Cell extraction method and device, computer equipment and storage medium

Abstract

The invention relates to a cell extraction method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring various cell images of cells to be sorted, and identifying fluorescent dots in each cell image through a deep learning model; determining the fluorescence signal intensity of the fluorescent dots in each cell image; and according to the fluorescence signal intensity of the fluorescent dots in the multiple cell images, absorbing cells meeting a preset sorting condition from the to-be-sorted cells. By adopting the method, the cell extraction accuracy can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of cell sorting technology, and in particular to a cell extraction method, device, computer equipment, storage medium and computer program product. Background Art

[0002] With the development of biomedical field, cell sorting technology has emerged. Cell sorting technology refers to the separation and processing of a specific cell subpopulation from a mixed cell sample according to the characteristics of the cells, so as to realize the functional analysis of specific cells.

[0003] In traditional technology, cell extraction is performed through threshold segmentation technology or template matching technology to achieve cell sorting. However, threshold segmentation technology is difficult to give a suitable threshold for all backgrounds and cell morphologies, and template matching technology is also difficult to accurately match cells, resulting in low accuracy of cell extraction. Summary of the invention

[0004] Based on this, it is necessary to provide a cell extraction method, apparatus, computer equipment, storage medium and computer program product that can improve the accuracy of cell extraction in response to the above technical problems.

[0005] In a first aspect, the present application provides a cell extraction method. The method comprises:

[0006] Obtain multiple cell images of cells to be sorted, and identify the fluorescent spots in each cell image through a deep learning model;

[0007] Determine the fluorescence signal intensity of each fluorescent spot in each cell image;

[0008] According to the fluorescence signal intensity of the fluorescent spots in the multiple cell images, the cells meeting the preset sorting conditions are picked up from the cells to be sorted.

[0009] In one embodiment, identifying fluorescent spots in each cell image by a deep learning model includes:

[0010] Identify the location data of cells to be sorted in each cell image through a deep learning model;

[0011] The fluorescent spots of the cells to be sorted in each cell image are identified based on the position data.

[0012] In one embodiment, determining the fluorescence signal intensity of each fluorescent spot in each cell image comprises:

[0013] Intercepting the cells to be sorted in each cell image to obtain the cell image to be sorted corresponding to each cell image;

[0014] Grayscale analysis is performed on the fluorescent spots in the to-be-sorted cell images corresponding to each cell image to obtain the fluorescent signal intensity of the fluorescent spots in each cell image.

[0015] In one embodiment, acquiring multiple cell images of cells to be sorted includes:

[0016] Irradiating the cells to be sorted with excitation lights of multiple bands respectively, so that the cells to be sorted emit different fluorescence, and obtaining multiple fluorescence images of the cells to be sorted; wherein the multiple fluorescence images include DAPI fluorescence images, TRITC fluorescence images and FITC fluorescence images;

[0017] The plurality of fluorescent images are determined as a plurality of cell images.

[0018] In one embodiment, according to the fluorescence signal intensity of the fluorescent spots in the multiple cell images, extracting cells that meet the preset sorting conditions from the cells to be sorted includes:

[0019] For each cell image, the fluorescence signal intensity of the cell to be sorted is determined according to the average gray value of the fluorescence spots belonging to the same cell to be sorted;

[0020] Sort the fluorescence signal intensity of the cells to be sorted in each cell image;

[0021] Cells meeting preset sorting conditions are determined according to the sorting results corresponding to the multiple cell images, and cells meeting the preset sorting conditions are extracted from the cells to be sorted.

[0022] In one embodiment, the method further comprises:

[0023] Using a deep learning model, first position data of each cell in a first bright field cell image before cell extraction and second position data of each cell in a second bright field image after cell extraction are respectively identified;

[0024] comparing the first position data with the second position data;

[0025] If the alignment is consistent, the cell extraction fails and the cell extraction process is repeated.

[0026] In a second aspect, the present application also provides a cell extraction device. The device comprises:

[0027] The recognition module is used to obtain multiple cell images of cells to be sorted and identify the fluorescent spots in each cell image through a deep learning model;

[0028] A determination module, used for determining the fluorescence signal intensity of the fluorescent spot in each cell image;

[0029] The extraction module is used to extract cells that meet the preset sorting conditions from the cells to be sorted according to the fluorescence signal intensity of the fluorescent spots in the multiple cell images.

[0030] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0031] Obtain multiple cell images of cells to be sorted, and identify the fluorescent spots in each cell image through a deep learning model;

[0032] Determine the fluorescence signal intensity of each fluorescent spot in each cell image;

[0033] According to the fluorescence signal intensity of the fluorescent spots in the multiple cell images, the cells meeting the preset sorting conditions are picked up from the cells to be sorted.

[0034] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0035] Obtain multiple cell images of cells to be sorted, and identify the fluorescent spots in each cell image through a deep learning model;

[0036] Determine the fluorescence signal intensity of each fluorescent spot in each cell image;

[0037] According to the fluorescence signal intensity of the fluorescent spots in the multiple cell images, the cells meeting the preset sorting conditions are picked up from the cells to be sorted.

[0038] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0039] Obtain multiple cell images of cells to be sorted, and identify the fluorescent spots in each cell image through a deep learning model;

[0040] Determine the fluorescence signal intensity of each fluorescent spot in each cell image;

[0041] According to the fluorescence signal intensity of the fluorescent spots in the multiple cell images, the cells meeting the preset sorting conditions are picked up from the cells to be sorted.

[0042] The above-mentioned cell extraction method, device, computer equipment, storage medium and computer program product adopt a method combining deep learning and traditional image processing to realize the screening of cells to be sorted and the evaluation of fluorescence signal intensity. The fluorescent spots in the cells to be sorted are located by the deep learning model, and then the fluorescence signal intensity of the fluorescent spots is evaluated by image processing technology, and the fluorescence signal intensity of the located cells to be sorted is obtained, and then according to the fluorescence signal intensity of the cells to be sorted, the cells that meet the preset sorting conditions are extracted from the cells to be sorted. The problem of low cell positioning accuracy caused by the difficulty of distinguishing the fluorescence signal from impurities or background due to the large difference in fluorescence signals in template matching technology and threshold segmentation technology is solved, and the accurate positioning of cells is realized. On this basis, the signal strength is evaluated, which can greatly improve the accuracy of cell sorting. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A diagram showing an application environment of a cell extraction method in one embodiment;

[0044] Figure 2 A schematic diagram of a cell extraction method in one embodiment;

[0045] Figure 3 A schematic diagram of a process for identifying fluorescent spots in each cell image by a deep learning model in one embodiment;

[0046] Figure 4 is a structural block diagram of a cell extraction device in one embodiment;

[0047] Figure 5 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0049] Cell sorting technology is a technology that separates cells that meet preset requirements from cells to be sorted based on their characteristics. When it is necessary to perform functional analysis on cells that meet preset requirements, such as analyzing cell molecules through ELISA (enzyme linked immunosorbent assay) analysis of cell culture supernatant, detecting cell functions through cell co-culture, etc., it is necessary to obtain high-purity cells that meet preset requirements.

[0050] The cell sorting method provided in the embodiments of the present application can be applied to Figure 1In the application environment shown. The cell sorting device 102 communicates with the terminal 104 via a network. The cell sorting device 102 sends the acquired cell image data to the terminal 104, and the terminal 104 locates the cells to be sorted in the cell image data through a deep learning model, evaluates the fluorescence signal intensity of the cells to be sorted according to the cell image data, and extracts cells that meet the preset sorting conditions from the cells to be sorted according to the fluorescence signal intensity of the cells to be sorted. The terminal 104 can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, and image processing devices.

