JP6798905B2 - Abnormal index calculation device, abnormal index calculation method and abnormal index calculation program - Google Patents

Description

The present technology relates to an abnormality detection device, an abnormality detection method, and an abnormality detection program.

It is thought that nerve cells regulate the functions of various organs in living organisms, and in the central nervous system, they perform various information processing by forming a network of nerves and activating them. At present, with the development of cell culture technology, it has become possible to culture nerve cells extracted from a living body or nerve cells differentiated from pluripotent stem cells in a culture medium for a long period of time.

The activity of nerve cells is to transmit information using electrical signals, such as the mechanism of electrical signal generation in nerve cells and nerve fibers, the biochemical properties of the membrane surrounding nerve cells, and the response to nerve cells and nerve fibers. In order to elucidate the pharmacological action of a substance, the potential difference between the inside and outside of the membrane of this nerve cell is measured.

When cultured nerve cells are targeted for measurement, methods for measuring the action potential of nerve cells include (1) intracellular potential recording method, (2) patch electrode recording method, and (3) extracellular potential recording method. .. Since (1) and (2) are for inserting glass microelectrodes into cells or applying them to the cell surface for recording, it is possible to directly measure the membrane potential, but the procedure is difficult. It is difficult to perform long-term measurements such as damaging the cells themselves. On the other hand, the extracellular potential recording method (3) is to measure the potential change occurring around the cell by arranging the electrode near the cell, which is technically relatively easy and enables long-term measurement. Yes, and its application range is wide. In particular, the extracellular potential recording method using a multi-point planar electrode (MEA: Multiple Electrode Array) is used as a method for stable simultaneous measurement from a plurality of electrodes and a plurality of samples.

Conventionally, as a measuring device used for the extracellular potential recording method, one cell is used so as not to generate noise due to electromagnetic induction generated between signal lines connected to different microelectrodes arranged in different cells or cell groups. Alternatively, the signal line connected to the microelectrodes arranged in the cell group and the signal line connected to the microelectrodes arranged in other cells or cell groups are substantially separated and connected to the amplifier. Those have been proposed (for example, Patent Document 1).

Japanese Unexamined Patent Publication No. 2014-124133

In order to analyze the cell activity pattern from the action potential data obtained by measuring the cell potential change, noise is excluded from the measured cell activity waveform data and the cell activity pattern is detected. In addition, in order to perform drug discovery screening and toxicity evaluation using cells, the difference in activity pattern depending on the presence or absence of physical stimulation, electrical stimulation or chemical stimulation caused by application of candidate samples is evaluated.

However, it has been difficult to detect abnormal potential fluctuations caused by stimulation, for example. In particular, it is difficult to mechanically detect unknown abnormal potential fluctuations, and it is difficult to detect unknown abnormalities from potential fluctuations that occur in new candidate samples for drug discovery screening and the like. ..

Therefore, an object of the present invention is to provide a technique for detecting a fluctuation pattern of a potential different from a predetermined standard from extracellular potential data.

The abnormality detection device according to the present invention can be exemplified as follows. That is, the abnormality detection device sets the potential value measured in the first period and the second period different from the first period with respect to the data representing the time-dependent change of the extracellular potential as the standard in the abnormality detection. Of the data representing the time course of the extracellular potential to be compared with the standard using a prediction model generated by learning the relationship with the measured potential value while changing the first period and the second period. A prediction processing unit that obtains a predicted value of the potential measured in the fourth period corresponding to the second period based on the value of the potential measured in the third period corresponding to the first period, and a fourth It is provided with an abnormality detection unit that calculates an index value according to the magnitude of the difference based on the difference between the measured value of the potential measured in the period and the predicted value.

By doing so, it becomes possible to learn and predict the characteristics of the time-dependent change of the extracellular potential as a standard, and detect whether or not the measured value is different from the standard based on the difference from the predicted value. You will be able to. That is, it is possible to provide a technique for detecting a fluctuation pattern of a potential different from the standard from the extracellular potential data.

In addition, the prediction processing unit uses the potential value measured in the third period or the feature amount based on the potential value as an input, and all the nodes on the input side are connected to all the nodes on the output side to the fully connected layer and each bond. The feature amount or predicted value of the potential measured in the fourth period may be obtained based on the associated predetermined coefficient. For example, in this way, a configuration including a so-called fully connected layer can be adopted.

In addition, the prediction processing unit inputs the potential value measured in the third period or the feature amount based on the potential value, and inputs the feature amount or the predicted value of the potential measured in the fourth period by a predetermined convolutional neural network. You may ask for it. For example, it is possible to capture the characteristics of potential values that change over time.

Further, the prediction processing unit receives a feature amount based on the potential value measured in the third period as an input, and obtains a predicted value of the potential measured in the fourth period by a predetermined deconvolutional neural network. You may. In particular, when the feature amount is generated by convolution from the potential value, the output value can be generated by a symmetric neural network by so-called deconvolution.

In addition, the value of the extracellular potential is measured via a plurality of electrodes arranged so as to be in contact with the cell, and the prediction processing unit is a value of a plurality of potentials measured at the same time by different electrodes or a feature amount based on the value. May be used to obtain the predicted value of the potential. In this way, it is possible to make a prediction based on the synchronic characteristics of the fluctuation pattern of the potential generated at the same time in at least a part of the cell.

In addition, the value of the extracellular potential is measured via one or more electrodes arranged so as to be in contact with the cell, and the prediction processing unit is measured by the same electrode at different time points within the third period. The predicted value of the potential may be obtained by using the value of the potential of or the feature amount based on the potential value of. In this way, it is possible to make a prediction based on the characteristics of the change over time in the fluctuation pattern of the potential detected at each electrode.

