[go: up one dir, main page]

WO2019159247A1 - Dispositif de calcul, programme d'analyse et procédé d'analyse - Google Patents

Dispositif de calcul, programme d'analyse et procédé d'analyse Download PDF

Info

Publication number
WO2019159247A1
WO2019159247A1 PCT/JP2018/005026 JP2018005026W WO2019159247A1 WO 2019159247 A1 WO2019159247 A1 WO 2019159247A1 JP 2018005026 W JP2018005026 W JP 2018005026W WO 2019159247 A1 WO2019159247 A1 WO 2019159247A1
Authority
WO
WIPO (PCT)
Prior art keywords
feature
distribution
correlation
value
feature amount
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/JP2018/005026
Other languages
English (en)
Japanese (ja)
Inventor
伸一 古田
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nikon Corp
Original Assignee
Nikon Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nikon Corp filed Critical Nikon Corp
Priority to PCT/JP2018/005026 priority Critical patent/WO2019159247A1/fr
Publication of WO2019159247A1 publication Critical patent/WO2019159247A1/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12M—APPARATUS FOR ENZYMOLOGY OR MICROBIOLOGY; APPARATUS FOR CULTURING MICROORGANISMS FOR PRODUCING BIOMASS, FOR GROWING CELLS OR FOR OBTAINING FERMENTATION OR METABOLIC PRODUCTS, i.e. BIOREACTORS OR FERMENTERS
    • C12M1/00—Apparatus for enzymology or microbiology
    • C12M1/34—Measuring or testing with condition measuring or sensing means, e.g. colony counters
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis

Definitions

  • the present invention relates to a calculation device, an analysis program, and an analysis method.
  • a calculation device that calculates the correlation between the elements using the feature amounts of the elements constituting the cells, the element of the cells based on an image obtained by imaging the cells.
  • a feature amount extraction unit that extracts a plurality of feature amounts each time, a distribution calculation unit that calculates a feature amount distribution of the element from the feature amounts extracted by the feature amount extraction unit, the change in the feature amount, and the feature
  • a correlation calculation unit that calculates a correlation between the elements from the distribution of the amount.
  • an analysis program for causing a computer to calculate a correlation between the elements using feature amounts of the elements constituting the cells, and based on an image obtained by imaging the cells,
  • a feature quantity extracting step for extracting a plurality of feature quantities for each element of the cell;
  • a distribution computing step for computing a distribution of the feature quantities of the elements from the feature quantities extracted by the feature quantity extracting step; and a change in the feature quantities
  • a correlation calculation step of calculating a correlation between the elements from the distribution of the feature amount.
  • an analysis method for calculating a correlation between the elements using feature amounts of elements constituting the cells, wherein the elements of the cells are based on an image obtained by imaging the cells.
  • a feature quantity extracting means for extracting a plurality of feature quantities for each; a distribution computing means for computing a feature quantity distribution of the element from the feature quantities extracted by the feature quantity extracting means; the change in the feature quantity and the feature.
  • a calculation device for calculating a correlation between the elements using the variable of the element, wherein the variable extraction unit extracts a plurality of variables for each element, and is extracted by the variable extraction unit
  • a calculation unit comprising: a distribution calculation unit that calculates a distribution of the variable of the element from the variable to be calculated; and a correlation calculation unit that calculates a correlation between the elements from the change of the variable and the distribution of the variable.
  • FIG. 1 is a diagram illustrating an example of a configuration of a microscope observation system 1 according to an embodiment of the present invention.
  • the microscope observation system 1 performs image processing on an image acquired by imaging a cell or the like. In the following description, an image acquired by imaging a cell or the like is also simply referred to as a cell image.
  • the microscope observation system 1 includes a calculation device 10, a microscope device 20, and a display unit 30.
  • the microscope apparatus 20 is a biological microscope and includes an electric stage 21 and an imaging unit 22.
  • the electric stage 21 can arbitrarily operate the position of the imaging target in a predetermined direction (for example, a certain direction in a horizontal two-dimensional plane, a vertical direction, or an axial rotation direction).
  • the imaging unit 22 includes an imaging element such as a charge-coupled device (CCD) and a complementary MOS (CMOS), and images an imaging target on the electric stage 21.
  • the microscope apparatus 20 may not include the electric stage 21 and may be a stage in which the stage does not operate in a predetermined direction.
  • the microscope apparatus 20 includes, for example, a differential interference microscope (DIC), a phase contrast microscope, a fluorescence microscope, a confocal microscope, a super-resolution microscope, a two-photon excitation fluorescence microscope, and a light sheet microscope. , A light field microscope, a holographic microscope, an optical coherence tomography (OCT), and the like.
  • the microscope apparatus 20 images the culture vessel placed on the electric stage 21. Examples of the culture container include a well plate WP and a slide chamber.
  • the microscope apparatus 20 captures transmitted light that has passed through the cells as an image of the cells by irradiating the cells cultured in the many wells W of the well plate WP with light.
  • the microscope apparatus 20 can acquire images such as a transmission DIC image of a cell, a phase difference image, a dark field image, and a bright field image. Furthermore, by irradiating the cell with excitation light that excites the fluorescent substance, the microscope apparatus 20 captures fluorescence emitted from the biological substance as an image of the cell. Further, the microscope apparatus 20 captures light emission or phosphorescence from a luminescent substance in the cell as an image of the cell.
  • cells are dyed while they are alive, and time-lapse imaging is performed to acquire a cell change image after cell stimulation.
  • a cell image is acquired by expressing a fluorescent fusion protein or staining a cell with a chemical reagent or the like while it is alive.
  • the cells are fixed and stained to obtain a cell image.
  • the fixed cells stop metabolizing. Therefore, in order to observe changes with time in fixed cells after stimulating the cells, it is necessary to prepare a plurality of cell culture containers seeded with the cells. For example, there may be a case where it is desired to observe the change of the cell after the first time and the change of the cell after the second time different from the first time by applying stimulation to the cells. In this case, after stimulating the cells and passing the first time, the cells are fixed and stained to obtain a cell image.
  • a cell culture container different from the cells used for the observation at the first time is prepared, and after stimulating the cells for a second time, the cells are fixed and stained to obtain a cell image.
  • the time-dependent change in a cell can be estimated by observing the change of the cell in 1st time, and the change of the cell in 2nd time.
  • the number of cells used for observing the intracellular change between the first time and the second time is not limited to one. Therefore, images of a plurality of cells are acquired at the first time and the second time, respectively. For example, if the number of cells for observing changes in the cells is 1000, 2000 cells are photographed at the first time and the second time. Therefore, in order to acquire details of changes in cells with respect to a stimulus, a plurality of cell images are required at each timing of imaging from the stimulus, and a large amount of cell images are acquired.
  • the microscope apparatus 20 captures, as the above-described cell image, luminescence or fluorescence from the coloring material itself taken into the biological material, or luminescence or fluorescence generated when the substance having the chromophore is bound to the biological material. May be.
  • the microscope observation system 1 can acquire a fluorescence image, a confocal image, a super-resolution image, and a two-photon excitation fluorescence microscope image.
  • the method of acquiring the cell image is not limited to the optical microscope.
  • an electron microscope may be used as a method for acquiring a cell image.
  • an image obtained by a different method may be used to acquire the correlation. That is, the type of cell image may be selected as appropriate.
  • the cells in this embodiment are, for example, primary culture cells, established culture cells, tissue section cells, and the like.
  • the sample to be observed may be observed using an aggregate of cells, a tissue sample, an organ, an individual (animal, etc.), and an image containing the cells may be acquired.
  • the state of the cell is not particularly limited, and may be a living state or a fixed state.
  • the state of the cell may be “in-vitro”. Of course, you may combine the information of the living state and the fixed information.
  • the cells may be treated with chemiluminescent or fluorescent protein (for example, chemiluminescent or fluorescent protein expressed from an introduced gene (such as green fluorescent protein (GFP))) and observed.
  • chemiluminescent or fluorescent protein for example, chemiluminescent or fluorescent protein expressed from an introduced gene (such as green fluorescent protein (GFP)
  • the cells may be observed using immunostaining or staining with chemical reagents. You may observe combining them. For example, it is possible to select a photoprotein to be used according to the type for discriminating the intracellular nuclear structure (eg, Golgi apparatus).
  • pretreatment for analyzing correlation acquisition such as a means for observing these cells and a method for staining cells, may be appropriately selected according to the purpose.
  • cell dynamic information is obtained by the most suitable method for obtaining the dynamic behavior of the cell
  • information on intracellular signal transmission is obtained by the optimum method for obtaining the intracellular signal transmission. It doesn't matter.
  • These pre-processing selected according to the purpose may be different.
  • the well plate WP has one or a plurality of wells W.
  • the well plate WP has 8 ⁇ 12 96 wells W as shown in FIG.
  • the number of well plates WP is not limited to this, and 6 ⁇ 8 48 wells W, 6 ⁇ 4 24 wells W, 3 ⁇ 4 12 wells W, 2 ⁇ 3 6 wells W , 384 wells W of 12 ⁇ 32 or 1536 wells W of 32 ⁇ 48 may be provided.
  • Cells are cultured in wells W under certain experimental conditions. Specific experimental conditions include temperature, humidity, culture period, elapsed time since stimulation was applied, type and intensity of stimulation applied, concentration, amount, presence or absence of stimulation, induction of biological characteristics, etc. Including.
  • the stimulus is, for example, a physical stimulus such as electricity, sound wave, magnetism, or light, or a chemical stimulus caused by administration of a substance or a drug.
  • Biological characteristics include the stage of cell differentiation, morphology, number of cells, behavior of molecules in cells, morphology and behavior of organelles, behavior of each form, structure of nucleus, behavior of DNA molecules, etc. It is a characteristic to show.
  • FIG. 2 is a block diagram illustrating an example of a functional configuration of each unit included in the calculation apparatus 10 of the present embodiment.
  • the calculation device 10 is a computer device that analyzes an image acquired by the microscope device 20.
  • the calculation device 10 analyzes the correlation between the elements using the feature amounts of the elements constituting the cell.
  • elements constituting the cell are also referred to as components.
  • the calculation device 10 includes a calculation unit 100, a storage unit 200, and a result output unit 300.
  • the image processed by the calculation device 10 is not limited to an image captured by the microscope device 20, and for example, an image stored in advance in the storage unit 200 included in the calculation device 10 or an external storage (not shown). It may be an image stored in advance in the apparatus.
  • the calculation unit 100 functions when the processor executes a program stored in the storage unit 200. Also, some or all of the functional units of the arithmetic unit 100 may be configured by hardware such as LSI (Large Scale Integration) or ASIC (Application Specific Integrated Circuit).
