Cell inspection device, cell inspection method, program, and recording medium
阅读说明：本技术 细胞检查装置、细胞检查方法、程序和记录介质 (Cell inspection device, cell inspection method, program, and recording medium ) 是由 平野明成 于 2018-12-20 设计创作，主要内容包括：本发明提供细胞检查装置、细胞检查方法、程序和记录介质。本发明的一种方式的细胞检查装置包括：计测培养液的阻抗的阻抗传感器；存储系数的存储部,从细胞的培养开始到死亡的培养期间内的规定期间分类为多个期间,该系数用于对分类的多个期间的每一个,使用阻抗,推定在规定期间内在培养液中存活的活细胞数；以及活细胞数推定部,获取阻抗,使用获取的阻抗和存储部存储的每一个期间的系数中的至少一个来推定活细胞数。(The invention provides a cell inspection apparatus, a cell inspection method, a program, and a recording medium. A cell inspection apparatus according to an aspect of the present invention includes: an impedance sensor for measuring the impedance of the culture solution; a storage unit for storing a coefficient for estimating the number of living cells that have survived in a culture medium for a predetermined period, using an impedance for each of a plurality of classified periods, the predetermined period being a period from the start of culture of the cells to the death of the cells; and a living cell number estimating unit that acquires the impedance and estimates the number of living cells using at least one of the acquired impedance and the coefficient for each period stored in the storage unit.)
1. A cell inspection apparatus characterized by comprising:
an impedance sensor for measuring the impedance of the culture solution;
a storage unit for storing a coefficient for estimating the number of living cells that have survived in the culture medium in each of a plurality of classified periods using the impedance, the predetermined period being a period from the start of cell culture to the death of the cells during the culture; and
and a living cell number estimating unit that acquires the impedance and estimates the number of living cells using at least one of the acquired impedance and the coefficient for each of the periods stored in the storage unit.
2. The cell inspection apparatus according to claim 1, wherein the storage unit classifies the predetermined period into a plurality of periods based on the impedance measured by the impedance sensor, information on the number of living cells, and a time at which the impedance is measured, and stores the coefficient for each of the classified periods.
3. The cell inspection apparatus according to claim 1 or 2,
the living cell number estimating unit performs principal component analysis on the impedance, the information on the number of living cells, and an elapsed time from the start time of the predetermined period when the impedance and the information on the number of living cells are measured during a period from the start of culture to death of the cells,
the living cell number estimating unit classifies the predetermined period into a plurality of groups based on the result of the principal component analysis,
the living cell number estimating unit obtains the coefficient for each period of the classification based on a result of the principal component analysis,
the living cell number estimating unit stores the coefficient obtained for each period of the classification in the storage unit.
4. The cell inspection apparatus according to claim 3, wherein the living cell number estimating unit obtains the coefficient using at least one of a linear regression, a partial least squares regression, and a quadratic or more regression.
5. The cell inspection apparatus according to claim 3 or 4, wherein the living cell number estimating unit increases the elapsed time by a predetermined time, obtains a difference or ratio between the change in impedance and the change in the number of living cells, obtains a case where the difference or ratio is outside a predetermined range as a divergence point between an nth (n is an integer of 1 or more) period and an n +1 th period, and classifies the nth period based on the obtained divergence point.
6. The cell inspection apparatus according to any one of claims 1 to 5, wherein the living cell number estimation unit obtains boundary information including a bifurcation point in the plurality of periods and a capacitance of the impedance at the bifurcation point, and stores the boundary information in the storage unit.
7. The cell inspection apparatus according to claim 6,
the living cell number estimating unit determines which period the acquired capacitance of the impedance corresponds to based on the boundary information stored in the storage unit,
the living cell number estimating unit estimates the number of living cells using the acquired impedance and the coefficient corresponding to the determined period.
8. A cell inspection method in a cell inspection apparatus, the cell inspection apparatus comprising: an impedance sensor; a storage unit for storing a coefficient for estimating the number of living cells that have survived in the culture medium during a predetermined period from the start of cell culture to the death of the cells, using the impedance measured by the impedance sensor for each of the plurality of classified periods; and a living cell number estimating unit,
the cell inspection method is characterized in that,
measuring the impedance of the culture solution during the predetermined period by using the impedance sensor,
the impedance is acquired by the living cell number estimating unit, and the number of living cells is estimated using at least one of the acquired impedance and the coefficient for each of the periods stored in the storage unit.
9. A program executed by a computer of a cell inspection apparatus having: an impedance sensor; a storage unit for storing a coefficient for estimating the number of living cells that have survived in the culture medium during a predetermined period from the start of cell culture to the death of the cells, using the impedance measured by the impedance sensor for each of the plurality of classified periods; and a living cell number estimating unit,
the program is characterized in that it is,
the computer of the cell inspection apparatus acquires the impedance of the culture solution measured by the impedance sensor by executing the program, and estimates the number of living cells using at least one of the acquired impedance and the coefficient for each of the periods stored in the storage unit.
10. A computer-readable recording medium characterized in that the program according to claim 9 is recorded.
Technical Field
The invention relates to a cell inspection apparatus, a cell inspection method, a program, and a recording medium.
