two are weighted and summed according to the fusion weight to obtain the compensation recommendation value for this cycle. The drift compensation amount is obtained by taking the arithmetic mean of the compensation recommendation values ​​for the most recent N inspection cycles of this borehole position, where N is an integer between 3 and 7. The drift compensation amount is then superimposed on the focal plane of subsequent fluorescence acquisition. For each aperture, the starting reference focal plane of the task and the focal plane of the current effective reference participating in the drift calculation are recorded independently. The current effective reference is updated according to a predetermined refresh strategy, and the drift compensation amount is reset according to the updated reference.

2. The method for compensating for long-term focal plane drift in fluorescent live cells as described in claim 1, characterized in that, The steps of sampling multiple focal planes at the aperture edge in bright-field mode and determining the observation focal plane at the aperture edge based on sharpness evaluation include: Using the current compensated fluorescence acquisition focal plane as the central reference point, multiple brightfield images are acquired at equal intervals within the preset focal plane range; The sharpness of the region of interest at the aperture edge of the bright field image is evaluated to obtain the sharpness value corresponding to each focal plane; The focal plane corresponding to the maximum sharpness is taken as the coarse peak position. When the coarse peak position is not at both ends of the sampling range, the sharpness values ​​of the coarse peak position and its adjacent sampling points are subjected to quadratic curve fitting, and the focal plane corresponding to the vertex of the fitted curve is taken as the focal plane of the aperture edge observation in this period. When the coarse peak position is at both ends of the sampling range, the sampling range is reduced and the step size is decreased to supplement the sampling, and the aperture edge observation focal plane is refitted to obtain the focal plane.

3. The method for compensating for long-term focal plane drift in fluorescent live cells as described in claim 2, characterized in that, The region of interest at the edge of the hole is located by locating the annulus on the hole wall using a geometric edge detection algorithm, and the annulus with a radius equal to a preset proportion of the hole radius is taken as the region of interest at the edge of the hole; the sharpness evaluation is to calculate the Laplacian variance of the grayscale image within the region of interest at the edge of the hole.

4. The method for compensating for long-term focal plane drift in fluorescent live cells as described in claim 1, characterized in that, The steps of updating the current valid benchmark and resetting the drift compensation amount according to the predetermined refresh strategy include: Record the starting reference focal plane of the task and the focal plane of the current effective reference that participates in the drift calculation independently for each aperture position; When the timer period arrives, the sharpest focal plane is confirmed by sampling the edge of the bright field aperture, the current effective reference is updated to the focal plane, and the drift compensation is reset to zero or the residual is checked. When the absolute value of the drift compensation exceeds the preset upper limit threshold, the continuous fluorescence clarity is lower than the threshold, or the aperture edge and temperature drift estimation continuously exceed the tolerance, the current effective benchmark is forcibly refreshed. For well sites that have not been monitored for a long time, the sampling inspection benchmark is refreshed according to a preset ratio.

5. The method for compensating for long-term focal plane drift in fluorescent live cells as described in claim 1, characterized in that, It also includes a low-dose fluorescence verification step: When the deviation between the observed drift at the aperture edge and the predicted drift at the temperature drift exceeds a first threshold, the change in the compensation recommendation value for the current period relative to the previous period exceeds a second threshold, or the fluorescence clarity for a consecutive preset number of frames is lower than a preset proportion of the historical average, fluorescence verification is triggered. Multiple verification images are acquired at equal intervals within a preset range near the focal plane corresponding to the current drift compensation amount, using an excitation dose lower than that used in the formal fluorescence acquisition. The drift compensation amount of this well is corrected based on the sharpness fitting peak position of the verification image in the cell region.

6. A system for implementing the long-term focal plane drift compensation method for fluorescent live cells according to any one of claims 1 to 5, characterized in that, include: The bright-field imaging unit is configured to acquire multiple images of the hole edge within a preset focal plane range in a bright-field manner during the inspection cycle. Temperature reading unit, configured to acquire temperature change information of the incubator; Temperature drift calibration storage unit, configured to store pre-calibrated temperature drift coefficient; The aperture edge analysis unit is configured to determine the aperture edge observation focal plane based on the sharpness evaluation function and compare it with the current effective benchmark to obtain the aperture edge observation drift. The drift fusion unit is configured to comprehensively evaluate the quality of aperture edge detection based on multiple observable indicators, and assign fusion weights to the aperture edge observed drift and temperature drift predicted drift based on the evaluation results. The two are weighted and merged according to the fusion weights and then the drift compensation is obtained by moving average. The drift compensation is then superimposed on the focal plane of subsequent fluorescence acquisition. The reference refresh unit is configured to update the current valid reference according to a predetermined refresh strategy and reset the drift compensation amount based on the updated reference. The borehole position compensation recording unit is configured to maintain reference information and compensation history according to the borehole position.

7. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to perform the long-term focal plane drift compensation method for fluorescent live cells according to any one of claims 1 to 5.

Images