[0051] In one embodiment, Figure 2 As shown, a cell extraction method is provided, which is applied to Figure 1 The terminal in is used as an example to illustrate, including the following steps:

[0052] Step 202: Acquire multiple cell images of cells to be sorted, and identify fluorescent spots in each cell image through a deep learning model.

[0053] Among them, the cells to be sorted can be cells that have been specifically screened by fluorescent dyes. Cell image data refers to an image containing the cells to be sorted obtained by imaging the cells to be sorted. Multiple cell images refer to fluorescent images of cells that have been specifically screened by different fluorescent dyes. Fluorescent images refer to images with fluorescent characteristics that are taken when the cells to be sorted are illuminated by bright ambient light and a filter channel is used in the imaging of the cells to be sorted.

[0054] Specifically, the imaging mechanism of the cell sorting device images the cells to be sorted, obtains multiple cell images of the cells to be sorted, and sends the multiple cell images to the terminal. The terminal calls the deep learning model, inputs the multiple cell images into the deep learning model respectively, performs target recognition on each cell image, determines the location area of ​​the cells to be sorted in each cell image, and thus identifies the fluorescent points of the cells to be sorted according to the location area of ​​the cells to be sorted for each cell image.

[0055] Step 204, determining the fluorescence signal intensity of the fluorescence point in each cell image.

[0056] The grayscale analysis of each cell image is performed using image processing technology to obtain the grayscale value of the fluorescent point in each cell image, and the fluorescence signal intensity of the fluorescent point is evaluated based on the grayscale value.

[0057] Step 206 , according to the fluorescence signal intensities of the fluorescence spots in the multiple cell images, cells meeting the preset sorting conditions are extracted from the cells to be sorted.

[0058] Specifically, for each cell image, the average grayscale value of the fluorescent spots belonging to the same cell to be sorted is calculated. The fluorescent signal intensity of each cell to be sorted is determined based on the relationship between the grayscale value and the fluorescent signal intensity and the average grayscale value of the fluorescent spots in each cell to be sorted. According to the fluorescent signal intensity of the cells to be sorted in multiple cell images, cells that meet the preset sorting conditions are determined and absorbed.

[0059] Optionally, the cell sorting device includes a main body, a selection mechanism, a loading mechanism, an imaging mechanism and a control mechanism. The main body is used to support and accommodate other mechanisms in the cell sorting device, forming a storage space for the selection mechanism, the loading mechanism, the imaging mechanism and the control mechanism. The loading mechanism is used to carry the parts to be sorted, and the parts to be sorted are used to accommodate the cells to be sorted. The imaging mechanism is used to image the cells to be sorted and obtain cell images of the cells to be sorted, and send the cell images to the terminal. The terminal displays the cell image, and identifies the fluorescent points in each cell image through a deep learning model, determines the fluorescence signal intensity of the fluorescent points in each cell image, and determines the cells that meet the preset sorting conditions based on the fluorescence signal intensity of the fluorescent points in multiple cell images. Then, the selection mechanism is used to absorb and recover the cells that meet the preset sorting conditions in the parts to be sorted, so as to achieve the sorting of cells that meet the preset sorting conditions. The control mechanism is used to control the various mechanisms of the cell sorting device to perform cell sorting operations.

[0060] In the above-mentioned cell extraction method, a method combining deep learning and traditional image processing is used to realize the screening of cells to be sorted and the evaluation of fluorescence signal intensity. The fluorescent spots in the cells to be sorted are located by the deep learning model, and then the fluorescence signal intensity of the fluorescent spots is evaluated by image processing technology, and the fluorescence signal strength of the located cells to be sorted is obtained, and then the cells that meet the preset sorting conditions are extracted from the cells to be sorted according to the fluorescence signal intensity of the cells to be sorted. The problem of low cell positioning accuracy caused by the difficulty in distinguishing the fluorescence signal from impurities or background due to the large difference in fluorescence signals in template matching technology and threshold segmentation technology is solved, and the accurate positioning of cells is realized. On this basis, the signal strength is evaluated, which can greatly improve the accuracy of cell sorting. In addition, the cells that meet the preset sorting conditions are automatically selected from the cells to be sorted by the cell sorting device, and the cells are separated and purified, which has the advantages of multi-species, high throughput, multiple sorting, short sorting cycle, and high sorting accuracy. Compared with traditional manual selection, sorting cells through cell sorting equipment can reduce the interference of human factors, ensure the consistency of cells during the cell sorting process, accurately obtain single cells with complete morphology and high activity, improve sorting efficiency, and achieve high-throughput sorting of single cells.

[0061] In one embodiment, obtaining multiple cell images of cells to be sorted includes: irradiating the cells to be sorted with excitation light of multiple bands respectively, so that the cells to be sorted emit different fluorescence, and obtaining multiple fluorescence images of the cells to be sorted; wherein the multiple fluorescence images include DAPI fluorescence images, TRITC fluorescence images and FITC fluorescence images; and determining the multiple fluorescence images as multiple cell images.

[0062] Specifically, the imaging mechanism of the cell sorting device includes a fluorescent component, an optical path component and a detection component. The fluorescent component and the optical path component are arranged on the same side of the cells to be sorted. The fluorescent component is used to make the cells to be sorted emit fluorescence to obtain a fluorescent image of the cells to be sorted. The optical path component is used to form an optical path of the imaging mechanism, which is configured with optical elements for forming the optical path to achieve imaging of the cells to be sorted. The detection component is used to obtain the fluorescent image of the cells to be sorted obtained by the optical path component.

[0063] The fluorescence component includes an excitation light source and a filter unit. Among them, the excitation light source refers to a light source that can make the cells to be sorted emit fluorescence. The excitation light emitted by the excitation light source irradiates the cells to be sorted and makes the cells to be sorted emit fluorescence. For example, the excitation light source can be a xenon lamp, a mercury lamp, various lasers such as an argon laser, and a light-emitting diode. The filter unit is the core component of the imaging mechanism to realize fluorescence imaging. The excitation light source is filtered by the filter unit so that the excitation light source emits excitation light of the corresponding band, so that the cells to be sorted emit fluorescence, and the fluorescence passes through the optical path component, and the fluorescence image generated by the optical path component is transmitted to the detection component, and the detection component obtains the fluorescence image of the cells to be sorted. The detection component is connected to the terminal, and the detection component sends the fluorescence image of the cells to be sorted to the terminal.

[0064] Optionally, there may be multiple filter units, and the filter units are used to excite the culture medium to which a preset type of fluorescent dye is added for specific screening of the cells to be sorted to emit fluorescence. Among them, the preset types of fluorescent dyes may include fluorescent dyes capable of generating blue fluorescence, fluorescent dyes capable of generating green fluorescence, and fluorescent dyes capable of generating red fluorescence. For example, the fluorescent dye capable of generating blue fluorescence may be DAPI fluorescent dye, the fluorescent dye capable of generating red fluorescence may be TRITC fluorescent dye, and the fluorescent dye capable of generating green fluorescence may be FITC fluorescent dye.