  • the calculation unit 100 includes a cell image acquisition unit 101, a feature amount extraction unit 102, a distribution calculation unit 103, a feature amount arrangement unit 104, and a correlation calculation unit 105.
  • the cell image acquisition unit 101 acquires the cell image captured by the imaging unit 22. That is, the cell image acquisition unit 101 acquires a cell image in which cells are captured. The cell image acquisition unit 101 supplies the acquired cell image to the feature amount extraction unit 102.
  • the cell image acquired by the cell image acquisition unit 101 includes a plurality of images in which the cell culture state is captured in time series, and a plurality of images in which cells are cultured under various experimental conditions.
  • the feature amount extraction unit 102 extracts a plurality of types of feature amounts of the cell image supplied by the cell image acquisition unit 101. That is, the feature amount extraction unit 102 extracts a plurality of feature amounts for each element of the cell based on the image obtained by capturing the cell.
  • This feature amount includes the brightness of the cell image, the cell area in the image, the dispersion and shape of the brightness of the cell image in the image, and the like. That is, the feature amount includes a feature derived from information acquired from a captured cell image.
  • the feature amount extraction unit 102 uses a plurality of acquired images and extracts information about luminance in the image as a feature amount.
  • the feature amount extraction unit 102 extracts, as a feature amount, that a position showing a luminance higher than a predetermined value in the image changes with respect to the stimulus. In addition, for example, the feature amount extraction unit 102 extracts, as a feature amount, that a value indicating luminance in an image changes in response to a stimulus. In addition, for example, when the change in the luminance value with respect to the stimulus is observed in the image and the change in the luminance value is different from the change in the other luminance values, the feature amount extraction unit 102 determines the different luminance values. May be extracted as a feature amount.
  • the change in luminance value with respect to the stimulus is a time change after the stimulus is applied to the cell. That is, the feature amount extraction unit 102 may extract a change in luminance value in time series or a change in position indicating luminance as a feature amount. As described above, the change in the feature amount is a time change with respect to the stimulus applied to the cell.
  • the feature amount extraction unit 102 may acquire information other than luminance from the image and use it for feature amount extraction.
  • the feature quantity extraction unit 102 may acquire color information together with brightness from an image, and extract only a change in brightness indicating a specific color as a feature quantity.
  • the feature amount extraction unit 102 is not limited to the luminance in the image, and may extract a cell shape from the image and extract a change in the shape of the cell as a feature amount.
  • the luminance value indicating a predetermined component increases with time, even if the luminance value does not change by applying a stimulus, the state where the stimulus is applied and the state where the stimulus is not applied Changes in luminance values are different. Therefore, it is good also as a feature-value changing that the change of the value of a brightness differs in the state which added the stimulus, and the state which does not add a stimulus.
  • the feature quantity extraction unit 102 observes each of a plurality of images taken at a predetermined time interval, thereby energetic cells that are less affected by cell contraction, heartbeat cycle, cell movement speed, and stimulation. Changes in the degree of aggregation of nuclear chromatin, which is an indicator of healthy and dying cells, the rate of change in the number and length of neuronal processes, the number of synapses in nerve cells, neural activity such as changes in membrane potential, intracellular Extract dynamic features such as changes in calcium concentration, secondary messenger activity, organelle morphology, intracellular molecular behavior, nuclear morphology, nuclear structure behavior, DNA molecule behavior, etc. Also good. These feature quantity extraction methods use, for example, Fourier transform, wavelet transform, and time differentiation, and use a moving average for noise removal. The feature amount extraction unit 102 supplies the extracted plural types of feature amounts to the distribution calculation unit 103.
  • the distribution calculation unit 103 acquires a plurality of types of feature amounts supplied by the feature amount extraction unit 102. That is, the distribution calculation unit 103 calculates the distribution of the feature amounts of the elements constituting the cell from the feature amounts extracted by the feature amount extraction unit 102.
  • the feature amount extraction unit 102 extracts a feature amount for each cell, for example. Therefore, the feature amount extracted by the feature amount extraction unit 102 for each cell may be different for each cell. For example, when the feature amount extraction unit 102 extracts a luminance value of a predetermined protein for each cell from a plurality of cells after a predetermined time after the stimulus is applied to the cell, the feature amount extraction unit 102 The brightness of a given protein is extracted from a plurality of different cells.
  • the distribution calculation unit 103 calculates the spread of the distribution for each acquired feature amount. For example, the distribution calculation unit 103 calculates the spread of the luminance distribution of the protein after a predetermined time after the stimulation is applied to the cell. The spread of the distribution can vary depending on the time elapsed since the cells were stimulated. Furthermore, the distribution calculation unit 103 calculates the spread of the luminance distribution of the protein at a plurality of times when the predetermined time is changed. In addition, when the feature amount extraction unit 102 extracts feature amounts related to different types of proteins, the distribution calculation unit 103 calculates a feature amount distribution for each type. The distribution calculation unit 103 calculates a value including the spread of the feature amount distribution from the feature amounts extracted by the feature amount extraction unit 102.
  • the distribution calculation unit 103 calculates the spread of the feature quantity distribution based on the standard deviation of the feature quantity distribution as an example.
  • the distribution calculation unit 103 may calculate the spread of the feature amount distribution based on a quantile such as a quartile of the feature amount distribution.
  • the distribution calculation unit 103 determines, for each of a plurality of types of feature quantities, a predetermined time that has elapsed since the stimulation was applied to the cells based on the distribution of feature quantities for each predetermined time that has elapsed since the stimulation was applied to the cells.
  • a representative value of each distribution is calculated.
  • the representative value of the distribution is, for example, an average value of the distribution.
  • the median value or the mode value may be used as the representative value of the distribution.
  • the distribution calculated by the distribution calculation unit 103 may be a range of feature value values extracted by the feature value extraction unit 102.
  • the distribution calculation unit 103 calculates the feature amount distribution by using the number of cells of each value of the feature amount together with the range of the feature amount value. . Accordingly, the distribution calculation unit 103 calculates the distribution of the feature amount using the number of cells for each value of the feature amount. The distribution calculation unit 103 calculates a histogram using the feature value and the number of cells. In this case, the distribution calculation unit 103 may calculate the feature value range using only the maximum value and the minimum value of the feature value, for example, as the feature value distribution.
  • the range of the maximum value and the minimum value of the brightness value is characterized. It does not matter as distribution of quantity.
  • a distribution using the number of proteins indicating the luminance value together with the luminance value of the protein may be used.
  • the distribution calculation unit 103 supplies the feature distribution unit 104 with the calculated distribution spread and the representative value of the distribution every predetermined time after the stimulation is applied to the cells.
  • the information indicating the spread of the feature amount distribution is also referred to as distribution information.
  • the distribution calculation unit 103 extracts a plurality of values as feature amounts from the feature amount distribution.
  • the distribution calculation unit 103 may use an average value, a median value, or a mode value as a representative value of the distribution according to the type of experiment.
  • the distribution calculation unit 103 may use a standard deviation or a quantile as the spread of the distribution according to the type of experiment.
  • the feature quantity array unit 104 acquires a representative value and a spread of the feature quantity distribution supplied by the distribution calculation unit 103.
  • the distribution calculation unit 103 supplies representative values and spreads of a plurality of types of feature quantity distributions to the feature quantity array unit 104
  • the distribution calculation unit 103 sets the representative value of the feature quantity distribution for each type of feature quantity.
  • the spread is supplied to the feature quantity array unit 104.
  • the feature amount arrangement unit 104 represents the representative of the distribution. Get value and spread.
  • the feature quantity arraying unit 104 arranges the feature quantities at predetermined time intervals after the stimulation is applied to the cells, based on the acquired distribution spread and the representative value of the distribution.
  • the value of the feature amount arranged every predetermined time after the stimulation is applied to the cell includes a plurality of values in addition to the representative value of the feature amount distribution.
  • the value indicating the distribution of the feature amount includes a value in which the spread of the distribution of the feature amount is added in addition to the representative value.
  • the spread is a value generated from the feature amount distribution, and is generated by the feature amount array unit 104. Details of how the feature quantity array unit 104 generates the spread of the distribution will be described later.
  • the feature quantity arraying unit 104 supplies information on the feature quantities arranged in time series to the correlation calculation unit 105 based on the spread of the distribution of the feature quantities calculated by the distribution calculation unit 103.