This application claims priority to patent application 2018-005648 filed in japan on day 17 of month 1, 2018, the contents of which are incorporated herein by reference.
Background
In cell culture, the linear relation between the number of living cells and the capacitance is strong in the first half of culture when the survival rate is high. Therefore, a method of measuring the number of living cells in a vessel during culture using an impedance sensor can be generally used for cell culture in a liquid. However, although this method is consistent in the first half of the culture where a high survival rate is maintained, the linear relationship may disappear in the second half of the culture where the survival rate decreases, and the estimated value may deviate from the actual measured value of the sample.
For the case of a deviation between the estimated value and the actual measured value of the sample, the following method is proposed: the data acquired by the impedance sensor is processed by software in a computer to correct the data to obtain an estimated value (see, for example, patent document 1).
In the technique described in patent document 1, the volume of living cells during culture is used as correct data. Viable cell volume was measured by flow cytometry (flow cytometry). In the culture data used in the technique described in patent document 1, the volume of living cells and the estimated value of capacitance data from each acquired frequency are the same in the first half of the culture, but are different in the second half of the culture.
Therefore, in the technique described in patent document 1, data acquired by the impedance sensor is divided into capacitance data for each frequency, and the data is normalized by using the maximum value and the minimum value. In the technique described in patent document 1, a correction value for the volume of living cells is obtained from the ratio of the area of the integral of the first half and the area of the integral of the second half, with the frequency based on the deviation divided into the first half and the second half, the frequency being taken as the abscissa, and the normalized value being taken as the ordinate. The technique described in patent document 1 is applied to culture of cells in which the amount of deviation between the measured value and the estimated value and the area ratio have a correlation, using data acquired by an impedance sensor.
Disclosure of Invention
One aspect of the present invention provides a cell inspection apparatus, a cell inspection method, a program, and a recording medium, which can estimate the number of living cells with high accuracy during a predetermined period of a culture period from inoculation to death until a certain survival rate or less in cell culture.
A cell inspection device (2) according to one embodiment of the present invention includes: an impedance sensor (
In the cell inspection device according to one aspect of the present invention, the storage unit may classify the predetermined period into a plurality of periods based on the impedance measured by the impedance sensor, information on the number of living cells, and a time at which the impedance is measured, and store the coefficient for each of the classified periods.
In the cell inspection device according to one aspect of the present invention, the living cell number estimation unit may perform principal component analysis on the impedance, the information on the number of living cells, and an elapsed time from a start point of the predetermined period when the impedance and the information on the number of living cells are measured during a period from a start of culture of the cells to death of the cells, the living cell number estimation unit may classify the predetermined period into a plurality of groups based on a result of performing the principal component analysis, the living cell number estimation unit may obtain the coefficient for each of the classified groups based on a result of performing the principal component analysis, and the living cell number estimation unit may store the coefficient obtained for each of the classified groups in the storage unit.
In the cell inspection apparatus according to one aspect of the present invention, the living cell number estimating unit may calculate the coefficient using at least one of a linear regression, a partial least squares regression, and a quadratic or more regression.
In the cell inspection apparatus according to one aspect of the present invention, the living cell number estimating unit may increase the elapsed time by a predetermined time, calculate a difference or ratio between the change in impedance and the change in the number of living cells, calculate a difference or ratio between the difference or ratio outside a predetermined range as a divergence point between an nth (n is an integer equal to or greater than 1) period and an n +1 th period, and classify the nth period based on the calculated divergence point.
In the cell inspection apparatus according to one aspect of the present invention, the living cell number estimation unit may obtain boundary information including a bifurcation point between the plurality of periods and a capacitance of the impedance at the bifurcation point, and store the boundary information in the storage unit.
In the cell inspection apparatus according to one aspect of the present invention, the living cell number estimating unit may determine which period the acquired capacitance of the impedance corresponds to based on the boundary information stored in the storage unit, and the living cell number estimating unit may estimate the number of living cells using the acquired impedance and the coefficient corresponding to the determined period.
A cell inspection method according to an aspect of the present invention is a cell inspection method in a cell inspection apparatus including: an impedance sensor; a storage unit for storing a coefficient for estimating the number of living cells that have survived in the culture medium during a predetermined period from the start of cell culture to the death of the cells, using the impedance measured by the impedance sensor for each of the plurality of classified periods; and a living cell number estimating unit that measures an impedance of the culture solution during the predetermined period by using the impedance sensor, acquires the impedance by using the living cell number estimating unit, and estimates a number of living cells by using at least one of the acquired impedance and a coefficient for each of the periods stored in the storage unit.
One aspect of the present invention provides a program for a cell inspection apparatus including: an impedance sensor; a storage unit for storing a coefficient for estimating the number of living cells that have survived in the culture medium during a predetermined period from the start of cell culture to the death of the cells, using the impedance measured by the impedance sensor for each of the plurality of classified periods; and a living cell number estimating unit that causes a computer of the cell inspection apparatus to acquire an impedance of the culture solution measured by the impedance sensor and estimate the number of living cells using at least one of the acquired impedance and a coefficient for each of the periods stored in the storage unit.