[0065] The type of filter unit corresponds to the type of fluorescent dye. The filter unit may include a first filter unit, a second filter unit, and a third filter unit. After the excitation light passes through the first filter unit, the cells to be sorted emit a first fluorescence, after the excitation light passes through the second filter unit, the cells to be sorted emit a second fluorescence, and after the excitation light passes through the third filter unit, the cells to be sorted emit a third fluorescence. Specifically, the cells to be sorted are specifically screened by adding a culture medium containing a fluorescent dye capable of generating blue fluorescence, a fluorescent dye capable of generating green fluorescence, and a fluorescent dye capable of generating red fluorescence. Short-wavelength light will excite and generate long-wavelength light. Therefore, ultraviolet excitation generates blue light, blue light excitation generates green light, and green light excitation generates red light. The cells to be sorted cultured with a fluorescent dye capable of generating blue fluorescence use an ultraviolet filter unit, and the excitation light causes the cells to be sorted to emit blue fluorescence, and the blue fluorescence passes through an optical path component, and the optical path component generates a DAPI fluorescent image. The cells to be sorted that have been cultured with a fluorescent dye capable of generating green fluorescence use a blue light filter unit, and the excitation light causes the cells to be sorted to emit green fluorescence. The green fluorescence passes through the optical path component, and the optical path component generates a green fluorescence image, i.e., a FITC fluorescence image. The cells to be sorted that have been cultured with a fluorescent dye capable of generating red fluorescence use a green light filter unit, and the excitation light causes the cells to be sorted to emit red fluorescence. The red fluorescence passes through the optical path component, and the optical path component generates a TRITC fluorescence image. The cells to be sorted are specifically screened with a preset type of fluorescent dye, which helps to more accurately select cells that meet the preset requirements. It can be understood that different filter units are selected according to different fluorescent dyes, and the types of fluorescent dyes and filter units are not limited.

[0066] Furthermore, the fluorescence component also includes a filter switching unit. By switching different filter units through the filter switching unit, excitation light of different wavelengths is obtained after passing through different filter units, so that the cells to be sorted emit different fluorescence, so that the imaging mechanism can meet the culture requirements of the cells to be sorted in various situations, and increase the applicability of the cell sorting device.

[0067] In this embodiment, the cells to be sorted are irradiated with excitation light of multiple bands to make the cells to be sorted emit different fluorescence, so as to obtain multiple fluorescence images of the cells to be sorted, and the cells to be sorted can be accurately located according to the fluorescence signal characteristics of the cells.

[0068] In one embodiment, Figure 3 As shown, the deep learning model is used to identify the fluorescent spots in each cell image:

[0069] Step 302: Identify the location data of the cells to be sorted in each cell image through a deep learning model.

[0070] Step 304, identifying the fluorescent spots of the cells to be sorted in each cell image according to the position data.

[0071] The terminal inputs the acquired multiple cell images into the deep learning model respectively, and extracts the fluorescence signal features of the cell image data through the deep learning model. Among them, the fluorescence signal features may include the outline of the cell to be sorted. Thus, the outline of the sorted cell is determined as its position data. For example, the position data may be the position coordinates of the cell to be sorted. Then, the position and number of the fluorescent spots in the cell to be sorted are identified according to the position data of the cell to be sorted.

[0072] Optionally, the deep learning model can be trained by the fluorescence signal characteristics of a large number of cells, and can accurately identify the fluorescence signals to locate the cells.

[0073] In this embodiment, the position data of the cells to be sorted in the cell image data are extracted from the deep learning model, and the fluorescent points of the cells to be sorted are identified based on the position data, which can fully utilize the advantages of deep learning technology to achieve accurate cell positioning.

[0074] In one embodiment, determining the fluorescence signal intensity of the fluorescence points in each cell image includes: intercepting the cells to be sorted in each cell image to obtain the cell images to be sorted corresponding to each cell image; performing grayscale analysis on the fluorescence points in the cell images to be sorted corresponding to each cell image to obtain the fluorescence signal intensity of the fluorescence points in each cell image.

[0075] The image of cells to be sorted refers to a cell image captured with the position data of the cells to be sorted as the boundary.

[0076] After locating the cells to be sorted in the cell image data, the cells to be sorted are intercepted to obtain the images of the cells to be sorted. For each cell image to be sorted, grayscale analysis is performed on the cell image to be sorted. Specifically, the cell image to be sorted is denoised; the denoised cell image to be sorted is threshold segmented; the segmented cell image to be sorted is grayscale analyzed to obtain the fluorescence signal intensity of the fluorescent spot.

[0077] Specifically, the denoising process can be an image preprocessing method known to those skilled in the art, which is not limited here. Denoising is used to remove noise points in the image of cells to be sorted. The denoised image of cells to be sorted is subjected to threshold segmentation to separate the cells to be sorted from the background in the image, so that the image presents an obvious black and white effect. Among them, threshold segmentation can also be called binarization. Afterwards, the grayscale values ​​of the fluorescent points in the segmented image of cells to be sorted are calculated to obtain the grayscale values ​​of the fluorescent points, and the average grayscale value of the fluorescent points belonging to the same cell to be sorted is calculated. According to the relationship between the grayscale value and the fluorescence signal intensity and the average grayscale value of the fluorescent points in each cell to be sorted, the fluorescence signal intensity of each cell to be sorted is determined.

[0078] In this embodiment, by intercepting the image of the cells to be sorted, the analysis range of the fluorescence signal intensity can be narrowed, which is conducive to improving the efficiency of the fluorescence signal intensity analysis. The image of the cells to be sorted is denoised to avoid the influence of noise points, and the denoised image of the cells to be sorted is threshold segmented. After that, the grayscale analysis of the image of the cells to be sorted is performed to obtain the fluorescence signal intensity of the cells to be sorted. The threshold segmentation can separate the cells to be sorted in the image from the background, so that the image presents an obvious black and white effect, which is conducive to improving the accuracy of the cell fluorescence signal intensity.

[0079] In one embodiment, according to the fluorescence signal intensity of the fluorescence spots in the multiple cell images, extracting the cells that meet the preset sorting conditions from the cells to be sorted includes: for each cell image, determining the fluorescence signal intensity of the cell to be sorted according to the average gray value of the fluorescence spots belonging to the same cell to be sorted; sorting the fluorescence signal intensity of the cells to be sorted in each cell image; determining the cells that meet the preset sorting conditions according to the sorting results corresponding to the multiple cell images, and extracting the cells that meet the preset sorting conditions from the cells to be sorted.

[0080] Specifically, for each cell image, i.e., fluorescent image, the average gray value of the fluorescent spots belonging to the same cell to be sorted is calculated. According to the relationship between the gray value and the fluorescent signal intensity and the average gray value of the fluorescent spots in each cell to be sorted, the fluorescent signal intensity of each cell to be sorted is determined. Thereby, the fluorescent signal intensities of a plurality of cells to be sorted are sorted, and the cells that meet the preset sorting conditions are determined according to the sorting results corresponding to the various cell images, and the suction and release parts on the picking mechanism are controlled to absorb the cells that meet the preset sorting conditions. Wherein, the fluorescent signal intensity of each cell to be sorted includes DAPI fluorescent value, TRITC fluorescent value and FITC fluorescent value. The sorting results corresponding to the various cell images include DAPI fluorescent value sorting results, TRITC fluorescent value sorting results and FITC fluorescent value sorting results.

[0081] Optionally, impurities without cell nuclei are removed according to the DAPI fluorescence value sorting results, and cells with TRITC fluorescence values ​​less than a preset low value among the total cells are marked according to the TRITC fluorescence value sorting results, and sorted according to the FITC fluorescence value; if the FITC fluorescence value of the marked cells is greater than the preset high value, it is determined that the cells meet the preset sorting conditions.