  • the correlation calculation unit 105 acquires the feature quantity array supplied by the feature quantity array unit 104. In the present embodiment, it is possible to acquire a temporal change in the feature amount with respect to the stimulus by arranging the feature amount at every predetermined time after the stimulus is applied to the cell.
  • the feature quantity array is a feature quantity arranged based on a predetermined order. In the present embodiment, the feature quantity array is a feature quantity arranged in order of a predetermined time after the stimulation is applied to the cell. Therefore, in the present embodiment, the correlation calculation unit 105 acquires a temporal change in the feature amount for each component.
  • the correlation calculation unit 105 calculates the correlation between the constituent elements from the correlation between the feature quantities of the constituent elements of the cell using the acquired feature quantity array.
  • the correlation calculation unit 105 calculates the correlation between the constituent elements based on the temporal change of the feature amount with respect to the stimulus and the spread of the feature amount distribution calculated by the distribution calculation unit 103. That is, the correlation calculation unit 105 calculates the correlation between elements from the change in the feature value and the distribution of the feature value. Here, the correlation calculation unit 105 calculates the correlation between the elements based on the extracted plurality of feature amounts.
  • the correlation calculation unit 105 calculates a network indicating each correlation based on the calculated correlation. That is, the correlation calculation unit 105 generates a correlation diagram (in this example, a network) related to the correlation between the elements calculated by the correlation calculation unit 105. A method by which the correlation calculation unit 105 calculates the network will be described later.
  • the correlation calculation unit 105 supplies the calculated correlation and the calculated network to the result output unit 300.
  • the result output unit 300 outputs the correlation and network supplied by the correlation calculation unit 105 to the display unit 30. That is, the result output unit 300 displays the correlation diagram (in this example, a network) on the display device.
  • the result output unit 300 may output the correlation and network supplied by the correlation calculation unit 105 to an output device other than the display unit 30, a storage device, or the like.
  • the display unit 30 displays the correlation output from the result output unit 300 and the network.
  • FIG. 3 is a flowchart illustrating an example of a calculation procedure of the calculation unit 100 according to the present embodiment. Note that the calculation procedure shown here is an example, and the calculation procedure may be omitted or added.
  • the cell image acquisition unit 101 acquires a cell image (step S10).
  • This cell image includes images of a plurality of types of biological tissues having different sizes such as genes, proteins, and organelles. That is, an image obtained by imaging cells includes a plurality of cells.
  • the cell image includes cell shape information. Since the cell image includes phenotype, metabolite, protein, and gene information, the calculation device 10 can analyze the correlation between them.
  • the feature amount extraction unit 102 extracts the cell image included in the cell image acquired in step S10 for each cell (step S20).
  • the feature amount extraction unit 102 extracts a cell image by performing image processing on the cell image.
  • the feature amount extraction unit 102 extracts a cell image by performing image contour extraction, pattern matching, and the like.
  • the feature quantity extraction unit 102 determines the cell type for the cell image extracted in step S20 (step S30). Further, the feature amount extraction unit 102 determines the constituent elements of the cells included in the cell image extracted in step S20 based on the determination result in step S30 (step S40).
  • the cell components include cell organelles (organelles) such as cell nucleus, lysosome, Golgi apparatus, mitochondria, and proteins constituting organelles.
  • the cell type is determined, but the cell type may not be determined. In this case, if the type of cell to be introduced is determined in advance, the information may be used. Of course, the type of cell need not be specified.
  • the feature quantity extraction unit 102 extracts the feature quantity of the image for each cell component determined in step S40 (step S50).
  • the feature amount includes a luminance value of the pixel, an area of a certain area in the image, a variance value of the luminance of the pixel, and the like. Further, there are a plurality of types of feature amounts according to the constituent elements of the cells.
  • the feature amount of the image of the cell nucleus includes the total luminance value in the nucleus, the area of the nucleus, and the like.
  • the feature amount of the cytoplasm image includes the total luminance value in the cytoplasm, the area of the cytoplasm, and the like.
  • the feature amount of the image of the whole cell includes the total luminance value in the cell, the area of the cell, and the like.
  • the feature amount of the mitochondrial image includes the fragmentation rate.
  • the feature amount extraction unit 102 may extract the feature amount by normalizing it to a value between 0 (zero) and 1, for example.
  • the feature amount extraction unit 102 may extract feature amounts based on information on experimental conditions for cells associated with cell images. For example, in the case of a cell image captured when an antibody is reacted with a cell, the feature amount extraction unit 102 may extract a characteristic amount that is unique when the antibody is reacted. In addition, in the case of a cell image captured when cells are stained or when fluorescent proteins are added to cells, the feature quantity extraction unit 102 is used when the cells are stained or when fluorescent proteins are added to the cells A characteristic amount peculiar to each may be extracted.
  • the storage unit 200 may include an experimental condition storage unit 202.
  • the experimental condition storage unit 202 stores information on experimental conditions for cells associated with cell images for each cell image.
  • the experimental condition information includes, for example, cell conditions, image acquisition conditions, cell processing conditions, and the like.
  • Cell conditions include, for example, the type of cell, whether it is a control cell or an inhibitory cell.
  • the conditions at the time of image acquisition include, for example, imaging conditions such as the type of microscope apparatus used and the magnification at the time of image acquisition.
  • the treatment conditions for the cells include, for example, staining conditions when the cells are stained, types of stimulation applied to the cells, and the like. If the experiment condition storage unit 202 does not store the experiment conditions, the experiment conditions may be input using an input unit (not shown).
  • An input unit (not shown) includes, for example, a touch panel, a mouse, or a keyboard.
  • information may be obtained from another device.
  • the microscope apparatus 20 may obtain information on experimental conditions.
  • information on experimental conditions may be obtained from a public database or literature.
  • the captured image may be compared with an image included in a public database or literature, the type of cell included in the captured image may be specified, and the information may be used.
  • FIG. 4 is a diagram illustrating an example of a feature amount extraction result by the feature amount extraction unit 102 of the present embodiment.
  • the feature amount extraction unit 102 extracts a plurality of feature amounts for the protein P1 for each cell and for each time.
  • the feature amount extraction unit 102 extracts feature amounts for N cells from the cell C1 to the cell CN.
  • the feature amount extraction unit 102 extracts feature amounts for seven times from time 1 to time 7.
  • the feature quantity extraction unit 102 extracts K types of feature quantities from the feature quantity k1 to the feature quantity kK.
  • the feature amount extraction unit 102 extracts feature amounts in the directions of the three axes.
  • an axis in the cell direction is described as axis Nc
  • an axis in the time direction as axis N
  • an axis in the feature quantity direction as axis d1.
  • a plurality of cells are included in an image obtained by capturing cells, and the feature amount extraction unit 102 extracts a plurality of feature amounts for each element from the plurality of cells.
  • the K types of feature quantities from the feature quantity k1 to the feature quantity kK are combinations of feature quantities for the protein P1.
  • the types and combinations of feature quantities may differ.
  • the feature amount extraction unit 102 supplies the feature amount extracted in step S50 to the distribution calculation unit 103.
  • the correlation calculation unit 105 calculates the correlation between the constituent elements based on the temporal change of the feature amount with respect to the stimulus and the spread of the feature amount distribution calculated by the distribution calculation unit 103. (Step S60). The details of the process in which the correlation calculation unit 105 calculates the correlation will be described with reference to FIG. The correlation calculation unit 105 calculates a network representing the correlation using the correlation calculated in step S60.
  • FIG. 5 is a flowchart showing an example of detailed processing in step S60 shown in FIG.
  • the distribution calculation unit 103 calculates, for each of a plurality of types of feature amounts supplied by the feature amount extraction unit 102, feature amount distribution information for each predetermined time that has elapsed since the stimulation was applied to the cells. (Step S601). Here, when a plurality of types of feature amounts are extracted by the feature amount extraction unit 102, the distribution calculation unit 103 calculates the spread of the distribution and the representative value of the distribution for each feature amount.
  • the feature quantity arraying unit 104 arranges the feature quantities in a time series of a predetermined time after the stimulation is applied to the cells (step S602). Furthermore, in the present embodiment, the feature amounts are arranged every predetermined time after the stimulation is applied to the cells. In the present embodiment, since the distribution information of the feature amount is extracted from the distribution calculation unit 103, when the feature amount is arranged every predetermined time after the stimulation is applied to the cell, the distribution spread and The feature amount having the representative value of the distribution is arranged. Here, when the distribution calculation unit 103 calculates the spread of the distribution and the representative value of the distribution for each feature amount, the feature amount arrangement unit 104 applies the stimulus to the cell for each feature amount. The feature amounts are arranged in a time series of a predetermined time.
  • the correlation calculation unit 105 calculates the correlation between the feature values using the distribution spread for each predetermined time and the representative value of the distribution supplied from the feature value array unit 104 after the stimulus is applied to the cells. (Step S603).