A recording medium according to an aspect of the present invention is a computer-readable program recording medium having the program recorded thereon.
According to one aspect of the present invention, the number of living cells can be estimated with high accuracy in a predetermined period of a culture period from inoculation to death to a survival rate or less in cell culture.
Drawings
FIG. 1 is a diagram showing an example of the configuration of the cell inspection system according to the present embodiment.
Fig. 2 is a diagram showing an example of the results of the principal component analysis of capacitance and viable cell density according to the present embodiment.
Fig. 3 is an example of classifying the results of the principal component analysis shown in fig. 2 into three transition states.
Fig. 4A is a diagram showing the result of principal component analysis in the first transition state I of fig. 2.
Fig. 4B is a diagram showing the result of principal component analysis in the second transition state II of fig. 2.
Fig. 4C is a diagram showing the result of principal component analysis in the third transition state III of fig. 2.
Fig. 5 is a flowchart showing an example of the processing procedure performed by the living cell number estimating unit in the learning operation mode according to the present embodiment.
Fig. 6 is a diagram showing an example of information stored in the storage unit according to the present embodiment.
Fig. 7 is a flowchart showing an example of processing procedures performed by the living cell number estimating unit in estimating the operation mode according to the present embodiment.
Fig. 8 is a diagram showing an example of the result of estimating the number of living cells after learning according to the present embodiment.
FIG. 9 is a diagram showing an example of the number of live cells estimated over the entire period of culture according to the present embodiment and the number of live cells actually measured off-line.
Fig. 10 is a diagram showing an example of a deviation between the estimated value and the measured value in the comparative example.
Detailed Description
Embodiments of the present invention will be described below with reference to the drawings.
Fig. 1 is a diagram showing an example of the configuration of a cell inspection system 1 according to the present embodiment. As shown in fig. 1, the cell inspection system 1 includes: a cell inspection apparatus 2, a culture tank 3, and a cell counter 4. The cell inspection apparatus 2 includes: a
The
The culture vessel 3 contains a culture medium containing cells. The culture tank 3 is provided with a
The Cell counter 4 samples the culture solution, stains living cells, and measures the living Cell Density (VCD; visual Cell Density) [ cells/mL ] in the culture solution. The timing at which the
In the learning operation mode, the cell inspection apparatus 2 acquires information indicating the density of living cells measured by the cell counter 4 and a measurement value measured by the
The cell inspection device 2 acquires the measurement value measured by the
The probe 21 includes a measurement probe and an impedance change sensor. The impedance change sensor is a sensor for measuring the impedance of a culture solution containing cells. The measured value measured by the impedance change sensor includes impedances at a plurality of different frequencies within a predetermined frequency range (for example, a range of 10kHz to 100 MHz). The probe 21 measures the impedance of the culture solution, and outputs the measured value to the sensor unit 22. The probe 21 may have an amplifier for amplifying a measured value to be measured.
The sensor unit 22 separates the measurement value output from the probe 21 into two components, i.e., a capacitance (dielectric constant) and an induced conductance (conductivity), by a known method, and outputs the separated capacitance and conductance to the control unit 24. The sensor unit 22 may have an amplifier unit for amplifying a measurement value measured by the probe. The capacitor contains a frequency component of 10kHz to 100MHz, for example.
The operation unit 23 is, for example, a touch panel sensor, an operation button, or the like provided on the
The control unit 24 outputs information indicating that the operation mode is the learning operation mode or information indicating that the operation mode is the estimation operation mode to the operation unit 26 based on the operation result output from the operation unit 23. The controller 24 controls the stirrer, the heater, and the like of the culture tank 3 in accordance with the culture start instruction and the set values of the culture conditions to start the culture. The controller 24 starts the time measurement by using the timer 241 when the culture is started. In the learning operation mode and the estimation operation mode, the control unit 24 acquires the measurement values output from the sensor unit 22 at predetermined time intervals by online processing. In the learning operation mode and the estimation operation mode, the control unit 24 outputs information indicating the elapsed time when the measurement value is acquired, to the calculation unit 26 in association with the measurement value. The control unit 24 may store the number of living cells, the measured value, and the elapsed time in the storage unit 27. In the learning operation mode, the control unit 24 acquires information indicating the density of living cells output from the cell counter 4 at each timing of acquiring the measurement value of the sensor unit 22 by off-line processing. In the learning operation mode, the control unit 24 outputs information indicating the elapsed time when the viable cell density was acquired, to the calculation unit 26 in association with the viable cell density. The control unit 24 may store the number of living cells and the elapsed time in the storage unit 27. In the estimation of the operation mode, the control unit 24 acquires information indicating the estimated number of living cells output from the calculation unit 26. The information indicating the estimated number of living cells output from the calculation unit 26 includes the elapsed time from the start of culture. The control unit 24 generates an image of the estimation result of the number of cells based on the information indicating the estimated number of living cells output from the calculation unit 26, and displays the generated image of the estimation result of the number of cells on the
The