[0082] Optionally, the selection mechanism includes a mounting assembly, a selection moving assembly, and a recovery assembly mounted on the main body of the cell sorting device, wherein the mounting assembly is used to mount the suction and release member, the selection moving assembly can drive the suction and release member to move so that the suction and release member is close to or away from the loading mechanism, and the recovery assembly is connected to the suction and release member so that the suction and release member can absorb cells that meet the preset sorting conditions. This arrangement can realize the automatic absorption of cells that meet the preset sorting conditions and improve the efficiency of cell sorting.

[0083] Furthermore, the recovery component includes a driving unit, a connecting piece and a passage connecting piece. The suction and release piece includes a hollow pipeline. Under the control of the control mechanism, the selection moving component drives the suction and release port of the suction and release piece to move to a set position. At this time, there is a preset distance between the suction and release port and the cells that meet the preset sorting conditions. The control mechanism controls the driving unit to generate a pressure difference at the suction and release port of the suction and release piece, thereby sucking the cells that meet the preset sorting conditions into the hollow pipeline, and completing the absorption of the cells that meet the preset sorting conditions. Among them, the preset distance is a distance that enables the suction and release piece to absorb the cells that meet the preset sorting conditions into the hollow pipeline.

[0084] Optionally, the selection mechanism further includes a selection lighting unit, which is connected to the mounting assembly and located at an end of the suction and placement member away from the first carrier, and the light emitted by the selection lighting unit can pass through the suction and placement member. The setting of the selection lighting unit can realize the calibration of the suction and placement member and obtain the position of the suction and placement member, so that the suction and placement member can accurately absorb cells that meet the preset requirements, thereby improving the accuracy of cell sorting.

[0085] In this embodiment, the fluorescence signal intensity of the cell to be sorted is determined according to the average grayscale value of the fluorescence spots belonging to the same cell to be sorted, and based on the ranking results of the fluorescence signal intensities of the cells to be sorted in multiple cell images, cells that meet the preset sorting conditions are extracted from the cells to be sorted. Based on the multiple fluorescence signal intensities of the cells, cells that meet the preset sorting conditions can be selected more accurately, thereby greatly improving the accuracy of cell sorting.

[0086] In one embodiment, the method also includes: using a deep learning model to respectively identify first position data of each cell in a first bright field cell image before cell extraction and second position data of each cell in a second bright field image after cell extraction; comparing the first position data with the second position data; if the comparison is consistent, the cell extraction fails, and the cell extraction process is repeated.

[0087] Among them, the bright field image refers to an image with normal color captured when the cells to be sorted are illuminated by bright ambient light and no filter channel is used in the imaging of the cells to be sorted.

[0088] The cell sorting device can also obtain a first bright field cell image before cell extraction, and send the first bright field cell image to the terminal. The terminal inputs the first bright field cell image into the deep learning model to identify the first position data of each cell in the first bright field cell image.

[0089] When the control suction and release component completes the cell extraction, the second bright field cell image is obtained through the cell sorting device and sent to the terminal. The terminal inputs the second bright field cell image into the deep learning model to identify the second position data of each cell in the second bright field cell image. The first position data is compared with the second position data. If the comparison is consistent, it indicates that the cell absorption has failed and the cell extraction process needs to be repeated. If the second position data lacks the data in the first position data, that is, the comparison is inconsistent, it indicates that the cell existed before the extraction and does not exist after the extraction, and the cell absorption is successful.

[0090] Optionally, the bright field image acquisition process is as follows: the imaging mechanism of the cell sorting device also includes an imaging lighting unit. The part to be sorted is placed between the imaging lighting unit and the optical path component. The imaging lighting unit provides a bright field lighting source for the part to be sorted and the optical path component. The cells to be sorted are irradiated with transmitted light or reflected light. The transmitted light or reflected light enters the optical path component after passing through the part to be sorted, and then reaches the detection component through the optical path component. The optical path component transmits the generated bright field image to the detection component, and the detection component obtains the bright field image data of the cells to be sorted. The detection component is connected to the terminal, and the detection component sends the bright field image of the cells to be sorted to the terminal for display.

[0091] In one embodiment, the imaging lighting unit includes an imaging light source, a shading member and a light-transmitting member. The selection mechanism of the cell sorting device includes a suction and release member for sucking and releasing cells that meet the preset sorting conditions. The light emitted by the imaging light source passes through the shading member and the light-transmitting member to the cells to be sorted, and illuminates the cells to be sorted along the periphery of the suction and release member. It is understandable that the position of the imaging light source can be set according to actual conditions.

[0092] Optionally, the imaging light source is a ring light source, which provides bright and uniform illumination for a large visual field. Further, the imaging light source may include but is not limited to an LED light source. The ring LED light source may adopt a whole ring light source, or may be obtained by circumferentially evenly arranging a plurality of light sources, so that the imaging light source irradiates the radial periphery of the suction and release member to the cell area to be sorted.

[0093] Optionally, the brightness of the imaging light source can be adjusted, so that the imaging light source is suitable for more cell sorting scenarios and facilitates brightness adjustment of image data.

[0094] In one embodiment, when the to-be-sorted member containing the to-be-sorted cells moves to a preset area of ​​the cell sorting device, the objective lens in the optical path assembly images the to-be-sorted cells in the preset area to obtain cell image data.

[0095] Specifically, the loading mechanism includes a loading assembly, and the loading assembly includes a first loading member and a second loading member. The first loading member is used to accommodate the placement of the to-be-sorted object containing the cells to be sorted, and the second loading member is used to place the first loading member. The second loading member can move relative to the base of the main body under the drive of the loading moving assembly, so that the to-be-sorted object moves relative to the base of the main body to a preset area.

[0096] The optical path assembly includes an objective lens and an imaging lens. The objective lens has a certain focal length. When the to-be-sorted member containing the to-be-sorted cells moves to a preset area of ​​the cell sorting device, the objective lens images the to-be-sorted cells in the preset area. For example, the objective lens can be a 10x objective lens. It is understandable that the magnification of the objective lens can be set according to actual needs and is not limited here.

[0097] Optionally, the optical path component may include multiple objective lenses to obtain image data of the cells to be sorted at different magnifications.

[0098] In one of the embodiments, during the imaging process, the optical path component adjusts the focal length between the objective lens and the object to be sorted so that the objective lens is focused and the cells to be sorted are imaged.

[0099] Specifically, the optical path assembly also includes an objective lens moving unit, which is used to focus the objective lens and change the focal length between the objective lens and the object to be sorted, so as to make the image of the cells to be sorted clear and enable the imaging mechanism to obtain clear image information of the cells to be sorted.

[0100] In one embodiment, the detection component includes a bright field image data detector and a fluorescent image data detector. The bright field image data detector is used to detect image data based on the transmitted light or reflected light of the imaging light source through the cells to be sorted. The fluorescent image data detector is used to detect fluorescent image data based on the cells to be sorted.

[0101] The specific operation of obtaining the bright field image data of the cells to be sorted is as follows: turn on the imaging lighting unit, so that the light emitted by the imaging light source enters the bright field image data detector through the cells to be sorted, the objective lens, the filter unit and the first imaging lens in the imaging lens, and the bright field image data detector transmits the acquired image data of the cells to be sorted to the terminal. According to the image data displayed by the terminal, the objective lens moving unit is controlled to move the objective lens to make the image data of the cells to be sorted clear.