  • the correlation between the constituent elements is calculated from the correlation between the feature quantities.
  • the correlation calculation unit 105 calculates the correlation between the feature amounts by using the spread of the feature amount at each elapsed time from the stimulus and the representative value of the distribution for each feature amount. Therefore, the correlation calculation unit 105 uses the feature quantity distribution of the first constituent element constituting the cell and the feature quantity distribution of the second constituent element constituting the cell to use the feature of the first constituent element.
  • the correlation calculation unit 105 When the correlation between the quantity and the feature quantity of the second component is calculated and it is determined that there is a correlation, it is determined that there is a correlation between the first component and the second component.
  • the correlation calculation unit 105 generates a matrix X based on a plurality of types of feature amounts extracted by the feature amount array unit 104, and calculates a partial correlation coefficient of the matrix X by the Graphic Lasso method. Note that the method by which the correlation calculation unit 105 calculates the partial correlation coefficient is not limited to the Graphic Lasso method, and any calculation method may be used.
  • FIG. 6 is a diagram illustrating an example of a change in feature amount over time of a stimulated cell.
  • the feature amount A, the feature amount B, and the feature amount C at four times of time t1, time t2, time t3, and time t4 are illustrated.
  • the representative values of the distributions are shown with error bars indicating the spread of each distribution.
  • the size of the error bar indicating the spread of the distribution is, for example, twice the standard deviation of the distribution, and the middle point of the error bar coincides with the representative value.
  • the spread of each distribution of the feature quantity A, the feature quantity B, and the feature quantity C increases in the order of the feature quantity A, the feature quantity B, and the feature quantity C.
  • a small distribution spread that is, a small error bar indicates a high reliability of the feature amount. Therefore, in the feature amount A, the feature amount B, and the feature amount C, the reliability decreases in the order of the feature amount A, the feature amount B, and the feature amount C.
  • the size of the error bar can vary from time to time. That is, the distribution of the feature amount can vary depending on the time elapsed since the cell was stimulated.
  • FIG. 7 is a diagram illustrating an example of a correlation between feature quantities having different distributions according to the present embodiment.
  • the correlation between the feature quantity 1 and the feature quantity 2 the correlation between the feature quantity 3 and the feature quantity 4, and the feature quantity 5 and the feature quantity 6 are illustrated. Correlation with is shown.
  • the distribution of the feature quantity 1 and the distribution of the feature quantity 2 do not spread.
  • the distribution of the feature quantity 3 and the distribution of the feature quantity 4 have the same size spread.
  • the distribution of the feature quantity 5 and the distribution of the feature quantity 6 have the same size spread.
  • the spread of the distribution of each feature value is indicated by using a circle having a radius as a standard deviation.
  • the distribution of the feature quantity 3 and the distribution of the feature quantity 4 is smaller than the distribution of the feature quantity 5 and the distribution of the feature quantity 6.
  • the correlation calculation unit 105 calculates the correlation between elements higher as the spread of the feature amount distribution after the predetermined time with respect to the stimulus is smaller. That is, the correlation calculation unit 105 calculates a higher correlation strength value between elements as the spread of the feature amount distribution after a predetermined time with respect to the stimulus is smaller. Thus, the correlation calculation unit 105 calculates the strength of correlation between elements based on the distribution.
  • FIG. 8 is a diagram illustrating an example of a change in feature amount over time of a cell to which the distribution spread of this embodiment is added.
  • the representative value of the distribution of the feature amount 7 at the four times of time t1, time t2, time t3, and time t4 is shown together with an error bar indicating the spread of the distribution of each time.
  • the feature quantity arraying unit 104 extracts a feature quantity spread other than the representative value from the feature quantity distribution.
  • the feature value is arranged in the error bar range indicating the feature value distribution in addition to the feature value corresponding to the black circle indicating the representative value. Has been.
  • one value is extracted from the feature value in the error bar that is larger than the representative value and smaller than the representative value, and one value is extracted from the feature value in the error bar. .
  • one value of the feature amount in the error bar that is larger than the representative value is set as the first variance value.
  • One value of the feature amount in the error bar that is smaller than the representative value is set as the second variance value.
  • the first variance value uses the largest value among the error bars.
  • the second variance value is the smallest value among the error bars.
  • the first variance value and the second variance value extracted by the feature quantity array unit 104 at the four times of time t1, time t2, time t3, and time t4 are the distribution of the feature quantity 7. Displayed with representative values. Accordingly, in FIG. 8B, two values other than the representative value can be extracted from the feature value distribution at each time based on the feature value distribution information.
  • FIG. 9 is a diagram illustrating an example of a correlation between feature quantities in which a first variance value and a second variance value are added in addition to the representative values of the present embodiment.
  • the representative value of the distribution of the feature quantity 8, the first variance value of the feature quantity 8, the second variance value of the feature quantity 8, the representative value of the distribution of the feature quantity 9, the first variance value of the feature quantity 9, and the feature quantity 9 A correlation with the second variance value is shown as a correlation between the feature quantity 8 and the feature quantity 9.
  • the correlation between the feature quantity 8 and the feature quantity 9 extracted from four different times with respect to the stimulus is shown.
  • the intersection of the representative value of the feature amount 8 and the representative value of the feature amount 9 that has passed a predetermined time with respect to the stimulus is represented by a black circle.
  • FIG. 9A in the Cartesian coordinate system in which the value of the feature value 9 is the Y axis and the value of the feature value 8 is the X axis, the representative value of the feature value 8 and the feature value 9 that have passed for a predetermined time with respect to the stimulus.
  • a pair with a representative value is represented as a black circle.
  • the feature amount 8 the first variance value and the second variance value are calculated for each representative value as shown in FIG. 8B.
  • the difference between the representative value of the feature value 8 and the first variance value is equal to the difference between the representative value of the feature value 8 and the second variance value.
  • the difference between the representative value of the feature value 9 and the first variance value is equal to the difference between the representative value of the feature value 9 and the second variance value. Further, the difference between the representative value of the feature value 8 and the first variance value is equal to the difference between the representative value of the feature value 9 and the first variance value. Therefore, in the present embodiment, the correlation between the feature quantity 8 and the feature quantity 9 is an area surrounded by the first variance value and the second variance value of the feature quantity 8 and the feature quantity 9, respectively. In A), the area is indicated by a chain line around the black circle.
  • the correlation between the feature quantity 8 and the feature quantity 9 is an example of a correlation when there is a positive correlation between the feature quantity 8 and the feature quantity 9.
  • the correlation calculation unit 105 calculates the representative value of the feature quantity 8, the first variance value, the second variance value, the representative value of the feature quantity 9, and the first variance. The value and the second variance value are used.
  • the first variance value and the second variance value are used in addition to the representative points of the feature quantities, but the present invention is not limited to this.
  • the first variance value of one feature quantity, the second variance value of the other feature quantity, the second variance value of one feature, and the first variance value of the other feature quantity And may be used to calculate the correlation.
  • the intersection of the second variance value of the feature value 8 and the first variance value of the feature value 9, the first variance value of the feature value 8, and A positive correlation may be calculated using the intersection of the second variance values of the feature amount 9.
  • FIG. 10 is a diagram illustrating an example of a positive correlation in which the first variance value and the second variance are added in addition to the representative value of the present embodiment.
  • a correlation with the second variance value is shown as a correlation between the feature quantity 10 and the feature quantity 11.
  • the distribution of the feature quantity 10 and the distribution of the feature quantity 11 have different sizes of variance at each time.
  • the correlation between the feature quantity 10 and the feature quantity 11 is an example of a correlation when there is a positive correlation between the feature quantity 10 and the feature quantity 11.
  • FIG. 11 is a diagram illustrating an example of a negative correlation in which the first variance value and the second variance are added in addition to the representative value of the present embodiment.
  • a correlation with the second variance value is shown as a correlation between the feature quantity 12 and the feature quantity 13.
  • the distribution of the feature quantity 12 and the distribution of the feature quantity 13 have distribution spreads of different sizes at each time.
  • the correlation between the feature quantity 12 and the feature quantity 13 is an example of a correlation when there is a negative correlation between the feature quantity 12 and the feature quantity 13.
  • the feature amounts 1 to 12 are feature amounts of elements constituting the cell.
  • the elements constituting the cell include the first element (for example, the element corresponding to the feature amount 1) and the second element (for example, the element corresponding to the feature amount 2).