In the learning operation mode, the classification unit 261 of the calculation unit 26 acquires, in real time, the measurement value output by the control unit 24 and information indicating the elapsed time measured when the measurement value was acquired, from the start of culture to the end of culture. In the learning operation mode, the classification unit 261 acquires information indicating the density of living cells and information indicating the elapsed time measured when the information indicating the density of living cells is acquired, by offline processing at the intermittent time. In the learning operation mode, the classification unit 261 stores the acquired measurement value, the viable cell density, and information indicating the elapsed time during which the measurement value and the viable cell density were acquired in the storage unit 27 in association with each other during a period from the start of culture to the end of culture. In learning the operation pattern, the classification unit 261 performs principal component analysis using the stored measurement value of the period from the start of culture to the end of culture, the viable cell density, and information indicating the elapsed time. The classification unit 261 uses the capacitance value among the measurement values. Therefore, the number of data used by the classification unit 261 for principal component analysis is 18 (frequency) × 1 lot if 18 points are in the frequency range 287kHz to 20MHz, for example. Batch 1 refers to the period from the start of culture to the end of culture. In learning the operation pattern, the classification unit 261 obtains boundary information for classifying the measured value and the number of living cells based on the result of the principal component analysis. In learning the operation mode, the classification unit 261 classifies the results of the principal component analysis based on the obtained boundary information, and obtains information indicating the measured value and the viable cell density for each transition state included in the classification. In the learning operation mode, the classification unit 261 stores the obtained boundary information, the measured value included in each transition state, and the information indicating the density of living cells in the storage unit 27. In learning the operation pattern, the estimation unit 262 of the calculation unit 26 calculates coefficients for each transition state of the classification by linear regression, Partial Least Squares regression (PLS), or regression of two or more degrees, for example, with respect to the relationship between the main component of the capacitance and the density of living cells based on the obtained boundary information, and stores the calculated coefficients in the storage unit 27.
When estimating the operation mode, the estimation unit 262 of the calculation unit 26 acquires the measurement value output by the control unit 24 and information indicating the elapsed time measured when the measurement value was acquired, during the period from the start of culture to the end of culture. When estimating the operation mode, the estimation unit 262 determines which transition state the acquired capacitance corresponds to, with respect to the value of the capacitance among the measured values, using the boundary information stored in the storage unit 27. The state of the transition will be described later. When estimating the operation mode, the estimation unit 262 estimates the number of living cells by correcting the measured capacitance using the coefficient based on the result of the determination. The estimation unit 262 outputs information indicating the estimated number of living cells to the control unit 24 in association with information indicating the elapsed time measured when the measurement value was acquired.
The storage unit 27 stores the cell names, the set values of the culture conditions, and the like. In the learning operation mode, the storage unit 27 stores the measured value, the density of living cells, the elapsed time, and the like. The storage unit 27 stores boundary information, coefficients of the boundaries, and the like, which are determined when the operation pattern is learned. The storage unit 27 stores a threshold value of the separation probability r used when classifying the transition state. The separation probability r will be explained later.
< learning operation mode >
Next, an example of processing performed by the classification unit 261 and the estimation unit 262 during learning operation mode will be described with reference to fig. 2 to 4C. The ratio of the measured value of capacitance and the relationship between the viable cell density and the variation in capacitance for each frequency were varied simultaneously, assuming that the period of time during which no variation was made was stable. That is, when the capacitance curve is within a certain range, a certain relationship between the viable cell density and the capacitance is established, and therefore, the detection can be performed by a single method. When the relationship between the density of living cells and the capacitance changes, the ratio of the capacitance per frequency changes, and the change can be detected. Based on this assumption, the entire period of the incubation is divided into stable transition states of the capacitance, and an estimation method and a coefficient can be determined for each of the divided transition states.
Fig. 2 is a diagram showing an example of the results of the principal component analysis of capacitance and viable cell density according to the present embodiment. In fig. 2, the horizontal axis represents the first principal component, and the vertical axis represents the second principal component. The shading of each point indicates the elapsed time during the culture, and the lighter the elapsed time, the shorter the elapsed time, i.e., the first half of the culture. Among the densities of the respective points, the thicker the thickness, the longer the elapsed time, i.e., the latter half of the culture.
As shown in fig. 2, the results of the principal component analysis are regarded as triangles surrounded by the substantially straight lines of reference numerals g1 to g 3. In the case of such principal component analysis results, it is possible to mathematically judge to which side of reference numerals g1 to g3 the data point belongs.
When learning the operation pattern, the classification unit 261 classifies the principal component analysis result into three transition states using, for example, logistic regression. The classification unit 261 prepares capacitance data and positive de-labels "first transition state I", "second transition state II", and "third transition state III" at each time. The classification unit 261 obtains a boundary vector between the labels by regression. The data represented by the substantially straight line shown by reference character g1 is the first transition state I. The data represented by the substantially straight line shown by reference character g2 is the second transition state II. The data represented by the substantially straight line indicated by reference character g3 is the third transition state III.
The first transition state I, the second transition state II, and an example of the determination method will be described. In the following description, the time at which the culture starts is 0, and the elapsed time is t. The classification unit 261 obtains the ratio of the change in viable cell density from time 0 to t to the change in capacitance [ { viable cell density (t) -viable cell density (0) }/{ capacitance (t) -capacitance (0) }. The measurement interval is, for example, 1 hour, and the classification unit 261 increases the time t every 1 hour to sequentially obtain the ratios. That is, the classification unit 261 obtains a ratio after an elapsed time of 1 hour, a ratio after 2 hours, a ratio after 3 hours, and the like. The classification unit 261 obtains a difference or ratio between the ratio at the time when the elapsed time is 1 hour and the ratio at the time when the elapsed time is 2 hours, and determines whether the difference is within a predetermined range or whether the ratio is within a predetermined range. When the difference or the ratio is within the predetermined range, the classification unit 261 determines that the linear relationship continues. When the difference or ratio is outside the predetermined range, the classification unit 261 determines that the linear relationship is not continued and determines that the difference or ratio is a branch point between the first transition state I and the second transition state II.
The classification unit 261 determines the ratio of the change in the number of living cells from the starting point (bifurcation point) to time t to the change in the second component in the main component analysis, using the bifurcation point as the starting point of the second transition state II. The classification unit 261 obtains a difference or ratio between the ratio when the elapsed time is 1 hour and the ratio when the elapsed time is 2 hours, and determines whether the difference is within a predetermined range or whether the ratio is within a predetermined range. When the difference or ratio is outside the predetermined range, the classification unit 261 determines that the linear relationship is not continued and determines that the difference or ratio is a branch point between the second transition state II and the third transition state III. The method of calculating the bifurcation point is an example, and is not limited thereto. When the transition state is four or more, the classification unit 261 further obtains a divergence point between the nth transition state and the (n + 1) th transition state.
The boundary information will be explained.
The boundary information includes the above-described bifurcation point and the capacitance included in the bifurcation point. The classification unit 261 determines the capacitance included in the branch point to determine the transition state. The classification unit 261 obtains a coefficient for each transition state. When the transition state is three types, the coefficients of the transition state include a first coefficient used in the first transition state I, a second coefficient used in the second transition state II, and a third coefficient used in the third transition state. The classification unit 261 classifies the transition state using the separation probability r. The classification unit 261 calculates a separation probability r for each data based on the coefficient and the capacitance. For example, when the two transition states are classified, the classification unit 261 determines the first transition state I and the second transition state II as follows: when the separation probability r is less than 0.5, it is determined as the first transition state I, and when the separation probability r is 0.5 or more, it is determined as the second transition state II.
Fig. 3 is a result of separating transition states using a coefficient for determination.
If the transition state separation is performed on the data points shown in fig. 2 based on the reference numerals g1 to g3, it is as shown in fig. 3.
Fig. 3 is an example of classifying the results of the principal component analysis shown in fig. 2 into three transition states. In fig. 3, the horizontal axis represents the first principal component, and the vertical axis represents the second principal component.
In the example shown in fig. 3, the data points are classified into three transition states, a first transition state I, a second transition state II, and a third transition state III, by line segments g11, g12, and g 13.
Next, the classification unit 261 confirms the transition state after the division and the relationship between the viable cell density and the capacitance during the culture period in each transition state. As shown in fig. 3, the three transition states are divided, and the relationship between the main component and the viable cell density due to the capacitance in each transition state is shown in fig. 4A to 4C.
Fig. 4A to 4C are diagrams in which the results of the principal component analysis in fig. 2 are divided into three transition states. Fig. 4A is a principal component analysis result in the first transition state I. The horizontal axis represents the first principal component, and the vertical axis represents the viable cell density [ cells/mL ].
Fig. 4B is the principal component analysis result in the second transition state II. The horizontal axis represents the second principal component, and the vertical axis represents the viable cell density [ cells/mL ]. Fig. 4C is the principal component analysis result in the third transition state III. The horizontal axis represents the first principal component, and the vertical axis represents the viable cell density [ cells/mL ].
The estimating unit 262 obtains the relationship between the capacitance and the number of living cells by, for example, linear regression or partial least squares regression of the first principal component with respect to fig. 4A. The estimating unit 262 obtains the relationship between the capacitance and the number of living cells by, for example, linear regression or partial least squares regression of the second principal component with respect to fig. 4B. The estimating unit 262 obtains the relationship between the capacitance and the number of living cells by, for example, two or more regressions of the first principal component with respect to fig. 4C. The estimation unit 262 stores the obtained coefficient in the storage unit 27 in association with the transition state.
Here, the main component is a frequency component of the capacitance. For example, in the case where the frequency range of the capacitor is 287kHz to 20MHz, the frequency of the first principal component is 600kHz, and the frequency of the second principal component is 287kHz to 20 MHz.
The estimating unit 262 obtains the coefficient of the first transition state I by linear regression using a first coefficient for determining the first transition state I and the second transition state II, for example, the capacitance and the viable cell density of 600 kHz.
The estimating unit 262 obtains the coefficient of the second transition state II by partial least squares regression using the second coefficient for determining the second transition state II and the third transition state III, for example, the capacitance of 287kHz to 20MHz and the viable cell density.
The estimating unit 262 obtains the coefficient of the third transition state III by quadratic regression using the second coefficient for determining the second transition state II and the third transition state III, for example, the capacitance of 600kHz and the viable cell density.