The contents described in the means for solving the problems can be combined as much as possible without departing from the problems and technical ideas of the present invention. Further, the content of the means for solving the problem can be provided as a device such as a computer or a system including a plurality of devices, a method executed by the computer, or a program executed by the computer. The program can also be run on the network. Further, a recording medium for holding the program may be provided.

According to the present invention, it is possible to provide a technique for detecting a fluctuation pattern of a potential different from a predetermined standard from extracellular potential data .

It is a perspective view which shows an example of the measuring apparatus. It is a top view which shows an example of a chip. It is a functional block diagram which shows an example of the machine learning apparatus. It is a graph which shows an example of the value of the extracellular potential of the measured nerve cell. It is a device block diagram which shows an example of a computer. It is a processing flow diagram which shows an example of the machine learning processing. It is a processing flow diagram which shows an example of the processing of an operation stage. It is a graph which shows typically the change of the potential value for explaining the prediction of the extracellular potential value and the abnormality detection. It is a figure which shows an example of the structure of a neural network. It is a figure for demonstrating an example of the structure of a convolution layer. It is a figure for demonstrating an example of the structure of a fully connected layer. It is a table which shows the result of the Example which applied a compound to a nerve cell. It is a figure for demonstrating the relationship between the learning (prediction) original period and the learning (prediction) target period. It is a figure for demonstrating the relationship between the learning (prediction) original period and the learning (prediction) target period related to the modification. It is a figure for demonstrating the relationship between the learning (prediction) original period and the learning (prediction) target period related to the modification.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. The embodiment is an example, and the present invention is not limited to the following configuration.

<Device configuration>
FIG. 1 is a perspective view showing an example of a measuring device according to the present embodiment. The measuring device 1 includes a chip holder 11, a potential measuring device 12, a monitor 13, and a machine learning device 14. On the chip holder 11, a chip 111 having a plurality of electrodes for measuring the extracellular potential from the cell group is placed. The cell may be any cell that can acquire an extracellular potential, and examples thereof include nerve cells and muscle cells (for example, cardiomyocytes). The potential measuring device 12 amplifies the potential measured via the electrode of the chip 111 by the built-in amplifier, and measures the difference between the amplified potential and a predetermined reference potential (ground potential). Further, the potential value measured by the potential measuring device 12 is output to the machine learning device 14, and the machine learning device 14 records the measured potential value. At this time, the potentials measured at each of the plurality of electrodes are recorded, for example, in association with the measurement time so that they can be processed in synchronization. Further, the machine learning device 14 learns the characteristics of the change in potential from the measurement data by using DNN (Deep Neural Network), and the future potential from the measurement data by using the prediction model generated by learning the characteristics. Predict changes in. That is, the machine learning device 14 learns the characteristics of the change in potential measured from the cell group in an arbitrary state, which is used as a reference (also referred to as “standard”) in the abnormality detection process. In the present embodiment, the potential change in the normal state without external stimulus is used as "reference data" for the cell group, and the characteristics of the reference data are learned. The external stimulus means, for example, a physical, electrical or chemical stimulus. When applying the candidate sample, of the series of extracellular potential data before and after the application of the candidate sample, the data before the application of the candidate sample may be used as the reference data. Further, when a solvent for dissolving a candidate sample or a plurality of samples or an external stimulus is applied, data having an external stimulus to which a sample other than the sample to be evaluated or the stimulus is applied may be treated as reference data. In addition, the machine learning device 14 compares the value of the potential measured from the cell group in some state, which the user compares with the above-mentioned reference, with the predicted value predicted based on the above-mentioned model. The machine learning device 14 calculates the predicted value using, for example, the extracellular potential value of the cell group to which some candidate sample is applied and the predicted model, and is based on the difference between the predicted value and the measured value of the extracellular potential. , Detect changes that differ from the reference data. The monitor 13 appropriately displays the measured extracellular potential value and predicted value, information on the difference between them, the detected abnormal value, and the like.

FIG. 2 is a plan view showing an example of the chip (probe) 111. Tissue sections are placed or cells are cultured in the central region of chip 111. Further, 16 measuring electrodes 1111 (also simply referred to as “electrodes”) are provided in the central region of the chip 111, and the potential generated in the contact region of the cell group can be measured. In addition, a reference electrode 1112 for measuring the ground potential is provided around the region where the cell group is placed. The measuring electrode 1111 and the reference electrode 1112 are connected to the potential measuring device 12 via the wiring 1113. A plurality of chips 111 as shown in FIG. 2 may be placed on the chip holder 11 so that potential data can be measured from a plurality of cell groups. The number of electrodes may be one or more, and the shape and the like related to the arrangement when a plurality of electrodes are provided are not limited to the example of FIG.

FIG. 3 is a functional block diagram showing an example of the machine learning device 14. The machine learning device 14 of FIG. 3 includes a potential data storage unit 141, a machine learning unit 142, a prediction model storage unit 143, a prediction processing unit 144, a prediction result storage unit 145, a verification processing unit 146, and anomaly detection. A unit 147 is provided. The machine learning device 14 divides the input signal sequence of the continuously measured potentials into predetermined periods (sections), and features the relationship between the potential values in a predetermined number of sections and the potential values in subsequent sections. To learn. Further, based on the difference between the potential value predicted for a certain section and the measured value measured by the potential measuring device 12, an abnormal value different from the change in potential at the reference time is detected.