  • a correlation between the first element and the second element is calculated using the distribution of the feature quantity of the first element and the distribution of the feature quantity of the second element.
  • the distribution calculation unit 103 extracts a plurality of values as the feature quantity of the first element from the distribution of the feature quantity of the first element, and extracts a plurality of values as the feature quantity of the second element from the distribution of the feature quantity of the second element. To do.
  • the calculation device 10 calculates the correlation between the first element and the second element using the extracted feature quantities of the first elements and the extracted feature quantities of the second elements.
  • the feature amount of the first element includes a predetermined value, a value larger than the predetermined value, and a value smaller than the predetermined value
  • the feature amount of the second element is larger than the predetermined value and the predetermined value. Value and a value smaller than a predetermined value.
  • the calculation device 10 includes a value of a feature amount that satisfies a condition that is greater than a predetermined value for the feature amount of the first element and that is less than the predetermined value for the feature amount of the second element; A positive correlation between the first element and the second element is obtained by using a feature quantity value that satisfies a condition that is smaller than the predetermined value in the feature quantity of the first element and that is greater than the predetermined value in the feature quantity of the second element. calculate.
  • the calculation device 10 includes a value of a feature amount that satisfies a condition that is greater than a predetermined value for the feature amount of the first element and that is greater than the predetermined value for the feature amount of the second element.
  • the negative correlation between the first element and the second element is calculated using the feature quantity value that satisfies a condition that is smaller than the predetermined value for the feature quantity of one element and that is smaller than the predetermined value for the feature quantity of the second element.
  • the predetermined value is a median value or an average value of the distribution of the feature amount.
  • FIG. 12 is a diagram illustrating an example of a feature amount matrix for each cell to which the first variance value and the second variance value of the present embodiment are added.
  • the matrix X is a matrix having an axis N in the row direction and an axis d in the column direction.
  • the (3 * i + 1) -th column of the matrix X (where i 0 to d ⁇ 1, and 3 * i means that 3 is multiplied by i, and so on).
  • each element of the matrix X is indicated by the average value and standard deviation of the cell population, but statistics such as the median value and the mode value may be used instead of the average value.
  • a quantile such as a quartile may be used instead of the standard deviation.
  • a matrix X of feature values for each cell may be used.
  • FIG. 13 is a diagram illustrating an example of the correlation and partial correlation coefficient calculated by the correlation calculation unit 105.
  • each of the feature quantity A, feature quantity B, feature quantity C, feature quantity D, feature quantity E, and feature quantity F at eight times indicating the time elapsed since the stimulation was applied to the cell. Representative values of the distribution are shown with error bars indicating the extent of each distribution.
  • the size of the error bar of the feature amount A, the feature amount B, the feature amount C, and the feature amount E is 0.1
  • the size of the error bar of the feature amount D is 0.2
  • the feature amount The size of the error bar for F is 0.3.
  • the feature amount A and the feature amount B are large at time 1.
  • the feature quantity C and the feature quantity D are large at time 3.
  • the feature quantity E and feature quantity F are large at time 6.
  • the value of the partial correlation coefficient between the feature amount A and the feature amount B is 0.167.
  • the value of the partial correlation coefficient between the feature quantity C and the feature quantity D is 0.135.
  • the value of the partial correlation coefficient between the feature quantity E and the feature quantity F is 0.098.
  • the value of the partial correlation coefficient between other feature amounts is 0 (zero).
  • the partial correlation coefficient between the feature quantity A and the feature quantity B is calculated as the correlation between the feature quantity A and the feature quantity B having both error bar sizes of 0.1.
  • the partial correlation coefficient between the feature quantity C and the feature quantity D is a correlation between the feature quantity C having an error bar size of 0.1 and the feature quantity D having an error bar size of 0.2. It is calculated.
  • the partial correlation coefficient between the feature quantity C and the feature quantity D is compared with the value of the partial correlation coefficient between the feature quantity E and the feature quantity F, the partial correlation coefficient between the feature quantity C and the feature quantity D Is larger than the value of the partial correlation coefficient between the feature quantity E and the feature quantity F.
  • the partial correlation coefficient between the feature quantity E and the feature quantity F is calculated as the correlation between the feature quantity E having an error bar size of 0.1 and the feature quantity F having an error bar size of 0.3. It is a thing. That is, the smaller the error bar of the feature amount distribution, the higher the correlation value. Thus, the correlation calculation unit 105 calculates a higher correlation value as the spread of the feature amount distribution is smaller.
  • FIG. 14 is a diagram illustrating an example of the partial correlation coefficient calculated by the correlation calculation unit 105.
  • the network illustrated in FIG. 14 includes node A, node B, node C, node D, node E, and node F, and edges that connect the respective nodes.
  • the node indicates a feature amount used for edge calculation.
  • the edge indicates a correlation between nodes connected by the edge, that is, a correlation between feature amounts.
  • the magnitude of the partial correlation coefficient between nodes is shown by adding shades to edges. The greater the partial correlation coefficient between nodes, the darker the edges.
  • an edge when the correlation calculation unit 105 calculates the correlation using only the representative value of the feature amount distribution is referred to as an original edge.
  • the edge when the correlation calculation unit 105 calculates the correlation using the information on the feature quantity distribution together with the representative value of the feature quantity distribution is referred to as an edge with reliability.
  • the reliability means high reliability of the feature amount.
  • the edge with reliability includes information related to the distribution of the feature amount of the element. In other words, the network shown in FIG. 14, that is, the correlation diagram includes information regarding the distribution of the feature amounts of the elements.
  • a deviation indicating an original edge corresponding to an edge AB connecting the node A and the node B, an edge CD connecting the node C and the node D, and an edge EF connecting the node E and the node F is shown.
  • the value of the correlation coefficient is 1.00.
  • the value of the edge with reliability indicating edge AB is 1.00.
  • the value of the edge with reliability indicating the edge CD is 0.81.
  • the value of the edge with reliability indicating the edge EF is 0.59.
  • the value of the partial correlation coefficient between other nodes is 0, and there is no edge. Therefore, the shading is lighter in the order of edge AB, edge CD, and edge EF.
  • the original edge and the edge with reliability are standardized so that the maximum value is 1.00.
  • the value of the partial correlation coefficient indicating the edge AB, the value of the partial correlation coefficient indicating the edge CD, and the value of the partial correlation coefficient indicating the edge EF are the partial correlation coefficients shown in FIG. Are the values obtained by dividing and normalizing by the partial correlation coefficient value 0.167 between the feature value A and the feature value B, which is the maximum value of.
  • FIG. 15 is a diagram illustrating another example of the partial correlation coefficient calculated by the correlation calculation unit 105.
  • FIG. 15 partial correlation coefficients among the feature amount A, the feature amount B, the feature amount C, the feature amount D, the feature amount E, and the feature amount F are shown.
  • the thickness of each edge is displayed thicker as the value of the partial correlation coefficient is larger.
  • an edge connecting nodes corresponding to these feature quantities is not displayed.
  • the calculation device 10 includes the feature amount extraction unit 102, the distribution calculation unit 103, and the correlation calculation unit 105, and uses the feature amounts of the elements constituting the cell, and uses the feature amounts of the elements. Analyze the correlation.
  • the feature amount extraction unit 102 extracts a plurality of feature amounts for each element of the cell based on the image obtained by capturing the cell.
  • the distribution calculation unit 103 calculates the distribution of element feature amounts from the feature amounts extracted by the feature amount extraction unit 102.
  • the feature quantity array unit 104 can calculate information on the feature quantity distribution by arranging the feature quantities of the constituent elements together with the information on the feature quantity distribution calculated by the distribution calculation unit 103. .
  • the correlation calculation unit 105 changes the feature amount of the constituent elements constituting the cells with respect to the stimulus after the predetermined time with respect to the stimulus and the constituent elements of the constituent elements with respect to the stimulus after the predetermined time with respect to the stimulus calculated by the distribution calculation unit 103. Based on the distribution of the feature amount, the correlation of the elements constituting the cell is calculated. With this configuration, the correlation between the elements can be calculated in consideration of the reliability of the feature amount data.
  • the correlation calculation unit 105 calculates the strength of the correlation between elements based on the distribution of the feature values of the elements constituting the cell. With this configuration, the correlation between feature quantities can be calculated based on the reliability of the feature quantity data.
  • the correlation calculation unit 105 calculates a higher correlation strength value between the elements as the distribution spread is smaller. With this configuration, the higher the reliability of the feature amount data, the higher the correlation between the feature amounts.
  • the spread of the element feature amount distribution is calculated based on the standard deviation of the distribution.
  • the spread of the distribution of the feature amount of the element is calculated based on the quantile of this distribution.