The method of obtaining the coefficient is an example, but is not limited thereto.
In the examples shown in fig. 2 to 4C, the results of the principal component analysis are divided into three types, but the present invention is not limited to this. The number of divisions may be two or four or more based on the result obtained by the principal component analysis. The estimation unit 262 may calculate the coefficient for each transition state of the division.
Fig. 5 is a flowchart showing an example of the processing procedure performed by the living cell number estimating unit 28 in the learning operation mode according to the present embodiment. Fig. 5 is an example of classifying the results of principal component analysis into n (n is an integer of 2 or more).
(step S1) the control unit 24 measures the elapsed time from the start of culture when the measured value is acquired by acquiring the measured value output by the sensor unit 22 through on-line processing. Next, the control unit 24 outputs information indicating the measured value and the elapsed time to the calculation unit 26. Next, the classification unit 261 acquires information indicating the measurement value and the elapsed time output by the control unit 24. Next, the classification unit 261 stores the acquired information indicating the measured value and the elapsed time in the storage unit 27.
(step S2) the control unit 24 determines whether or not the measurement is completed. For example, the operation unit 23 detects that the measurement is completed by the operator, and the control unit 24 completes the measurement. If the control unit 24 determines that the measurement has not been completed (step S2; no), the process returns to step S1. When determining that the measurement is completed (step S2; yes), the control unit 24 advances the process to step S3.
(step S3) the control unit 24 acquires information indicating the viable cell density in the culture solution from the start to the end of the culture by offline processing at each timing measured by the
(step S4) the classification unit 261 performs principal component analysis using the measured value of the period from the start of culture to the end of culture, the viable cell density, and the elapsed time stored in the storage unit 27.
(step S5) the classification unit 261 obtains boundary information for classifying the measurement value and the viable cell density, for example, using logistic regression, based on the result of the principal component analysis. Next, the classification unit 261 classifies the results of the principal component analysis based on the obtained boundary information, and obtains information indicating the measured value and the viable cell density for each transition state included in the classification.
Based on the obtained boundary information, the estimation unit 262 obtains a coefficient for each transition state of the classification, for example, by linear regression, partial least squares regression, or regression of two or more orders, with respect to the relationship between the main component of the capacitance and the density of living cells (step S6). Next, the estimation unit 262 stores the obtained coefficient in the storage unit 27.
In this way, the learning operation mode is ended.
In the above example, the example in which the viable cell density is obtained by the off-line process is described, but it may be obtained by the on-line process. The cell inspection apparatus 2 may perform the learning by performing the learning according to the cells or culture conditions. In this case, the cell inspection apparatus 2 may store the boundary information and the coefficient in the storage unit 27 in association with each other according to the cells or the culture conditions.
In the above example, the example in which the coefficient is obtained according to the transition state has been described, but a relational expression may be used.
In the above example, the description has been given of the case where the living cell density is measured by the cell sensor in the learning operation mode, but the present invention is not limited to this. The living cell density or the number of living cells may be measured by other sensors or measuring instruments.
In fig. 2 to 5, an example in which the period from the start of cell culture to death is classified into a plurality of transition states is described, but the present invention is not limited thereto. The classification unit 261 may perform principal component analysis at any time during the cell culture period to classify the cells into transition states. In this case, the estimating unit 262 may calculate the coefficient for each of the classified transition states at any time during the cell culture period.
Fig. 5 illustrates an example in which the principal component analysis is performed after the measurement is completed and the coefficient is obtained for each transition state (period) of the classification, but the present invention is not limited to this. The classification unit 261 may analyze the principal component of the measurement values acquired from the sensor unit 22 and the cell counter 4 from the time when the culture is started to the elapsed time or from the start time of the predetermined period to the elapsed time, for example, at each timing when the measurement values are acquired, and estimate the coefficients. When switching to a transition state with a different measured value, the classification unit 261 stores the final coefficient of the transition state before switching in the storage unit 27, and estimates the coefficient of the new transition state after initializing the coefficient. That is, the classification unit 261 may classify the transition state and estimate the coefficient on line.
The estimation unit 262 may estimate the number of living cells using the transition state classified online and the estimated coefficient as described above.
In the above example, the number of living cells is estimated using one of the coefficients set for each transition state, but the present invention is not limited to this. For example, the estimation unit 262 may estimate the number of living cells by multiplying the coefficient of the first transition state by 60% and multiplying the coefficient of the second transition state by 40% in switching between the first transition state I and the second transition state II. Thus, according to the present embodiment, the transition state can be continuously switched between the transition state and the transition state.
An example of information stored in the storage unit 27 according to the learning operation pattern will be described.
Fig. 6 is a diagram showing an example of information stored in the storage unit 27 according to the present embodiment. As shown in fig. 6, the storage unit 27 stores coefficients for estimating the number of living cells in accordance with the state of transition.
< estimated operation mode >
Next, an example of processing performed by the estimated operation mode estimating unit 262 will be described with reference to fig. 7. Fig. 7 is a flowchart showing an example of processing procedures performed by the living cell number estimating unit 28 in estimating the operation mode according to the present embodiment. Fig. 7 is an example of classifying the result of principal component analysis into n (n is an integer of 2 or more) in the learning operation mode.
