Centrifugal control method and system based on artificial intelligence analysis of exosomes
Through the centrifugal control method based on artificial intelligence, the centrifugal parameters are adjusted in combination with exosome categories and real-time weight, and the status of centrifugal equipment is monitored, the problems of inaccurate exosome centrifugal parameters and insufficient monitoring of equipment status in the existing technology are solved, achieving high efficiency , accurate exosome separation and equipment status monitoring.
Patent Information
- Application Number
- CN202410671724.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-27
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-05-27
AI Technical Summary
The prior art fails to effectively combine exosomes' own characteristics and artificial intelligence technology in exosome centrifugal detection, resulting in inaccurate centrifugal parameters and wasted exosome samples.
Using artificial intelligence-based centrifugal control methods, we can obtain exosome categories and real-time weight, adjust centrifugal parameters, and monitor the status of the centrifugal equipment in real time to ensure that the centrifugal equipment is in a normal state and avoid wasting samples.
It realizes the acquisition of precise centrifugal parameters in combination with exosomes' own characteristics and artificial intelligence technology, ensures effective separation of exosomes, avoids sample waste, and improves the efficiency and accuracy of the centrifugal process.
Smart Images

Figure CN118588281B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of centrifugal detection technology, and more specifically, to a centrifugal control method and system based on artificial intelligence analysis of exosomes. Background Art
[0002] Exosomes are nano-scale vesicles with a lipid bilayer membrane structure that are actively released by most cells. Initially, exosomes were considered to be waste products discharged by cells and were not taken seriously. Later, more and more studies believed that exosomes are enriched with a variety of bioactive molecules such as nucleic acids, proteins and lipids, and can be transferred from donor cells to recipient cells to transmit intracellular information. They are important participants in intercellular communication. One of the most important links in auxiliary diagnosis is the centrifugal detection of exosomes. Artificial intelligence technology refers to the use of theories, methods and technologies in related fields such as computer science and machine learning to simulate the ability of human intelligence and achieve various complex tasks. In the existing technology, in the centrifugal detection of exosomes, there are very few technologies that can be combined with artificial intelligence technology to achieve accurate separation of exosomes.
[0003] For example, the Chinese patent with publication number CN115710575A provides an exosome extraction method and application. The patent promotes the release of exosomes in platelets by gradient centrifugation of platelet-rich plasma, thereby achieving the purpose of fully extracting exosomes from platelets. Although the above patent centrifuges exosomes, it still has the following problems:
[0004] Although the above-mentioned patent performs centrifugation operation, it does not specifically combine the characteristics of exosomes themselves and artificial intelligence technology to obtain more accurate centrifugation parameters, and during the centrifugation process, it does not monitor the status of the centrifugation equipment, resulting in the subsequent waste of exosome samples.
[0005] In view of this, the present invention proposes a centrifugal control method and system based on artificial intelligence analysis of exosomes to solve the above problems. Summary of the invention
[0006] In order to overcome the above-mentioned defects of the prior art, the present invention provides a centrifugal control method and system based on artificial intelligence analysis of exosomes.
[0007] To achieve the above object, the present invention provides the following technical solutions:
[0008] Centrifugal control method based on artificial intelligence analysis of exosomes, including:
[0009] S10: obtaining the exosome category of the hth exosome to be separated, and obtaining a corresponding initial centrifugation parameter set based on the exosome category;
[0010] S20: obtaining the real-time weight of the hth exosome to be separated, inputting the initial centrifugation parameter set and the real-time weight into a pre-built parameter adjustment model, and obtaining a target centrifugation parameter set;
[0011] S30: separating the hth exosome to be separated by adopting the target centrifugal parameter set based on the centrifugal device, obtaining electromagnetic field energy, and judging whether the centrifugal device is in a normal state based on the electromagnetic field energy;
[0012] S40: When the centrifugal device is in a normal state, h+1=h is set, and the above S10-S30 are repeated until h=f, and the centrifugation is terminated, and f>h>0;
[0013] S50: When the centrifugal device is not in a normal state, a warning message is sent.
[0014] Furthermore, the exosome category includes blood exosomes and urine exosomes, and the method for obtaining the exosome category of the hth exosome to be separated includes:
[0015] A real-time surface image of the h-th exosome to be separated is obtained, and the real-time surface image is input into a pre-built exosome classification model to obtain the exosome category of the h-th exosome to be separated.
[0016] Furthermore, the method for constructing the secretory body classification model includes:
[0017] Obtain r groups of data, where r is a positive integer greater than 1, and the data include historical surface images and historical oosome categories. The historical surface images and historical oosome categories are used as sample sets, and the sample sets are divided into training sets and test sets. A classifier is constructed, and the historical surface images in the training set are used as input data, and the historical oosome categories in the training set are used as output data. The classifier is trained to obtain an initial classifier, and the initial classifier is tested using the test set, and a classifier that meets the preset accuracy is output as a oosome classification model.
