Centrifugal system and cell processing method and device

Abstract

The invention relates to a centrifugal system and a cell processing method and device. The centrifugal system comprises a sample chamber, a mixing and resuspending chamber, a collecting chamber and a waste liquid chamber; the sample chamber is used for placing a preset reagent; a first channel for enabling liquid in the sample chamber to enter the mixing resuspension chamber at a preset centrifugal speed is connected between the mixing resuspension chamber and the sample chamber, and the mixing resuspension chamber is used for mixing a preset reagent entering from the first channel and a to-be-detected sample liquid input from the input channel at the mixing centrifugal speed to obtain a mixed liquid; separating the mixed liquid to obtain cell precipitate and waste liquid; the collecting chamber is used for collecting cell precipitates; the waste liquid chamber is used for collecting waste liquid. According to the centrifugal structure, automatic centrifugal separation can be achieved, and meanwhile damage to cellular morphology and interference to other cells are reduced.

Description

Technical Field

[0001] The present application relates to the field of biomedical technology, and in particular to a centrifugal system, a cell processing method and a device. Background Art

[0002] The "gold standard" for bladder cancer diagnosis and recurrence monitoring is cystoscopy, but cystoscopy is invasive and increases the risk of other urinary system complications. Traditional imaging (ultrasound, computed tomography, etc.) often finds tumors in the middle and late stages, with poor treatment effects and poor prognosis. How to make early diagnosis and recurrence monitoring of urinary system transitional cell carcinoma has become an urgent problem that needs to be solved.

[0003] As one of the research methods of cell molecular genetics, fluorescence in situ hybridization (FISH) technology has been used in recent years to detect chromosome number and structural aberrations. In related technologies, FISH technology is used to examine chromosomes in exfoliated cells, but the sensitivity of this method is low, and the detection process is prone to loss of cell morphology, resulting in low accuracy of cell detection. Summary of the invention

[0004] Based on this, it is necessary to provide a centrifugal system, a cell processing method and a device that can improve the accuracy of cell detection in order to solve the above technical problems.

[0005] In a first aspect, the present application provides a centrifugal system, the centrifugal system comprising: a sample chamber, a mixing and resuspension chamber, a collection chamber, and a waste liquid chamber;

[0006] The sample chamber is used to place preset reagents;

[0007] The mixing and resuspension chamber is connected to the sample chamber with a first channel for allowing the liquid in the sample chamber to enter the mixing and resuspension chamber at a preset centrifugal speed, and is used to mix the preset reagent entering the first channel and the sample liquid to be detected input from the input channel at the mixing centrifugal speed to obtain a mixed liquid, and separate the mixed liquid to obtain a cell precipitate and a waste liquid;

[0008] The collection chamber is used to collect the cell precipitation; the waste liquid chamber is used to collect the waste liquid.

[0009] In one embodiment, the preset reagent includes a first reagent and a second reagent; the sample chamber includes a first sample chamber and a second sample chamber, the mixing centrifugal speed includes a first mixing centrifugal speed and a second mixing centrifugal speed; the sample liquid to be detected enters the mixing and resuspension chamber before the preset reagent; the mixing and resuspension chamber includes a first mixing and resuspension chamber and a second mixing and resuspension chamber; wherein,

[0010] The first mixing and resuspension chamber is used to mix and centrifuge the sample liquid to be detected at the first mixing centrifugal speed, or to mix and centrifuge the sample liquid to be detected and the first reagent to obtain a first mixed liquid, so that the waste liquid chamber collects the waste liquid of the first mixed liquid through the second channel, and collects the first cell precipitate of the first mixed liquid into the second mixing and resuspension chamber through the second channel;

[0011] The second mixing and resuspension chamber is used to mix the first cell pellet with the second reagent in the second sample chamber at the second mixing centrifugal speed to obtain a second mixed liquid, so that the collecting chamber collects the second cell pellet in the second mixed liquid, and the waste liquid of the second mixed liquid is collected through the waste liquid chamber.

[0012] In one embodiment, the preset centrifugal speed includes a first centrifugal speed, and the system further includes:

[0013] The first sample chamber is used to store the first reagent;

[0014] The first channel is used to introduce the first reagent into the first mixing and resuspension chamber under the action of the first centrifugal speed after the sample solution to be detected is input through the input channel.

[0015] In one embodiment, the preset centrifugal speed includes a second centrifugal speed, and the second sample chamber includes a plurality of sample sub-chambers and third channels corresponding to each sample sub-chamber;

[0016] The plurality of sample subchambers are used to store the second reagent;

[0017] The third channel is used to introduce the second reagent into the second mixing and resuspension chamber under the action of the second centrifugal speed.

[0018] In one embodiment, the collection chamber is a multi-layer filter membrane structure; the pore size of the multi-layer filter membrane structure decreases from top to bottom, and the pore size of the bottom filter membrane of the multi-layer filter membrane structure is smaller than the diameter of the desired cells; the multi-layer filter membrane structure includes a first filter membrane, a second filter membrane and a third filter membrane; the first filter membrane is used to retain a first impurity of a first diameter; the second filter membrane is used to retain a second impurity of a second diameter; the third filter membrane is used to collect the cell precipitate and filter a third impurity of a third diameter;

[0019] The first diameter is greater than the second diameter, and the second diameter is greater than the third diameter.

[0020] In one embodiment, the waste liquid chamber includes a first waste liquid chamber and a second waste liquid chamber;

[0021] The first waste liquid chamber is connected to the first mixing and resuspension chamber with a fourth channel for allowing the waste liquid in the first mixing and resuspension chamber to enter the waste liquid chamber at a third centrifugal speed;

[0022] The second waste liquid chamber is used to collect the third impurities with the third diameter.

[0023] In one embodiment, a fifth channel is connected between the collection chamber and the second mixing and resuspension chamber, which allows the second mixed liquid to enter the collection chamber under the action of a fourth centrifugal speed.

[0024] In a second aspect, the present application further provides a method for processing cells, which is applied to the centrifugal system as described in the first aspect; the method comprises:

[0025] Injecting the sample liquid to be detected from the input channel into the first mixing and resuspension chamber for pretreatment to obtain the pretreated sample liquid to be detected; injecting the first reagent into the first mixing and resuspension chamber through the first channel under the action of the first centrifugal speed; mixing and centrifuging the first reagent and the pretreated sample liquid to be detected under the action of the first mixing centrifugal speed to obtain a first mixed liquid; when the centrifugal speed of the centrifugal system matches the preset speed threshold, activating the fourth channel, sucking the waste liquid of the first mixed liquid into the first waste liquid chamber through the fourth channel, and collecting the first cell precipitate of the first mixed liquid into the second mixing and resuspension chamber;

[0026] Under the action of the second centrifugal speed, the second reagent is injected into the second mixing and resuspension chamber, and the first cell pellet and the second reagent are mixed and centrifuged under the action of the second mixing centrifugal speed to obtain a second mixed liquid; when the centrifugal speed of the centrifugal system reaches a fourth centrifugal speed, the second mixed liquid is injected into the collecting chamber through the fifth channel, the second cell pellet of the second mixed liquid is collected through the collecting chamber, and the second waste liquid of the second mixed liquid is collected through the waste liquid chamber.

[0027] In one embodiment, the method further comprises:

[0028] The second cell precipitation is rinsed with a cell buffer to obtain a cell suspension of the sample solution to be detected; the cell suspension is imaged by coherent Raman scattering using one or more Raman shifts to obtain a coherent Raman scattering image corresponding to each Raman shift;

[0029] Using an image segmentation network model, performing single-cell image segmentation processing on the coherent Raman scattering image to obtain a plurality of single-cell coherent Raman scattering images;

[0030] Using a feature extraction model, single-cell feature extraction is performed on the coherent Raman scattering image of each single cell in the coherent Raman scattering images of the multiple single cells to obtain a single-cell feature map corresponding to the coherent Raman scattering image of each single cell;

[0031] Combining the single cell feature maps according to the target object to obtain a single cell feature map group corresponding to the target object;

[0032] A preset classification model is used to perform feature analysis on the single-cell feature map group to obtain the abnormal probability of the target object; based on the abnormal probability and a preset abnormal probability threshold, the abnormal condition of the target object is determined.

[0033] In one embodiment, the single cell feature map includes cell morphological characteristics and cell metabolic characteristics; the cell morphological characteristics include at least one or more of cell area, cell eccentricity, cell fitted circle diameter, cell major axis length, cell minor axis length, cell circumference, cell true area / fitted circle area, cell nuclear area, cytoplasm area, and cell nuclear-cytoplasm ratio; the cell metabolic characteristics include at least one or more of cell lipid droplet number, cell lipid droplet area, lipid droplet ratio of cell area, distance from lipid droplet to cell center, average distance from lipid droplet to cell center of gravity, and cell lipid droplet grayscale.

[0034] In one embodiment, the cell suspension is subjected to coherent Raman scattering imaging using one or more Raman shifts to obtain a coherent Raman scattering image corresponding to each Raman shift, including:

[0035] Input one or more groups of excitation light into a coherent Raman scattering imaging system respectively, wherein the Raman shift of each group of excitation light is different, and each group of excitation light includes a first excitation light and a second excitation light; the coherent Raman scattering imaging system includes at least: an optical path module and a signal acquisition module;

[0036] For each group of excitation light, the first excitation light and the second excitation light are processed in a colinear manner through the optical path module to obtain a combined optical path, and the combined optical path is used to scan the cell suspension to obtain a coherent Raman scattering signal; the coherent Raman scattering signal is processed through the signal acquisition module to obtain a coherent Raman scattering image.

[0037] In a third aspect, the present application further provides a cell processing device, including: the device is used in a centrifugal system, the device includes:

[0038] A first processing module is used to inject the sample liquid to be detected from the input channel into the first mixing and resuspension chamber for pretreatment to obtain the pretreated sample liquid to be detected; under the action of a first centrifugal speed, inject a first reagent into the first mixing and resuspension chamber through the first channel; under the action of the first mixing centrifugal speed, the first reagent and the pretreated sample liquid to be detected are mixed and centrifuged to obtain a first mixed liquid; when the centrifugal speed of the centrifugal system matches a preset speed threshold, the fourth channel is activated, and the waste liquid of the first mixed liquid is sucked into the first waste liquid chamber through the fourth channel, and the first cell precipitate of the first mixed liquid is collected into the second mixing and resuspension chamber;

[0039] The second processing module is used to inject the second reagent into the second mixing and resuspension chamber under the action of the second centrifugal speed, and mix and centrifuge the first cell precipitate and the second reagent under the action of the second mixing centrifugal speed to obtain a second mixed liquid; when the centrifugal speed of the centrifugal system reaches a fourth centrifugal speed, inject the second mixed liquid into the collecting chamber through the fifth channel, collect the second cell precipitate of the second mixed liquid through the collecting chamber, and collect the second waste liquid of the second mixed liquid through the waste liquid chamber.

[0040] In a fourth aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the method provided in the second aspect when executing the computer program.

[0041] In a fifth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the method provided in the second aspect above.

[0042] In a sixth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the steps in the method provided in the second aspect above.

[0043] The above-mentioned centrifugal system, cell processing method and device, by setting a sample chamber, a mixing and resuspension chamber, a collection chamber and a waste liquid chamber, the sample liquid to be tested is centrifuged and mixed in the mixing and resuspension chamber by adding preset reagents, and the generated waste liquid and cell sediment are separated to obtain the required cell sediment, thereby realizing automated centrifugal separation, achieving sample enrichment and impurity removal, while reducing damage to cell morphology and improving the accuracy of subsequent detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] FIG1 is a schematic structural diagram of a centrifugal system in one embodiment;

[0046] FIG2 is a schematic diagram of the structure of a centrifugal system in one embodiment;

[0047] FIG3 is a schematic diagram of the structure of a centrifugal system in one embodiment;

[0048] FIG4 is a schematic diagram of a multi-layer filter membrane structure in one embodiment;

[0049] FIG5 is a schematic diagram of a process for treating cells in one embodiment;

[0050] FIG6 is a schematic flow chart of a method for treating cells in one embodiment;

[0051] FIG7 is a schematic diagram of the structure of a coherent Raman scattering imaging system in one embodiment;

[0052] FIG8 is a schematic diagram of the structure of an image segmentation network model in one embodiment;

[0053] FIG9 is a schematic diagram of a coherent Raman scattering image of a single cell in one embodiment;

[0054] FIG10 is a schematic diagram of cell processing in one embodiment;

[0055] FIG11 is a schematic diagram of signal intensity under different Raman shifts in one embodiment;

[0056] FIG12 is a block diagram of a cell processing device according to an embodiment;

[0057] FIG. 13 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0058] 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.

[0059] In an exemplary embodiment, as shown in FIG. 1 and FIG. 2 , FIG. 1 is a schematic structural diagram of a centrifugal system in an embodiment, and the centrifugal system includes: a sample chamber, a mixing and resuspension chamber, a collection chamber, and a waste liquid chamber.

[0060] Sample chamber, used to place preset reagents.

[0061] The mixing and resuspension chamber is connected to the sample chamber and has a first channel for allowing the liquid in the sample chamber to enter the mixing and resuspension chamber at a preset centrifugal speed. The mixing and resuspension chamber is used to mix the preset reagent entering from the first channel and the sample liquid to be detected input from the input channel at the mixing centrifugal speed to obtain a mixed liquid, and separate the mixed liquid to obtain a cell precipitate and a waste liquid.

[0062] The collection chamber is used to collect cell sediments. The waste liquid chamber is used to collect waste liquid.

[0063] The preset reagent may be a reagent for decomposing impurities in the sample liquid to be detected, and the preset reagent may be one or more. The sample liquid to be detected may be urine, and the preset reagent may be a reagent that can decompose other impurities in urine, such as red blood cell lysis solution and other reagents. The sample chamber may include multiple subchambers for storing different types of preset reagents respectively; the sample chamber may also include a chamber for placing different preset reagents in sequence, which is not limited here.

[0064] Specifically, as shown in FIG1 , the sample chamber can be connected to the mixing and resuspension chamber, and the preset reagent in the sample chamber is injected into the mixing and resuspension chamber through the first channel, and mixed with the sample liquid to be detected in the mixing and resuspension chamber to obtain a mixed liquid. The mixed liquid is separated in the first mixing and resuspension chamber to obtain a cell precipitate and a waste liquid. The separation method can be to separate the waste liquid and the cell precipitate in the mixed liquid by setting a corresponding centrifugal speed for the centrifugal system, and the specific separation method is not limited here.

[0065] Specifically, the cell precipitate obtained after the separation of the mixed liquid can be collected by a collection chamber or other container so as to perform subsequent treatment on the collected cell precipitate. The waste liquid after the separation of the mixed liquid can be collected by the waste liquid chamber, and the collection method can be to introduce the waste liquid into the waste liquid chamber through a siphon channel, or the waste liquid after the cell precipitate is filtered through the collection chamber by the mixed liquid can be collected into the waste liquid chamber. As shown in Figures 1 and 2, Figure 2 is a schematic diagram of the structure of a centrifugal system in an embodiment. How to collect the cell precipitate into the collection chamber or the waste liquid into the waste liquid chamber is not limited here, and can be adjusted according to the actual application scenario.

[0066] Specifically, a material that is relatively easy to puncture can be designed at the bottom of the sample chamber. For example, the material that is relatively easy to puncture can be aluminum foil, and spikes are designed at corresponding positions of the aluminum foil, and springs are arranged on the plane where the spikes are located and at the bottom of the sample chamber. Under the action of the centrifugal speed, the spring can be deformed, and the aluminum foil is punctured under the action of the spikes, and the preset reagent in the sample chamber can flow into the mixing and resuspension chamber.

[0067] For example, in this embodiment, the centrifugal systems in FIG. 1 and FIG. 2 can be superimposed, the cell precipitate in the collection chamber of FIG. 1 can be directly introduced into the mixing and resuspension chamber of FIG. 2, and the cell precipitate obtained in FIG. 1 can be processed in the next step. For example, the system of FIG. 1 can be used as a device for preliminary processing of the sample solution to be detected, and the cell precipitate after the preliminary processing can be directly introduced into a device for further processing of the cell precipitate. As shown in FIG. 3, FIG. 3 is a schematic diagram of the structure of a centrifugal system in an embodiment.

[0068] Specifically, the centrifugal system can be used in a centrifuge, and the centrifuge provides the required centrifugal speed for the centrifugal system, thereby reducing application costs.

[0069] In summary, this embodiment sets up a sample chamber, a mixing and resuspension chamber, a collection chamber and a waste liquid chamber, and the sample liquid to be tested is centrifuged and mixed in the mixing and resuspension chamber by adding preset reagents, and the generated waste liquid and cell precipitate are separated to obtain the required cell precipitate, thereby realizing automated centrifugal separation, reducing damage to cell morphology, and improving the accuracy of subsequent detection.

[0070] In an exemplary embodiment, the mixing and resuspension chamber includes a first mixing and resuspension chamber and a second mixing and resuspension chamber; wherein the first mixing and resuspension chamber is used to mix and centrifuge the sample liquid to be detected at a first mixing centrifugal speed, or to mix and centrifuge the sample liquid to be detected with a first reagent to obtain a first mixed liquid, so that the waste liquid chamber collects the waste liquid of the first mixed liquid through the waste liquid chamber, and collects the first cell precipitate of the first mixed liquid into the second mixing and resuspension chamber through the second channel; the second mixing and resuspension chamber is used to mix the first cell precipitate with the second reagent in the second sample chamber at a second mixing centrifugal speed to obtain a second mixed liquid, so that the collecting chamber collects the second cell precipitate in the second mixed liquid, and collects the waste liquid of the second mixed liquid through the waste liquid chamber.

[0071] Wherein, the preset reagent includes a first reagent and a second reagent, and the types of the first reagent and the second reagent can be different or the same. The sample chamber includes a first sample chamber and a second sample chamber, and the first sample chamber and the second sample chamber can both be used to place the preset reagent. The first sample chamber and the second sample chamber only represent different positions in the centrifugal system. The mixed centrifugal speed includes a first mixed centrifugal speed and a second mixed centrifugal speed, and the first mixed centrifugal speed is less than the second mixed centrifugal speed. The sample liquid to be detected needs to enter the mixing and resuspension chamber before the preset reagent. The second cell precipitate is the cells required in the extracted sample liquid to be detected.

[0072] Specifically, the centrifugal system shown in FIG3 includes a mixing and resuspension chamber including a first mixing and resuspension chamber 31 and a second mixing and resuspension chamber 32. The first mixing and resuspension chamber 31 and the second resuspension chamber are located at different positions of the centrifugal structure. The first mixing and resuspension chamber 31 can receive the sample liquid to be detected injected through the input channel, and then mix and centrifuge the sample liquid to be detected at a mixed centrifugal speed to complete the pretreatment of the sample liquid to be detected, so that the waste liquid chamber can collect the waste liquid generated through the channel. Under the action of the first centrifugal speed, the first reagent can be injected into the first mixing and resuspension chamber 31 through the first channel. Under the action of the first mixing centrifugal speed, the first mixing and resuspension chamber 31 mixes and centrifuges the first reagent with the sample liquid to be detected; the waste liquid chamber can collect the waste liquid generated through the set channel; the cell precipitate can pass through the second channel, and the first cell precipitate of the first mixed liquid is collected to the second mixing and resuspension chamber 32.

[0073] The second mixing and resuspension chamber 32 is used to receive the first cell precipitate, and when the centrifugal speed reaches the preset centrifugal speed, receive the second reagent in the second sample chamber. At the second mixing centrifugal speed, the second reagent is mixed and centrifuged with the first cell precipitate to obtain a second mixed liquid. The second mixed liquid can flow into the collection chamber through the channel. The collection chamber retains the second cell precipitate and discharges other waste liquid into the waste liquid chamber.

[0074] In an exemplary embodiment, the centrifugal system further includes: a first sample chamber 34 for storing a first reagent; and a first channel for introducing the first reagent into a first mixing and resuspension chamber under the action of a first centrifugal speed after the sample liquid to be tested is input through the input channel 33.

[0075] Specifically, the first sample chamber 34 can be used to store the first reagent, and the first reagent can be a red blood cell lysate. The first channel can establish a connection between the first sample chamber 34 and the first mixing and resuspension chamber under the action of the first centrifugal speed, inject the first reagent in the first sample chamber 34 into the first mixing and resuspension chamber, and perform preliminary treatment on the first reagent and the sample liquid to be detected. Exemplarily, as shown in Figure 3, the bottom of the first sample chamber 34 can be made of aluminum foil, and spikes are set at the position corresponding to the aluminum foil. The position of the spike is designed to be connected to the bottom of the first sample chamber 34 with a spring. Under the action of the first centrifugal speed, the spike can pierce the aluminum foil, establish the first channel between the first sample chamber 34 and the first mixing and resuspension chamber, and introduce the first reagent from the first sample chamber 34 into the first mixing and resuspension chamber.

[0076] In an exemplary embodiment, the second sample chamber 35 includes a plurality of sample subchambers and third channels corresponding to each sample subchamber; the plurality of sample subchambers are used to store the second reagent; and the third channel is used to introduce the second reagent into the second mixing and resuspension chamber under the action of the second centrifugal speed.

[0077] Specifically, the second sample chamber 35 may include a plurality of sub-sample chambers, each of which may store a second reagent. The second reagent may be one or more different reagents. Under the action of the second centrifugal speed, the third channel establishes a connection between the second sample chamber 35 and the third channel, and introduces the reagent in each sub-sample chamber into the second mixing and resuspension chamber in sequence. The centrifuge can sequentially puncture the aluminum foil of each sub-sample chamber by designing different second centrifugal speeds, so that the second reagent in each sub-sample chamber is sequentially introduced into the second mixing and resuspension chamber. Exemplarily, as shown in FIG3 , the second sample chamber 35 in FIG3 includes two sub-sample chambers, and NaHCO3 (sodium bicarbonate) and PBS (phosphate buffered saline) may be placed in the two sub-sample chambers in sequence. At most one reagent is placed in each sub-sample chamber according to specific application requirements, which is not limited here. Aluminum foil is set at the bottom of the second sample chamber 35, and spikes are set below the corresponding position of the aluminum foil. Under the action of the second centrifugal speed, the spikes sequentially pierce the aluminum foil of the sub-sample chamber where NaHCO3 is located and the aluminum foil of the sub-sample chamber where PBS is located to form a third channel, and the second reagent is sequentially introduced from each sub-sample chamber into the second mixing and resuspension chamber.

[0078] In an exemplary embodiment, the collection chamber 36 is a multi-layer filter membrane structure; the pore size of the multi-layer filter membrane structure decreases from top to bottom, and the pore size of the bottom filter membrane of the multi-layer filter membrane structure is smaller than the diameter of the required cells; the multi-layer filter membrane structure includes a first filter membrane, a second filter membrane and a third filter membrane; the first filter membrane is used to retain the first impurities of the first diameter; the second filter membrane is used to retain the second protein flocs; the third filter membrane is used to collect the second cell precipitate and filter the impurities of the third diameter.

[0079] The first diameter is greater than the second diameter, the second diameter is greater than the third diameter, and the second cell sediment is smaller than the second diameter and greater than the third diameter. The first impurity and the second impurity are protein floccules of different diameters. The third impurity may be impurities such as bacteria that meet the third diameter range.

[0080] In one example, as shown in FIG4 , FIG4 is a schematic diagram of the structure of a multilayer filter membrane structure in an embodiment. The collection chamber 36 is a multilayer filter membrane structure. The multilayer filter membrane structure is fixed to the tube wall of the centrifugal system and is detachable. The pore size of each layer of the multilayer filter membrane structure is set according to the diameter of the impurities to be filtered and the diameter of the required cells, and the impurities larger than the diameter of the required cells are filtered before reaching the bottom filter membrane, that is, the pore size of the bottom filter membrane should be smaller than the diameter of the required cells and larger than the diameter of the impurities to be filtered, and the required cells are collected on the bottom filter membrane, providing a basis for subsequent detection of the type of cells, and further improving the accuracy of cell detection.

[0081] For example, as shown in Table 1, Table 1 is the diameters of common cells in one embodiment.

[0082] Table 1

[0083]

[0084] The multi-layer filter membrane structure is divided into three layers of filter membranes, wherein the first filter membrane has a pore size of 80 μm, which is used to filter out the first impurity of the first diameter. The second filter membrane has a pore size of 40 μm, which is used to filter out the second impurity of the second diameter. The third filter membrane has a pore size of 3 μm, which is used to filter out bacteria and collect the desired cells. It should be understood that the above values ​​are only used for example and cannot limit the specific pore size of the multi-layer filter membrane structure.

[0085] In an exemplary embodiment, the waste liquid chamber includes a first waste liquid chamber 37 and a second waste liquid chamber 38; the first waste liquid chamber 37 is connected to the first mixing and resuspension chamber by a fourth channel 39 for allowing the waste liquid of the first mixing and resuspension chamber to enter the waste liquid chamber at a third centrifugal speed; the second waste liquid chamber 38 is used to collect third impurities of a third diameter.

[0086] Specifically, as shown in FIG3 , the fourth channel 39 can be a siphon channel, which is composed of inner and outer channels, the inner siphon channel is used to remove the waste liquid generated when the sample liquid to be detected is pretreated, and the outer siphon channel is used to remove the waste liquid of the first mixed liquid. The end of the fourth channel 39 connected to the first waste liquid chamber 37 is closed with aluminum foil, and a spike is set at the lower end of the aluminum foil. When the centrifugal speed is the third centrifugal speed, the spike can pierce the aluminum foil, and the connection between the first waste liquid chamber 37 and the first mixing and resuspension chamber is established through the fourth channel 39. When the centrifugal speed drops to 0, the siphon channel is activated, and the waste liquid of the first mixed liquid in the first mixing and resuspension chamber can be siphoned to the first waste liquid chamber 37, so as to separate the waste liquid and cell precipitation of the first mixed liquid.

[0087] Specifically, the second waste liquid chamber 38 can collect waste liquid after the second mixed liquid passes through the collection chamber 36 .

[0088] In an exemplary embodiment, a fifth channel is connected between the collection chamber 36 and the second mixing and resuspension chamber 32 for allowing the second mixed liquid to enter the collection chamber 36 under the fourth centrifugal speed.

[0089] Specifically, as shown in FIG3 , the bottom of the second mixing and resuspension chamber 32 can be set to aluminum foil, and spikes are set at the lower end of the corresponding position of the aluminum foil, and the spikes can pierce the aluminum foil under the action of the fourth centrifugal speed. At the same time, the fifth channel can establish a connection between the second mixing and resuspension chamber and the collection chamber, and introduce the second mixed liquid in the second mixing and resuspension chamber 32 into the collection chamber 36, so as to achieve the enrichment of the second cell precipitation in the collection chamber 36.

[0090] Specifically, the speed values ​​of the first centrifugal speed, the second centrifugal speed, the third centrifugal speed and the fourth centrifugal speed increase successively. The specific speed values ​​can be set according to the actual application scenario and are not limited here, so as to realize the automatic flow of liquid in one chamber into the next chamber in sequence.

[0091] In an exemplary embodiment, as shown in FIG5 , FIG5 is a schematic flow chart of a cell processing method provided by an embodiment, which is applied to a centrifugal system as in the above-mentioned embodiment. The cell processing method comprises:

[0092] Step 501, injecting the sample liquid to be detected from the input channel into the first mixing and resuspension chamber for pretreatment to obtain the pretreated sample liquid to be detected; under the action of the first centrifugal speed, injecting the first reagent through the first channel into the first mixing and resuspension chamber; under the action of the first mixing centrifugal speed, the first reagent and the pretreated sample liquid to be detected are mixed and centrifuged to obtain a first mixed liquid; when the centrifugal speed of the centrifugal system matches the preset speed threshold, activating the fourth channel, sucking the waste liquid of the first mixed liquid into the first waste liquid chamber through the fourth channel, and collecting the first cell precipitate of the first mixed liquid into the second mixing and resuspension chamber.

[0093] Step 502: Under the action of the second centrifugal speed, the second reagent is injected into the second mixing and resuspension chamber, and the first cell pellet and the second reagent are mixed and centrifuged under the action of the second mixing centrifugal speed to obtain a second mixed liquid; when the centrifugal speed of the centrifugal system reaches a fourth centrifugal speed, the second mixed liquid is injected into the collecting chamber through the fifth channel, the second cell pellet of the second mixed liquid is collected through the collecting chamber, and the second waste liquid of the second mixed liquid is collected through the waste liquid chamber.

[0094] The preset speed threshold may be 0, and when the centrifugal speed drops to 0, it may match the preset speed threshold.

[0095] In one example, the cell processing method is applied to a centrifugal system as shown in FIG3 , and a urine sample is used as a sample to be tested. The specific processing flow is as follows:

[0096] (1) 30 ml of urine sample was injected into the first mixing and resuspension chamber, 1.5 ml of red blood cell lysis solution was injected into the first sample chamber, 0.1 ml of NaHCO3 was injected into the first sub-sample chamber, and 1 ml of PBS was injected into the second sub-sample chamber.

[0097] (2) Centrifuge structure: Centrifuge at 3000 rpm for 10 min.

[0098] (3) Centrifuge at 3500 rpm for 5 seconds. Compress the spring and puncture the aluminum foil at the bottom of the fourth channel.

[0099] (4) Reduced to 0 rpm, the siphon channel is opened, and the waste liquid of the urine sample enters the first waste liquid chamber through the siphon effect.

[0100] (5) Centrifuge at 3700 rpm for 5 seconds. The spring is compressed, puncturing the aluminum foil of the first sample chamber, and the red blood cell lysis chamber flows into the first mixing and resuspension chamber.

[0101] (6) Repeat the mixing process for 15 min (centrifugation at 1000 rpm for 10 s followed by centrifugation at 2500 rpm for 10 s, repeated) to lyse the red blood cells in the urine sample.

[0102] (7) Centrifuge at 3000 rpm for 10 min. Centrifuge at 3900 rpm for 5 s. Compress the spring and puncture the aluminum foil of the fourth channel.

[0103] (8) Reduced to 0 rpm, the fourth channel is opened, and the waste liquid of the first mixed liquid enters the first waste liquid chamber through siphon action.

[0104] (9) Centrifuge at 4100 rpm for 1 min. The spring is compressed, piercing the aluminum foil at the bottom of the first mixing and resuspension chamber, and the sample sediment (first cell pellet) flows into the second mixing and resuspension chamber.

[0105] (10) Perform a rotation process (centrifuge at 400 rpm for 10 s and then reduce to 0 rpm), puncture the aluminum foil at the bottom of the first subsample chamber, and release NaHCO3 (sodium bicarbonate) into the second mixing and resuspension chamber.

[0106] (11) Repeat the mixing process for 15 min to digest the crystals in the urine sample.

[0107] (12) Perform a rotation process to puncture the second sub-sample chamber and release PBS (phosphate buffered saline) into the second mixing and resuspension chamber.

[0108] (13) Repeat the mixing process for 1 min to obtain a second mixed liquid.

[0109] (14) Centrifuge at 4500 rpm for 3 min, puncture the aluminum foil at the bottom of the second mixing and resuspension chamber, and accelerate the mixed sample to flow through the multi-layer filter membrane under the action of centrifugal force. Defoliated cells, protein flocs and other substances are collected by the multi-layer filter membrane structure, and the waste liquid (including digested red blood cells, digested crystals and bacteria) enters the second waste liquid chamber.

[0110] (15) Reduce to 0 rpm, restore the centrifuge tube, collect the desired cell sample (second cell pellet) on the third filter membrane and take it out.

[0111] In an exemplary embodiment, as shown in FIG6 , FIG6 is a schematic flow chart of a cell processing method provided by an embodiment, and the cell processing method further includes:

[0112] Step 601: rinse the second cell precipitate with a cell buffer to obtain a cell suspension of a sample solution to be detected.

[0113] Wherein, the cell buffer can be PBS buffer.

[0114] Specifically, the second cell pellet was washed away with PBS. The collected cell suspension was concentrated to 1*10 7 cell / ml, and pipette 3 μl of cell suspension onto a glass slide for the next step of imaging.

[0115] Step 602: Use one or more Raman shifts to perform coherent Raman scattering imaging on the cell suspension to obtain a coherent Raman scattering image corresponding to each Raman shift.

[0116] Specifically, the present embodiment may use a coherent Raman scattering imaging system to detect the cells to be detected, as shown in FIG7 , which is a schematic diagram of the structure of a coherent Raman scattering imaging system, using a first excitation light and a second excitation light to irradiate the urine exfoliated cell specimen; in response to the first excitation light and the second excitation light irradiating the urine exfoliated cell specimen, receiving reflected light having at least one Raman shift characteristic peak from the urine exfoliated cell specimen (cell suspension) to obtain the coherent Raman scattering image of the urine exfoliated cell specimen. The first excitation light and the second excitation light each have a predetermined frequency; the first excitation light and the second excitation light are both pulsed lasers; the first excitation light and the second excitation light are both pulsed lasers with a certain frequency difference, and the frequency difference may be 2800~3100cm -1 , repetition frequency is 10~300MHz, and pulse width is 10fs~20ps.

[0117] Specifically, the coherent Raman scattering imaging system in this embodiment is a stimulated Raman scattering imaging system, which collects stimulated Raman scattering (SRS) signals, wherein the SRS signals include stimulated Raman gain signals and stimulated Raman loss signals. In addition to the stimulated Raman scattering imaging system, the coherent Raman scattering imaging system also includes a coherent anti-Stokes Raman scattering (CARS) imaging system. In the imaging system, the common characteristics of CARS and SRS are: 1. Both require two beams of excitation light with a certain frequency difference to excite the sample to generate a coherent Raman scattering signal. 2. Both require an optical path device composed of components such as a reflector, a dichroic mirror, and a scanning galvanometer, so that the two beams of excitation light overlap in space and time, and realize scanning imaging of the sample. However, for SRS signal acquisition, the signal light is collected by a device such as a photodetector, and the excitation light needs to be modulated and the collected signal needs to be demodulated to realize the collection of stimulated Raman loss or stimulated Raman gain signals. The excitation light can be modulated by using an optoelectronic modulator, and the signal can be demodulated by using a lock-in amplifier. The coherent anti-Stokes Raman scattering signal can be collected by a photomultiplier tube (PMT) and converted into an electrical signal. At the same time, polarization detection, pump and Stokes field time delay detection, and backward detection CARS can be used to suppress the non-resonant background signal in the coherent anti-Stokes Raman scattering signal.

[0118] Step 603: Use the image segmentation network model to perform single-cell image segmentation processing on the coherent Raman scattering image to obtain coherent Raman scattering images of multiple single cells.

[0119] Among them, the image segmentation network model can be a comprehensive neural network segmentation model of deep supervision and transfer learning. The neural network segmentation model of deep supervision + transfer learning mainly has the following two features: 1. Pre-trained single-cell segmentation model. The currently public pre-trained models mainly include Cellpose (cell segmentation algorithm), Stardist (an open source computer vision library) and DeepCell (deep cell), etc. 2. Deep supervision mechanism, adding additional network branches between certain intermediate hidden layers in the pre-trained model. The network branches can introduce additional supervision information, which can help the model better learn the feature representation of the input data and improve the performance and generalization ability of the model.

[0120] Specifically, a neural network segmentation model of deep supervision + transfer learning is used to perform single cell image segmentation on the coherent Raman scattering image under the at least one Raman shift to obtain coherent Raman scattering images of multiple single cells. For example, a Raman shift of 2930 cm-1 The SRS images were used for single cell segmentation, and the pre-trained Cellpose + deep supervision segmentation method was used to segment urine exfoliated cells. Compared with using only the Cellpose model, this segmentation method combined with deep supervision can significantly improve the segmentation effect of urine exfoliated cells.

[0121] As shown in FIG8 , FIG8 is a schematic diagram of the structure of an image segmentation network model, in which the Cellpose model and the deep supervision model are combined. The deep supervision network branch combines part of the feature map of the output of the intermediate hidden layer of the Cellpose model in the figure with the output of the Cellpose model as the final output, so as to supervise the learning of the Cellpose output and the output of the intermediate hidden layer. Each intermediate hidden layer of the Cellpose model is composed of two residual blocks, and each residual block is composed of two convolutions with a convolution kernel size of 3×3. Among them, the Cellpose model includes an encoder and a decoder, wherein the encoder includes a first downsampling layer, a second downsampling layer, a third downsampling layer and a fourth downsampling layer; the decoder includes a first upsampling layer, a second upsampling layer, a third upsampling layer and a fourth upsampling layer. The feature map is encoded in the upsampling layer and the downsampling layer to obtain the corresponding feature map. The downsampling layer is connected to the corresponding upsampling layer through an attention mechanism, so that the feature map of the downsampling layer is sent to the corresponding upsampling layer. For example, the output feature map of the first downsampling layer can be sent to the fourth sampling layer through the attention mechanism, and so on. The deep supervision model includes the first convolutional layer, the second convolutional layer and the third convolutional layer; the first convolutional layer receives the output result of the second upsampling layer and performs convolution processing on the output result, and the output result after convolution processing is input to the third convolutional layer; the second convolutional layer receives the output result of the third upsampling layer and performs convolution processing to obtain the output result, and the output result is input to the third convolutional layer. In the third convolutional layer, the output result of the Cellpose model, the output result of the first convolutional layer and the output result of the second convolutional layer are combined to obtain the coherent Raman scattering images of multiple single cells. Specifically, the output result of each layer of the decoder calculates the loss with the manually annotated mask, and the loss is optimized through back propagation. The model parameters can be optimized, and the speed of model fitting can be further improved.

[0122] As shown in Figure 9, Figure 9 is a schematic diagram of a coherent Raman scattering image of a single cell provided in one embodiment. The top three figures in Figure 9 are coherent Raman scattering images of a single cell obtained using only the Cellpose model; the bottom three figures in Figure 9 are coherent Raman scattering images of a single cell obtained using the Cellpose + deep supervision network model.

[0123] In one example, in a urine exfoliated cell specimen, the shapes and sizes of different cells vary greatly. Neither the traditional watershed segmentation algorithm nor the segmentation method using a neural network based on transfer learning can achieve an ideal segmentation effect. To this end, in this embodiment, on the basis of a neural network segmentation model based on transfer learning, a deep supervision mechanism is introduced, and a network branch is added to some intermediate hidden layers of the deep neural network to supervise the backbone network, which is beneficial to improving the performance and generalization ability of the model. This method can not only effectively segment small and round cells, but also perform good segmentation on cells with more peculiar shapes (irregular shapes, small roundness).

[0124] In urine exfoliated cell specimens, the shapes and sizes of different cells vary greatly. Neither the traditional watershed segmentation algorithm nor the segmentation method based on the transfer learning neural network can achieve an ideal segmentation effect. To this end, this embodiment introduces a deep supervision mechanism based on the transfer learning-based neural network segmentation model, and adds a network branch to some intermediate hidden layers of the deep neural network to supervise the backbone network, which is conducive to improving the performance and generalization ability of the model.

[0125] Step 604: Using the feature extraction model, single-cell feature extraction is performed on the coherent Raman scattering image of each single cell in the coherent Raman scattering images of multiple single cells to obtain a single-cell feature map corresponding to the coherent Raman scattering image of each single cell.

[0126] Among them, the feature extraction model includes but is not limited to threshold segmentation algorithm, spot detection algorithm, deep neural network algorithm, etc. The single cell feature map includes cell morphological characteristics and cell metabolic characteristics; cell morphological characteristics include but are not limited to one or more of cell area, cell eccentricity, cell fitting circle diameter, cell major axis length, cell minor axis length, cell circumference, cell real area / fitted circle area, cell nuclear area, cytoplasm area, and cell nuclear-cytoplasm ratio; cell metabolic characteristics include but are not limited to one or more of cell lipid droplet number, cell lipid droplet area, lipid droplet to cell area ratio, distance from lipid droplet to cell center, average distance from lipid droplet to cell center of gravity, and cell lipid droplet grayscale. Morphological characteristics are mainly extracted by binary images obtained by single cell segmentation, while metabolic characteristics need to be extracted using coherent Raman scattering images.

[0127] Specifically, single cell feature extraction is performed on the coherent Raman scattering image of each single cell in the coherent Raman scattering images of the multiple single cells to obtain a single cell image feature map corresponding to the coherent Raman scattering image of the single cell. In order to extract the lipid droplet features, lipid droplet segmentation is performed on the coherent Raman scattering image of each single cell in the coherent Raman scattering images of the multiple single cells to obtain a coherent Raman scattering image of lipid droplets in each single cell region in the coherent Raman scattering image of the single cell. Algorithms such as threshold segmentation and spot detection methods can be used for lipid droplet segmentation, and deep learning models such as YOLO can also be used for lipid droplet segmentation.

[0128] As shown in FIG10 , FIG10 is a schematic diagram of cell processing in one embodiment. Specifically, based on the Raman shift of 2930 cm -1 The SRS image is used for single cell segmentation (the single cell segmentation model is the Cellpose+ deep supervision model) to obtain a binary image of the single cell. The morphological features of each cell are extracted based on the binary image of the single cell, such as cell area, eccentricity, cell perimeter, etc. Based on the binary image of the single cell, the Raman shift of 2850cm -1 and Raman shift 2930cm -1 The SRS images at the locations are used to extract the metabolic characteristics of single cells, such as the number of lipid droplets, lipid droplet area, lipid droplet area ratio, etc. The specific methods are as follows: 1. Based on the binary image of a single cell, the Raman shift of the corresponding single cell is 2850 cm -1 and Raman shift 2930cm -1 2. Based on the Raman shift of the single cell at 2850 cm -1 and Raman shift 2930 cm -1 The SRS images at the positions are used to calculate the cell images of the lipid channel and the cell images of the protein channel by using matrix operations, as shown in FIG11 . FIG11 is a schematic diagram of signal intensity under different Raman shifts in one embodiment, and the solid line in the figure represents the Raman shift of 2850 cm -1 Signal image, the dotted line in the figure represents the Raman shift of 2930cm -1 3. Based on the cell image of the lipid channel and the cell image of the protein channel, metabolic characteristics such as the cell lipid / protein content ratio, the protein signal intensity in the cytoplasm region, and the lipid signal intensity in the cytoplasm region can be extracted. In particular, in the embodiment of the present invention, based on the Raman shift of 2850cm -1 The YOLOv5 model is used to segment lipid droplets in the SRS image.

[0129] Exemplarily, the expression of matrix operation can be:

[0130]

[0131] Among them, P1 represents the lipid channel signal, P2 represents the protein channel signal, and S1 represents the 2850cm -1 The signal of S2 is 2930cm -1 signal, A1, A2, S1, and S2 represent different coefficients respectively.

[0132] Step 605: Combine the single-cell feature maps according to the target object to obtain a single-cell feature map group corresponding to the target object.

[0133] Among them, the target object is the person to be tested.

[0134] Specifically, the single cell image feature maps from the same target object are combined to obtain the image feature map of the target object. The image feature map of the target object is input into a pre-trained machine learning classifier to determine whether the target object has cancer.

[0135] Step 606: Use a preset classification model to perform feature analysis on the single-cell feature map group to obtain the abnormal probability of the target object.

[0136] Among them, the preset classification models include but are not limited to one or more of support vector machine classifiers, linear discriminant classifiers, K-neighbor classifiers, logistic regression classifiers, random forest decision tree classifiers, gradient boosting classifiers, virtual neural network classifiers, deep learning convolutional neural network classifiers, contrastive learning classifiers, and multiple instance learning classifiers.

[0137] Specifically, the principal component analysis algorithm is used to reduce the dimension of the image feature map of the target object to obtain 10 principal components, namely PC1 to PC10. The 10 principal components are input into the distributed gradient boosting library (eXtreme Gradient Boosting, XGboost) classifier based on multi-instance learning for classification to determine whether the target object has cancer.

[0138] Step 607: Determine the abnormal situation of the target object based on the abnormal probability and the preset abnormal probability threshold.

[0139] Specifically, this embodiment can use leave-one-out cross-validation to obtain a preset abnormality probability threshold. For example, 80 samples can be designed. Specifically, among all the data, each time the model is trained, a group of data will be left as a test set, and the remaining data will be used as a training set. Then each piece of data will be used as a test set, and the performance of the model will be evaluated according to the results of each test set to obtain the preset abnormality probability threshold.

[0140] Specifically, when the abnormal probability is greater than the preset abnormal probability threshold, the target object is determined to be abnormal. When the abnormal probability is less than the preset abnormal probability threshold, the target object is normal. The sensitivity of the case classification method is 88.46%, the specificity is 88.88%, the accuracy is 86.67%, and the area under the curve (AUC) is 0.955.

[0141] In an exemplary embodiment, coherent Raman scattering imaging is performed on the cell suspension using one or more Raman shifts to obtain a coherent Raman scattering image corresponding to each Raman shift, including:

[0142] One or more groups of excitation lights are respectively input into the coherent Raman scattering imaging system; for each group of excitation lights, the first excitation light and the second excitation light are collinearly processed by the optical path module to obtain a combined optical path, and the combined optical path is used to scan the cell suspension to obtain a coherent Raman scattering signal; the coherent Raman scattering signal is processed by the signal acquisition module to obtain a coherent Raman scattering image.

[0143] The Raman shift of each group of excitation light is different, and each group of excitation light includes a first excitation light and a second excitation light; the first excitation light can be a pump light, and the second excitation light can be a Stokes light, and the Raman shift is determined by the frequency difference between the first excitation light and the second excitation light. The coherent Raman scattering imaging system includes at least: an optical path module and a signal acquisition module; the optical path module includes at least a lens, a dichroic mirror, a two-dimensional galvanometer assembly, and an objective lens.

[0144] Specifically, as shown in FIG7 , first, a picosecond laser with a high repetition rate is used to emit a first excitation light and a second excitation light, wherein the first excitation light includes a pump light (the dotted line portion in FIG7 ), and the second excitation light includes a Stokes light (the thick solid line portion in FIG7 ); the wavelength of the emitted pump light can be tuned within the range of 700-960 nm (in this embodiment, the wavelength is 796.8 nm during imaging, corresponding to the Raman shift characteristic peak of 2850 cm -1), the wavelength of Stokes light is fixed at 1031nm. Then, the first excitation light and the second excitation light are respectively emitted into the dichroic mirror through the lens and the reflector, so that the pump light and the Stokes light overlap in space and time to obtain a combined light path. The combined light path is emitted into the objective lens through the scanning two-dimensional galvanometer assembly. Finally, the combined light beam is focused on the cell suspension (urine exfoliated cell specimen) through the objective lens, and the specific molecular resonance in the cell suspension is induced to produce stimulated Raman loss and gain signals, and then the coherent Raman scattering signal is obtained. The coherent Raman scattering signal is collected by another water immersion objective (condenser) of the same model. The signal acquisition module converts the coherent Raman scattering signal into an electrical signal, and stores it in the computer device in the signal acquisition module as a coherent Raman scattering image. For example, according to different collected signals, for SRS signals, an optoelectronic modulator or the like can be used to modulate the SRS signals, and a phase-locked amplifier or the like can be used to demodulate the signals to obtain SRS images; for CARS signals, a photomultiplier tube (PMT) can be used to collect and convert them into electrical signals, and polarization detection, pump and Stokes field time delay detection, backward detection CARS and other methods can be used to suppress non-resonant background signals in coherent anti-Stokes Raman scattering signals. It should be understood that the signal acquisition module can be adjusted accordingly according to different signals, and the signal acquisition module is not specifically limited here.

[0145] In an exemplary embodiment, a preset classification model is used to perform feature analysis on the single cell feature map group to obtain the abnormal probability of the target object, including:

[0146] The single-cell feature map group is subjected to dimensionality reduction processing to obtain the single-cell feature map group after dimensionality reduction processing, and the single-cell feature map group after dimensionality reduction processing is input into a preset classification model to obtain the abnormal probability of the target object.

[0147] Alternatively, clustering is performed on the single-cell feature map group to obtain a clustered single-cell feature map group, and the clustered single-cell feature map group is input into a preset classification model to obtain an outlier value corresponding to the target.

[0148] Alternatively, the single-cell feature map group is subjected to dimensionality reduction and clustering processing to obtain a processed single-cell feature map group, and the processed single-cell feature map group is input into a preset classification model to obtain an outlier value corresponding to the target.

[0149] Among them, the dimensionality reduction algorithms include but are not limited to principal component analysis algorithm, linear discriminant analysis algorithm, multidimensional scaling algorithm, isometric mapping algorithm, local linear embedding algorithm, Laplace eigenmapping algorithm, local preservation projection algorithm, t-distributed stochastic neighbor embedding (t-SNE) algorithm. Clustering algorithms include but are not limited to k-means clustering algorithm, hierarchical clustering algorithm, self-organizing graph clustering algorithm, fuzzy clustering algorithm, Gaussian mixture model clustering algorithm, and expectation maximization algorithm.

[0150] Specifically, before inputting the image feature map of the target object into a pre-trained machine learning classifier, the image feature map of the target object may be subjected to dimensionality reduction to remove noise or redundant information in the image feature map of the target object.

[0151] Specifically, before inputting the image feature map of the target object into a pre-trained machine learning classifier, a cluster analysis can also be performed on the image feature map of the target object to obtain the number of different types of cells in the target object and identify cell populations that are important for determining whether the target object has cancer.

[0152] 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.

[0153] Based on the same inventive concept, the embodiment of the present application also provides a cell processing device for implementing the cell processing 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 the one or more cell processing device embodiments provided below can refer to the limitations of the cell processing method above, and will not be repeated here.

[0154] In an exemplary embodiment, as shown in FIG. 12 , a schematic diagram of the structure of a cell processing device is provided. The cell processing device 120 includes: a first processing module 121 and a second processing module 122, wherein:

[0155] The first processing module 121 is used to inject the sample liquid to be detected from the input channel into the first mixing and resuspension chamber for pretreatment to obtain the pretreated sample liquid to be detected; under the action of the first centrifugal speed, the first reagent is injected into the first mixing and resuspension chamber through the first channel; under the action of the first mixing centrifugal speed, the first reagent and the pretreated sample liquid to be detected are mixed and centrifuged to obtain a first mixed liquid; when the centrifugal speed of the centrifugal system matches the preset speed threshold, the fourth channel is activated, the waste liquid of the first mixed liquid is sucked into the first waste liquid chamber through the fourth channel, and the first cell precipitate of the first mixed liquid is collected in the second mixing and resuspension chamber.

[0156] The second processing module 122 is used to inject the second reagent into the second mixing and resuspension chamber under the action of the second centrifugal speed, and mix and centrifuge the first cell precipitate and the second reagent under the action of the second mixing centrifugal speed to obtain a second mixed liquid; when the centrifugal speed of the centrifugal system reaches a fourth centrifugal speed, the second mixed liquid is injected into the collecting chamber through the fifth channel, the second cell precipitate of the second mixed liquid is collected through the collecting chamber, and the second waste liquid of the second mixed liquid is collected through the waste liquid chamber.

[0157] In one example, the second processing module 122 is further used to rinse the second cell precipitate with a cell buffer to obtain a cell suspension of the sample solution to be detected;

[0158] Using one or more Raman shifts to perform coherent Raman scattering imaging on the cell suspension, respectively, to obtain a coherent Raman scattering image corresponding to each Raman shift;

[0159] Using an image segmentation network model, performing single-cell image segmentation processing on the coherent Raman scattering image to obtain a plurality of single-cell coherent Raman scattering images;

[0160] Using a feature extraction model, single-cell feature extraction is performed on the coherent Raman scattering image of each single cell in the coherent Raman scattering images of the multiple single cells to obtain a single-cell feature map corresponding to the coherent Raman scattering image of each single cell;

[0161] Combining the single cell feature maps according to the target object to obtain a single cell feature map group corresponding to the target object;

[0162] A preset classification model is used to perform feature analysis on the single-cell feature map group to obtain the abnormal probability of the target object; based on the abnormal probability and a preset abnormal probability threshold, the abnormal condition of the target object is determined.

[0163] In one example, the single-cell feature map includes cell morphological features and / or cell metabolic features; the cell morphological features include at least one or more of cell area, cell eccentricity, cell fitted circle diameter, cell major axis length, cell minor axis length, cell circumference, cell true area / fitted circle area, cell nuclear area, cytoplasm area, and cell nuclear-cytoplasm ratio; the cell metabolic features include at least one or more of cell lipid droplet number, cell lipid droplet area, lipid droplet ratio of cell area, distance from lipid droplet to cell center, average distance from lipid droplet to cell center of gravity, and cell lipid droplet grayscale.

[0164] In one example, the second processing module 122 is further used to input one or more groups of excitation light into the coherent Raman scattering imaging system, each group of excitation light has a different Raman shift, and each group of excitation light includes a first excitation light and a second excitation light; the coherent Raman scattering imaging system includes at least: an optical path module and a signal acquisition module;

[0165] For each group of excitation light, the first excitation light and the second excitation light are processed in a colinear manner through the optical path module to obtain a combined optical path, and the combined optical path is used to scan the cell suspension to obtain a coherent Raman scattering signal; the coherent Raman scattering signal is processed through the signal acquisition module to obtain a coherent Raman scattering image.

[0166] Each module in the above-mentioned cell processing 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.

[0167] In an exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be shown in FIG13. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. 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, a computer program and a database. 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 through a network connection. The computer program is executed by the processor to implement a cell processing method.

[0168] Those skilled in the art will understand that the structure shown in FIG. 13 is merely a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0169] In one embodiment, a processor-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.

[0170] 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.

[0171] 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 this 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.

[0172] 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.

[0173] 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 construed as limiting the scope of the present application. It should be noted that, for a person of ordinary skill in the art, several modifications 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 centrifugal system, characterized in that: The centrifugal system comprises: a sample chamber, a mixing and resuspension chamber, a collection chamber and a waste liquid chamber; The sample chamber is used to place preset reagents; The mixing and resuspension chamber is connected to the sample chamber with a first channel for allowing the liquid in the sample chamber to enter the mixing and resuspension chamber at a preset centrifugal speed, and is used to mix the preset reagent entering from the first channel and the sample liquid to be detected input from the input channel at the mixing centrifugal speed to obtain a mixed liquid, and separate the mixed liquid to obtain a cell precipitate and a waste liquid; The collection chamber is used to collect the cell precipitation; The waste liquid chamber is used to collect the waste liquid.

2. The centrifugal system according to claim 1, characterized in that The preset reagents include a first reagent and a second reagent; the sample chamber includes a first sample chamber and a second sample chamber, the mixing centrifugal speed includes a first mixing centrifugal speed and a second mixing centrifugal speed; the sample liquid to be detected enters the mixing and resuspension chamber before the preset reagents; the mixing and resuspension chamber includes a first mixing and resuspension chamber and a second mixing and resuspension chamber; wherein, The first mixing and resuspension chamber is used to mix and centrifuge the sample liquid to be detected at the first mixing centrifugal speed, or to mix and centrifuge the sample liquid to be detected and the first reagent to obtain a first mixed liquid, so that the waste liquid chamber collects the waste liquid of the first mixed liquid through the second channel, and collects the first cell precipitate of the first mixed liquid into the second mixing and resuspension chamber through the second channel; The second mixing and resuspension chamber is used to mix the first cell pellet with the second reagent in the second sample chamber at the second mixing centrifugal speed to obtain a second mixed liquid, so that the collecting chamber collects the second cell pellet in the second mixed liquid, and the waste liquid of the second mixed liquid is collected through the waste liquid chamber.

3. The centrifugal system according to claim 2, characterized in that: The preset centrifugal speed includes a first centrifugal speed, and the system further includes: The first sample chamber is used to store the first reagent; The first channel is used to introduce the first reagent into the first mixing and resuspension chamber under the action of the first centrifugal speed after the sample solution to be detected is input through the input channel.

4. The centrifugal system according to claim 2, characterized in that: The preset centrifugal speed includes a second centrifugal speed, and the second sample chamber includes a plurality of sample sub-chambers and third channels corresponding to each sample sub-chamber; The plurality of sample subchambers are used to store the second reagent; The third channel is used to introduce the second reagent into the second mixing and resuspension chamber under the action of the second centrifugal speed.

5. The centrifugal system according to claim 1, characterized in that: The collection chamber is a multi-layer filter membrane structure; the pore size of the multi-layer filter membrane structure decreases from top to bottom, and the pore size of the bottom filter membrane of the multi-layer filter membrane structure is smaller than the diameter of the required cells; The multi-layer filter membrane structure comprises a first filter membrane, a second filter membrane and a third filter membrane; the first filter membrane is used to retain first impurities of a first diameter; the second filter membrane is used to retain second impurities of a second diameter; the third filter membrane is used to collect second cell precipitation and filter third impurities of a third diameter; The first diameter is greater than the second diameter, and the second diameter is greater than the third diameter.

6. The centrifugal system according to claim 5, characterized in that The waste liquid chamber comprises a first waste liquid chamber and a second waste liquid chamber; The first waste liquid chamber is connected to the first mixing and resuspension chamber with a fourth channel for allowing the waste liquid in the first mixing and resuspension chamber to enter the first waste liquid chamber at a third centrifugal speed; The second waste liquid chamber is used to collect the third impurities with the third diameter.

7. The centrifugal system according to claim 2, characterized in that: A fifth channel is connected between the collection chamber and the second mixing and resuspension chamber, and allows the second mixed liquid to enter the collection chamber under the action of a fourth centrifugal speed.

8. A method for treating cells, characterized in that: Applicable to the centrifugal system according to any one of claims 1 to 7; the method comprising: Injecting the sample liquid to be detected from the input channel into the first mixing and resuspension chamber for pretreatment to obtain the pretreated sample liquid to be detected; injecting the first reagent into the first mixing and resuspension chamber through the first channel under the action of the first centrifugal speed; mixing and centrifuging the first reagent and the pretreated sample liquid to be detected under the action of the first mixing centrifugal speed to obtain a first mixed liquid; when the centrifugal speed of the centrifugal system matches the preset speed threshold, activating the fourth channel, sucking the waste liquid of the first mixed liquid into the first waste liquid chamber through the fourth channel, and collecting the first cell precipitate of the first mixed liquid into the second mixing and resuspension chamber; Under the action of the second centrifugal speed, the second reagent is injected into the second mixing and resuspension chamber, and the first cell pellet and the second reagent are mixed and centrifuged under the action of the second mixing centrifugal speed to obtain a second mixed liquid; when the centrifugal speed of the centrifugal system reaches a fourth centrifugal speed, the second mixed liquid is injected into the collection chamber through the fifth channel, the second cell pellet of the second mixed liquid is collected through the collection chamber, and the second waste liquid of the second mixed liquid is collected through the waste liquid chamber.

9. The method according to claim 8, characterized in that The method further comprises: The second cell precipitation is washed with a cell buffer to obtain a cell suspension of the sample solution to be detected; Using one or more Raman shifts to perform coherent Raman scattering imaging on the cell suspension, respectively, to obtain a coherent Raman scattering image corresponding to each Raman shift; Using an image segmentation network model, performing single-cell image segmentation processing on the coherent Raman scattering image to obtain a plurality of single-cell coherent Raman scattering images; Using a feature extraction model, single-cell feature extraction is performed on the coherent Raman scattering image of each single cell in the coherent Raman scattering images of the multiple single cells to obtain a single-cell feature map corresponding to the coherent Raman scattering image of each single cell; Combining the single cell feature maps according to the target object to obtain a single cell feature map group corresponding to the target object; A preset classification model is used to perform feature analysis on the single-cell feature map group to obtain the abnormal probability of the target object; based on the abnormal probability and a preset abnormal probability threshold, the abnormal condition of the target object is determined.

10. The method according to claim 9, characterized in that The single cell feature map includes cell morphological characteristics and / or cell metabolic characteristics; the cell morphological characteristics at least include one or more of cell area, cell eccentricity, cell fitted circle diameter, cell major axis length, cell minor axis length, cell circumference, cell true area / fitted circle area, cell nuclear area, cytoplasm area, and cell nuclear-cytoplasm ratio; the cell metabolic characteristics at least include one or more of cell lipid droplet number, cell lipid droplet area, lipid droplet ratio of cell area, distance from lipid droplet to cell center, average distance from lipid droplet to cell center of gravity, and cell lipid droplet grayscale.

11. The method according to claim 9, characterized in that The method of using one or more Raman shifts to perform coherent Raman scattering imaging on the cell suspension to obtain a coherent Raman scattering image corresponding to each Raman shift comprises: Input one or more groups of excitation light into a coherent Raman scattering imaging system respectively, wherein the Raman shift of each group of excitation light is different, and each group of excitation light includes a first excitation light and a second excitation light; the coherent Raman scattering imaging system includes at least: an optical path module and a signal acquisition module; For each group of excitation light, the first excitation light and the second excitation light are processed in a colinear manner through the optical path module to obtain a combined optical path, and the combined optical path is used to scan the cell suspension to obtain a coherent Raman scattering signal; the coherent Raman scattering signal is processed through the signal acquisition module to obtain a coherent Raman scattering image.

12. A cell processing device, characterized in that: Applicable to the centrifugal system according to any one of claims 1 to 7, the device comprising: A first processing module is used to inject the sample liquid to be detected from the input channel into the first mixing and resuspension chamber for pretreatment to obtain the pretreated sample liquid to be detected; under the action of a first centrifugal speed, inject a first reagent into the first mixing and resuspension chamber through the first channel; under the action of the first mixing centrifugal speed, the first reagent and the pretreated sample liquid to be detected are mixed and centrifuged to obtain a first mixed liquid; when the centrifugal speed of the centrifugal system matches a preset speed threshold, the fourth channel is activated, and the waste liquid of the first mixed liquid is sucked into the first waste liquid chamber through the fourth channel, and the first cell precipitate of the first mixed liquid is collected into the second mixing and resuspension chamber; The second processing module is used to inject the second reagent into the second mixing and resuspension chamber under the action of the second centrifugal speed, and mix and centrifuge the first cell precipitate and the second reagent under the action of the second mixing centrifugal speed to obtain a second mixed liquid; when the centrifugal speed of the centrifugal system reaches a fourth centrifugal speed, inject the second mixed liquid into the collecting chamber through the fifth channel, collect the second cell precipitate of the second mixed liquid through the collecting chamber, and collect the second waste liquid of the second mixed liquid through the waste liquid chamber.

Images