(step S101) the control unit 24 acquires the measurement value output from the sensor unit 22 by on-line processing, and counts the elapsed time from the start of culture when the measurement value is acquired. Next, the control unit 24 outputs information indicating the measured value and the elapsed time to the calculation unit 26. Next, the classification unit 261 acquires information indicating the measurement value and the elapsed time output by the control unit 24. The classification unit 261 may store the acquired information indicating the measurement value and the elapsed time in the storage unit 27.
(step S102) the estimating unit 262 determines which transition state the acquired capacitance corresponds to, with respect to the value of the capacitance among the measured values, using the boundary information stored in the storage unit 27.
(step S103) the estimating unit 262 determines whether the capacitance belongs to the first transition state, the second transition state,. cndot.. cndot., and the n-th transition state, as a result of the determination. When determining that the capacitance belongs to the first transition state (step S103; first transition state), the estimating unit 262 advances the process to step S104. When determining that the capacitance belongs to the second transition state (step S104; second transition state), the estimating unit 262 advances the process to step S105. When determining that the capacitance belongs to the n-th transition state (step S103; n-th transition state), the estimating unit 262 advances the process to step S106.
(step S104) the estimation unit 262 corrects the measured value using the coefficient of the first transition state stored in the storage unit 27 to estimate the number of living cells. For example, the estimation unit 262 multiplies the capacitance of the measurement value by a coefficient to obtain an estimation value of the number of living cells. In the present embodiment, the number of living cells is estimated by correcting the measured value using the coefficient of the first transition state stored in the storage unit 27 as described above, and this is referred to as a first estimation method. After the processing, the estimating unit 262 advances the processing to step S107.
(step S105) the estimation unit 262 corrects the measured value using the coefficient of the second transition state stored in the storage unit 27 to estimate the number of living cells. In the present embodiment, the number of living cells is estimated by correcting the measured value using the coefficient of the second transition state stored in the storage unit 27 as described above, and this is referred to as a second estimation method. After the processing, the estimating unit 262 advances the processing to step S107.
(step S106) the estimation unit 262 corrects the measured value using the coefficient of the n-th transition state stored in the storage unit 27 to estimate the number of living cells. In the present embodiment, the number of living cells is estimated by correcting the measured value using the coefficient of the nth transition state stored in the storage unit 27 as described above, and this is referred to as an nth estimation method. After the processing, the estimating unit 262 advances the processing to step S107.
(step S107) the estimating unit 262 outputs information indicating the estimated number of living cells to the control unit 24 in association with information indicating the elapsed time. Next, the control unit 24 generates image information using the information indicating the estimated number of living cells and the information indicating the elapsed time output from the estimation unit 262, and causes the
(step S108) the control unit 24 determines whether or not the measurement is completed. The operation unit 23 detects that the measurement is completed by the operator, and the control unit 24 completes the measurement, for example. When determining that the measurement has not been completed (step S108; no), the control unit 24 returns the process to step S101. When determining that the measurement is completed (step S108; yes), the control unit 24 terminates the process.
The estimation unit 262 can estimate the number of living cells using the coefficients for each transition state classified for any period of the cell culture period.
The capacitance used for estimation may use a range of frequencies or data of frequencies suitable for the transition state. The frequency range or frequency used may be selected to best balance linearity and stability in the measured frequency.
Next, an example of the results of estimation after learning in the above-described manner will be described.
Fig. 8 is a diagram showing an example of the result of estimating the number of living cells after learning in the present embodiment. In FIG. 8, the horizontal axis represents time, and the left vertical axis represents the estimated viable cell density [ cells/mL ]. The value of the time on the horizontal axis is the number of acquired data points of elapsed time from the start of measurement, and if converted to units of date and time, 2000[ points ] ≈ 56[ hours ]. The right vertical axis 1 indicates the first transition state, the right vertical axis 2 indicates the second transition state, and the right vertical axis 3 indicates the third transition state.
The triangle marks indicate the measured viable cell density. The broken line g21 represents the density of viable cells as estimated by the first estimation method. The chain line g22 represents the density of viable cells estimated by the second estimation method. The solid line g23 represents the density of viable cells estimated by the third estimation method. The broken line g24 represents the state of transition of the classification.
As shown by reference character g24 in fig. 8, the determination result between the first transition state and the second transition state and the determination result between the second transition state and the third transition state have a period during which both transition states vibrate, but the estimated values are close to each other, and therefore, the influence of the determination on the estimated result is small regardless of which transition state is determined.
Next, an example of the number of living cells estimated during the entire period of culture by applying the present embodiment will be described.
FIG. 9 is a diagram showing an example of the number of living cells estimated throughout the culture period of the present embodiment and the number of living cells actually measured in an off-line state. In FIG. 9, the horizontal axis represents time, and the vertical axis represents the measured value and the estimated value [ cells/mL ] of the viable cell density. The value of the time on the horizontal axis is the number of acquired data points of elapsed time from the start of measurement, and if converted to units of date and time, 2000[ points ] ≈ 56[ hours ]. Reference character g31 is an estimated value, and reference character g32 is an actual value.
In fig. 9, the estimated value has a discontinuous portion, but the transition state is switched as in fig. 8. As shown in fig. 9, if the measured values are positive solution data, the number of living cells can be estimated with high accuracy over the entire period in the present embodiment.
A comparative example in which the estimated value and the measured value deviate from each other will be described.
In the method of the related art described in patent document 1, the deviation in the second half of the culture may become small and the deviation in the first half of the culture may become large. In contrast, in the comparative example, the correction of the correction formula of the method of the related art described in patent document 1 is performed so as to suppress the correction amount in the first half of the incubation and secure the correction amount in the second half of the incubation. In the case of performing the correction in the above-described manner, although there is a culture state that can be appropriately corrected, as shown in fig. 10, a difference may occur between the estimated value and the actual measured value in the first half of the culture.
Fig. 10 is a diagram showing an example of a deviation between an estimated value and an actual measurement value according to a comparative example. In fig. 10, the horizontal axis represents elapsed time, and the vertical axis represents Viable Cell Density (VCD). The triangular markers are the measured values of viable cell density. The reference character g41 is an estimated value in the first half of the culture in the comparative example. The reference character g42 is an estimated value in the second half of the culture in the comparative example.
In the example shown in FIG. 10, the deviation in the first half of the culture occurred relatively low compared to the measured value.
That is, as shown in comparative example, even if the relational expression of the method of the related art described in patent document 1 is corrected, the correspondence of no correction for the first half of the culture and correction for the second half of the culture cannot be performed in a single method.
In contrast, in the present embodiment, a method (estimation using a coefficient) for separating the transition states from the start of culture to the end of culture and appropriately correcting the transition states is applied. Thus, according to the present embodiment, even when the culture conditions are different and the tendency to deviate is different, the number of living cells can be estimated with high accuracy over the entire period from seeding to death to a certain survival rate or less in cell culture (in particular, CHO cells). In addition, in the present embodiment, an optimal estimation method can be automatically selected for each transition state, and an effect is obtained that data other than capacitance data used in estimation is not necessary. In the present embodiment, each estimation method in each transition state of the classification is suitable only for a transition state in which the behavior of the capacitance is regarded as stable, and therefore, it is possible to generally estimate with high accuracy.
In the above example, the period is classified by performing principal component analysis or the like on the capacitance and viable cell density in the impedance measured by the
In the above example, the control unit 24 and the calculation unit 26 are separated, but the control unit 24 and the calculation unit 26 may be integrally configured.
As described above, according to the present embodiment, the number of living cells can be estimated with high accuracy in a predetermined period of a culture period from inoculation to death to a certain survival rate or less in cell culture.
According to the present embodiment, the number of living cells can be estimated with high accuracy even in the second half of the culture in which the number of living cells decreases by classifying the predetermined period into a plurality of transition states, setting a coefficient for each transition state, and estimating the number of living cells using the coefficient.
According to the present embodiment, for example, in switching between the first transition state and the second transition state, since the estimation is performed using the coefficient of the first transition state and the coefficient of the second transition state, the number of living cells can be estimated with high accuracy even in switching between transition states.
In the present embodiment, the coefficient is obtained using at least one of a linear regression, a partial least squares regression, and a quadratic or more regression, and thus the amount of calculation is small.
In the present embodiment, the state of transition is classified by principal component analysis using impedance and the number of live cells actually measured, and therefore the state of transition can be classified with high accuracy.
In the present embodiment, the transition state during actual measurement is determined from the probability distribution calculated for each data based on the coefficient and the capacitance, and therefore, the transition state can be determined with high accuracy and the number of living cells can be estimated with high accuracy.
The whole or a part of the processing executed by the living cell number estimating unit 28 may be performed by recording a program for realizing the whole or a part of the functions of the living cell number estimating unit 28 in a computer-readable recording medium, and causing a computer system to read and execute the program recorded in the recording medium. The term "computer system" as used herein is intended to include hardware such as an OS, peripheral devices, and the like. The "computer system" also includes a WWW system provided with a homepage providing environment (or display environment). The "computer-readable recording medium" refers to a storage device such as a flexible disk, a magneto-optical disk, a removable medium such as a ROM or a CD-ROM, or a hard disk incorporated in a computer system. The "computer-readable recording medium" also includes a recording medium that holds a program for a certain period of time, such as a volatile memory (RAM) in a computer system serving as a server or a client when the program is transmitted via a network such as the internet or a communication line such as a telephone line.
The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. The "transmission medium" for transmitting the program is a medium having a function of transmitting information, such as a network (communication network) such as the internet or a communication line (communication line) such as a telephone line. The above-described program may be a part for realizing the functions. Further, the above function may be realized by a combination with a program already recorded in the computer system, so-called differential file (differential program).
While the embodiments for carrying out the present invention have been described above with reference to the embodiments, the present invention is not limited to the embodiments, and various modifications and substitutions can be made without departing from the spirit and scope of the present invention.
- 上一篇：一种医用注射器针头装配设备
- 下一篇：根据流体阻抗的流体性质确定