  • the distribution calculation unit 103 extracts a plurality of values as feature amounts from the distribution of the feature amounts of the elements, and the correlation calculation unit 105 is based on the plurality of feature amounts extracted by the distribution calculation unit 103. To calculate the correlation between the elements. With this configuration, the reliability of the correlation between elements can be improved as compared with the case where one value is extracted as a feature quantity from the distribution of the feature quantity of the element and the correlation between the elements is calculated.
  • the elements constituting the cell include the first element and the second element
  • the correlation calculation unit 105 calculates the distribution of the feature quantity of the first element and the feature quantity of the second element.
  • the correlation between the first element and the second element is calculated using the distribution.
  • the distribution calculation unit 103 extracts a plurality of values as the feature values of the first element from the distribution of the feature values of the first element, and the distribution of the feature values of the second element A plurality of values are extracted as feature quantities of the second element, and the feature quantities of the plurality of first elements extracted and the feature quantities of the plurality of second elements extracted are used to calculate the first element and the second element. Calculate the correlation. With this configuration, the correlation between elements can be easily calculated based on a small amount of data reflecting the distribution of data for each feature amount.
  • the representative value of the feature quantity distribution is the median value and / or the average value of the distribution.
  • the change in the feature value is a change in time with respect to the stimulus applied to the cell.
  • the spread of the distribution may vary depending on the time that has elapsed since the cells were stimulated.
  • the calculation device 10 includes an image acquisition unit (cell image acquisition unit 101) that acquires a cell image in which cells are imaged.
  • image acquisition unit 101 acquires a cell image in which cells are imaged.
  • the calculation device 10 includes a correlation calculation unit 105 that generates a network that is a correlation diagram regarding the correlation between elements constituting the cells calculated by the correlation calculation unit 105, that is, an image generation unit.
  • a correlation calculation unit 105 that generates a network that is a correlation diagram regarding the correlation between elements constituting the cells calculated by the correlation calculation unit 105, that is, an image generation unit.
  • the correlation diagram includes information on the distribution of the feature amount of the element.
  • a plurality of cells are included in an image obtained by capturing cells, and a plurality of feature amounts are calculated for each element from the plurality of cells.
  • a correlation output unit 105 that is, a result output unit 300 that displays on the display device a network that is a correlation diagram based on the correlation of feature amounts generated by the image generation unit, that is, a display control unit is provided.
  • FIG. 16 is a flowchart illustrating another example of the calculation procedure of the calculation unit 100 according to the present embodiment.
  • the processing from step S10 to step S603 and step S70 is the same as the processing from step S10 to step S70 shown in FIG. 3 and from step S601 to step S603 shown in FIG. The description is omitted.
  • the correlation calculation unit 105 determines whether or not the calculated correlation value is appropriate (step S604).
  • the feature amount extracted from the abnormal cell image spreads the feature amount distribution. May become larger than usual.
  • the correlation value between the components calculated by the correlation calculation unit 105 becomes smaller than an appropriate value.
  • the storage unit 200 may include a feature amount storage unit 201.
  • the feature quantity storage unit 201 stores the lowest value predicted as a correlation value for each correlation between feature quantities.
  • the minimum value predicted as the correlation value may be determined based on past experiment records and the experience of the user of the calculation device 10.
  • the lowest value predicted as the correlation value may be determined based on the correlation value calculated by the calculation device 10.
  • step S604 determines that all the calculated correlation values are larger than the minimum predicted value as the correlation value stored in the feature amount storage unit 201 (step S604: YES).
  • the calculation is performed.
  • the correlation and network are supplied to the result output unit 300.
  • the correlation calculation unit 105 determines that there is a calculated correlation value that is smaller than the minimum value of predicted values stored as the correlation value stored in the feature amount storage unit 201 (step S604: NO)
  • an abnormal signal that is a signal indicating that the calculated correlation value is abnormal is supplied to the result output unit 300.
  • the result output unit 300 when acquiring the abnormal signal supplied by the correlation calculation unit 105, causes the display unit 30 to output a message indicating that there is an invalid correlation (step S605).
  • the user of the calculation device 10 selects a cell image to be excluded from the cell images acquired by the cell image acquisition unit 101 (step S606). ).
  • a user of the calculation device 10 inputs an operation signal indicating a cell image to be excluded from the operation unit (not shown) to the calculation device 10.
  • the feature quantity extraction unit 102 excludes the cell image and extracts the cell image included in the remaining cell image for each cell based on the operation signal indicating the cell image to be excluded input from the operation unit (not shown). (Step S20).
  • the feature quantity extraction unit 102 may specify a cell image from which a feature quantity whose feature quantity is larger than a predetermined value is specified, and supply a list indicating the specified cell image to the result output unit 300.
  • the result output unit 300 may cause the display unit 30 to output a list of cell images from which feature amounts are extracted that are greater than a predetermined value.
  • the predetermined value is a value such as twice or three times the spread of the feature amount distribution calculated by the distribution calculation unit 103, for example.
  • the distribution calculation unit 103 calculates the spread of the distribution of the feature value, and adds the calculated spread to the representative value of the distribution or subtracts the distribution spread data from the representative value of the distribution.
  • the distribution calculation unit 103 may generate the spread by adding a value obtained by multiplying the spread by a coefficient having a predetermined value to the representative value of the distribution or subtracting it from the representative value of the distribution.
  • the calculation device 10 adjusts the influence of the representative value on the correlation value and the influence of the standard deviation of the feature amount distribution on the correlation value. Can do.
  • the distribution of the feature amount has a distribution spread because a plurality of cells are included in the cell image, and the calculation device 10 takes in the influence of the spread of the distribution to calculate the feature amount.
  • the type of spread of the distribution that the calculation apparatus 10 incorporates when calculating the correlation between the feature amounts is not limited to this.
  • the calculation device 10 may calculate a correlation between feature amounts using an error due to an experimental condition instead of the standard deviation of the distribution described in the present embodiment.
  • the error due to experimental conditions is, for example, an error due to cell staining.
  • the feature amount correlation is calculated using the temporal change of the feature amount with respect to the stimulus applied to the cell, but the change used for the feature amount correlation calculation is not limited to time.
  • the correlation between the feature amounts may be calculated using a change in the feature amount with respect to the amount of the drug solution related to the stimulus applied to the cell.
  • the correlation between the feature amounts of the elements constituting the cell is used as an example using an image obtained by imaging the cell, but the present invention is not limited to this.
  • the feature quantity related to the cell may be extracted using a technique other than an image.
  • a feature amount extracted by Western blotting may be used.
  • the analysis apparatus calculates a correlation between elements using a variable of the element.
  • the analysis device may include a variable extraction unit, a distribution calculation unit, and a correlation calculation unit.
  • the variable extraction unit extracts a plurality of variables for each element.
  • the distribution calculation unit calculates the distribution of the variable of the element from the variable extracted by the variable extraction unit.
  • the correlation calculation unit calculates the correlation between the elements from the change in the variable and the distribution of the variable.
  • the correlation between the feature quantities of the cells is calculated, but the target of the correlation calculation is not limited to the feature quantities of the cells.
  • the correlation between elements is calculated using element variables.
  • the variable distribution of the element is calculated by extracting a plurality of variables for each element.
  • the correlation between elements is calculated from the change of the element and the distribution of variables.
  • the elements are temperature and snow cover
  • the element variables are the average temperature and the number of snow days
  • the correlation between elements is the correlation between temperature and snow cover.
  • a program for executing each process of the calculation device 10 in the embodiment of the present invention is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read into a computer system and executed.
  • the various processes described above may be performed.
  • the “computer system” referred to here may include an OS and hardware such as peripheral devices. Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
  • the “computer-readable recording medium” means a flexible disk, a magneto-optical disk, a ROM, a writable nonvolatile memory such as a flash memory, a portable medium such as a CD-ROM, a hard disk built in a computer system, etc. This is a storage device.
  • the “computer-readable recording medium” refers to a volatile memory (for example, DRAM (Dynamic) in a computer system serving as a server or a client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. Random Access Memory)), etc. that hold a program for a certain period of time.
  • 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 refers to 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 program may be for realizing a part of the functions described above. Furthermore, what can implement