[0018] Furthermore, the historical surface image includes an actual historical surface image and a synthetic historical surface image, and the method for generating the historical surface image includes:
[0019] Using the actual historical surface image to train a discriminator in the target generative adversarial network to obtain a trained discriminator;
[0020] The trained discriminator guides the generator in the target generative adversarial network to generate a synthetic historical surface image similar to the actual historical surface image.
[0021] Furthermore, the method for constructing the parameter adjustment model includes:
[0022] A sample data set is obtained, wherein the sample data set includes a historical weight, a historical initial centrifugal parameter set, and a historical target centrifugal parameter set. The sample data set is divided into a sample training set and a sample test set, and a regression network is constructed. The historical weight and the historical initial centrifugal parameter set in the sample training set are used as input data of the regression network, and the historical target centrifugal parameter set in the sample training set is used as output data of the regression network. The regression network is trained to obtain an initial regression network for predicting the target centrifugal parameter set. The initial regression network is tested using the sample test set, and a regression network with a smaller error value than a preset value is output as a parameter adjustment model.
[0023] Furthermore, the method for obtaining the historical target centrifugal parameter set includes:
[0024] The historical initial centrifugation parameter set is adjusted m times to obtain m historical intermediate centrifugation parameter sets, exosomes are centrifuged based on the m historical intermediate centrifugation parameter sets, a sedimentation data set of exosomes after centrifugation is obtained, the centrifugation coefficient is calculated based on the sedimentation data set, and the historical target centrifugation parameter set is determined from the m historical intermediate centrifugation parameter sets according to the centrifugation coefficient, and the sedimentation data set includes sedimentation rate, clear liquid transparency and marker concentration.
[0025] Further, the method for calculating the centrifugal coefficient based on the sedimentation data set includes:
[0026]
[0027] Where Cft is the centrifugal coefficient, SDE re is the real-time precipitation rate, SDE sd is the standard precipitation rate, CLT re For real-time clarity of liquid, CLT sd For standard clear liquid transparency, COM re is the real-time marker concentration, COM sd is the standard marker concentration, u 1 、u 2 and u 3 are all weight factors.
[0028] Furthermore, the method for determining whether the centrifugal device is in a normal state based on the electromagnetic field energy includes:
[0029] The electromagnetic field energy during the centrifugal time is fitted to generate an electromagnetic field energy curve. The centrifugal time is segmented based on the slope of the electromagnetic field energy curve to generate segmentation results. Based on the segmentation results, it is judged whether the centrifugal equipment is in a normal state. The segmentation results include a startup phase, a working phase, and a termination phase.
[0030] Further, the method for judging whether the centrifugal device is in a normal state based on the segmentation result includes:
[0031] The electromagnetic field energy corresponding to the working stage in the segmented result is obtained, and the mean value of the electromagnetic field energy is calculated to determine whether the mean value of the electromagnetic field energy is within the standard energy range. If so, the centrifugal device is determined to be in a normal state; if not, the centrifugal device is determined to be not in a normal state.
[0032] A centrifugal control system based on artificial intelligence analysis of exosomes, which is used to implement the above-mentioned centrifugal control method based on artificial intelligence analysis of exosomes, comprises:
[0033] Data acquisition module: used to obtain the exosome category of the hth exosome to be separated, and obtain the corresponding initial centrifugation parameter set based on the exosome category;
[0034] Data correction module: used to obtain the real-time weight of the hth exosome to be separated, input the initial centrifugation parameter set and the real-time weight into the pre-built parameter adjustment model, and obtain the target centrifugation parameter set;
[0035] A state judgment module is used to separate the hth exosome to be separated based on the target centrifugal parameter set of the centrifugal device, and obtain electromagnetic field energy, and judge whether the centrifugal device is in a normal state based on the electromagnetic field energy;
[0036] Cycle module: when the centrifugal device is in a normal state, h+1=h is set, and the above data acquisition module, data correction module and state judgment module are repeatedly executed until h=f, then the centrifugation is terminated, f>h>0;
[0037] Warning module: used to send warning information when the centrifugal device is not in normal state.
[0038] An electronic device comprises a power supply, an interface, a keyboard, a memory, a central processing unit, and a computer program stored in the memory and executable on the central processing unit. When the processor executes the computer program, the centrifugal control method based on artificial intelligence analysis of exosomes is implemented. The interface comprises a network interface and a data interface. The network interface comprises a wired or wireless interface, and the data interface comprises an input or output interface.
[0039] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed, implements the above-mentioned centrifugation control method based on artificial intelligence analysis of exosomes.
[0040] Compared with the prior art, the present invention has the following beneficial effects:
[0041] The present invention first obtains the vesicle category of the hth exosome to be separated, obtains the corresponding initial centrifugation parameter set based on the vesicle category, and then obtains the real-time weight of the hth exosome to be separated, inputs the initial centrifugation parameter set and the real-time weight into a pre-constructed parameter adjustment model, obtains a target centrifugation parameter set, separates the hth exosome to be separated by adopting the target centrifugation parameter set based on a centrifugal device, and obtains electromagnetic field energy, determines whether the centrifugal device is in a normal state based on the electromagnetic field energy, and determines whether to perform the next centrifugation work or send a warning message according to the state of the centrifugal device. In this way, the present invention combines the characteristics of the exosomes themselves and artificial intelligence technology to obtain more accurate centrifugation parameters, and monitors the state of the centrifugal device throughout the process to avoid the subsequent waste of exosome samples. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1 This is a flow chart of the centrifugal control method for analyzing exosomes based on artificial intelligence in the present invention;
[0043] Figure 2 A schematic diagram of a computer-readable storage medium in the present invention. DETAILED DESCRIPTION
[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0045] Example 1
[0046] See also Figure 1 As shown, this embodiment discloses a centrifugal control method for analyzing exosomes based on artificial intelligence, comprising:
[0047] S10: obtaining the exosome category of the hth exosome to be separated, and obtaining a corresponding initial centrifugation parameter set based on the exosome category;
[0048] It should be noted that the exosome category includes blood exosomes and urine exosomes. Exosomes are small vesicles secreted by cells. They contain rich biological molecules, such as proteins, nucleic acids and lipids. These molecules can provide information about tumor cells. Usually, exosomes are widely present in easily accessible biological fluids such as human blood and urine. The corresponding exosomes are obtained by filtering blood and urine. There are many methods for determining the exosome category. This embodiment discloses a method for determining by image.
[0049] The method for obtaining the exosome type of the hth exosome to be separated includes:
[0050] Obtaining a real-time surface image of the hth exosome to be separated, inputting the real-time surface image into a pre-built exosome classification model, and obtaining the exosome category of the hth exosome to be separated;
[0051] The methods for constructing the secretory body classification model include:
[0052] Obtain r groups of data, where r is a positive integer greater than 1, and the data includes historical surface images and historical oosome categories. The historical surface images and historical oosome categories are used as sample sets, and the sample sets are divided into training sets and test sets. A classifier is constructed, and the historical surface images in the training set are used as input data, and the historical oosome categories in the training set are used as output data. The classifier is trained to obtain an initial classifier, and the initial classifier is tested using the test set, and a classifier that meets a preset accuracy is output as a oosome classification model, and the classifier is preferably a naive Bayes model or a support vector machine model;
[0053] It should be understood that the size of exosomes ranges widely, usually from tens to hundreds of nanometers. Exosomes in urine are usually small, with diameters generally between 30 and 200 nanometers, and are round or oval in shape. Exosomes in blood are more diverse in size, including small exosomes and large exosomes. Therefore, the size of the vesicles in the image can be used to distinguish the types of exosomes. However, in reality, it is difficult to collect a large number of historical surface images of exosomes. Therefore, it is necessary to generate a synthetic historical surface image that is similar to the actual historical surface image generated by the target adversarial network.
[0054] Exemplarily, the historical surface image includes an actual historical surface image and a synthetic historical surface image, and the method for generating the historical surface image includes:
[0055] Using the actual historical surface image to train a discriminator in the target generative adversarial network to obtain a trained discriminator;
[0056] Instructing the generator in the target generative adversarial network to generate a synthetic historical surface image similar to the actual historical surface image through the trained discriminator;
[0057] It should be noted that Generative Adversarial Networks (GANs) is a generative model proposed by deep learning. Its goal is to generate realistic data samples through adversarial learning between two network models. GANs consists of two main components: the generator and the discriminator. These two components compete with each other, and through adversarial training, the quality of samples generated by the generator is gradually improved.
[0058] In this embodiment, the discriminator can be regarded as a trained binary classification classifier, and the generator includes several upsampling layers, several feature extraction layers and several feature gating layers. The generator first obtains the features of the actual historical surface image, generates an initial historical surface image through the features, and then gives the initial historical surface image to the discriminator. The discriminator classifies the initial historical surface image and generates a classification result. The classification result includes true and false. The discriminator then feeds the classification result back to the generator, and iterates repeatedly. Through the continuous confrontation between the generator and the discriminator, a synthetic historical surface image with high similarity is generated. The synthetic historical surface image and the actual historical surface image are used to generate a historical surface image set. The historical surface image set is used to train the secretory body classification model. By amplifying the training samples, overfitting in the secretory body classification model training is avoided.
[0059] In this embodiment, the initial centrifugation parameter set includes but is not limited to centrifugation speed, centrifugal force and centrifugation time. It can be understood that the initial centrifugation parameter set refers to a centrifugation parameter set pre-set according to the secretory body type. Different secretory body types lead to different initial centrifugation parameter sets;
[0060] S20: obtaining the real-time weight of the hth exosome to be separated, inputting the initial centrifugation parameter set and the real-time weight into a pre-built parameter adjustment model, and obtaining a target centrifugation parameter set;
[0061] Furthermore, the real-time weight of the exosomes to be separated can be directly obtained through the weight sensor provided by the separation device;
[0062] The construction method of the parameter adjustment model includes:
[0063] Acquire a sample data set, wherein the sample data set includes a historical weight, a historical initial centrifugal parameter set, and a historical target centrifugal parameter set, divide the sample data set into a sample training set and a sample test set, construct a regression network, use the historical weight and the historical initial centrifugal parameter set in the sample training set as input data of the regression network, use the historical target centrifugal parameter set in the sample training set as output data of the regression network, train the regression network, obtain an initial regression network for predicting the target centrifugal parameter set, test the initial regression network using the sample test set, output a regression network with a value less than a preset error as a parameter adjustment model, and the regression network is preferably a neural network model;
[0064] It should be noted that different types of exosomes in the above-mentioned medium lead to different initial centrifugation parameter sets. The weight of the same exosomes will also affect the centrifugation parameters. Therefore, the real-time weight is required to correct the initial centrifugation parameter set. The above-mentioned historical target centrifugation parameter set is obtained through continuous optimization of experiments. Those skilled in the art can determine the effect of exosome centrifugation by detecting the centrifuged exosomes and obtaining the data of the exosomes after centrifugation. For example:
[0065] Methods for obtaining historical target centrifugal parameter sets include:
[0066] Adjust the historical initial centrifugation parameter set m times to obtain m historical intermediate centrifugation parameter sets, centrifuge the exosomes based on the m historical intermediate centrifugation parameter sets respectively, obtain the sedimentation data set of the exosomes after centrifugation, calculate the centrifugation coefficient based on the sedimentation data set, and determine the historical target centrifugation parameter set from the m historical intermediate centrifugation parameter sets according to the centrifugation coefficient, the sedimentation data set includes sedimentation rate, clear liquid transparency and marker concentration;
[0067] Methods for calculating the centrifugal coefficient based on a sedimentation data set include:
[0068]
[0069] Where Cft is the centrifugal coefficient, SDE re is the real-time precipitation rate, SDE sd is the standard precipitation rate, CLT re For real-time clarity of liquid, CLT sd For standard clear liquid transparency, COM re is the real-time marker concentration, COM sd is the standard marker concentration, u 1 、u 2 and u 3 All are weight factors;
[0070] It should be noted that the higher the above precipitation rate, the better the exosome centrifugation effect, that is, the exosomes can be precipitated to the bottom of the centrifuge tube faster during the centrifugation process, and the supernatant is clear and transparent, with basically no suspended matter or precipitation, indicating that the exosomes are effectively separated. By measuring the concentration of exosome markers (such as CD63, CD9, etc.) in the supernatant after centrifugation, the centrifugation effect of the exosomes can be evaluated. The lower the concentration, the better the exosome centrifugation effect. Therefore, in this embodiment, the centrifugation coefficient is proportional to the centrifugation effect, that is, the larger the centrifugation coefficient, the better the exosome centrifugation effect. It can be understood that the standard precipitation rate, standard clear liquid transparency and standard marker concentration in the above are all preset according to expert experience;
[0071] In the above experimental process, m historical intermediate centrifugal parameter sets are first obtained. It can be understood that some parameters in the m historical intermediate centrifugal parameter sets are the same as the historical initial centrifugal parameter set. The method for determining the historical target centrifugal parameter set from the m historical intermediate centrifugal parameter sets according to the centrifugal coefficient can be to use the historical intermediate centrifugal parameter corresponding to the maximum value of the m centrifugal coefficients as the historical target centrifugal parameter set;
[0072] S30: separating the hth exosome to be separated by adopting the target centrifugal parameter set based on the centrifugal device, obtaining electromagnetic field energy, and judging whether the centrifugal device is in a normal state based on the electromagnetic field energy;
[0073] In this embodiment, the electromagnetic field energy refers to the electromagnetic field energy inside the motor in the separation device. It can be understood that the separation device is usually driven by the internal motor to perform the separation action, and the electromagnetic field energy inside the motor can be directly obtained by the Hall sensor;
[0074] The method for judging whether the centrifugal device is in a normal state based on the electromagnetic field energy includes:
[0075] Fitting the electromagnetic field energy during the centrifugal time to generate an electromagnetic field energy curve, segmenting the centrifugal time based on the slope of the electromagnetic field energy curve to generate segmentation results, and judging whether the centrifugal device is in a normal state based on the segmentation results, wherein the segmentation results include a startup phase, a working phase, and a termination phase;
[0076] The method for judging whether the centrifugal device is in a normal state based on the segmentation results includes:
[0077] Obtaining the electromagnetic field energy corresponding to the working stage in the segmented result, and calculating the mean value of the electromagnetic field energy, judging whether the mean value of the electromagnetic field energy is within the standard energy interval, if so, judging that the centrifugal device is in a normal state, and if not, judging that the centrifugal device is not in a normal state;
[0078] It should be noted that the reason why the electromagnetic field energy is obtained in this embodiment but the current or voltage is not obtained is that the centrifugal time can be segmented more accurately through the electromagnetic field energy, and when the motor is started, a large starting current is usually applied to overcome inertia and friction, which will cause the electromagnetic field energy of the motor to increase. When the motor needs to be stopped, the control system will reduce the electromagnetic field energy of the motor, so that the motor gradually slows down and eventually stops. The slope of the electromagnetic field energy curve represents the rate of change of the electromagnetic field energy. Therefore, whether in the startup stage or in the termination stage, the slope of the electromagnetic field energy curve is at a large value, and only in the working stage, the electromagnetic field energy of the motor will be maintained at a relatively constant level. Therefore, the electromagnetic field energy curve is approximately a straight line at this time. In this embodiment, the startup stage and the termination stage are first excluded, and the electromagnetic field energy in the working stage is used to determine whether there is a problem with the centrifugal device, so that the centrifugal device can be detected more accurately;
[0079] S40: When the centrifugal device is in a normal state, h+1=h is set, and the above S10-S30 are repeated until h=f, and the centrifugation is terminated, and f>h>0;
[0080] S50: When the centrifugal device is not in a normal state, a warning message is sent;
[0081] In this embodiment, the exosome category of the hth exosome to be separated is first obtained, and the corresponding initial centrifugation parameter set is obtained based on the exosome category, and then the real-time weight of the hth exosome to be separated is obtained, and the initial centrifugation parameter set and the real-time weight are input into the pre-constructed parameter adjustment model to obtain the target centrifugation parameter set, and the hth exosome to be separated is separated by adopting the target centrifugation parameter set based on the centrifugal device, and the electromagnetic field energy is obtained, and it is determined based on the electromagnetic field energy whether the centrifugal device is in a normal state, and according to the state of the centrifugal device, it is determined whether to perform the next centrifugation work or send a warning message. In this way, this embodiment combines the characteristics of the exosomes themselves and the artificial intelligence technology to obtain more accurate centrifugation parameters, and monitors the state of the centrifugal device throughout the process to avoid the subsequent waste of exosome samples.
[0082] Example 2
[0083] This embodiment provides a centrifugal control system based on artificial intelligence analysis of exosomes on the basis of embodiment 1, including:
[0084] Data acquisition module: used to obtain the exosome category of the hth exosome to be separated, and obtain the corresponding initial centrifugation parameter set based on the exosome category;
[0085] It should be noted that the exosome category includes blood exosomes and urine exosomes. Exosomes are small vesicles secreted by cells. They contain rich biological molecules, such as proteins, nucleic acids and lipids. These molecules can provide information about tumor cells. Usually, exosomes are widely present in easily accessible biological fluids such as human blood and urine. The corresponding exosomes are obtained by filtering blood and urine. There are many methods for determining the exosome category. This embodiment discloses a method for determining by image.
[0086] The method for obtaining the exosome type of the hth exosome to be separated includes:
[0087] Obtaining a real-time surface image of the hth exosome to be separated, inputting the real-time surface image into a pre-built exosome classification model, and obtaining the exosome category of the hth exosome to be separated;
[0088] The methods for constructing the secretory body classification model include:
[0089] Obtain r groups of data, where r is a positive integer greater than 1, and the data includes historical surface images and historical oosome categories. The historical surface images and historical oosome categories are used as sample sets, and the sample sets are divided into training sets and test sets. A classifier is constructed, and the historical surface images in the training set are used as input data, and the historical oosome categories in the training set are used as output data. The classifier is trained to obtain an initial classifier, and the initial classifier is tested using the test set, and a classifier that meets a preset accuracy is output as a oosome classification model, and the classifier is preferably a naive Bayes model or a support vector machine model;
[0090] It should be understood that the size of exosomes ranges widely, usually from tens to hundreds of nanometers. Exosomes in urine are usually small, with diameters generally between 30 and 200 nanometers, and are round or oval in shape. Exosomes in blood are more diverse in size, including small exosomes and large exosomes. Therefore, the size of the vesicles in the image can be used to distinguish the types of exosomes. However, in reality, it is difficult to collect a large number of historical surface images of exosomes. Therefore, it is necessary to generate a synthetic historical surface image that is similar to the actual historical surface image generated by the target adversarial network.
[0091] Exemplarily, the historical surface image includes an actual historical surface image and a synthetic historical surface image, and the method for generating the historical surface image includes:
[0092] Using the actual historical surface image to train a discriminator in the target generative adversarial network to obtain a trained discriminator;
[0093] Instructing the generator in the target generative adversarial network to generate a synthetic historical surface image similar to the actual historical surface image through the trained discriminator;
[0094] It should be noted that Generative Adversarial Networks (GANs) is a generative model proposed by deep learning. Its goal is to generate realistic data samples through adversarial learning between two network models. GANs consists of two main components: the generator and the discriminator. These two components compete with each other, and through adversarial training, the quality of samples generated by the generator is gradually improved.
[0095] In this embodiment, the discriminator can be regarded as a trained binary classification classifier, and the generator includes several upsampling layers, several feature extraction layers and several feature gating layers. The generator first obtains the features of the actual historical surface image, generates an initial historical surface image through the features, and then gives the initial historical surface image to the discriminator. The discriminator classifies the initial historical surface image and generates a classification result. The classification result includes true and false. The discriminator then feeds the classification result back to the generator, and iterates repeatedly. Through the continuous confrontation between the generator and the discriminator, a synthetic historical surface image with high similarity is generated. The synthetic historical surface image and the actual historical surface image are used to generate a historical surface image set. The historical surface image set is used to train the secretory body classification model. By amplifying the training samples, overfitting in the secretory body classification model training is avoided.
[0096] In this embodiment, the initial centrifugation parameter set includes but is not limited to centrifugation speed, centrifugal force and centrifugation time. It can be understood that the initial centrifugation parameter set refers to a centrifugation parameter set pre-set according to the secretory body type. Different secretory body types lead to different initial centrifugation parameter sets;
[0097] Data correction module: used to obtain the real-time weight of the hth exosome to be separated, input the initial centrifugation parameter set and the real-time weight into the pre-built parameter adjustment model, and obtain the target centrifugation parameter set;
[0098] Furthermore, the real-time weight of the exosomes to be separated can be directly obtained through the weight sensor provided by the separation device;
[0099] The construction method of the parameter adjustment model includes:
[0100] Acquire a sample data set, wherein the sample data set includes a historical weight, a historical initial centrifugal parameter set, and a historical target centrifugal parameter set, divide the sample data set into a sample training set and a sample test set, construct a regression network, use the historical weight and the historical initial centrifugal parameter set in the sample training set as input data of the regression network, use the historical target centrifugal parameter set in the sample training set as output data of the regression network, train the regression network, obtain an initial regression network for predicting the target centrifugal parameter set, test the initial regression network using the sample test set, output a regression network with a value less than a preset error as a parameter adjustment model, and the regression network is preferably a neural network model;
[0101] It should be noted that different types of exosomes in the above-mentioned medium lead to different initial centrifugation parameter sets. The weight of the same exosomes will also affect the centrifugation parameters. Therefore, the real-time weight is required to correct the initial centrifugation parameter set. The above-mentioned historical target centrifugation parameter set is obtained through continuous optimization of experiments. Those skilled in the art can determine the effect of exosome centrifugation by detecting the centrifuged exosomes and obtaining the data of the exosomes after centrifugation. For example:
[0102] Methods for obtaining historical target centrifugal parameter sets include:
[0103] Adjust the historical initial centrifugation parameter set m times to obtain m historical intermediate centrifugation parameter sets, centrifuge the exosomes based on the m historical intermediate centrifugation parameter sets respectively, obtain the sedimentation data set of the exosomes after centrifugation, calculate the centrifugation coefficient based on the sedimentation data set, and determine the historical target centrifugation parameter set from the m historical intermediate centrifugation parameter sets according to the centrifugation coefficient, the sedimentation data set includes sedimentation rate, clear liquid transparency and marker concentration;
[0104] Methods for calculating the centrifugal coefficient based on a sedimentation data set include:
[0105]
[0106] Where Cft is the centrifugal coefficient, SDE re is the real-time precipitation rate, SDE sd is the standard precipitation rate, CLT re For real-time clarity of liquid, CLT sd For standard clear liquid transparency, COM re is the real-time marker concentration, COM sd is the standard marker concentration, u 1 、u 2 and u 3 All are weight factors;
[0107] It should be noted that the higher the above precipitation rate, the better the exosome centrifugation effect, that is, the exosomes can be precipitated to the bottom of the centrifuge tube faster during the centrifugation process, and the supernatant is clear and transparent, with basically no suspended matter or precipitation, indicating that the exosomes are effectively separated. By measuring the concentration of exosome markers (such as CD63, CD9, etc.) in the supernatant after centrifugation, the centrifugation effect of the exosomes can be evaluated. The lower the concentration, the better the exosome centrifugation effect. Therefore, in this embodiment, the centrifugation coefficient is proportional to the centrifugation effect, that is, the larger the centrifugation coefficient, the better the exosome centrifugation effect. It can be understood that the standard precipitation rate, standard clear liquid transparency and standard marker concentration in the above are all preset according to expert experience;
[0108] In the above experimental process, m historical intermediate centrifugal parameter sets are first obtained. It can be understood that some parameters in the m historical intermediate centrifugal parameter sets are the same as the historical initial centrifugal parameter set. The method for determining the historical target centrifugal parameter set from the m historical intermediate centrifugal parameter sets according to the centrifugal coefficient can be to use the historical intermediate centrifugal parameter corresponding to the maximum value of the m centrifugal coefficients as the historical target centrifugal parameter set;
[0109] A state judgment module is used to separate the hth exosome to be separated based on the target centrifugal parameter set of the centrifugal device, and obtain electromagnetic field energy, and judge whether the centrifugal device is in a normal state based on the electromagnetic field energy;
[0110] In this embodiment, the electromagnetic field energy refers to the electromagnetic field energy inside the motor in the separation device. It can be understood that the separation device is usually driven by the internal motor to perform the separation action, and the electromagnetic field energy inside the motor can be directly obtained by the Hall sensor;
[0111] The method for judging whether the centrifugal device is in a normal state based on the electromagnetic field energy includes:
[0112] Fitting the electromagnetic field energy during the centrifugal time to generate an electromagnetic field energy curve, segmenting the centrifugal time based on the slope of the electromagnetic field energy curve to generate segmentation results, and judging whether the centrifugal device is in a normal state based on the segmentation results, wherein the segmentation results include a startup phase, a working phase, and a termination phase;
[0113] The method for judging whether the centrifugal device is in a normal state based on the segmentation results includes:
[0114] Obtaining the electromagnetic field energy corresponding to the working stage in the segmented result, and calculating the mean value of the electromagnetic field energy, judging whether the mean value of the electromagnetic field energy is within the standard energy interval, if so, judging that the centrifugal device is in a normal state, and if not, judging that the centrifugal device is not in a normal state;
[0115] It should be noted that the reason why the electromagnetic field energy is obtained in this embodiment but the current or voltage is not obtained is that the centrifugal time can be segmented more accurately through the electromagnetic field energy, and when the motor is started, a large starting current is usually applied to overcome inertia and friction, which will cause the electromagnetic field energy of the motor to increase. When the motor needs to be stopped, the control system will reduce the electromagnetic field energy of the motor, so that the motor gradually slows down and eventually stops. The slope of the electromagnetic field energy curve represents the rate of change of the electromagnetic field energy. Therefore, whether in the startup stage or in the termination stage, the slope of the electromagnetic field energy curve is at a large value, and only in the working stage, the electromagnetic field energy of the motor will be maintained at a relatively constant level. Therefore, the electromagnetic field energy curve is approximately a straight line at this time. In this embodiment, the startup stage and the termination stage are first excluded, and the electromagnetic field energy in the working stage is used to determine whether there is a problem with the centrifugal device, so that the centrifugal device can be detected more accurately;
[0116] Cycle module: when the centrifugal device is in a normal state, h+1=h is set, and the above data acquisition module, data correction module and state judgment module are repeatedly executed until h=f, then the centrifugation is terminated, f>h>0;
[0117] Warning module: used to send warning information when the centrifugal device is not in normal state.
[0118] Example 3
[0119] The present embodiment discloses an electronic device, including a power supply, an interface, a keyboard, a memory, a central processing unit, and a computer program stored in the memory and executable on the central processing unit. When the central processing unit executes the computer program, the centrifugal control method based on artificial intelligence analysis of exosomes provided by the above methods is implemented. The interface includes a network interface and a data interface. The network interface includes a wired or wireless interface, and the data interface includes an input or output interface.
[0120] Since the electronic device introduced in this embodiment is an electronic device used to implement the centrifugal control method based on artificial intelligence analysis of exosomes in the embodiment of this application, based on the centrifugal control method based on artificial intelligence analysis of exosomes introduced in the embodiment of this application, the technical personnel of this field can understand the specific implementation of the electronic device of this embodiment and its various variations, so how the electronic device implements the method in the embodiment of this application is not described in detail here. As long as the technical personnel of this field implement the electronic device used in the centrifugal control method based on artificial intelligence analysis of exosomes in the embodiment of this application, it belongs to the scope of protection of this application.
[0121] Example 4
[0122] like Figure 2As shown, this embodiment discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, the centrifugal control method based on artificial intelligence analysis of exosomes is implemented.
[0123] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters, weights and thresholds in the formula are set by technicians in this field according to actual conditions.
[0124] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through a wired network or a wireless network. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD) or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.
[0125] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.
[0126] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0127] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.
[0128] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0129] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
[0130] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.
[0131] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A centrifugal control method for analyzing exosomes based on artificial intelligence, characterized in that: include: S10: obtaining the exosome category of the hth exosome to be separated, and obtaining a corresponding initial centrifugation parameter set based on the exosome category; S20: obtaining the real-time weight of the hth exosome to be separated, inputting the initial centrifugation parameter set and the real-time weight into a pre-built parameter adjustment model, and obtaining a target centrifugation parameter set; S30: separating the hth exosome to be separated by adopting the target centrifugal parameter set based on the centrifugal device, obtaining electromagnetic field energy, and judging whether the centrifugal device is in a normal state based on the electromagnetic field energy; S40: When the centrifugal device is in a normal state, h+1=h is set, and the above S10-S30 are repeated until h=f, then the centrifugation is terminated, and both f and h are greater than 0; S50: When the centrifugal device is not in a normal state, a warning message is sent; The method for constructing the parameter adjustment model includes: A sample data set is obtained, wherein the sample data set includes a historical weight, a historical initial centrifugal parameter set, and a historical target centrifugal parameter set. The sample data set is divided into a sample training set and a sample test set, and a regression network is constructed. The historical weight and the historical initial centrifugal parameter set in the sample training set are used as input data of the regression network, and the historical target centrifugal parameter set in the sample training set is used as output data of the regression network. The regression network is trained to obtain an initial regression network for predicting the target centrifugal parameter set. The initial regression network is tested using the sample test set, and a regression network with a smaller error value than a preset value is output as a parameter adjustment model.
2. The centrifugal control method for analyzing exosomes based on artificial intelligence according to claim 1, characterized in that: The exosome categories include blood exosomes and urine exosomes, and the method for obtaining the exosome category of the hth exosome to be separated includes: A real-time surface image of the h-th exosome to be separated is obtained, and the real-time surface image is input into a pre-built exosome classification model to obtain the exosome category of the h-th exosome to be separated.
3. The centrifugal control method for analyzing exosomes based on artificial intelligence according to claim 2, characterized in that: The method for constructing the secretory body classification model includes: Obtain r groups of data, where r is a positive integer greater than 1, and the data include historical surface images and historical oosome categories. The historical surface images and historical oosome categories are used as sample sets, and the sample sets are divided into training sets and test sets. A classifier is constructed, and the historical surface images in the training set are used as input data, and the historical oosome categories in the training set are used as output data. The classifier is trained to obtain an initial classifier, and the initial classifier is tested using the test set, and a classifier that meets the preset accuracy is output as a oosome classification model.
4. The centrifugal control method for analyzing exosomes based on artificial intelligence according to claim 3, characterized in that: The historical surface image is generated by a preset target generative adversarial network, the historical surface image includes an actual historical surface image and a synthetic historical surface image, and the method for generating the historical surface image includes: Using the actual historical surface image to train a discriminator in the target generative adversarial network to obtain a trained discriminator; The trained discriminator guides the generator in the target generative adversarial network to generate a synthetic historical surface image similar to the actual historical surface image.
5. The centrifugal control method for analyzing exosomes based on artificial intelligence according to claim 4, characterized in that: Methods for obtaining historical target centrifugal parameter sets include: The historical initial centrifugation parameter set is adjusted m times to obtain m historical intermediate centrifugation parameter sets, exosomes are centrifuged based on the m historical intermediate centrifugation parameter sets, a sedimentation data set of exosomes after centrifugation is obtained, the centrifugation coefficient is calculated based on the sedimentation data set, and the historical target centrifugation parameter set is determined from the m historical intermediate centrifugation parameter sets according to the centrifugation coefficient, and the sedimentation data set includes sedimentation rate, clear liquid transparency and marker concentration.
6. The centrifugal control method for analyzing exosomes based on artificial intelligence according to claim 5, characterized in that: The method for calculating the centrifugal coefficient based on the sedimentation data set comprises: Where Cft is the centrifugal coefficient, SDE re is the real-time precipitation rate, SDE sd is the standard precipitation rate, CLT re For real-time clarity of liquid, CLT sd For standard clear liquid transparency, COM re is the real-time marker concentration, COM sd is the standard marker concentration, u1, u2 and u3 are weight factors.
7. The centrifugal control method for analyzing exosomes based on artificial intelligence according to claim 1, characterized in that: The method for judging whether the centrifugal device is in a normal state based on electromagnetic field energy includes: The electromagnetic field energy during the centrifugal time is fitted to generate an electromagnetic field energy curve. The centrifugal time is segmented based on the slope of the electromagnetic field energy curve to generate segmentation results. Based on the segmentation results, it is judged whether the centrifugal equipment is in a normal state. The segmentation results include a startup phase, a working phase, and a termination phase.
8. The centrifugal control method for analyzing exosomes based on artificial intelligence according to claim 7, characterized in that: The method for judging whether the centrifugal device is in a normal state based on the segmentation result comprises: The electromagnetic field energy corresponding to the working stage in the segmented result is obtained, and the mean value of the electromagnetic field energy is calculated to determine whether the mean value of the electromagnetic field energy is within the standard energy range. If so, the centrifugal device is determined to be in a normal state; if not, the centrifugal device is determined to be not in a normal state.
9. A centrifugal control system for analyzing exosomes based on artificial intelligence, which is used to implement the centrifugal control method for analyzing exosomes based on artificial intelligence according to any one of claims 1 to 8, characterized in that: include: Data acquisition module: used to obtain the exosome category of the hth exosome to be separated, and obtain the corresponding initial centrifugation parameter set based on the exosome category; Data correction module: used to obtain the real-time weight of the hth exosome to be separated, input the initial centrifugation parameter set and the real-time weight into the pre-built parameter adjustment model, and obtain the target centrifugation parameter set; A state judgment module is used to separate the hth exosome to be separated based on the target centrifugal parameter set of the centrifugal device, and obtain electromagnetic field energy, and judge whether the centrifugal device is in a normal state based on the electromagnetic field energy; Cycle module: when the centrifugal device is in a normal state, h+1=h is set, and the above data acquisition module, data correction module and state judgment module are repeatedly executed until h=f, then the centrifugation is terminated, f>h>0; Warning module: used to send warning information when the centrifugal device is not in normal state.
Citation Information
Patent Citations
Extraction method and application of exosome
CN115710575A
Water cooling type ductile iron pipe centrifuging machine and control system thereof
CN109158569A
Method and device for mechanical seal fault diagnosis of centrifugal pump based on sensor-free monitoring technology
CN110848165A