FIG. 4 is a graph showing an example of the value of the extracellular potential measured by the plurality of measuring electrodes 1111 of the measuring device 1. In the example of FIG. 4, the passage of time is shown in the direction of the arrow along the horizontal axis, and in the vertical direction, a graph showing the potential values detected by the eight measurement electrodes 1111 shows the measurement timing. It is shown to synchronize. Specifically, the potential value is recorded as a sequence of numerical values that are continuously measured at predetermined sampling intervals.

Further, in the present embodiment, the measured value of the extracellular potential in a predetermined number of sections is used for learning as a teacher value with respect to the measured value of the extracellular potential in the subsequent section. For example, the scale of the broken line that divides the waveform of FIG. 4 for each predetermined period on the time axis represents an example of the boundary that divides the graph into predetermined sections. The potential sampling interval measured by the measurement electrode 1111 does not have to coincide with a predetermined section. For example, a plurality of measured values may be acquired within the section divided by the scale in FIG. Good. Further, in the present embodiment, the above-mentioned reference data is used as a training data set used for creating a prediction model and an evaluation data set used for evaluating the created prediction model. In addition, in the operation stage of the created prediction model, a predetermined candidate sample is applied to the cell group, and it is determined whether or not a change different from the reference time appears in the potential value measured before and after the application.

The potential data storage unit 141 stores the potential data as shown in FIG. In the present embodiment, it is assumed that the reference data and the potential measurement data (also referred to as “test data”) after applying the candidate sample to the cell group are stored in the potential data storage unit 141 in advance. The test data is the value of the extracellular potential of the cell in any state for the user to contrast with the reference data described above. Regarding the test data, it is assumed that the identification information for identifying the applied candidate sample and the information indicating the time when the candidate sample is applied are also retained.

The machine learning unit 142 performs machine learning using a neural network (NN: Neural Network) and a training data set. Specifically, the machine learning unit 142 has a feature amount calculation unit 1421 and a learning processing unit 1422. The feature amount calculation unit 1421 performs a process based on the NN having a predetermined configuration on the input potential data, and performs a process of calculating the feature amount of the change in the potential. Specifically, so-called convolution layers (also called "convolution layers" or "convolution NNs"), full connection (FC: Full Connection) layers (also called "full convolution layers" or "full convolution NNs"), and deconvolution. A (deconvolution) layer (also referred to as a "deconvolution layer" or a "deconvolution NN") is used. Then, the characteristic amount of the change with time of the potential data measured by one electrode, the potential data of each electrode measured by a plurality of electrodes, or the potential data measured by at least a part of the plurality of electrodes can be calculated. , Calculate the feature quantity of the co-temporal change of the potential at the same time, which is measured at least a part of the plurality of electrodes. The learning processing unit 1422 updates the value of the coefficient applied in each layer of the NN based on the difference between the teacher value held by the potential data storage unit 141 and the feature amount output by the NN.

The prediction model storage unit 143 stores parameters such as the NN configuration adopted in the present embodiment and the weighting coefficient generated by the machine learning unit 142. The configuration of the NN, parameters, etc. shall be referred to as a "prediction model".

The prediction processing unit 144 predicts future potential changes by using the evaluation data and test data stored in the potential data storage unit 141 and the prediction model stored in the prediction model storage unit 143. The prediction result storage unit 145 stores the predicted value of the potential predicted by the prediction processing unit 144. Further, the verification processing unit 146 compares the result of performing the prediction processing using the evaluation data with the actually measured value, and evaluates the performance of the prediction processing. In addition, the abnormality detection unit 147 compares the predicted value, which is the result of performing the prediction process using the test data, with the measured value, and based on the magnitude of these differences, changes the potential different from the reference time. To detect.

The machine learning unit 142 is a so-called computer, and when the processor executes a predetermined program, it functions as the machine learning unit 142, the prediction processing unit 144, and the verification processing unit 146 described above. Further, the potential data storage unit 141, the prediction model storage unit 143, and the prediction result storage unit 145 are specifically realized by a main storage device or an auxiliary storage device provided in a computer.

<Device configuration>
FIG. 5 is a device configuration diagram showing an example of a computer. The machine learning device 14 is, for example, a computer as shown in FIG. The computer 1000 shown in FIG. 5 includes a CPU (Central Processing Unit) 1001, a main storage device 1002, an auxiliary storage device 1003, and a communication I.
It includes an F (Interface) 1004, an input / output IF (Interface) 1005, a drive device 1006, and a communication bus 1007. The CPU 1001 performs processing and the like according to the present embodiment by executing a program (also referred to as "software" or "application"). The main storage device 1002 caches the programs and data read by the CPU 1001 and expands the work area of the CPU. Specifically, the main storage device is a RAM (Random Access Memory), a ROM (Read Only Memory), or the like. The auxiliary storage device 1003 stores a program executed by the CPU 1001 and setting information used in the present embodiment. Specifically, the auxiliary storage device 1003 is an HDD (Hard-disk Drive), an SSD (Solid State Drive), a flash memory, or the like. The main storage device 1002 and the auxiliary storage device 1003 include a potential data storage unit 141, a prediction model storage unit 143, and a prediction result storage unit 145.
Work as. Although a plurality of storage units are shown in FIG. 3 for convenience of explanation, one storage device or a plurality of storage devices may be physically used. The communication IF1004 transmits / receives data to / from another computer. The machine learning device 14 may receive potential data from a computer (not shown) connected via the communication IF 1004. Specifically, the communication IF1004 is a wired or wireless network card or the like. The input / output IF1005 is connected to an input / output device, receives input from the user, and outputs information to the user. Specifically, the input / output device is a keyboard, a mouse, a display, a touch panel, or the like. The drive device 1006 reads data recorded on a storage medium such as a magnetic disk, a magneto-optical disk, or an optical disk, or writes data to the storage medium. Then, the above-mentioned components are connected by the communication bus 1007. It should be noted that a plurality of these components may be provided, or some components (for example, the drive device 1006) may not be provided. Further, the input / output device may be integrally configured with the computer. Further, a program executed in the present embodiment is provided via a portable storage medium that can be read by the drive device 1006, a portable auxiliary storage device 1003 such as a flash memory, a communication IF 1004, and the like. It may be. Then, the CPU 1001 executes the program to operate the computer as described above as the machine learning device 14 shown in FIG.