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Zoology (AREA)
  • Wood Science & Technology (AREA)
  • Organic Chemistry (AREA)
  • Biotechnology (AREA)
  • General Engineering & Computer Science (AREA)
  • Genetics & Genomics (AREA)
  • Microbiology (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Sustainable Development (AREA)
  • Medicinal Chemistry (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Apparatus Associated With Microorganisms And Enzymes (AREA)
  • Investigating Or Analysing Biological Materials (AREA)

Abstract

L'invention concerne un dispositif de calcul pour calculer une corrélation entre des éléments configurant une cellule à l'aide de valeurs de caractéristiques des éléments, ledit dispositif comprenant : une partie d'extraction de valeur de caractéristique pour extraire une pluralité de valeurs de caractéristique pour chacun des éléments d'une cellule sur la base d'une image capturée de la cellule ; une partie de calcul de distribution pour calculer une distribution des valeurs de caractéristiques des éléments à partir des valeurs de caractéristiques extraites par la partie d'extraction de valeur de caractéristique ; et une partie de calcul de corrélation pour calculer une corrélation entre les éléments à partir d'un changement des valeurs de caractéristiques et de la distribution des valeurs de caractéristiques.
PCT/JP2018/005026 2018-02-14 2018-02-14 Dispositif de calcul, programme d'analyse et procédé d'analyse Ceased WO2019159247A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/JP2018/005026 WO2019159247A1 (fr) 2018-02-14 2018-02-14 Dispositif de calcul, programme d'analyse et procédé d'analyse

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2018/005026 WO2019159247A1 (fr) 2018-02-14 2018-02-14 Dispositif de calcul, programme d'analyse et procédé d'analyse

Publications (1)

Publication Number Publication Date
WO2019159247A1 true WO2019159247A1 (fr) 2019-08-22

Family

ID=67619779

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2018/005026 Ceased WO2019159247A1 (fr) 2018-02-14 2018-02-14 Dispositif de calcul, programme d'analyse et procédé d'analyse

Country Status (1)

Country Link
WO (1) WO2019159247A1 (fr)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2017109860A1 (fr) * 2015-12-22 2017-06-29 株式会社ニコン Appareil de traitement d'image
JP2017207900A (ja) * 2016-05-18 2017-11-24 大日本印刷株式会社 立体物造形用データ出力規制装置

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2017109860A1 (fr) * 2015-12-22 2017-06-29 株式会社ニコン Appareil de traitement d'image
JP2017207900A (ja) * 2016-05-18 2017-11-24 大日本印刷株式会社 立体物造形用データ出力規制装置

Similar Documents

Publication Publication Date Title
Wang et al. Live-cell imaging and analysis reveal cell phenotypic transition dynamics inherently missing in snapshot data
Sità et al. A deep-learning approach for online cell identification and trace extraction in functional two-photon calcium imaging
US8320655B2 (en) Process and system for analyzing the expression of biomarkers in cells
US11321836B2 (en) Image-processing device, image-processing method, and image-processing program for setting cell analysis area based on captured image
US11379983B2 (en) Analysis device, analysis program, and analysis method
JP6756339B2 (ja) 画像処理装置、及び画像処理方法
Delestro et al. In vivo large-scale analysis of Drosophila neuronal calcium traces by automated tracking of single somata
JP6818041B2 (ja) 解析装置、解析方法、及びプログラム
Cunha et al. Machine learning in microscopy–insights, opportunities and challenges
Knyshov et al. Pretrained convolutional neural networks perform well in a challenging test case: identification of plant bugs (Hemiptera: Miridae) using a small number of training images
US20200372652A1 (en) Calculation device, calculation program, and calculation method
JP6777147B2 (ja) 画像選択装置、画像選択プログラム、演算装置、及び表示装置
WO2018193612A1 (fr) Dispositif de calcul de corrélation, procédé de calcul de corrélation, et programme de calcul de corrélation
WO2018066039A1 (fr) Dispositif d'analyse, procédé d'analyse et programme
WO2019159247A1 (fr) Dispositif de calcul, programme d'analyse et procédé d'analyse
JP6999118B2 (ja) 画像処理装置
WO2018109826A1 (fr) Dispositif d'analyse, programme d'analyse et procédé d'analyse
WO2020070885A1 (fr) Dispositif de détermination, programme de détermination et procédé de détermination
WO2018122908A1 (fr) Dispositif d'analyse, programme d'analyse et procédé d'analyse
WO2020090089A1 (fr) Dispositif, procédé et programme de détermination
Cai et al. Compressive streak microscopy for fast sampling of fluorescent reporters of neural activity
Pujol Masiá Interpreting killer mechanisms of gamma-delta T-cells with Machine Learning
KR20230040873A (ko) 조인트 히스토그램 기반 형광 신호 분리 방법 및 장치
Hoeppner A biologist's perspective on computer vision
WO2018142570A1 (fr) Dispositif de traitement d'image, dispositif d'analyse, procédé de traitement d'image, programme de traitement d'image et dispositif d'affichage

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18906431

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18906431

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP