Method for rapidly detecting ultralow-temperature storage efficiency based on differential scanning calorimetry
By constructing a multi-dimensional dataset and using machine learning algorithms to establish a data association model, the problem of insufficient preprocessing adaptability in the evaluation of cryogenic preservation efficiency was solved, quantitative prediction and online updates were realized, and the accuracy and applicability of cryogenic preservation were improved.
Patent Information
- Application Number
- CN202511987087.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-26
- Publication Date
- 2026-02-17
AI Technical Summary
Existing technologies lack quantitative models for evaluating the efficiency of cryopreservation, making it difficult to dynamically adapt to the preservation needs of new species or novel tissues. Furthermore, the lack of an effective online update mechanism means that pretreatment schemes rely on trial and error based on experience and cannot provide a basis for prediction before freezing.
A multi-dimensional dataset is constructed, and a data association model is trained using machine learning algorithms. A nonlinear mapping relationship is established through basic characteristic parameters, initial thermodynamic response, and preprocessing parameters. An online update mechanism is introduced to dynamically expand the applicable scope and output a customized optimal preprocessing scheme.
It enables quantitative prediction based on historical data, solves the problem of insufficient preprocessing adaptability, improves the accuracy and applicability of preprocessing schemes, and has comparability and repeatability across batches and operators.
Smart Images
Figure CN121540754A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of plant cryopreservation detection technology, specifically relating to a method for rapid detection of cryopreservation efficiency based on differential scanning calorimetry. Background Technology
[0002] In the field of plant germplasm resource conservation and biodiversity preservation, cryopreservation technology is widely used due to its ability to maintain cell viability stably for extended periods. The core of this technology lies in achieving complete cessation of metabolic activity in biological materials through a liquid nitrogen environment, thereby preventing genetic variation and physiological degradation. However, different plant materials exhibit significantly different sensitivities to freezing damage due to differences in fundamental characteristics such as cell structure, water state, and membrane permeability, making it difficult to apply universal pretreatment protocols to all species or tissue types. To improve preservation success rates, it is usually necessary to optimize pretreatment parameters for specific materials, such as sucrose concentration, dehydration time, and the type of osmotic regulator, to control ice crystal formation and glass transition behavior.
[0003] Current technologies for evaluating cryopreservation efficiency largely rely on biological refolding indicators after thawing, such as callus regeneration rate, bud regeneration number, or cell viability. While these methods reflect the final preservation effect, they suffer from drawbacks such as long cycles, high costs, and strong subjectivity, and cannot provide predictive evidence before freezing. Some studies have attempted to introduce differential scanning calorimetry (DSC) to assess the thermodynamic behavior of samples during cooling and heating, indirectly evaluating the degree of intracellular ice formation and solution stability through parameters such as crystallization enthalpy, melting enthalpy, and glass transition temperature. However, current DSC applications are mostly limited to qualitative or semi-quantitative analysis, lacking quantitative models that systematically correlate material fundamental properties, initial thermodynamic responses, and pretreatment parameters. This results in pretreatment schemes still primarily relying on empirical trial and error for determination.
[0004] Furthermore, although existing studies have explored using single differential scanning calorimetry parameters to set thresholds to assess preservation effectiveness, they have not fully considered the impact of intrinsic differences in plant materials on thermodynamic responses, nor have they established an intelligent mapping mechanism from initial measurement data to optimal pretreatment strategies. Simultaneously, existing methods typically employ fixed pretreatment procedures, making it difficult to dynamically adapt to the preservation needs of new species or novel tissues, and they lack an effective online update mechanism to integrate newly added experimental data and continuously optimize predictive capabilities. Summary of the Invention
[0005] In view of this, in order to solve the problems mentioned in the background technology, a method for rapid detection of cryogenic preservation efficiency based on differential scanning calorimetry is proposed.
[0006] The objective of this invention can be achieved through the following technical solution: This invention provides a method for rapidly detecting cryogenic preservation efficiency based on differential scanning calorimetry, comprising: S1, constructing a multi-dimensional dataset based on historical data, and training a data association model using a machine learning algorithm; the historical data includes sample data of various plant materials, the sample data including basic characteristic parameters, initial differential scanning calorimetry measurement data without preprocessing, parameters of the preprocessing scheme adopted, and optimized differential scanning calorimetry measurement data after preprocessing and cryogenic preservation.
[0007] S2. The basic characteristic parameters of various plant materials and their initial measurement data by differential scanning calorimetry are used as inputs and imported into the data association model to output customized optimal ultra-low temperature preservation pretreatment scheme parameters.
[0008] S3. The various plant materials are processed according to the customized optimal cryogenic preservation pretreatment parameters to obtain various plant tissue solutions, and differential scanning calorimetry is used to measure the crystallization enthalpy and melting enthalpy of the plant tissue solutions.
[0009] S4. Compare the enthalpy value with the set enthalpy value threshold to determine whether the ultra-low temperature preservation efficiency meets the standard.
[0010] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The present invention constructs a structured multi-dimensional dataset containing the basic characteristics of plant materials, initial thermodynamic response, preprocessing parameters and optimized enthalpy performance, and uses normalization and supervised learning algorithms to establish a nonlinear mapping relationship between input and output, so that the generation of preprocessing schemes no longer depends on experience trial and error, but is based on quantitative prediction driven by historical effective data, thus solving the problem of insufficient preprocessing adaptability caused by ignoring the intrinsic differences of materials in traditional methods.
[0011] 2. This invention introduces an online update mechanism based on newly added valid samples. The model parameters are fine-tuned through an incremental learning algorithm, and a validation set backtesting threshold is set to control the risk of model degradation. This enables the system to continuously absorb experimental data from new species or new tissue types and dynamically expand its applicability without losing its historical predictive ability.
[0012] 3. This invention uses the enthalpy value of differential scanning calorimetry as the core criterion and strictly limits the test conditions, sample preparation methods, data processing rules and multi-peak judgment logic, so that the thermodynamic parameters have comparability and repeatability across batches and operators. Attached Figure Description
[0013] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0014] Figure 1 This is a flowchart illustrating the implementation steps of the method of the present invention. Detailed Implementation
[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0016] Please see Figure 1 As shown, this invention provides a method for rapidly detecting cryogenic preservation efficiency based on differential scanning calorimetry. The specific steps are as follows: S1, construct a multi-dimensional dataset based on historical data and train a data association model using a machine learning algorithm; the historical data includes sample data of various plant materials, and the sample data includes basic characteristic parameters, initial differential scanning calorimetry measurement data without preprocessing, parameters of the preprocessing scheme adopted, and optimized differential scanning calorimetry measurement data after preprocessing and cryogenic preservation.
[0017] In one feasible embodiment of the present invention, the specific construction process of the multidimensional dataset includes: collecting various plant material samples, uniformly removing surface impurities, and recording the cell type, water content and cell membrane permeability of each sample as basic characteristic data.
[0018] It should be noted that the plant material samples should be fresh and undamaged.
[0019] For each plant material sample, an untreated plant tissue solution was prepared, and the initial enthalpy of crystallization and the initial enthalpy of melting were measured using a differential scanning calorimeter according to a pre-set temperature control rule. These were recorded as the initial measurement data of differential scanning calorimetry, i.e., the initial DSC data.
[0020] It should be noted that the pre-set temperature control rules of the differential scanning calorimeter remain strictly consistent during the initial detection and storage efficiency verification stages.
[0021] For each plant material sample, different combinations of sucrose concentration, dehydration time, and osmotic regulator type were used for pretreatment, resulting in multiple combinations of pretreatment parameters.
[0022] Differential scanning calorimetry (DSC) was performed again on the plant material samples after processing with various pretreatment scheme parameter combinations to obtain the corresponding optimized DSC data. The pretreatment scheme parameter combinations that were selected from these data and whose enthalpy values were not higher than the set enthalpy value threshold were recorded as the effective pretreatment scheme parameter combinations.
[0023] The basic characteristic parameters, the initial measurement data of differential scanning calorimetry, the parameters of each preprocessing scheme, and the optimized measurement data of differential scanning calorimetry are collectively referred to as the sample data of plant materials, and are denoted as the multidimensional dataset.
[0024] Specifically, all data is integrated into a structured multi-dimensional dataset based on the field structure of basic characteristic data, initial DSC data, preprocessing scheme parameter combination, and optimized DSC data, for subsequent model training.
[0025] In one feasible embodiment of the present invention, the specific process of training the data association model using a machine learning algorithm includes: normalizing the basic characteristics of plant materials and the initial DSC data in the multidimensional dataset to eliminate dimensional differences.
[0026] It should be noted that the normalization process specifically includes normalization of continuous data and normalization of discrete data; the normalization of continuous data such as moisture content and initial enthalpy value can be performed using the Z-score normalization method; the normalization of discrete data such as cell type and osmotic regulator type can be performed using the one-heat coding method to completely eliminate dimensional differences and numerical bias of categorical variables.
[0027] The normalized basic characteristics of plant materials and initial DSC data are used as input variables, and the corresponding optimal combination of parameters for cryopreservation pretreatment schemes is used as output variables to construct a supervised learning task.
[0028] Machine learning algorithms are selected to model the nonlinear mapping relationship between input and output variables. These machine learning algorithms include, but are not limited to, support vector machines, random forests, and neural networks.
[0029] Specifically, the support vector machine is suitable for high-dimensional sparse data scenarios; the random forest is suitable for multi-feature interaction and anti-overfitting requirements; the neural network is suitable for complex nonlinear relationship modeling; the optimal algorithm can be selected according to the dimensionality and distribution characteristics of the actual data.
[0030] Cross-validation is used to train and validate the constructed model to ensure its generalization ability on unseen samples.
[0031] Models that have been trained and whose cross-validation accuracy meets the standard (≥90%) are solidified as data association models.
[0032] In one feasible embodiment of the present invention, the specific contents of the pre-set temperature control rules include: the starting temperature of the differential scanning calorimeter is -100℃, the ending temperature is 50℃, the heating rate is 5℃ / min, the nitrogen atmosphere flow rate is 50mL / min, and the sample mass is controlled at 8–12mg.
[0033] It should be noted that the purity of nitrogen in the nitrogen atmosphere is ≥99.99%; the sample mass can be accurately weighed using a balance with a mass of 0.0001 g / L, and the sample is evenly spread in a special aluminum crucible to avoid local accumulation.
[0034] In one feasible embodiment of the present invention, the combination of parameters of the multiple pretreatment schemes specifically includes: the sucrose concentration gradient is any one of 0.2 mol / L, 0.4 mol / L, 0.6 mol / L, 0.8 mol / L, and 1.0 mol / L.
[0035] The dehydration time gradient is specifically any one of 30 minutes, 60 minutes, 90 minutes, and 120 minutes.
[0036] The specific type of osmotic regulator is any one of glycerol, dimethyl sulfoxide, trehalose, polyethylene glycol 4000, or proline, and different types of osmotic regulators are encoded as discrete categorical variables during the training of the data association model.
[0037] It should be noted that different types of permeation regulators are converted into discrete categorical variables using a unique encoding method. The encoding rule is that each permeation regulator corresponds to an independent feature dimension, which is marked as 1 if it exists, and marked as 0 otherwise.
[0038] By arbitrarily combining the sucrose concentration gradient, dehydration time gradient, and osmotic conditioner type mentioned above, a variety of pretreatment scheme parameter combinations can be obtained.
[0039] It should be noted that during the experiment, a basic solution was first prepared according to the corresponding sucrose concentration, then a specified type of osmotic regulator was added and stirred evenly. After standing for 10 minutes to ensure the stability of the solution system, the plant material was dehydrated.
[0040] It should also be noted that all effective combinations of parameters, such as the type of osmotic regulator, sucrose concentration, and dehydration time, as well as the raw data from DSC detection and regeneration growth, are marked as components of the optimal cryopreservation pretreatment scheme parameters for the plant material and included in the multidimensional dataset for model training.
[0041] This invention constructs a structured, multi-dimensional dataset containing the basic properties of plant materials, initial thermodynamic response, preprocessing parameters, and optimized enthalpy performance. It also employs normalization and supervised learning algorithms to establish a nonlinear mapping relationship between input and output. This allows the generation of preprocessing schemes to no longer rely on empirical trial and error, but rather on quantitative prediction driven by historical effective data. This solves the problem of insufficient preprocessing adaptability caused by traditional methods ignoring the intrinsic differences of materials.
[0042] In one feasible embodiment of the present invention, the data association model needs to be updated online after training. The specific process includes: whenever a new plant material completes the full-process verification and obtains an effective result that the optimal combination of parameters for cryopreservation pretreatment scheme and the corresponding enthalpy value is not higher than the set enthalpy value threshold, the basic characteristics of the plant material, the initial DSC data, and the optimal combination of parameters for cryopreservation pretreatment scheme are added as new samples to the original multidimensional dataset.
[0043] It should be noted that the specific steps of the full-process verification include: pretreatment → cryogenic storage → thawing → DSC detection.
[0044] The expanded dataset is re-normalized, and the data association model is fine-tuned using an incremental learning algorithm.
[0045] It should be noted that the incremental learning algorithm is an online gradient descent or elastic weight consolidation method, which ensures that the model retains its predictive ability for historical plant materials while absorbing new knowledge.
[0046] The online gradient descent method is suitable for scenarios where the number of new samples is small and the data distribution is not significantly different from that of historical samples; the elastic weight consolidation method is suitable for scenarios where it is necessary to retain the predictive ability of historical plant materials and where the new samples are new species / new tissue types.
[0047] It should also be noted that the updated model needs to be backtested on the validation set. If the prediction accuracy drops by no more than 2%, the original model will be replaced as the new data association model. This mechanism enables the model to adapt to the rapid adaptation needs of new species or new tissue types, improving the universality and practicality of the method.
[0048] The reserved validation set consists of 20% independent samples from the original multidimensional dataset, which are not included in the initial training of the model or subsequent incremental training.
[0049] This invention introduces an online update mechanism based on newly added valid samples. It fine-tunes the model parameters through an incremental learning algorithm and sets a validation set backtesting threshold to control the risk of model degradation. This enables the system to continuously absorb experimental data from new species or new tissue types and dynamically expand its applicability without losing its historical predictive ability.
[0050] S2. The basic characteristic parameters of various plant materials and their initial measurement data by differential scanning calorimetry are used as inputs and imported into the data association model to output customized optimal ultra-low temperature preservation pretreatment scheme parameters.
[0051] In one feasible embodiment of the present invention, the specific implementation process of step S2 includes: taking fresh plant tissue samples of various plant materials, cleaning and removing surface impurities, grinding them into a uniform slurry, and preparing an untreated plant tissue solution.
[0052] The plant tissue solution was dispensed into a special aluminum crucible for differential scanning calorimetry (DSC), sealed, and placed inside the DSC. Based on a pre-set temperature control rule, heat flow scanning was performed to obtain the enthalpy of crystallization and the enthalpy of melting.
[0053] Simultaneously, cell type identification was performed on plant tissue samples from the same batch, and the cell type was determined by microscopic observation or flow cytometry.
[0054] The moisture content of plant tissue samples was determined by drying and weighing, and the cell membrane permeability was determined by fluorescent dye penetration test.
[0055] Specifically, the moisture content of plant tissue samples was determined by drying and weighing at 105℃ for 2 hours, and the cell membrane permeability was determined by the fluorescein diacetate (FDA) fluorescent dye penetration test.
[0056] The above measurement results were recorded and formatted in a unified manner, serving as the basic characteristic parameters of the plant material to be tested and its initial measurement data for differential scanning calorimetry.
[0057] The basic characteristic parameters and the initial measurement data of differential scanning calorimetry are used as inputs and imported into the data association model to output customized optimal cryogenic preservation pretreatment parameters.
[0058] Specifically, the above measurement results are recorded in a structured format of sample number - basic characteristic parameters - initial DSC data.
[0059] S3. The various plant materials are processed according to the customized optimal cryogenic preservation pretreatment parameters to obtain various plant tissue solutions, and differential scanning calorimetry is used to measure the crystallization enthalpy and melting enthalpy of the plant tissue solutions.
[0060] In one feasible embodiment of the present invention, the specific implementation process of step S3 includes: preparing a sucrose solution of a corresponding concentration according to the optimal cryogenic preservation pretreatment scheme parameter combination, adding a corresponding type of osmotic regulator to obtain a standard solution corresponding to the optimal cryogenic preservation pretreatment scheme parameter combination, stirring evenly and letting it stand for 10 minutes for later use to ensure the stability of the solution system.
[0061] The plant material to be tested is immersed in the standard solution and dehydrated for the corresponding time under controlled environmental temperature and humidity conditions.
[0062] Specifically, the plant material is completely immersed in the above mixed solution, with the liquid level at least 1 cm above the material surface. The material is then dehydrated for a specified duration by oscillation at a speed of 120 r / min under a constant environment of 25 ± 1 °C and 60% relative humidity.
[0063] After dehydration, the plant material is transferred to liquid nitrogen for cryopreservation at ultra-low temperatures.
[0064] Specifically, after the dehydration process is completed, the plant material is quickly transferred to a special cryogenic container for liquid nitrogen within 30 seconds and immersed below the liquid nitrogen surface for ultra-low temperature cryopreservation. The preservation time is set according to the experimental requirements.
[0065] After preservation, the sample was thawed and re-prepared into a plant tissue solution, which was then placed in a special crucible for differential scanning calorimetry (DSC) and measured using DSC to obtain the DSC curve of the plant tissue solution. The enthalpy of the cooling stage (from the supercooling point of the aqueous solution to the completion of crystallization) and the enthalpy of the melting endothermic peak (the melting of ice crystals) were obtained by integral calculation.
[0066] Specifically, after preservation, the sample is removed from liquid nitrogen and immediately placed in a constant temperature water bath at 37-40℃ to thaw quickly until completely melted. The liquid nitrogen residue and solution traces on the sample surface are then gently wiped dry with sterile filter paper.
[0067] Differential scanning calorimetry was performed again under the same temperature control program as the initial DSC test.
[0068] In a feasible embodiment of the present invention, the specific process of integrating the DSC curve to obtain the enthalpy value of the crystallization exothermic peak in the cooling stage and the enthalpy value of the melting endothermic peak in the heating stage includes: after baseline correction of the DSC curves of multiple plant tissue solutions, the peak area integration method is used to calculate the enthalpy value; the starting and ending points of the peak area integration are defined as the intersection points of the baseline and the tangents on both sides of the thermal effect peak, and the calculation is completed using the standard integration software supporting the instrument.
[0069] And the calculation range of the enthalpy value is limited to the exothermic peak and endothermic peak between -80°C and -20°C. If there are multiple thermal effect peaks in this temperature range for the sample, the absolute value difference of each peak is determined. If the difference is greater than the set difference threshold, the single-peak enthalpy value with the largest absolute value is taken as the judgment basis; otherwise, the sample is re-prepared and the DSC test is carried out to exclude the multi-peak interference caused by sample non-uniformity.
[0070] It should be noted that the correction temperature range in the baseline correction is the same as the calculation range of the enthalpy value, that is, -80°C to -20°C.
[0071] It should also be noted that before each DSC test, the instrument needs to be calibrated for temperature and heat flow using a high-purity indium standard substance, and the calibration error is controlled within ±0.1°C and ±1%.
[0072] This standardized test process ensures the comparability and repeatability of DSC data obtained in different batches and by different operators, providing a reliable data basis for model input.
[0073] S4. Compare the enthalpy value with the set enthalpy value threshold to judge whether the cryopreservation efficiency meets the standard.
[0074] In a feasible embodiment of the present invention, the specific process of obtaining the renaturation state evaluation index includes: comparing the enthalpy value with the set enthalpy value threshold. If the enthalpy value is not higher than the set enthalpy value threshold, the cryopreservation efficiency is recorded as meeting the standard.
[0075] The set enthalpy value threshold is determined by statistical quantiles based on the optimized measurement data of differential scanning calorimetry corresponding to the effective pretreatment schemes in the multi-dimensional dataset.
[0076] It should be noted that the determination method of the set enthalpy value threshold includes: extracting all optimized DSC enthalpy values corresponding to the effective pretreatment schemes from the multi-dimensional dataset, which includes the crystallization enthalpy value and the melting enthalpy value.
[0077] For a single plant type, the 95% quantile of the effective enthalpy values of this type is taken, that is, 95% of the effective sample enthalpy values are lower than this value; for the general threshold of multiple plant types, the 90% quantile of all effective sample enthalpy values is taken.
[0078] This invention uses the enthalpy value of differential scanning calorimetry as the core criterion and strictly limits the test conditions, sample preparation methods, data processing rules and multi-peak judgment logic, so that the thermodynamic parameters have comparability and repeatability across batches and operators.
[0079] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.
[0080] Those skilled in the art will recognize that the algorithmic steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.
[0081] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.
[0082] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
[0083] 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 within the protection scope of the present invention.
Claims
1. A method for rapid detection of the efficiency of cryopreservation based on differential scanning calorimetry, characterized by: The method comprises the following steps: S1. Constructing a multi-dimensional data set based on historical data and training a data correlation model using a machine learning algorithm; the historical data comprises sample data of multiple plant materials, and the sample data comprises basic characteristic parameters, initial differential scanning calorimetry (DSC) measurement data without pretreatment, adopted pretreatment scheme parameters, and optimized DSC measurement data after pretreatment and ultra-low temperature preservation; S2. Inputting the obtained basic characteristic parameters and initial DSC measurement data of the multiple plant materials into the data correlation model as inputs, and outputting customized optimal ultra-low temperature preservation pretreatment scheme parameters; S3. Processing the multiple plant materials according to the customized optimal ultra-low temperature preservation pretreatment scheme parameters, obtaining multiple plant tissue solutions, and measuring the plant tissue solutions by DSC to obtain the crystallization enthalpy and melting enthalpy of the plant tissue solutions; S4. Comparing the enthalpy with a set enthalpy threshold to determine whether the ultra-low temperature preservation efficiency meets the standard.
2. The method for rapid detection of the efficiency of cryopreservation based on differential scanning calorimetry according to claim 1, characterized in that: The specific construction process of the multi-dimensional data set comprises the following steps: Collecting multiple plant material samples, uniformly removing surface impurities, and recording the cell type, water content and cell membrane permeability of each sample as basic characteristic data; Preparing plant tissue solutions without pretreatment for each plant material sample, and using a differential scanning calorimeter to measure the initial crystallization enthalpy and initial melting enthalpy according to a pre-set temperature control rule, which are recorded as initial DSC measurement data; For each plant material sample, different combinations of sucrose concentration, dehydration time and osmotic regulator type are used for pretreatment to form multiple pretreatment scheme parameter combinations; Performing DSC test on the plant material samples treated by the pretreatment scheme parameter combinations to obtain corresponding optimized DSC measurement data, and selecting each pretreatment scheme parameter combination corresponding to an enthalpy not higher than a set enthalpy threshold as an effective pretreatment scheme parameter combination; The basic characteristic parameters, initial DSC measurement data, pretreatment scheme parameters and optimized DSC measurement data are collectively referred to as sample data of the plant materials, and are recorded as a multi-dimensional data set.
3. The method for rapid detection of the efficiency of cryopreservation based on differential scanning calorimetry according to claim 2, characterized in that: The specific process of training the data correlation model using a machine learning algorithm comprises the following steps: Normalizing the plant material basic characteristics and initial DSC data in the multi-dimensional data set to eliminate dimensional differences; Using the normalized plant material basic characteristics and initial DSC data as input variables, and using the corresponding optimal ultra-low temperature preservation pretreatment scheme parameter combinations as output variables to construct a supervised learning task; Selecting a machine learning algorithm to model the nonlinear mapping relationship between the input variables and the output variables, wherein the machine learning algorithm includes but is not limited to support vector machines, random forests and neural networks; Using a cross-validation method to train and verify the constructed model to ensure the generalization ability of the model on unseen samples; Solidifying the trained model with a cross-validation accuracy that meets the standard as a data correlation model.
4. The method for rapid detection of the efficiency of cryopreservation based on differential scanning calorimetry according to claim 2, characterized in that: The specific content of the pre-set temperature control rule includes: The initial temperature of the differential scanning calorimeter is -100℃, the final temperature is 50℃, the temperature rise rate is 5℃ / min, the nitrogen atmosphere flow rate is 50mL / min, and the sample mass is controlled at 8-12mg.
5. The method for rapid detection of the efficiency of cryopreservation based on differential scanning calorimetry according to claim 2, characterized in that: The specific combination of the plurality of pretreatment scheme parameters includes: The specific sucrose concentration gradient is any one of 0.2mol / L, 0.4mol / L, 0.6mol / L, 0.8mol / L, and 1.0mol / L; The specific dehydration time gradient is any one of 30 minutes, 60 minutes, 90 minutes, and 120 minutes; The specific type of osmotic regulator is any one of glycerol, dimethyl sulfoxide, trehalose, polyethylene glycol 4000, or proline, and in the data correlation model training process, different osmotic regulator types are coded as discrete categorical variables; Any combination of the above sucrose concentration gradient, dehydration time gradient, and osmotic regulator type is used to obtain a plurality of pretreatment scheme parameter combinations.
6. The method for rapid detection of the efficiency of cryopreservation based on differential scanning calorimetry according to claim 1, characterized in that: After the data correlation model is trained, it also needs to be updated online, and the specific process includes: Whenever a new plant material completes the full-process verification and obtains an effective optimal ultra-low temperature preservation pretreatment scheme parameter combination and an effective result with a corresponding heat value not higher than a set heat value threshold, the basic characteristics, initial DSC data, and optimal ultra-low temperature preservation pretreatment scheme parameter combination of the plant material are added to the original multidimensional dataset as new samples; The extended dataset is re-normalized, and the data correlation model is parameter fine-tuned using incremental learning algorithms.
7. The method for rapid detection of the efficiency of cryopreservation based on differential scanning calorimetry according to claim 1, characterized in that: The specific implementation process of step S2 includes: Take fresh plant tissue samples of a plurality of plant materials, wash and remove surface impurities, grind them into uniform slurry, and prepare an untreated plant tissue solution; The plant tissue solution is divided into differential scanning calorimeter special aluminum crucibles, sealed, and placed in a differential scanning calorimeter, and based on the pre-set temperature control rule, the heat flow scanning is performed to obtain the crystallization heat value and the melting heat value; Synchronously identify the cell type of the same batch of plant tissue samples, and determine the cell type by microscopic observation or flow cytometry; Determine the water content of the plant tissue sample by the drying weighing method, and determine the cell membrane permeability by the fluorescent dye penetration experiment; Record and format the above measurement results as the basic characteristic parameters and the initial differential scanning calorimetry measurement data of the plant material to be tested; The basic characteristic parameters and the initial differential scanning calorimetry measurement data are input into the data correlation model as inputs, and the customized optimal ultra-low temperature preservation pretreatment scheme parameters are output.
8. The method for rapid detection of the efficiency of cryopreservation based on differential scanning calorimetry according to claim 1, characterized in that: The specific implementation process of step S3 includes: According to the optimal ultra-low temperature preservation pretreatment scheme parameter combination, prepare a sucrose solution with the corresponding concentration, and add an osmotic regulator of the corresponding type to obtain a standard solution corresponding to the optimal ultra-low temperature preservation pretreatment scheme parameter combination, stir uniformly, and stand for 10 minutes for standby; Immerse the plant material to be tested in the standard solution and perform dehydration treatment for the corresponding dehydration time under controlled environmental temperature and humidity conditions; After the dehydration treatment is completed, the plant material is transferred to liquid nitrogen for ultra-low temperature freezing preservation; After the preservation is completed, the sample is thawed and prepared into a plant tissue solution again, loaded into a differential scanning calorimeter special crucible, and subjected to differential scanning calorimetry measurement to obtain a DSC curve of the plant tissue solution, and the DSC curve is integrated to obtain a crystallization exothermic peak enthalpy value in the cooling stage and a melting endothermic peak enthalpy value in the heating stage.
9. The method for rapid detection of the efficiency of cryopreservation based on differential scanning calorimetry according to claim 8, characterized in that: The specific process of integrating the DSC curve to obtain the crystallization exothermic peak enthalpy value in the cooling stage and the melting endothermic peak enthalpy value in the heating stage includes: After the DSC curves of the various plant tissue solutions are baseline corrected, the peak area integration method is used to calculate the enthalpy value; the starting point and the ending point of the peak area integration are defined as the intersection of the baseline and the tangent lines on both sides of the thermal effect peak, and the calculation is completed using the standard integral software provided with the instrument; And the enthalpy value calculation range is limited to the exothermic peak and the endothermic peak between-80℃ and-20℃, if the sample has multiple thermal effect peaks in this temperature range, the absolute value difference of each peak is determined, if the difference is greater than the set difference threshold, the single-peak enthalpy value with the largest absolute value is taken as the determination basis; otherwise, the sample is prepared again and subjected to DSC testing.
10. The method for rapid detection of the efficiency of cryopreservation based on differential scanning calorimetry according to claim 1, characterized in that: The specific implementation process of step S4 includes: Compare the enthalpy value with the set enthalpy value threshold, if the enthalpy value is not higher than the set enthalpy value threshold, the ultra-low temperature preservation efficiency is recorded as meeting the standard; The set enthalpy value threshold is based on the differential scanning calorimetry optimized measurement data corresponding to the effective pretreatment scheme in the multidimensional data set, and is determined by statistical quantile.