[0102] The specific operation of obtaining the fluorescence image data of the cells to be sorted is as follows: turn off the imaging light source, turn on the excitation light source, and the light emitted by the excitation light source passes through the filter unit and the objective lens to the cells to be sorted, so that the cells to be sorted that have been specifically screened by the fluorescent dye emit fluorescence, and the fluorescence enters the fluorescence image data detector through the objective lens, the filter unit and the second imaging lens in the imaging lens. The fluorescence image data detector transmits the fluorescence image data of the cells to be sorted to the terminal. The image data obtained by the terminal controls the objective lens moving unit to move the position of the objective lens to make the image data clear. Replace different filter units through the filter switching unit, repeat the above operation, and transmit different fluorescence image data of the cells to be sorted to the terminal. The terminal can find cells that meet the preset sorting conditions based on the fluorescence image data.

[0103] Optionally, the switching between the imaging light source and the excitation light source can be achieved by a control mechanism in the cell sorting device, or by manual switching.

[0104] Optionally, the objects to be sorted can be automatically moved under the control of a control mechanism, and the fluorescent component is provided with different filter units. The filter units can be automatically switched by a filter switching unit, and automatic focusing can be achieved by an objective lens moving unit. The automatic scanning of the imaging mechanism can be achieved, and a large number of cells to be sorted can be automatically scanned and imaged, which greatly improves the working efficiency of the cell sorting equipment and expands its application range.

[0105] In this embodiment, the cell positions before and after cell extraction are identified through a deep learning model, and whether the cells are successfully aspirated is determined by comparing the cell positions, thereby ensuring the accuracy of cell aspiration and avoiding the problem of missed cell selection.

[0106] In one embodiment, before acquiring multiple cell images of cells to be sorted, it is necessary to calibrate the position of the suction and placement components in the selection mechanism. The calibration process includes the following steps:

[0107] Acquire the image data of the suction and placement component, and determine the clarity of the image data of the suction and placement component through a calibration program; move the suction and placement component according to the clarity through the calibration program to obtain an suction and placement component image with optimal clarity; adjust the position data of the suction and placement component in the suction and placement component image through the calibration program until the preset conditions are met, and store the current position data of the suction and placement component.

[0108] The image data of the suction and placement member refers to the image data projected by the suction and placement member.

[0109] Specifically, the imaging mechanism in the cell sorting device acquires the image data of the suction and placement member, and sends the acquired image data of the suction and placement member to the terminal. The terminal displays the image data of the suction and placement member, and determines the clarity of the image data through a calibration program.

[0110] A hollow pipeline is provided in the suction and release member to enable cells to enter the suction and release member and to allow the sucked cells to be stored in the suction and release member. The size of the suction and release member can be gradually reduced along the radial direction of the suction and release member to form a suction and release port at one end of the suction and release member, and the size of the suction and release port matches the size of the cells to be sucked. During suction, the cells enter the hollow pipeline through the suction and release port, and during release, the cells leave the suction and release port through the hollow pipeline. The design of the suction and release port facilitates accurate and effective cell suction and improves the sorting efficiency of the cell sorting device. The material and size of the suction and release member can be selected according to the size of the cells to be sorted, such as a 1 mm glass capillary tube is selected for the suction and release member.

[0111] Furthermore, the picking mechanism includes an installation component, a picking movement component, a picking lighting unit, a cleaning component and a recycling component. The installation component is used for installing and fixing the suction and release component to prevent the suction and release component from falling off during use. Correspondingly, the installation component can also be used for replacing the suction and release component. The picking movement component is used to drive the suction and release component to move relative to the main body. The picking movement component includes a picking drive component, a picking transmission component and a picking movement component. Under the drive of the picking drive component and the rotation of the picking transmission component, the picking movement component drives the suction and release component to move. The cleaning component is used for cleaning the suction and release component before use, so that the activity of the cells can be guaranteed during the cell absorption and release process without damaging the normal state of the cells. The recycling component is used for the absorption and release of cells that meet the preset sorting conditions.

[0112] The selection mechanism is used to select cells that meet the preset sorting conditions from the objects to be sorted. The automated setting of the selection mechanism realizes automatic cell absorption, improves the efficiency of cell sorting, avoids damage to cells caused by manual operation, is conducive to high-throughput cell sorting, and can further realize single cell sorting.

[0113] The selection lighting unit is connected to the mounting assembly and is located at one end of the suction and placement member away from the loading mechanism. The selection lighting unit includes a selection light source connected to the control mechanism. The light emitted by the selection light source can pass through the hollow pipe of the suction and placement member to form a projection of the suction and placement member, which is convenient for positioning the suction and placement member.

[0114] In one of the embodiments, acquiring image data of the suction and placement component includes: when light emitted by a selected light source passes through a hollow tube of the suction and placement component to form a projection of the suction and placement component, the projection of the suction and placement component is identified by an imaging structure to obtain image data of the suction and placement component; acquiring image data of the suction and placement component sent by the imaging mechanism and displaying it.

[0115] Specifically, the control mechanism controls the selected light source to emit light, and the light passes through the hollow pipe of the suction and release part and passes through the suction and release part through the suction and release port. The suction and release part projection is obtained according to the light passing through the suction and release port, and the image data (such as a circular aperture) of the suction and release part projection is obtained through the imaging mechanism, and the image data of the suction and release part projection obtained is sent to the terminal for display. The terminal performs image edge detection on the image data of the suction and release part through a calibration program to obtain the clarity of the image data of the suction and release part. The purpose of edge detection is to extract edge information in the image, that is, the boundary between pixels in the image. It is understandable that the method of image edge detection can be an edge detection method known to those skilled in the art, such as Sobel, Prewitt, Roberts, etc.

[0116] Through the calibration program, according to the clarity of the image data of the suction and placement object, the selection moving component is controlled to move the suction and placement object, the clarity of the image data of the suction and placement object is determined again, and the steps of moving the suction and placement object and obtaining the image data of the suction and placement object are repeated until the suction and placement object image with optimal clarity is obtained.

[0117] The preset conditions refer to the position calibration conditions of the suction and placement parts.

[0118] After adjusting the position of the suction and release member to obtain the suction and release member image with the best clarity, the position data of the suction and release member is adjusted through the calibration program until the suction and release member is located at the preset position in the image. For example, the preset position can be the cross cursor. The suction and release member is displayed as a circular aperture in the image, that is, the middle position of the circular aperture can be adjusted to the cross cursor. At this time, the current position data of the suction and release member is marked and stored. In the subsequent cell sorting process, based on this position, the control mechanism controls the selection moving component to drive the suction and release member to move, so as to absorb cells that meet the preset sorting conditions.

[0119] In this embodiment, the clarity of the image data of the suction and release member is determined through a calibration program, and the suction and release member is moved according to the clarity through the calibration program to obtain an suction and release member image with optimal clarity. The position data of the suction and release member in the suction and release member image is adjusted through the calibration program until the preset conditions are met, and the current position data of the suction and release member is stored. The position of the suction and release member is automatically calibrated through the calibration program, so that the suction and release member can accurately and quickly absorb cells that meet the preset sorting conditions, thereby improving the efficiency and accuracy of cell sorting. In addition, the method can also avoid damage to the suction and release member and cells that meet the preset sorting conditions by manual operation, ensure the biological activity of cells that meet the preset sorting conditions, and achieve high-throughput cell sorting.