A part of the functional block illustrated in FIG. 3 may be shared by a plurality of computers, or a part of a data set to be processed may be processed in parallel by a plurality of computers. Further, the computer may provide a so-called cloud service on the network.

<Machine learning process>
FIG. 6 is a processing flow diagram showing an example of machine learning processing executed by the measuring device 1. First, the measuring device 1 performs a process in the learning stage using the training data among the potential data (S1). In this step, the predicted value is calculated for each predetermined section based on the potential value of the previous section for the potential data using the NN and the predetermined coefficient having the configuration described later, and the calculated predicted value and the section are calculated. The process of updating the coefficient based on the magnitude of the difference from the measured value of is repeated. Specifically, a parameter is obtained so that the sum of the squared values of the difference between the predicted value and the measured value becomes small.

After that, the measuring device 1 performs the processing in the model evaluation stage using the evaluation data among the potential data (S2). In this step, the predicted value of the potential is calculated for each predetermined section with respect to the potential data using the coefficients obtained in NN and S1 having the configuration described later, and it is evaluated whether or not the value sufficiently matches the teacher value. According to the prediction model evaluated to be in good agreement with the teacher value in this step, it can be expected that future potential changes can be predicted with the same performance for similar cells.

In addition, the measuring device 1 uses the test data of the potential data to perform processing in the operation stage (S3). The processing in the operation stage will be described in detail with reference to FIG.

FIG. 7 is a processing flow diagram showing an example of processing in the operation stage executed by the measuring device 1. The abnormality detection unit 147 first calculates a predicted value of test data using the coefficients obtained by NN and S1 having a predetermined configuration (S11). In this step, the potential value of the subsequent section is predicted by using the measured value of the section in the previous stage and the prediction model described above.

Further, the abnormality detection unit 147 calculates the difference between the predicted value and the measured value (S12). The difference in this step is calculated as, for example, the squared value of the difference between the predicted value and the measured value, which is averaged for each predetermined section (also referred to as “index value”).

The index value is not limited to the above example, and a value corresponding to the magnitude of the difference between the predicted value and the measured value can be adopted. As the index value, for example, the absolute value of the difference between the predicted value and the measured value may be calculated by averaging the values for each predetermined section. Further, a value obtained by averaging the power of the absolute value of the difference between the predicted value and the measured value for each predetermined section may be used. By adopting exponentiation, the degree of discrepancy between the predicted value and the measured value becomes clear.

Then, the abnormality detection unit 147 detects a change in potential different from the reference time based on the magnitude of the difference (S13). In this step, the relative magnitude of the index value in the section after the application of the candidate sample is evaluated based on the index value in the section before the application of the candidate sample. Specifically, it is normalized by dividing the index value in the section after application of the candidate sample by the index value in the section before application of the candidate sample. That is, it can be said that the closer the index value after normalization is to 1.00, the less the candidate sample is affected by the potential value measured before and after the application of the candidate sample. On the other hand, it can be detected that the larger or smaller the index value after normalization is based on 1.00, the greater the influence on the measured potential value of the application of the candidate sample. In S13, the abnormality detection unit 147 detects whether or not the index value after normalization based on the difference between the predicted value and the measured value is an abnormal value different from the reference value, for example, by comparing with a predetermined threshold value. As described above, the index value after normalization can be said to be a value indicating the degree of difference between the test data and the reference data. It should be noted that the degree of deviation from the standard is obtained even if the measured value of the extracellular potential related to the test data is not classified as abnormal or not, and the index value according to the magnitude of the difference between the test data and the reference data is calculated. It is useful because you can know.

FIG. 8 is a graph schematically showing changes in the potential value for explaining the prediction of the extracellular potential value and the detection of abnormality. 8 (1) to 8 (3) are waveforms showing potential values, respectively, and show changes in values with the passage of time toward the right. Further, the vertical broken line indicates an example of the timing for separating the period used for the prediction and the period to be predicted. In the present embodiment, as described above, by learning the characteristics of the change in potential at the reference time, as shown in FIG. 8 (1), the potential value (solid line waveform) during the period used for the prediction in the previous stage is followed. It becomes possible to calculate the predicted value (waveform of the broken line) of the potential during the period to be predicted. Further, as shown in FIG. 8 (2), when there is no difference between the predicted value calculated in the period to be predicted in FIG. 8 (1) and the measured value (the thick solid line waveform in FIG. 8 (2)). , The measured value is close to the standard value, and it can be detected that the activity of the cell group is not affected by the candidate sample. Further, as shown in FIG. 8 (3), when there is a difference between the predicted value calculated in the period to be predicted in FIG. 8 (1) and the measured value (the thick solid line waveform in FIG. 8 (3)). Depending on the size of the difference, the measured value is detected as abnormal, and it can be evaluated that the influence of the candidate sample on the activity of the cell group is large. As described above, in the present embodiment, since the abnormality is detected by the size of the difference between the predicted value and the measured value, the size of the influence of the unknown candidate sample on the cell group can be evaluated. Note that, unlike FIG. 8, when the waveform during the period used for prediction is an abnormal one, even if the predicted value is calculated by combining the waveform at the reference time with the learned prediction model, the predicted value and the measured value are usually obtained. And will never be close. Therefore, even in such a case, the abnormality can be detected.

Note that S1 to S3 shown in FIG. 6 do not need to be executed as continuous processing, are provided with a prediction model (that is, NN and coefficient) created in advance in S1 and perform only processing in the operation stage of S3. The measuring device 1 (“abnormality detecting device” according to the present invention) may be provided.

<Construction of neural network>
FIG. 9 is a diagram showing an example of the configuration of an NN including a plurality of layers (also referred to as “layers”). In the NN of FIG. 9, a plurality of conversion processes are performed between the input layer into which the measured potential value is input and the output layer in which the predicted result is shown. The configuration and processing of N1 to N5 existing between the input layer and the output layer is called an intermediate layer. Further, the machine learning unit 142 and the prediction processing unit 144 of the machine learning device 14 shown in FIG. 3 repeatedly use the NN shown in FIG. 9 in each of S1 to S3 shown in FIG. 6 to perform processing.

First, when the feature amount calculation unit 1421 of the machine learning device 14 receives the input of the potential data divided for each predetermined section, the feature amount calculation unit 1421 uses a predetermined number of potential data that are continuous in time series for each electrode and each section. The feature amount is calculated by the so-called convolution layer (N1). It is assumed that one section includes a plurality of potential data. Further, the range representing the predetermined number continuous in the time series described above is referred to as a "window" in N1. In N1, for example, the window is slid for each potential data to change the input value, and the convolution process is performed.

FIG. 10 is a diagram for explaining a convolution layer. As described above, in the convolution layer, a predetermined number of continuous potential data (FIG. 10: x) selected by the window is multiplied by a weighting coefficient W represented by a matrix of a predetermined size, and a feature amount (FIG. 10). : Y) is obtained. It is assumed that the horizontal arrows in FIG. 10 represent the flow of time in one section, and each of the scales shown on the horizontal arrows represents one potential data. In the convolution layer, a window containing a predetermined number of continuous potential data is slid backward for each potential data, for example, input data to be processed is selected, and the calculation of the feature amount is repeated. Further, the calculation of the feature amount by the convolution layer may be repeated with respect to the calculated feature amount. Further, for the feature amount calculated by the convolution layer, for example, an average pooling may be taken by so-called pooling to improve the generalization performance against time fluctuations. In the present embodiment, the measured value is processed by the convolution layer for each measurement electrode 1111 and for each section, so that the potential value measured at each electrode such as the slope of the waveform and the magnitude of the peak is within the section. It is possible to calculate the characteristic amount of the change with time in.

In addition, the feature amount calculation unit 1421 calculates a new feature amount (output value in N2) by the fully connected layer with respect to the feature amount (input value in N2) calculated in N1 of FIG. 9 (FIG. 9 :). N2). In N2, a new feature amount is calculated from the input value included for each electrode and each section based on a predetermined function.

FIG. 11 is a diagram for explaining a fully connected layer. FIG. 11 shows the relationship between the input value x and the output value y included in one section. In the fully connected layer, the input side unit x is processed according to the weighting coefficient W, and the value of the output side unit y is calculated. In N2, the feature amount (FIG. 10: y) calculated in N1 is the input value (FIG. 11: x). Further, all the units on the input side and all the units on the output side are combined, and a new feature amount y is obtained from the input value x. x is a vector having a predetermined number of elements and can be represented by a matrix. Further, y is also a vector having a predetermined number of elements, and is obtained by the following equation (1).
y = Wx + b ... (1)
Here, W is a weighting coefficient and is represented by a matrix of the number of elements of x × the number of elements of y. Further, b is a bias term and can be represented by the same number of vectors as y. With N2 as well, the feature amount for each section of the potential value measured at each electrode can be calculated.

Further, the feature amount calculation unit 1421 calculates the feature amount by the fully connected layer by inputting the feature amount calculated in N2 of FIG. 9 (FIG. 9: N3). In particular, when the chip 111 having a plurality of electrodes is used, the feature amount based on the fully connected layer is calculated by inputting the feature value based on the potential data measured in the same section by some or all the electrodes. May be good.

The structure of the fully connected layer is the same as that in FIG. Further, in N3, the potential data measured by different electrodes is processed by inputting the characteristic value as a result of performing the above-mentioned N1 to N2 processing. That is, in N3, the input value x in FIG. 11 is a characteristic value of some electrodes or all electrodes in the same section. Therefore, N3 can be used to calculate the synchronic features of the potential fluctuation patterns that occurred at the same time in at least a part of the cell group. Since cell potential fluctuations tend to occur all at once in a wide range of cell groups, the accuracy of prediction is improved by learning the characteristics of potential fluctuations that occurred at the same time in multiple electrodes, such as all electrodes. be able to.

In addition, the feature amount calculation unit 1421 calculates a new feature amount by the fully connected layer using the output value y of N3 (FIG. 9: N4). The processing of N4 is the same as that of N2. For example, a new feature amount is calculated from the included input values for each electrode and each section based on a predetermined function.

After that, the feature amount calculation unit 1421 calculates the potential by the deconvolution layer using the output value y of N4 for each electrode and each section (N5). In N5, the reverse processing of the convolution shown in FIG. 10 is performed. That is, the potential value (FIG. 10: x) is calculated from the input feature value (FIG. 10: y). The potential value calculated at this time is a predicted value of the section after the section to be processed.

In the learning stage processing (S1) of FIG. 6, the learning processing unit 1422 is based on the difference between the feature value and the teacher value output in N5, and the coefficient of the function used in N1 to N5 (for example, the weight described above). Correct the coefficient and the value of the bias term). That is, the coefficient is corrected so that the difference between the predicted value, which is the output value of N5, and the teacher value, which is the measured value, becomes small. Existing technologies such as backpropagation can be used to modify the coefficients. In S1, the calculation of the predicted value using the NN of FIG. 9 and the correction of the coefficient are repeated to obtain the coefficient for outputting the desired predicted value for the training data set.

In addition, the prediction model including the generated coefficients is used in S2 and S3 of FIG. In the model evaluation stage (S2) of FIG. 6, for example, using the generated prediction model and the evaluation data set, the prediction processing unit 144 outputs the predicted value of the potential, and the output predicted value and the measured value. The verification processing unit 146 evaluates the accuracy of the generated model based on the sum of the squares of the differences from the teacher value. Further, in the operation stage (S3) of FIG. 6, the prediction processing unit 144 calculates the predicted value of the potential before and after the application of the candidate sample by using the prediction model generated in S1 and the test data set. In addition, based on the size of the difference between the predicted value and the measured value, abnormalities in the activity of the cell group are detected, and the magnitude of the influence of the candidate sample on the cell group is evaluated. It should be noted that the determination as to whether or not it is abnormal may be performed by comparison with a predetermined threshold value. At this time, the threshold value may be set for each type of cell or type of candidate sample.

In each layer of the NN, an arbitrary activation function (transfer function) is appropriately used to enable non-linear separation, for example.

<Selection of electrodes to be evaluated>
When the chip 111 having a plurality of electrodes is used, the extracellular potential value is shown in the figure using the electrode in which the fluctuation of the cell potential, which appears as a rapid rise and fall of the potential in a short period of time, is measured before the application of the compound. The process of 6 may be performed. By adopting such an electrode, the accuracy of abnormality detection can be improved.

<Effect>
According to this embodiment, the value of the potential at the reference time can be predicted by learning the fluctuation pattern of the value of the extracellular potential at the reference time. In addition, it becomes possible to detect a fluctuation pattern of a potential different from the reference based on the difference between the predicted value and the measured value. In particular, it is possible to detect an unknown abnormality because the abnormality is detected not by detecting the characteristic of the change in potential at the time of a known abnormality but by the difference from the predicted value at the reference time. Therefore, in a situation such as drug discovery screening, it can be suitably used for the primary evaluation of the effect on the cell group by a candidate sample whose effect on the cell group is unknown.

<Example>
FIG. 12 is a table showing the results when the compound (candidate sample) is applied to rat nerve cells. In this example, an excitatory compound (Compound) is used as a candidate sample for cortical neurons (Cortex) and hippocampal neurons (Hippocampus), which are primary cultured rat cells (Cells).
4-Aminopyridine (4-AP), Aspirin, Gabardine
, Picrotoxin, Theophylline, and the inhibitory compound Triazolam. Each sample is before application (Before)
), When only the solvent (DMSO: Dimethyl sulfoxide) is applied, when the compound concentration is gradually increased (Drug Low to Drug High), and finally when tetrodotoxin (TTX: Tetrodotoxin) is applied extracellularly. Immediately after application, the potential value was measured for 10 minutes each. The application of the solvent was carried out to confirm the evaluation of the influence of the solvent. The application of tetrodotoxin was performed to suppress the firing of nerve cells and to confirm white noise and the like. In addition, as a value indicating the change in potential at the reference time, each prediction model in which the data of the section before application of the compound of each sample was trained was generated. (Fig. 6: S1). The predetermined section described above was set to 2000 milliseconds, and the sampling interval of the potential value was set to 0.05 milliseconds. In addition, the accuracy of the prediction was evaluated using the generated prediction model and the data of the section before the application of the compound, and it was confirmed that the extracellular potential value can be predicted with a certain accuracy (Fig. 6: S2). Then, the influence of each compound was evaluated based on the difference between the predicted value by the generated prediction model and the measured value before and after the application of the compound (FIG. 6: S3).

In this example, the index value after the normalization based on the sum of the squared values of the differences between the predicted value and the measured value for 8 minutes excluding 2 minutes after the start of each measurement was calculated for each compound and its concentration. If the decrease in the index value after application of tetrodotoxin is small based on the index value before application of the compound calculated from a certain measurement electrode 1111, the cells around the measurement electrode 1111 do not ignite much before application. It can be said that the ignition could not be measured due to insufficient cell adhesion or the like. In this embodiment, the value of the extracellular potential was measured using a chip 111 having a plurality of measurement electrodes 1111 and the index value after application of tetrodotoxin decreased by 3% or more based on the index value before application of the compound. The index values of the measurement electrodes 1111 were averaged, and the values are shown in FIG. As can be seen from FIG. 12, when the excitatory compound was applied, the index value after the normalization was higher than the standard 1.00. On the other hand, when the inhibitory compound was applied, the index value after the normalization was lower than the standard 1.00. That is, it can be said that the influence of the compound could be appropriately evaluated based on the difference between the predicted value and the measured value.

<Modification example 1>
The structure of the NN is not limited to the example shown in FIG. The above-mentioned NN has one or more intermediate layers between the input layer and the output layer, but the number of layers, the number of units belonging to each layer, and the like can be appropriately changed. Further, among the layers shown in FIG. 9, the NN may be composed of some layers, or the order of the layers may be changed.

For example, the NN composed of the input layer, the fully connected layer, and the output layer shown in FIG. 9 may be adopted. For example, in the fully coupled layer, the measured potential data is the input value x in FIG. 11 for each electrode and each section. Further, the predicted value of the potential in the subsequent section is obtained as the output value y. The value of the input layer and the value of the output layer are the same as those in the above-described embodiment. Further, the fully connected layer may have a configuration in which the configuration shown in FIG. 11 is repeated in a plurality of stages. Further, the fully coupled layer may calculate the converted feature amount using a plurality of potential values measured at the same time by different electrodes or a feature amount based on the potential value, or may be determined by the same electrode. The converted features may be calculated using the values of multiple potentials measured at different time points within the period of, or the features based on the values, or these layers may be repeated in multiple steps. Good.

<Modification 2>
Further, the convolution NN including the input layer, the convolution layer, the fully connected layer, and the output layer shown in FIG. 9 may be adopted. In this case, for example, the convolution layer as shown in FIG. 10 is processed in the same section with the same electrode. Further, for each electrode and each section, the output value of the convolution layer is set as the input value x, and the predicted value of the potential in the subsequent section is set as the output value y, and the fully coupled layer is processed. The value of the input layer and the value of the output layer are the same as those in the above-described embodiment. Further, the convolution layer and the fully connected layer may have a configuration in which the configurations shown in FIGS. 10 and 11, respectively, are repeated in a plurality of steps. Further, for the convolution layer and the fully coupled layer, the converted feature amount may be calculated using a plurality of potential values measured at the same time by different electrodes or a feature amount based on the potential value, or the same feature amount may be calculated. The converted features may be calculated using the values of multiple potentials measured at different time points within a predetermined period by the electrodes or the features based on the potential values, or these layers may be repeated in multiple steps. It may be.

<Modification example 3>
Further, the convolution NN composed of the input layer, the convolution layer, and the output layer shown in FIG. 9 may be adopted. In this case, for example, the convolution layer as shown in FIG. 10 is processed in the same section with the same electrode. The value of the input layer and the value of the output layer are the same as those in the above-described embodiment. Further, the convolution layer may have a configuration in which the configuration shown in FIG. 10 is repeated in a plurality of steps. Further, the convolution layer may calculate the converted feature amount using a plurality of potential values measured at the same time by different electrodes or a feature amount based on the potential value, or may be determined by the same electrode. The converted features may be calculated using the values of a plurality of potentials measured at different time points within the period or the features based on the potential values, or these layers may be repeated in a plurality of steps. ..

<Modification example 4>
Further, with respect to the above-described modifications 1 to 3, a convolution NN in which a deconvolution layer is added in front of the output layer may be adopted. In this case, for example, the above-mentioned deconvolution layer processing is performed on the feature amount based on the potential value measured in the same section at the same electrode. Further, the deconvolution layer may be repeated in a plurality of stages. Note that the reverse convolution layer may calculate the converted feature amount or predicted value using the feature amount based on the values of a plurality of potentials measured at the same time by different electrodes, or may use the same electrode. The converted feature amount or predicted value may be calculated using the feature amount based on the values of a plurality of potentials measured at different time points within a predetermined period, or these layers may be combined and repeated in a plurality of steps. You may do so.

<Modification 5>
Further, in the above-described embodiment, learning is performed using the change in potential in a later period as a teacher value with respect to the change in potential in a certain period. That is, for example, as shown in FIG. 13A, the potential value of the learning target period (also referred to as “second period”) is set by inputting the potential value of the learning source period (also referred to as “first period”). Machine learning is performed while sliding the two periods so that they are output. In the operation stage, the potential value of the prediction source period (also referred to as “third period”) corresponding to the first period is input, and the prediction target period (“fourth period”) corresponding to the second period is input. It will be possible to output the predicted value of the potential (also called the period of). However, the context and length of the learning (prediction) original period and the learning (prediction) target period are not limited to the above-described embodiments, and the potential of another period is subject to a change in the potential of one period. Can learn and predict changes in.

FIG. 13B is a diagram for explaining an example of learning and predicting a change in potential in a period before that change in potential in a certain period. The learning (prediction) source period and the learning (prediction) target period shown in FIG. 13B have the opposite context to the example of FIG. 13A. Even in this case, it is possible to detect a change in potential different from the reference time based on the difference between the predicted value and the measured value, and detect it as an abnormality.

FIG. 13C is a diagram for explaining an example of learning and predicting a change in potential during two discontinuous periods with respect to a change in potential between two periods. In FIG. 13C, a learning (prediction) target period is used as a reference, and a period having a predetermined length before and after the learning (prediction) target period is used as a learning (prediction) source period. Even in this case, it is possible to detect a change in potential different from the reference time based on the difference between the predicted value and the measured value, and detect it as an abnormality.

The learning (prediction) source period and the learning (prediction) target period do not have to be adjacent to each other. Further, the relationship between the learning (prediction) original period and the learning (prediction) target period described in the modified example 5 can be carried out in combination with the above-described embodiment and the modified examples 1 to 4.

When machine learning is performed while sliding the two periods in the embodiment or the present modification, the two periods may be periods in which the intervals between the start points or the end points are in a certain relationship. That is, it is possible to slide while keeping the relationship between the interval between the two periods and the length of each period constant. Further, in the order of machine learning, it is not necessary to slide the period to be processed in order to the adjacent period with the sliding width constant, and the combination of the two periods having a constant interval can be arranged in any direction and have an arbitrary size. You may slide it to the processing target. In the two periods, the potential value measured in one period is used as the input value to the input layer, and the potential value measured in the other period is used as the teacher value in the learning stage or the predicted value in the operation stage. It is used as an actual measurement value to calculate the difference between. At this time, the above-mentioned two periods may include a plurality of discontinuous periods as shown in FIG. 13C.

For the two periods, the start timing and length of the period to be predicted and the value of the potential measured in the period to be predicted may be predicted for a certain prediction source period. That is, the relationship between the interval between the two periods and the length of each does not have to be constant. In this case, the characteristics of the correlation period and the potential value measured there are learned with respect to the period of a certain learning source and the potential value measured there. That is, while changing the learning source period and the learning target period separately, it is possible to learn the period in which the potential value can be accurately predicted based on the potential value in the prediction source period and the potential value measured there. ..

As described above, when learning while changing the learning source period (first period) and the learning target period (second period), the learning source period (first period) and the learning target period Machine learning may be performed by changing the relationship between the interval with the period (second period) and the length of each while keeping it constant, or the period of the learning source and the period of the learning target are changed independently. However, machine learning may be performed. In each case, the prediction target period (fourth period) corresponding to the learning target period is based on the potential value measured in the prediction source period (third period) corresponding to the learning source period. It becomes possible to predict the value of the electric potential measured in.

<Modification 6>
Further, in the above-described embodiment, the extracellular potential value of the nerve cell is learned and predicted, but the extracellular potential value of the nerve cell can be learned and predicted. For example, for muscle cells (for example, myocardial cells), the extracellular potential value can be measured by the above-mentioned measuring device, and changes in the extracellular potential value can be learned and predicted as in the case of nerve cells. Abnormality can be detected. That is, it is possible to detect an abnormal change in potential different from a predetermined standard for various cells whose extracellular potential can be measured.

<Others>
The present invention includes a computer program that executes the above-mentioned processing and a computer-readable recording medium that records the program. The recording medium on which the program is recorded can perform the above-mentioned processing by causing the computer to execute the program.

Here, the computer-readable recording medium means a recording medium in which information such as data and programs is stored by electrical, magnetic, optical, mechanical, or chemical action and can be read from a computer. Among such recording media, those that can be removed from a computer include flexible disks, magneto-optical disks, optical disks, magnetic tapes, memory cards, and the like. Further, as the recording medium fixed to the computer, there are HDD, SSD (Solid State Drive), ROM and the like.

1: Measuring device 11: Chip holder 111: Chip 1111: Measuring electrode 1112: Reference electrode 1113: Wiring 12: Potential measuring device 13: Monitor 14: Machine learning device 141: Potential data storage unit 142: Machine learning unit 1421: Feature quantity Calculation unit 1422: Learning processing unit 143: Prediction model storage unit 144: Prediction processing unit 145: Prediction result storage unit 146: Verification processing unit 147: Abnormality detection unit

Claims (8)

Regarding the data representing the time course of the normal extracellular potential as a standard in abnormality detection, the potential value measured in the first period and the second period different from the first period were measured. Using a prediction model generated by learning the relationship with the potential value while changing the first period and the second period, the extracellular potential of the target to be judged whether it is abnormal or not is compared with the standard. Predicted value of the potential measured in the fourth period corresponding to the second period based on the value of the potential measured in the third period corresponding to the first period among the data representing the change with time of Prediction processing unit that finds
An index value calculation unit that calculates an index value according to the magnitude of the difference based on the difference between the measured value of the potential measured in the fourth period and the predicted value.
Abnormality index calculation device including. The prediction processing unit uses the potential value measured in the third period or the feature amount based on the potential value as an input, and all the nodes on the input side are connected to all the nodes on the output side to the fully connected layer and each bond. The abnormality index calculation device according to claim 1, wherein a feature amount or a predicted value of a potential measured in the fourth period is obtained based on an associated predetermined coefficient. The prediction processing unit receives the value of the potential measured in the third period or the feature amount based on the value as an input, and the feature amount or the predicted value of the potential measured in the fourth period by a predetermined convolutional neural network. The abnormality index calculation device according to claim 1 or 2. The prediction processing unit receives a feature amount based on the potential value measured in the third period as an input, and obtains a predicted value of the potential measured in the fourth period by a predetermined deconvolutional neural network. The abnormality index calculation device according to any one of Items 1 to 3. The value of the extracellular potential is measured via a plurality of electrodes arranged so as to be in contact with the cell, and the prediction processing unit is a plurality of values of the potential measured at the same time by different electrodes or features based on the same. The abnormality index calculation device according to any one of claims 1 to 4, wherein a predicted value of the potential is obtained using a quantity. The value of the extracellular potential was measured via one or more electrodes arranged in contact with the cell, and the predictive processing unit was measured by the same electrode at different time points within the third period. The abnormality index calculation device according to any one of claims 1 to 4, wherein a plurality of potential values or feature quantities based on the potential values are used to obtain a predicted value of the potential. The computer
Regarding the data representing the time course of the normal extracellular potential as a standard in abnormality detection, the potential value measured in the first period and the second period different from the first period were measured. Using a prediction model generated by learning the relationship with the potential value while changing the first period and the second period, the extracellular potential of the target to be judged whether it is abnormal or not is compared with the standard. Predicted value of the potential measured in the fourth period corresponding to the second period based on the value of the potential measured in the third period corresponding to the first period among the data representing the change with time of Seeking,
Based on the difference between the measured value of the potential measured in the fourth period and the predicted value, an index value corresponding to the magnitude of the difference is calculated.
Abnormality index calculation method. Regarding the data representing the time course of the normal extracellular potential as a standard in abnormality detection, the potential value measured in the first period and the second period different from the first period were measured. Using a prediction model generated by learning the relationship with the potential value while changing the first period and the second period, the extracellular potential of the target to be judged whether it is abnormal or not is compared with the standard. Predicted value of the potential measured in the fourth period corresponding to the second period based on the value of the potential measured in the third period corresponding to the first period among the data representing the change with time of Seeking,
An abnormality index calculation program that causes a computer to execute a process of calculating an index value according to the magnitude of the difference based on the difference between the measured potential value measured in the fourth period and the predicted value.

Images