[0120] In one embodiment, determining the clarity of the image data of the suction and placement object through a calibration program includes: performing edge detection on the image data of the suction and placement object to obtain the brightness distribution of multiple boundary points in the image data; determining the clarity of the image data of the suction and placement object based on a preset image clarity evaluation relationship and the brightness distribution of multiple boundary points.

[0121] The boundary points refer to the boundary points among the pixels in the image data.

[0122] The image data of the pickup and placement parts is grayed out through the calibration program to obtain a grayscale image, and the grayscale image is subjected to image edge detection to obtain the brightness distribution of the boundary points in the image data. The brightness distribution indicates the change in image clarity.

[0123] There is a corresponding relationship between the brightness distribution of the boundary points and the clarity. The clarity of the image data of the suction and placement part is determined by a calibration program based on a preset image clarity evaluation relationship and the brightness distribution of multiple boundary points obtained. For example, the preset image clarity evaluation relationship can be an image clarity evaluation function known to those skilled in the art, which can achieve clarity evaluation by brightness, and this embodiment does not limit this.

[0124] In this embodiment, edge detection is performed on the image data of the suction and placement object to obtain the brightness distribution of multiple boundary points in the image data. Since the brightness distribution can represent the change in clarity, accurate and reliable clarity of the suction and placement object image data can be obtained based on the preset image clarity evaluation relationship.

[0125] In one embodiment, moving the suction and placement member according to the clarity through a calibration program to obtain an image of the suction and placement member with optimal clarity includes: identifying the current position of the suction and placement member in the image data of the suction and placement member through a calibration program, and controlling the selection of a moving component to move the suction and placement member according to the current position of the suction and placement member; returning to the step of obtaining the image data of the suction and placement member until the image data of the suction and placement member reaches the optimal clarity, thereby obtaining an image of the suction and placement member with optimal clarity.

[0126] Specifically, the selection moving assembly comprises a selection driving member, a selection transmission member and a selection moving member. Under the driving of the selection driving member and the rotation of the selection transmission member, the selection moving member drives the suction and release member to move.

[0127] The suction and release member in the image data of the suction and release member is located by a calibration program to determine the current position of the suction and release member. According to the current position of the suction and release member, the control mechanism drives the selection drive member and the selection transmission member to rotate. Under the drive of the selection drive member and the rotation of the selection transmission member, the selection moving member drives the suction and release member to move. Return to the step of obtaining the image data of the suction and release member, and determine the clarity of the image data of the suction and release member by a calibration program until the clarity of the image data of the suction and release member is maximized, and the suction and release member image with the best clarity is obtained.

[0128] In this embodiment, the current position of the suction and placement component is identified through a calibration program, and the suction and placement component is moved according to the current position so that the image of the suction and placement component reaches the optimal clarity, which can ensure that the position of the suction and placement component can accurately absorb cells that meet the preset sorting conditions without damaging the cells that meet the preset sorting conditions.

[0129] In one embodiment, the selection mechanism also includes an installation component, which includes a fixing unit; the position data of the suction and placement component in the suction and placement component image is adjusted through a calibration program until a preset condition is met, and the current position data of the suction and placement component is stored, including: identifying the position data of the suction and placement component in the suction and placement component image through a calibration program; controlling the fixing unit according to the position data through the calibration program to fine-tune the suction and placement component so that the position data of the suction and placement component is located at the preset position of the suction and placement component image, and storing the current position data of the suction and placement component as a calibration value.

[0130] Specifically, the fixing unit may include a first fixing member and a second fixing member. The suction and release member in the image data of the suction and release member is located by a calibration program, and the position data of the suction and release member is determined. According to the current position of the suction and release member, the control mechanism controls the first fixing member and the second fixing member to fine-tune the suction and release member so that the position data of the suction and release member is located at a preset position of the suction and release member image. The preset position may be a cross cursor. At this time, the current position data of the suction and release member is marked, and the current position data is stored as a calibration value. In the subsequent cell sorting process, based on this position as a reference, the control mechanism controls the selection moving component to drive the suction and release member to move so as to absorb cells that meet the preset sorting conditions.

[0131] In this embodiment, the position data of the suction and placement component in the suction and placement component image is identified through a calibration program, and the fixed unit is controlled to perform fine adjustment on the suction and placement component according to the position data, so that the position data of the suction and placement component is located at a preset position of the suction and placement component image, which can further improve the calibration accuracy of the suction and placement component.

[0132] In one embodiment, when the biological culture chip after cell-specific screening is used as a to-be-sorted part, when the biological culture chip containing the to-be-sorted cells is moved to a preset area of ​​a cell sorting device, the objective lens in the imaging mechanism images the to-be-sorted cells in the preset area to obtain image data of the to-be-sorted cells, and the image data of the to-be-sorted cells is sent to a terminal for display. The preset area refers to an imaging area.

[0133] In this embodiment, since the biological culture chip can make full use of microscopic technology or nanotechnology to complete precise control of cell capture, fixation, and culture on the chip, it is beneficial to improve the accuracy of cell sorting.

[0134] In one embodiment, the biological culture chip includes a plurality of accommodating chambers, each accommodating chamber contains cells to be sorted, and two adjacent accommodating chambers are separated by a preset distance.

[0135] Specifically, the bio-culture chip may include a matrix and a accommodating cavity, wherein the matrix is ​​formed of a matrix material, the accommodating cavity is formed on the surface of the matrix, and the accommodating cavity is open on the surface of the matrix. The accommodating cavity defines a bio-culture space for culturing cell biomaterials. The structure and size of the accommodating cavity are not particularly limited, as long as it can provide a culture or growth amplification space for a single cell. The accommodating cavity may have a structure such as a cuboid, a cube, or a cylinder.

[0136] Optionally, the matrix material is a microporous soft matrix with good biocompatibility, cell culture medium compatibility and mechanical tension, which can more efficiently capture and culture cells (including single cells and multi-cells with controllable number of aggregations). The soft matrix is ​​easy to be affinity and fused with different extracellular matrix environments, providing cells with a uniform and fused three-dimensional culture space, thereby creating a microenvironment for the cells in the two-dimensional array that is conducive to survival, proliferation, migration, protein expression and secretion, stem cell differentiation, etc., as well as conducive to the detection of biological functions of array cells. It should be noted that the specific type of matrix material is not limited, and hard matrix materials such as glass, silicon wafers, plastics, etc. are also within the scope of protection of this application.

[0137] Optionally, the opening diameter of the accommodating cavity is smaller than the bottom diameter of the accommodating cavity. In this way, single cells can be easily captured for culture without cell escape during the culture process. For example, the opening diameter of the accommodating cavity is at most 80%, preferably 50%, of the bottom diameter of the accommodating cavity. The diameter refers to the diameter of the maximum area circle that the opening or bottom can accommodate. The opening diameter of the accommodating cavity is 8-25 microns and the depth is 15-35 microns. It should be noted that different opening diameters can be set for different cells. For example, if the biomaterial is a B cell, the opening diameter is 8-12 microns; if the biomaterial is a hybridoma cell or a tumor cell, the opening diameter is 15-25 microns.

[0138] Optionally, the cavities in the biological culture chip are discretely distributed. The cavities form a predetermined pattern, and the distance between two adjacent cavities is 10-100 microns. Thus, the culture of multiple single cells can be processed in batches. When the distance between the cavities is 10-100 microns, the cells cultured in the cavities will not interfere with each other between the cavities, thereby improving the uniformity of batch processing, improving the accuracy and efficiency of cell biological function detection, and facilitating high-throughput acquisition and analysis of cell signal data. The cavities on the biological culture chip can form an array, for example, there are tens of thousands of cavities or even hundreds of thousands and more cavities on a chip.

[0139] Optionally, because the biological culture chip is thin, soft and brittle, it is difficult to directly fix it in the cell sorting device, so the cell sorting device also includes a chip fixing component, and the chip fixing component can be placed on the loading mechanism. The chip fixing component includes a chip container, and the chip container is used to hold the biological culture chip. The biological culture chip is first fixed in the chip container, and then the chip container is placed on the loading mechanism of the cell sorting device.

[0140] Furthermore, in order to ensure the normal use of the biological culture chip in the cell sorting device, the cell sorting device also includes a fixing mechanism, which includes a chip fixing component and a loading and limiting component. The chip fixing component is used to fix the biological culture chip in the chip container, and the loading and limiting component is used to fix the chip container on the loading mechanism. The use of the biological culture chip improves the sorting efficiency and accuracy of the cell sorting device.

[0141] In one embodiment, the cell-specific screening process of the biological culture chip is as follows:

[0142] Place the chip in the chip holder. The biological culture chip needs to be stored in a low-temperature environment, such as in an environment of 0-5°C. Therefore, before use, the biological culture chip needs to be left at room temperature for a period of time to restore the temperature to room temperature. Because the biological culture chip is thin, the material is soft and brittle, it is usually placed in a chip holder. After disinfecting the outside of the chip holder where the biological culture chip is placed, transfer it to a sterile environment, remove the chip preservation solution with a pipette, and wash it several times with sterile 0.01M PBS (phosphate buffered solution). It should be noted that do not touch the biological culture chip during the cleaning process.

[0143] Add cells. Add the cells to be sorted prepared in advance to the chip holder containing the biological culture chip. Preferably, in order to ensure the best single-cell array efficiency, the ratio of the number of cells to the chip flux is between 5:1 and 10:1, and the volume of the cell suspension is 1.5 ml. If the chip flux is 100,000, the array is performed with a cell suspension containing 500,000 to 1,000,000 cells. Let it stand in the cell culture incubator for a preset time, such as 15 minutes, so that the cells to be sorted form an array in the biological culture chip, and the containing cavity of the biological culture chip is set as an array.

[0144] First separation. Use a pipette to recover the arrayed cell suspension, then wash it several times with 0.01M PBS buffer to remove residual cells. The chip holder can be tilted appropriately for washing. The whole process should be operated gently to separate the cells that have not formed an array. After washing, observe the array of the cell array chip under a microscope.

[0145] Prepare screening medium. Pre-select and prepare complete cell culture medium, add preset types of fluorescent dyes to the complete cell culture medium according to the required dilution multiples, shake at low speed, mix thoroughly, and add substances such as antigens pre-labeled with fluorescent dyes to the fully mixed complete culture medium according to a certain dilution ratio as needed, shake at low speed, mix thoroughly, and obtain screening culture medium.

[0146] Add screening medium. Add pre-configured screening medium to the chip behind the cell array, and then place the chip behind the cell array in a cell incubator for incubation. The substances in the screening medium are specifically combined with the secretory substances of the cells to be sorted after the array is completed. The incubation time is set according to different projects, such as 3-8 hours for cell line development projects and no less than 8 hours for antibody discovery.

[0147] Second separation: After the incubation is completed, the screening culture medium is recovered, and the chip is washed 3-5 times with 0.01M PBS buffer until all suspended cells are basically removed, and the cells to be sorted after being cultured with fluorescent dyes are obtained in the biological culture chip.

[0148] In one of the embodiments, after acquiring image data of cells to be sorted contained in the biological culture chip, the terminal locates the containing cavities in the image data of the cells to be sorted according to the positioning marks of each containing cavity pre-set in the biological culture chip through a deep learning model, and locates the cells to be sorted in the located containing cavities.

[0149] Specifically, the biological culture chip is provided with positioning marks, which are formed on the matrix. Thus, multiple cavities can be positioned by these positioning marks. The form of the positioning marks is not limited. For example, coordinate lines can be set to position each cavity. When cells that meet the preset sorting conditions are observed, each cavity in which cells that meet the preset sorting conditions are placed can be quickly positioned and recorded. This further facilitates the subsequent separation of cells or other components that meet the preset sorting conditions in the cavities.

[0150] In one embodiment, during the imaging process, the optical path component in the imaging mechanism adjusts the focal length between the objective lens and the biological culture chip to focus the objective lens and image the cells to be sorted. By changing the focal length between the objective lens and the biological culture chip, the imaging of the cells to be sorted is clear, and the imaging mechanism obtains clear image data of the cells to be sorted.

[0151] In one embodiment, the carrier assembly further includes a storage unit disposed on the first carrier. The storage unit is used to store and hold cells that meet the preset sorting conditions, so as to facilitate the subsequent culture of the cells that meet the preset sorting conditions. The cells that meet the preset sorting conditions can be placed in the storage unit after being sucked by the selection mechanism, and then the storage unit is transferred to the incubator to carry out the next culture of the cells that meet the preset sorting conditions.

[0152] In one embodiment, during the cell sorting process, the method for operating the cell sorting device comprises the following steps:

[0153] Mount the pick-and-place unit to the mounting assembly on the picking mechanism.

[0154] Calibrate the position of the suction and release member. Specifically, under the control of the control mechanism, turn on the selection light source, and the selection moving component drives the suction and release member to move until the image data projected by the suction and release member is clearly presented on the terminal. Use the first fixing member and the second fixing member to fine-tune the suction and release member so that the projection mark is located at the preset position of the projection mapping of the suction and release member. Then, the member to be sorted without the cells to be sorted is placed in the fixed groove of the first carrier. The selection moving component drives the suction and release member to move until the image data of the suction and release port is clearly presented on the terminal. Record the position information of the suction and release member at this time to complete the position calibration of the suction and release member.

[0155] The pick-and-place parts are cleaned for the first time by means of cleaning fluid.

[0156] Place the objects to be sorted and the storage unit. Place the objects to be sorted with the cells to be sorted in the loading part of the first loading part, pull the elastic member on one side of the loading part to the other side of the loading part, so that the elastic member is biased against the objects to be sorted, and the objects to be sorted are placed and fixed on the first loading part. Place the storage unit with an appropriate amount of culture fluid in the placement slot of the first loading part.

[0157] Find cells that meet the preset sorting conditions. Under the control of the control mechanism, turn on the imaging light source, and the objective lens moving unit drives the objective lens to move until the image data of the cells to be sorted to be observed is clearly presented on the terminal, and the bright field image data of the cells to be sorted is recorded. Turn off the imaging light source, turn on the excitation light source, and let the light emitted by the excitation light source pass through the first filter unit, so that the cells to be sorted emit a first fluorescence, record the first fluorescence image data of the cells to be sorted, control the filter switching unit to switch to the second filter unit, so that the cells to be sorted emit a second fluorescence, record the second fluorescence image data of the cells to be sorted, control the filter switching unit to switch to the third filter unit, so that the cells to be sorted emit a third fluorescence, and record the third fluorescence image data of the cells to be sorted. Then turn off the excitation light source and turn on the imaging light source. Drive the to-be-sorted member to move to the preset area and replace the cells to be sorted that need to be observed. Repeat the above process to record the bright field image data and fluorescence image data of the cells to be sorted in different areas. Process the acquired bright field image data and fluorescence image data of the cells to be sorted, and analyze and find cells that meet the preset sorting conditions.

[0158] Absorb and recover cells that meet the preset sorting conditions. Under the control of the control mechanism, the selection moving component drives the suction and release part to move until the suction and release part reaches a position where the cells can be absorbed, and the position is marked as the preset position to complete the pre-absorption. The selection moving component drives the suction and release part to move to the preset position, and controls the suction and release part to absorb cells into the hollow pipe of the suction and release part. The selection moving component drives the suction and release part to move in a direction away from the carrier component. The selection moving component drives the suction and release part to move so that there is a preset distance between the suction and release port and the storage part of the storage unit, and the preset distance is the distance from the suction and release part to the storage part that satisfies the preset requirements. The suction and release part is controlled to release the cells that meet the preset sorting conditions absorbed in the hollow pipe into the storage part. The selection moving component drives the suction and release part to move in a direction away from the carrier component. Repeat the above process so that different cells that meet the preset sorting conditions are absorbed, released, and recovered into different storage parts until the absorption and release of all cells that meet the preset sorting conditions in the part to be sorted are completed. The parts to be sorted are recovered and stored, and the storage unit is transferred to the incubator for further cultivation.

[0159] After the cell sorting is completed, the suction and placement parts are cleaned for the second time with cleaning solution.

[0160] In this embodiment, the cell partitioning device selectively absorbs the cells that meet the preset sorting conditions in the objects to be sorted, and releases them into the corresponding storage part of the storage unit. This process is automatically completed under the control of the control mechanism, thereby improving the efficiency of cell sorting, avoiding damage to the suction and release objects and cells that meet the preset sorting conditions due to manual operations, ensuring the biological activity of cells that meet the preset sorting conditions, and realizing high-throughput cell sorting.

[0161] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0162] Based on the same inventive concept, the embodiment of the present application also provides a cell extraction device for implementing the cell extraction method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more cell extraction device embodiments provided below can refer to the limitations on the cell extraction method above, and will not be repeated here.

[0163] In one embodiment, Figure 4 As shown, a cell extraction device is provided, including: an identification module 402, a determination module 404 and an extraction module 406, wherein:

[0164] The recognition module 402 is used to obtain multiple cell images of cells to be sorted and identify fluorescent spots in each cell image through a deep learning model.

[0165] The determination module 404 is used to determine the fluorescence signal intensity of the fluorescence points in each cell image.

[0166] The extraction module 406 is used to extract cells that meet the preset sorting conditions from the cells to be sorted according to the fluorescence signal intensity of the fluorescence points in the multiple cell images.

[0167] In one embodiment, the identification module 402 is further used to identify the position data of the cells to be sorted in each cell image through a deep learning model; and identify the fluorescent spots of the cells to be sorted in each cell image according to the position data.

[0168] In one embodiment, the determination module 404 is also used to intercept the cells to be sorted in each cell image to obtain the cell image to be sorted corresponding to each cell image; perform grayscale analysis on the fluorescent points in the cell image to be sorted corresponding to each cell image to obtain the fluorescence signal intensity of the fluorescent points in each cell image.

[0169] In one embodiment, the identification module 402 is also used to irradiate the cells to be sorted with excitation light of multiple bands respectively, so that the cells to be sorted emit different fluorescence, and obtain multiple fluorescence images of the cells to be sorted; wherein the multiple fluorescence images include DAPI fluorescence images, TRITC fluorescence images and FITC fluorescence images; and the multiple fluorescence images are determined as multiple cell images.

[0170] In one embodiment, the extraction module 406 is also used to determine the fluorescence signal intensity of the cell to be sorted for each cell image according to the average grayscale value of the fluorescence points belonging to the same cell to be sorted; sort the fluorescence signal intensities of the cells to be sorted in each cell image; determine the cells that meet the preset sorting conditions according to the sorting results corresponding to the multiple cell images, and extract the cells that meet the preset sorting conditions from the cells to be sorted.

[0171] In one embodiment, the apparatus further comprises:

[0172] The comparison module is used to identify the first position data of each cell in the first bright field cell image before cell extraction and the second position data of each cell in the second bright field image after cell extraction through a deep learning model; compare the first position data with the second position data; if the comparison is consistent, the cell extraction fails and the cell extraction process is repeated.

[0173] Each module in the above-mentioned cell extraction device can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0174] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 5As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. Wherein, the processor, the memory and the input / output interface are connected via a system bus, and the communication interface, the display unit and the input device are connected to the system bus via the input / output interface. Wherein, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a cell extraction method is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0175] Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0176] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0177] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0178] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0179] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0180] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0181] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are 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.

[0182] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A cell extraction method, characterized in that: The method comprises: Obtain multiple cell images of cells to be sorted, and identify the fluorescent spots in each cell image through a deep learning model; Determine the fluorescence signal intensity of each fluorescent spot in each cell image; According to the fluorescence signal intensity of the fluorescent spots in the multiple cell images, cells meeting the preset sorting conditions are extracted from the cells to be sorted.

2. The method according to claim 1, characterized in that The method of identifying the fluorescent spots in each cell image by a deep learning model includes: Identify the location data of cells to be sorted in each cell image through a deep learning model; The fluorescent spots of the cells to be sorted in each cell image are identified according to the position data.

3. The method according to claim 1, characterized in that Determining the fluorescence signal intensity of the fluorescent spot in each cell image comprises: Intercepting the cells to be sorted in each cell image to obtain the cell image to be sorted corresponding to each cell image; Grayscale analysis is performed on the fluorescent spots in the to-be-sorted cell images corresponding to each cell image to obtain the fluorescent signal intensity of the fluorescent spots in each cell image.

4. The method according to claim 1, characterized in that The obtaining of multiple cell images of cells to be sorted comprises: Irradiating the cells to be sorted respectively with excitation lights of multiple bands, so that the cells to be sorted emit different fluorescence, and obtaining multiple fluorescence images of the cells to be sorted; wherein the multiple fluorescence images include DAPI fluorescence images, TRITC fluorescence images and FITC fluorescence images; The plurality of fluorescent images are determined as a plurality of cell images.

5. The method according to claim 1, characterized in that The step of extracting cells meeting preset sorting conditions from the cells to be sorted according to the fluorescence signal intensities of the fluorescence spots in the multiple cell images comprises: For each cell image, the fluorescence signal intensity of the cell to be sorted is determined according to the average gray value of the fluorescence spots belonging to the same cell to be sorted; Sort the fluorescence signal intensity of the cells to be sorted in each cell image; Cells meeting preset sorting conditions are determined according to the sorting results corresponding to the multiple cell images, and the cells meeting the preset sorting conditions are extracted from the cells to be sorted.

6. The method according to claim 1, characterized in that The method further comprises: Using a deep learning model, first position data of each cell in a first bright field cell image before cell extraction and second position data of each cell in a second bright field image after cell extraction are respectively identified; comparing the first position data with the second position data; If the alignment is consistent, the cell extraction fails and the cell extraction process is repeated.

7. A cell extraction device, characterized in that: The device comprises: The recognition module is used to obtain multiple cell images of cells to be sorted and identify the fluorescent spots in each cell image through a deep learning model; A determination module, used for determining the fluorescence signal intensity of the fluorescent spot in each cell image; The extraction module is used to extract cells that meet the preset sorting conditions from the cells to be sorted according to the fluorescence signal intensity of the fluorescence points in the multiple cell images.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images