Evaluation method, image processing device, and evaluation system for evaluation object

Abstract

A method for assessing assessment targets according to one embodiment of the present invention comprises: a step for acquiring statistical information based on at least one color feature quantity for each of multiple assessment regions within a plate image that corresponds to an image of an assessment plate holding assessment targets in multiple wells provided to the plate; and a step for determining the color of the multiple assessment regions using the statistical information. The assessment targets include a test body. The multiple wells include specimen wells holding the assessment targets that further contain specimens. The plate image includes multiple well images that correspond to the multiple wells. Each of the assessment regions has at least one well image that corresponds to at least one of the wells.

Description

[Technical field]

[0001] The present invention relates to a method for evaluating an evaluation object, an image processing device, and a system for evaluating an evaluation object. [Background technology]

[0002] It is known that in biological safety evaluation, a specimen is added to a test specimen and the changes (e.g., color change) are observed. For example, in the Ames test, cells that have been genetically modified to prevent the synthesis of amino acids through genetic manipulation of amino acid synthesis genes are used as test specimens. In the Ames test, a specimen is added to the cells, and the cells are then cultured under certain conditions. If the specimen causes a mutation in the cell's amino acid synthesis gene, the cell will be able to synthesize amino acids again, so whether or not a mutation has occurred is evaluated by checking whether or not the cells are proliferating. In this case, for example, if an indicator that changes color as the cells proliferate is used, the cell proliferation can be evaluated by the color change.

[0003] Conventionally, in such safety evaluations, petri dishes were often used to hold the test specimen and the specimen. In this case, for example, when trying to evaluate the safety of the same test specimen by changing the concentration of the specimen, a petri dish was prepared for each concentration. Therefore, it was necessary to prepare a large amount of specimen. In response to this, in recent years, in order to reduce the amount of test specimen and specimen, for example, the Ames test using a plate with multiple wells as disclosed in Non-Patent Document 1, Non-Patent Document 2, and Non-Patent Document 3 has been known. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Fluckiger-Isler S., Kamber M., “Direct comparison of the Ames microplate format (MPF) test in liquid medium with the standard Ames pre-incubation assay on agar plates by use of equivocal to weakly positive test compounds.”, Mutation Research / Genetic Toxicology and Environmental Mutagenesis747, p36-45, 2012. [Non-Patent Document 2] Sui H., Kawakami K., Sakurai N., Hara T., Nohmi T., “Improvement and evaluation of high throughput fluctuation Ames test using 384-well plate with Salmonella typhimurium TA100 and TA98.”, Genes and Environment 31(2), p47-55, 2009. [Non-Patent Document 3] KamberM., Fluckiger-Isler S., Engelhardt G., Jaeckh R., Zeiger E., “Comparison of the Ames II and traditional Ames test responses with respect to mutagenicity, strain specificities, need for metabolism and correlation with rodent carcinogenicity”, Mutagenesis 24(4), p359-366, 2009. Summary of the Invention [Problem to be solved by the invention]

[0005] In biological safety evaluation, changes caused by adding a specimen to a test specimen (for example, proliferation when the test specimen is a cell, or color change when an indicator is used) have traditionally been evaluated visually by evaluators. Therefore, for example, when safety evaluation is performed using a plate with multiple wells, the burden on the evaluator increases, raising concerns about a decrease in the objectivity and accuracy of the evaluation.

[0006] SUMMARY OF THE PRESENT DISCLOSURE An object of the present invention is to provide an evaluation method, an image processing device, and an evaluation system for an evaluation object that enable more objective and accurate evaluation in biological safety evaluation. [Means for solving the problem]

[0007] A method for evaluating an evaluation object according to one aspect of the present invention comprises the steps of acquiring statistical information based on at least one color feature for each of a plurality of evaluation areas in a plate image corresponding to an image of an evaluation plate in which evaluation objects are held in a plurality of wells provided in the plate, and determining the color of the plurality of evaluation areas using the statistical information, wherein the evaluation object includes a test specimen, the plurality of wells have specimen wells holding the evaluation object further including a specimen, the plate image includes a plurality of well images corresponding to the plurality of wells, and each of the plurality of evaluation areas has at least one well image corresponding to at least one well.

[0008] In the evaluation method, statistical information based on at least one color feature is obtained for a plurality of evaluation regions in the plate image. The color of the evaluation region is determined using the statistical information obtained for the plurality of evaluation regions. Therefore, the color of the evaluation region can be determined objectively and with high accuracy. As a result, the safety of the sample can be evaluated objectively and with high accuracy based on the color of the evaluation region.

[0009] The at least one color feature may be two or more color features. An example of the two or more color features may be hue and saturation. The statistical information may be a frequency distribution with the two or more color features as variables.

[0010] The at least one color feature may be one color feature, and the statistical information may include at least one type of statistical amount for the one color feature.

[0011] In the step of determining colors of the plurality of evaluation regions, the colors of the plurality of evaluation regions may be determined based on a similarity of the statistical information.

[0012] The step of determining the colors of the evaluation areas includes a step of making a primary determination of the colors of the evaluation areas from among a plurality of color candidates based on the similarity of the statistical information, a step of acquiring first reference statistical information for at least one color feature amount for the plurality of color candidates based on the result of the primary determination, and a step of making a final determination of the colors of the evaluation areas using the similarity between the first reference statistical information and the statistical information, and in the step of acquiring the first reference statistical information, the first reference statistical information may be acquired based on the statistical information for at least one evaluation area that has been determined to be a color in the result of the primary determination, among the statistical information for the plurality of evaluation areas. In this case, since the determination is performed twice, the color of the evaluation area can be determined more accurately.

[0013] The plurality of evaluation regions are the plurality of well images, and the step of judging the colors of the plurality of evaluation regions includes a step of making a primary judgment of the colors of the plurality of well images from among a plurality of color candidates based on a similarity of the statistical information, a step of acquiring first reference statistical information for the at least one color feature amount for the plurality of color candidates based on a result of the primary judgment, a step of making a secondary judgment of the colors of the plurality of well images from among the plurality of color candidates using a similarity between the statistical information and the first reference statistical information, a step of acquiring second reference statistical information for the at least one color feature amount for the plurality of color candidates based on a result of the secondary judgment, and a step of making a tertiary judgment of the colors of the plurality of well images using the second reference statistical information, and the plate includes first to second wells each having at least one well of the plurality of wells. N (N is an integer of 2 or more) sections are set, and in the step of acquiring the second reference statistical information, in an i-th section image among the 1st to N-th section images corresponding to the 1st to N-th sections in the plate image, the second reference statistical information for the multiple color candidates is acquired based on statistical information of at least one well image determined to be the same color candidate among the statistical information for the multiple well images, and in the step of tertiary determining the colors of the multiple well images, the color of the well image belonging to the i-th section among the multiple well images may be determined based on the similarity between the second reference statistical information acquired for the i-th section image and statistical information of at least one well belonging to the i-th section image among the statistical information for the multiple well images, thereby finally determining the colors of the multiple well images. In this case, since the color is determined three times, the color of the evaluation area can be determined more accurately.

[0014] In the step of determining the colors of the evaluation regions, the colors of the evaluation regions may be determined based on a similarity between the reference statistical information for each of the color candidates, the reference statistical information being based on the at least one color feature, and the statistical information. In this case, the colors of the evaluation regions can be easily determined.

[0015] The evaluation method of the evaluation object according to one embodiment may further include a step of detecting a transparency change in the plurality of evaluation regions based on a similarity between the standard statistical information for detecting a transparency change based on the at least one color feature and the statistical information. For example, when a change occurs such as death of a cell or precipitation of a specimen, the transparency of the evaluation object changes. Therefore, the above method can evaluate, for example, whether or not a cell has died or whether or not a specimen has precipitated.

[0016] The plurality of wells further includes at least one solvent control well for a solvent control, and one of the plurality of evaluation regions is a solvent control region including at least one solvent control well image corresponding to the at least one solvent control well, and in the step of detecting a change in transparency of the plurality of evaluation regions, the statistical information of the solvent control region among the statistical information of the plurality of evaluation regions may be used as reference statistical information for detecting a change in transparency, and the change in transparency of the plurality of evaluation regions may be detected based on the similarity between the statistical information of the solvent control region and the remaining statistical information among the statistical information of the plurality of evaluation regions. In this case, the change in transparency is detected based on the statistical information of the solvent control region in the same plate image, so that the influence of test conditions (e.g., culture temperature conditions) can be reduced, and the change in transparency can be detected more accurately.

[0017] The method for evaluating an evaluation object according to an embodiment may further include a step of detecting the presence or absence of a change in shape of the test object held in the wells based on the well images, which makes it possible to detect colonies formed by cell proliferation when cells are used as the test object, for example.

[0018] The evaluation method of an evaluation object according to an embodiment may further include a step of detecting the presence or absence of foreign matter in the plurality of wells based on the plurality of well images. This makes it possible to detect foreign matter generated by cell death, for example, when cells are used as the test specimen. In this specification, a foreign matter is different from a structure generated by a change in shape of the test specimen. Therefore, the contents of the foreign matter are different from the contents of a structure generated by a change in shape of the test specimen.

[0019] The step of detecting the presence or absence of foreign matter in the plurality of wells includes a step of detecting the presence or absence of the foreign matter based on a first brightness threshold set for each of the plurality of well images, and a step of detecting the presence or absence of the foreign matter in the outer region based on a second brightness threshold set in an outer region surrounding an inner region corresponding to the bottom region of the well when viewed from the depth direction of the well in each of the plurality of well images, and the first brightness threshold and the second brightness threshold are different, and the area of ​​the bottom of each of the plurality of wells may be smaller than the area of ​​the opening on the opposite side to the bottom. The outer region is, for example, darker than the inner region of the well, so that it is difficult to detect foreign matter. However, in the above configuration, a brightness threshold is set for the outer region for detection, so that foreign matter located in the outer region can also be reliably detected.

[0020] The plate may be a plate determined to be proper by a plate determination method that determines a plate to be proper when at least one of the evaluation wells holds the foreign matter after a certain time has elapsed since an evaluation sample containing the test substance and the specimen selected and adjusted so as to cause the foreign matter to be generated in the evaluation wells of the plate to be evaluated. The term "plate determined to be proper" includes not only a plate that has been subjected to and determined by the plate determination method, but also a plate that should be determined when the determination method is applied.

[0021] In this case, the plate judged to be proper by the above plate judgment method is a plate capable of holding foreign matter. Therefore, foreign matter can be appropriately detected in the evaluation plate (specifically, the evaluation object) after a certain time has elapsed since the evaluation object was injected into the well, so that the evaluation object can be evaluated more accurately.

[0022] The plate may be a polystyrene plate with no surface coating, a polystyrene plate with a poly-D-lysine surface coating, a cyclic olefin copolymer plate with a tissue culture treated surface, or a cycloolefin polymer plate with no surface coating. The plate may be a plate having a glass plate and a resin layer provided on a surface of the glass plate, the resin layer being formed from a cyclic olefin copolymer or polystyrene, the resin layer having a plurality of through holes formed therein corresponding to the plurality of wells, and the areas of the surface exposed by the plurality of through holes being uncoated. Alternatively, the plate may be a plate having a glass plate and a resin layer provided on a surface of the glass plate, the resin layer being formed of a cyclic olefin copolymer or polystyrene, a plurality of through holes corresponding to the plurality of wells being formed in the resin layer, a coating layer being formed on the area of ​​the surface exposed by the plurality of through holes, and the material of the coating layer being fibronectin, collagen or laminin. Such a plate can be judged as suitable in the plate evaluation step.

[0023] A method for evaluating an evaluation object according to another aspect of the present invention comprises the steps of acquiring statistical information based on at least one color feature for each of a plurality of evaluation areas in a plate image corresponding to an image of an evaluation plate in which evaluation objects are held in a plurality of wells of the plate, and detecting a change in transparency of the plurality of evaluation areas based on reference statistical information for detecting a change in transparency based on the at least one color feature and a similarity to the statistical information, wherein the evaluation objects include test specimens, the plurality of wells have specimen wells holding the evaluation objects which further include a specimen, and each of the plurality of evaluation areas has at least one well image corresponding to at least one well.

[0024] In the evaluation method, statistical information based on at least one color feature is obtained for a plurality of evaluation regions in the plate image. The statistical information obtained for the plurality of evaluation regions is used to detect a change in transparency of the evaluation regions. Therefore, the change in transparency of the evaluation regions can be detected objectively and with high accuracy. As a result, the safety of the sample can be evaluated objectively and with high accuracy based on the change in transparency of the evaluation regions.

[0025] In one embodiment, the method for evaluating an evaluation object further includes an irradiation step of irradiating the evaluation plate with light, and in the irradiation step, the evaluation plate may be irradiated with light having a wavelength within at least one of the wavelength ranges of 430 nm to 460 nm and 520 nm to 620 nm.

[0026] In this case, the difference in luminance between wells with different states of the evaluation object becomes larger, and the difference in color, for example, becomes clearer, which makes it possible to more accurately evaluate the evaluation object.

[0027] The light irradiated onto the evaluation plate in the irradiating step may be light having a first wavelength selected from a wavelength range of 430 nm to 460 nm and a second wavelength selected from a wavelength range of 520 nm to 620 nm.

[0028] In the evaluation plate, there may be wells (specifically, evaluation objects) having colors close to colors represented by wavelengths in the wavelength range of 430 nm to 460 nm, and wells (specifically, evaluation objects) having colors close to colors represented by wavelengths in the wavelength range of 520 nm to 620 nm. Therefore, when the light L has the above-mentioned first wavelength and second wavelength, it is easy to ensure the brightness of the wells for both colors on the 430 nm to 460 nm side of the wavelength range and colors on the 520 nm to 620 nm side of the wavelength range. As a result, the difference in colors becomes clearer, making it possible to more accurately evaluate the evaluation objects.

[0029] The method may further include an imaging step of imaging the evaluation plate.

[0030] In the imaging step, light having a wavelength within at least one of the wavelength ranges of 430 nm to 460 nm and 520 nm to 620 nm may be selectively received. In this case, in the image obtained by imaging the evaluation plate, the brightness difference between wells having different states of the evaluation object becomes large, so that, for example, the color difference becomes clearer. As a result, the evaluation object can be evaluated more accurately.

[0031] In the imaging step, the evaluation plate may be imaged from the bottom side of the multiple wells. For example, when cells are used as test specimens, when the cells grow and form colonies, or when the cells die and foreign matter or the like is generated, they tend to sink to the bottom of the wells. Therefore, by imaging the evaluation plate from the bottom side of the multiple wells, the colonies, foreign matter, etc. can be detected more reliably.

[0032] An image processing apparatus according to yet another aspect of the present invention is configured to execute the evaluation method according to the present invention.

[0033] A system for evaluating evaluation objects according to yet another aspect of the present invention comprises a plate holding section for holding an evaluation plate having evaluation objects held in a plurality of wells provided on the plate, an imaging section for imaging the evaluation plate, and an image processing device for evaluating a plate image corresponding to the evaluation plate obtained based on the imaging results of the imaging section.

[0034] The imaging unit may be disposed on the bottom side of the plurality of wells. Effect of the Invention

[0035] According to the present invention, it is possible to provide an evaluation method, an image processing device, and an evaluation system for an evaluation object that enable more objective and highly accurate evaluation in biological safety evaluation. [Brief description of the drawings]

[0036] [Figure 1]FIG. 1 is a plan view of an evaluation plate used in a method for evaluating an evaluation object according to one embodiment. [Diagram 2] FIG. 2 is a schematic diagram of a cross section of a well. [Diagram 3] FIG. 3 is a schematic diagram of an evaluation system for an evaluation object according to one embodiment. [Figure 4] FIG. 4 is a functional block diagram of the calculation unit when performing the color judgment process. [Diagram 5] FIG. 5 is a diagram showing an example of an image of the evaluation plate (plate image) created by the plate image creation unit. [Figure 6] FIG. 6 is a diagram showing an example of an HSV image. [Figure 7] FIG. 7 is a diagram showing an example histogram of a frequency distribution created for a well. [Figure 8] FIG. 8 is a graph showing an example of a similarity distribution for purple and yellow created for the primary judgment. [Figure 9] FIG. 9 is a graph showing an example of a similarity distribution created for the secondary judgment. [Figure 10] FIG. 10 is a graph showing an example of the similarity distribution for each of the eight sections. [Figure 11] FIG. 11 is an example of a flowchart of an evaluation method using an example of a color judgment process. [Figure 12] FIG. 12 is a functional block diagram of the calculation unit when another example of the color judgment process is performed. [Figure 13] FIG. 13 is a graph showing an example of a similarity distribution for each of eight sections created for judging the color in another example of the color judgment process. [Figure 14] FIG. 14 is a functional block diagram of a calculation unit when creating a reference frequency distribution used in another example of the color judgment process. [Figure 15] FIG. 15 is an example of a flowchart of an evaluation method using another example of the color judgment process. [Figure 16] FIG. 16 is a functional block diagram of the calculation unit when performing the colony detection process. [Figure 17] FIG. 17 is a diagram for explaining the colony detection process performed by the calculation unit shown in FIG. [Figure 18] FIG. 18 is an example of a flowchart of an evaluation method using a colony detection process. [Figure 19] FIG. 19 is a functional block diagram of the calculation unit when performing the transparency change detection process. [Figure 20] FIG. 20 is a chart showing the similarity of the compartment frequency distributions among the eight compartments. [Figure 21] FIG. 21 is an example of a flowchart of an evaluation method using a transparency change detection process. [Figure 22] FIG. 22 is a functional block diagram of the calculation unit when performing the foreign object detection process. [Figure 23] FIG. 23 is a diagram for explaining the foreign object detection process performed by the calculation unit shown in FIG. [Figure 24] FIG. 24 is an example of a flowchart of an evaluation method using a foreign object detection process. [Diagram 25] FIG. 25 is a table showing the test conditions in Experimental Example 1. [Figure 26] FIG. 26 is a table showing the color judgment results in Experimental Example 1. [Figure 27] FIG. 27 is a table showing the color judgment results in Experimental Example 1. [Figure 28] FIG. 28 is a chart showing the color judgment results in Experimental Example 1. [Figure 29] FIG. 29 is a chart showing the color judgment results in Experimental Example 1. [Diagram 30] FIG. 30 is a chart showing the colony detection results in Experimental Example 1. [Diagram 31] FIG. 31 is a chart showing the colony detection results based on the images obtained by capturing images of the evaluation plate from the side opposite to the bottom of the wells in Experimental Example 1. [Diagram 32] FIG. 32 is a table showing the foreign matter detection results in Experimental Example 1. [Diagram 33]FIG. 33 is a chart showing the results of foreign matter detection based on images obtained by capturing images of the evaluation plate from the side opposite the bottom of the wells in Experimental Example 1. [Diagram 34] FIG. 34 is a schematic diagram of an observation device in Experimental Example 2. [Diagram 35] FIG. 35 is a diagram showing the spectral distribution by section based on the imaging results of the evaluation plate of Sample No. 1 in Experimental Example 2. [Diagram 36] FIG. 36 is a diagram showing the spectral distribution by section based on the imaging results of the evaluation plate of Sample No. 2 in Experimental Example 2. [Figure 37] FIG. 37 is a diagram showing the spectral distribution by section based on the imaging results of the evaluation plate of sample No. 3 in Experimental Example 2. [Figure 38] FIG. 38 is a diagram showing the spectral distribution by section based on the imaging results of the evaluation plate of sample No. 4 in Experimental Example 2. [Figure 39] FIG. 39 is a diagram showing the spectral distribution by section based on the imaging results of the evaluation plate of Sample No. 5 in Experimental Example 2. [Diagram 40] FIG. 40 is a graph showing the luminance difference between sections for each wavelength in Experimental Example 2. [Diagram 41] FIG. 41 is a cross-sectional view showing another example of the plate. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0037] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. The same elements are given the same reference numerals, and duplicated explanations will be omitted. The dimensional ratios of the drawings do not necessarily match those in the description.

[0038] In this embodiment, an embodiment for evaluating the safety of a specimen will be described. Specifically, a case in which the safety of a specimen is evaluated using the Ames test will be described.

[0039] First, an outline of the Ames test used in this embodiment will be described. In the Ames test, genetic manipulation is performed on amino acid synthesis genes, and cells that have been improved so that they cannot synthesize amino acids are used as test specimens. In the Ames test, a sample is added to the cells. The cells are then cultured under certain conditions. If a mutation occurs in the amino acid synthesis gene of the cells due to the sample, the cells will be able to synthesize amino acids again. Therefore, genotoxicity is evaluated by checking whether a mutation has occurred by confirming the presence or absence of cell proliferation.

[0040] Furthermore, the Ames test usually involves a cytotoxicity evaluation to confirm whether or not a sample has toxicity to cells.

[0041] In the present embodiment, an evaluation system, an image processing device, and an evaluation method applicable to genotoxicity evaluation and cytotoxicity evaluation by the Ames test will be described. In the following description of the embodiment based on the Ames test, the test specimen is a cell unless otherwise specified.

[0042] 1 is a plan view of an evaluation plate 10 used in an evaluation method according to one embodiment. The evaluation plate 10 has a plate 11. The plate 11 has a plurality of wells 13. In this embodiment, the plate 11 has 384 wells 13 arranged in a two-dimensional (16×24) array.

[0043] First to Nth sections (N is an integer of 2 or more) are virtually set on the plate 11. In this embodiment, as shown in FIG. 1, a case where N is 8, that is, eight sections are virtually set, will be described. Hereinafter, the eight sections will be referred to as sections Se1 to Se8. Sections Se1 to Se8 are regions including 4×12 (48) wells 13 (well groups). Section Se1 is a solvent control section, and section Se8 is a positive control section. The solvent control section functions as a negative control section.

[0044] The wells 13 are recesses (depressions) formed in the plate 11. The wells 13 hold evaluation objects 14. When the evaluation objects 14 held in the wells 13 belonging to the sections Se1 to Se8 are to be described separately, they will be referred to as evaluation objects 14a to 14h.

[0045] The evaluation objects 14a to 14h have a culture solution in which cells are dispersed (or suspended). The same amount of culture solution is held in all 384 wells 13. The amount of cells in the culture solution held in the wells 13 is also the same between the wells 13. Examples of cells include microorganisms having amino acid auxotrophic mutations, cells derived from organisms, etc. Examples of the microorganism having an amino acid auxotrophic mutation include histidine-auxotrophic Salmonella typhimurium (e.g., TA1535, TA1537, TA97, TA97a, TA98, TA100, TA92, TA94, TA102, TA104, TA1538, TA7001, TA7002, TA7003, TA7004, TA7005, TA7006, various YG strains, etc.), tryptophan-auxotrophic Escherichia coli (e.g., WP2, WP2uvrA, WP2 / pKM101, WP2uvrA / pKM101, etc.), and other microorganisms. Examples of the cells derived from organisms include cultured cells derived from mammals, cells derived from other vertebrates, and cells derived from insects. The cells included in the evaluation targets 14a to 14h may be cells in which a plurality of types of cells are mixed. An example of the culture medium is a phosphate buffer solution containing a small amount of amino acids.

[0046] The evaluation objects 14a to 14h further include an indicator. The amount of the indicator is the same for the evaluation objects 14a to 14h (i.e., the same for all wells 13). In this embodiment, the indicator is a pH indicator. An example of a pH indicator is bromocresol purple. In this embodiment, the color of the pH indicator is purple in the case of a negative result (when no mutation occurs in the cells), and is yellow in the case of a positive result (when a mutation occurs in the cells). Therefore, the color candidates in this embodiment are purple and yellow. In this specification, the color candidates are colors that the evaluation object is expected to show before and after a predetermined reaction, for example, when an indicator (e.g., a pH indicator) is used.

[0047] The evaluation objects 14b to 14g further include a liquid (specimen solution) in which a specimen is dissolved or uniformly dispersed in a solvent, in addition to the culture medium in which the cells are dispersed and the indicator. Therefore, the wells 13 in the sections Se2 to Se7 that hold the evaluation objects 14b to 14g are wells that hold specimens (specimen wells). The specimen is a substance (e.g., a chemical substance) whose safety (genotoxicity in this embodiment) is to be evaluated. Examples of the solvent include water and dimethyl sulfoxide (DMSO). The evaluation objects 14b to 14g are the same except that the concentrations of the specimens are different. The amount of specimen solution in the evaluation objects 14b in all wells 13 belonging to section Se2 is the same. The same is true for sections Se3 to Se7.

[0048] The evaluation object 14a contains a solvent control solution in addition to the culture solution containing the above-mentioned cells and an indicator. Therefore, the well 13 in section Se1 that holds the evaluation object 14a is a solvent control well. The solvent control solution is the solvent used in the specimen solution. However, the solvent control solution may be any liquid that does not affect the cells. Examples of the solvent control solution include water and DMSO. The amount of solvent control solution in the evaluation object 14a is the same in all wells 13 belonging to section Se1.

[0049] The evaluation target 14h includes a positive control solution in addition to the culture solution containing the above-mentioned cells and an indicator. Therefore, the well 13 of the section Se8 that holds the evaluation target 14h is a positive control well. The positive control solution is a solution containing a positive control compound and a solvent. The positive control compound may be any compound that is known to be positive for cells. Examples of the positive control compound include 9AA (9-Aminoacridine), 4-NQO (4-Nitroquinoline-N-oxide), 2-AA (2-Aminoanthracene), and 2-NF (2-Nitrofluorene). The solvent does not have to be the same as the solvent used in the sample solution. Examples of the solvent include water, DMSO, and the like. The amount of the positive control solution held in all wells 13 belonging to the section Se8, which is a positive section, is the same.

[0050] When examining the influence of the presence or absence of metabolic activation of a specimen, the evaluation objects 14a to 14h further include S9mix. S9mix is ​​a liquid obtained by adding a coenzyme to an animal liver extract. S9mix may be a liquid prepared by adding a cofactor involved in the electron transport system, such as NADPH, to the 9,000×g supernatant (S9) of liver homogenate.

[0051] The configuration of the well 13 will be further described with reference to FIG. 1 and FIG. 2. FIG. 2 is a schematic diagram of a cross section of a well. FIG. 2 shows a schematic diagram of an evaluation target 14. In this embodiment, the shape of the well 13 in plan view (shape seen from the depth direction of the well 13) is a square. An example of the length of one side of the opening of the well 13 is 3.0 mm to 4.0 mm. An example of the depth of the well 13 is 9.0 mm to 15.0 mm. A plate 11 having wells of such sizes is known as a microplate. As shown in FIG. 2, the well 13 tapers toward the bottom 13a. Therefore, the area of ​​the bottom 13a is smaller than the area of ​​the opening 13b of the well 13.

[0052] The evaluation plate 10 is a plate in which the evaluation subject 14 corresponding to each well 13 of the plate 11 is poured and subjected to a certain culture treatment, and then the plate is in a state in which the safety of the evaluation subject 14 can be evaluated.

[0053] Next, a description will be given of an evaluation system 20 used in a method for evaluating the safety of a sample held on the evaluation plate 10. Fig. 3 is a schematic diagram of the evaluation system 20 according to one embodiment. The evaluation system 20 includes an observation device 30 and an image processing device 40.

[0054] [Observation equipment] The observation device 30 has a stage (plate holder) 31, an illumination device 32, and an imaging unit 33. The stage 31 supports the evaluation plate 10. The evaluation plate 10 is placed on the stage 31 so that the bottom of the evaluation plate 10 (the bottom 13a side of the well 13) is located on the stage 31 side. The stage 31 is configured to pass the light L output from the illumination device 32 toward the imaging unit 33. For example, the material of the stage 31 may be optically transparent, or an opening or a window portion that transmits the light L may be formed in an area corresponding to an area in the evaluation plate 10 where the 384 wells 13 are formed. In this embodiment, the stage 31 is configured to be transportable in the direction of the white arrow in FIG. 3.

[0055] The illumination device 32 is disposed above the evaluation plate 10 (on the opposite side of the stage 31 as viewed from the evaluation plate 10). The illumination device 32 irradiates the evaluation plate 10 with light L. An example of the illumination device 32 is an LED illumination, and an example of the light L is light with a wavelength of 400 nm to 780 nm.

[0056] The imaging unit 33 receives light L transmitted through the evaluation plate 10 and images the evaluation plate 10. The imaging unit 33 obtains a color image (RGB image) that is substantially the same as when the evaluation plate 10 is viewed with the naked eye. The imaging unit 33 has an imager 33a and a light collecting unit 33b.

[0057] Examples of the imager 33a include a two-dimensional sensor (or camera) and a line sensor. Examples of the two-dimensional sensor include a CCD camera and a CMOS camera.

[0058] The light collecting unit 33b collects the light L and makes it incident on the image pickup device 33a. The light collecting unit 33b includes at least one lens. If the image pickup device 33a itself has a light collecting optical function, the image pickup device 33 does not need to have the light collecting unit 33b.

[0059] In this embodiment, the imaging area of ​​the imaging unit 33 is smaller than the size of the evaluation plate 10. In this case, the imaging unit 33 images the evaluation plate 10 in multiple batches. However, the imaging unit 33 may include multiple imaging units 33, and the imaging area of ​​the imaging unit 33 may be large enough to image the entire evaluation plate 10 at once.

[0060] The imaging unit 33 is connected to the image processing device 40 via wired or wireless communication, and outputs the acquired image (specifically, image data) to the image processing device 40. The image may be input from the imaging unit 33 to the image processing device 40 via a storage medium such as a DVD.

[0061] In the observation device 30, the positions of the illumination device 32 and the imaging unit 33 with respect to the evaluation plate 10 may be inverted. That is, in FIG. 3, the imaging unit 33 may be disposed above the evaluation plate 10, and the illumination device 32 may be disposed below the stage 31, and the evaluation plate 10 may be imaged.

[0062] A method for imaging the evaluation plate 10 using the observation device 30 will be described. When imaging the evaluation plate 10, the evaluation plate 10 is placed on the stage 31. While the stage 31 is moved in one direction, the evaluation plate 10 is irradiated with light L from the illumination device 32 (illumination step). In this state where the evaluation plate 10 is irradiated with light L, the evaluation plate 10 is imaged by the imaging unit 33 (imaging step).

[0063] In this embodiment, since the imaging area of ​​the imaging unit 33 is smaller than the size of the evaluation plate 10, a partial image of the evaluation plate 10 is obtained in one imaging. Therefore, in the imaging step, the imaging unit 33 is used to image the evaluation plate 10 multiple times so that the partial images are combined to form an image of the entire evaluation plate 10.

[0064] The imaging unit 33 inputs the acquired image to the image processing device 40.

[0065] [Image processing device] The image processing device 40 has an input unit 41, a calculation unit 42 (execution unit), and an output unit 43. The image processing device 40 is a computer. The image processing device 40 may further have a storage unit (memory) that is usually included in a computer. The image processing device 40 may be a personal computer that functions as the image processing device 40 by executing various programs, or may be a dedicated device for evaluation. The image processing device 40 functions as an evaluation device or evaluation support device that evaluates the evaluation plate 10.

[0066] The input unit 41 is an input unit to which (or which accepts) data of an image captured by the imaging unit 33 is input.

[0067] The calculation unit 42 creates a plate image P1 (see FIG. 5) based on the image input to the input unit 41. FIG. 5 is a drawing showing an example of the plate image P1. The plate image P1 is an image corresponding to the evaluation plate 10. In the plate image P1, the portions corresponding to the well 13 and the partitions S1 to Se8 are referred to as the well image 13p and the partition images Se1p to Se8p, respectively. The calculation unit 42 processes the plate image P1 to perform a color judgment process for the well 13. The calculation unit 42 may further perform at least one of a colony detection process in the well 13, a transparency change detection process in the well 13, and a foreign matter detection process in the well 13. In this embodiment, a case will be described in which the calculation unit 42 is also capable of performing the colony detection process, the transparency change detection process, and the foreign matter detection process.

[0068] In this embodiment, the "color of the well" refers to the color of the evaluation target 14 held in the well 13. The calculation unit 42 determines the color of the well 13 by determining the color of the well image 13p. In this embodiment, the "colony in the well" refers to a colony as a shape change that has occurred in the cell that is the evaluation target 14 held in the well 13. Since the colony is a shape change of the cell (specimen), it is a structure in which the boundary between the outline and the inside is not clear. In this embodiment, the "change in transparency of the well" refers to a change in transparency of the evaluation target 14 held in the well 13, and the calculation unit 42 determines the change in transparency of the well 13 by determining the change in transparency of the well image 13p. In this embodiment, the "foreign matter in the well" refers to a foreign matter that has occurred in the evaluation target 14 held in the well 13, and unlike a shape change of the cell (specimen) like a colony, it is a substance (or structure) that is separated from the surroundings of the foreign matter by an interface with a different refractive index. Therefore, the boundary between the outline and the inside of the foreign matter is clear (at least clearer than in the case of a colony).

[0069] When the input unit 41 outputs, for example, the plate image P1 to the storage unit, the calculation unit 42 accesses the storage unit to obtain the data to be processed. The calculation unit 42 will be described later.

[0070] The output unit 43 displays and outputs the results of the calculation unit 42 so that an evaluator can refer to the results. For example, the output unit 43 may output the results of the calculation unit 42 to an external display device, or the output unit 43 itself may have a display function. The output unit 43 may output and display data during the processing to the calculation unit 42.

[0071] As described above, in this embodiment, the calculation unit 42 is capable of carrying out color judgment processing, colony detection processing, transparency change detection processing, and foreign matter detection processing.

[0072] The operation of the calculation unit 42 when each process is performed will be described. Here, the case will be described where the imaging area of ​​the imaging unit 33 is smaller than the size of the evaluation plate 10, and the imaging unit 33 images different areas of the evaluation plate 10 multiple times in order to obtain an image of the entire evaluation plate 10.

[0073] [Color Judgment Processing] Regarding the color judgment process, a first color judgment process and a second color judgment process will be described. Statistical information based on at least one color feature is used for color judgment. The statistical information includes at least one type of statistical quantity. In this embodiment, a case will be described in which hue and saturation are used as two or more color feature quantities, and frequency distribution (statistical quantity) is used as the statistical information.

[0074] [First color judgment process] The calculation unit 42 when performing the first color judgment process is referred to as the calculation unit 42A. Fig. 4 is a functional block diagram of the calculation unit 42A. The calculation unit 42A has a plate image creation unit 42a, an HSV image creation unit 42b, a frequency distribution creation unit (statistical information acquisition unit) 42c, a first judgment unit 42d, a first reference creation unit (first reference acquisition unit) 42e, a second judgment unit 42f, a second reference creation unit (second reference acquisition unit) 42g, and a third judgment unit 42h.

[0075] The plate image creation unit 42a synthesizes a plurality of images (each of which is a partial image of the evaluation plate 10) input to the input unit 41 to create a plate image P1 corresponding to the evaluation plate 10, as shown in Fig. 5. The plate image P1 is a color image (RGB image).

[0076] The HSV image creation unit 42b converts the plate image P1 created by the plate image creation unit 42a into an HSV image P2. FIG. 6 is a diagram showing an example of the HSV image P2. The HSV image P2 shown in FIG. 6 is an image obtained by HSV converting the plate image P1 shown in FIG. 5. The conversion from an RGB image to an HSV image may be performed based on a known conversion formula (or a known program based thereon). Since the HSV image P2 is an image obtained by HSV converting the plate image P1, partial images in the HSV image P2 corresponding to the well 13 and the sections Se1 to Se8 are also referred to as the well image 13p and the section images Se1p to Se8p.

[0077] The frequency distribution creating unit 42c uses the HSV image P2 created by the HSV image creating unit 42b to create a frequency distribution with hue and saturation as variables for each well image 13p (in other words, well 13) (i.e., acquires statistical information as a frequency distribution with hue and saturation as variables). Therefore, the frequency distribution creating unit 42c creates multiple frequency distributions (384 frequency distributions in this embodiment) corresponding to all wells 13 in the evaluation plate 10. An example of a specific method of frequency distribution by the frequency distribution creating unit 42c will be described.

[0078] The frequency distribution creating unit 42c identifies the region of each well image 13p in the HSV image P2. For example, as shown by the dashed line in FIG. 6, an ROI (Region of Interest) may be set in the well image 13p. For illustrative purposes, FIG. 6 shows a state in which an ROI is set in one well image 13p. The ROI may be set as a region designated by the evaluator, or may be automatically set by the frequency distribution creating unit 42c based on, for example, data (shape, size, etc.) of the plate 11 input in advance, an image of the plate 11 with nothing in the well 13, brightness information of the HSV image, etc.

[0079] The frequency distribution creating unit 42c creates the frequency distribution by using image information (hue, saturation, etc.) of a plurality of pixels included in the specified well image 13p. In this embodiment, the frequency distribution is created by dividing the hue and saturation into 36 parts each and extracting 36×36=1296 feature amounts (the number of pixels in this embodiment).

[0080] Fig. 7 is a diagram showing an example of a histogram of the created frequency distribution. The horizontal axis of Fig. 7 indicates hue and saturation, and the vertical axis indicates the number of pixels. Fig. 7 is a histogram for one well image 13p.

[0081] The first determination unit 42d first calculates the similarity between the multiple frequency distributions created by the frequency distribution creation unit 42c. The similarity may be calculated using, for example, Euclidean distance, or a correlation coefficient. The similarity may be calculated using, for example, pattern matching. Deep learning may be used to calculate the similarity.

[0082] An example of calculating the similarity between a plurality of frequency distributions will be described. In this embodiment, in order to determine the color of the well image 13p (i.e., the color of the well 13) as purple or yellow, the similarity of the purple region (color candidate region) in each frequency distribution and the similarity of the yellow region (color candidate region) in each frequency distribution are calculated. The purple region and the yellow region can be set, for example, by an evaluator based on the hue and saturation in order to determine the color of the well image 13p. In the example of FIG. 7, for example, a hue of 20 to 35 and a saturation of 0 to 15 can be set as the purple region, and a hue of 0 to 15 and a saturation of 0 to 25 can be set as the yellow region.

[0083] Any one of the multiple well images 13p is referred to as a well image of interest. In this case, the similarity between the frequency distribution of the purple region in the frequency distribution corresponding to the well image of interest and the average frequency distribution of the purple region of all well images 13p whose sum of the frequencies of the purple region is equal to or greater than a certain value is taken as the similarity of the purple region of the well image of interest. Similarly, the similarity between the frequency distribution of the yellow region in the well image of interest and the average frequency distribution of the yellow region of all well images 13p whose sum of the frequencies of the yellow region is equal to or greater than a certain value is taken as the similarity of the yellow region of the well image of interest. As a result, a set of similarities of the purple region and the yellow region is obtained for each well image 13p, and a similarity distribution for purple and yellow is obtained.

[0084] Alternatively, the similarity between the frequency distribution corresponding to the well image of interest and the average frequency distribution of all well images 13p whose sum of the frequencies of the purple regions is equal to or greater than a certain value may be taken as the similarity of the well image of interest to purple. Similarly, the similarity between the frequency distribution corresponding to the well image of interest and the average frequency distribution of all well images 13p whose sum of the frequencies of the yellow regions is equal to or greater than a certain value may be taken as the similarity of the well image of interest to purple. In this case, similarity distributions for purple and yellow are also obtained.

[0085] FIG. 8 is a graph showing an example of a similarity distribution for purple and yellow. Specifically, the graph is obtained by mapping a set of the similarity of purple and yellow calculated for each well image 13p to the similarity of purple and yellow. The horizontal axis of FIG. 8 shows the similarity of purple, and the larger the number, the more similar it is to purple. The vertical axis shows the similarity of yellow, and the larger the number, the more similar it is to yellow.

[0086] The first determination unit 42d performs a primary determination of the color of each well image 13p as purple or yellow based on the acquired similarity distribution. For example, the well image 13p may be determined as purple or yellow using a threshold value for distinguishing between purple and yellow. The threshold value may be set by the evaluator by presenting the graph shown in FIG. 8 to the evaluator, or may be automatically set by the first determination unit 42d using a hierarchical clustering method such as the Ward method in cluster analysis, a non-hierarchical clustering method such as K-means, a linear discriminant method, or the like.

[0087] The first reference creation unit 42e creates a first reference frequency distribution for each of purple and yellow based on the frequency distribution (frequency distribution created by the frequency distribution creation unit 42c) of multiple well images 13p that are judged to be purple and yellow by the first judgment unit 42d, among all well images 13p contained in the evaluation plate 10.

[0088] In this embodiment, a first reference frequency distribution for purple is created based on an average frequency distribution obtained by averaging the frequency distributions of multiple well images 13p determined to be purple. Similarly, a first reference frequency distribution for yellow is created based on an average frequency distribution obtained by averaging the frequency distributions of multiple well images 13p determined to be yellow.

[0089] The second judgment unit 42f makes a secondary judgment of all well images 13p as purple or yellow based on the first reference frequency distributions for purple and yellow created by the first reference creation unit 42e and the frequency distributions of all well images 13p contained in the evaluation plate 10.

[0090] Specifically, the second judgment unit 42f calculates the similarity between the first reference frequency distribution for each of purple and yellow and the frequency distribution of all well images 13p. Since the first reference frequency distribution for each color is used, similarity can be calculated for the entire region without dividing it into purple and yellow regions, as in the example of the frequency distribution creation method in the frequency distribution creation unit 42c. A set of similarities for purple and yellow is obtained for each well image 13p, and a similarity distribution for purple and yellow is obtained.

[0091] 9 is a graph of the similarity distribution obtained by the second determination unit 42f. The horizontal and vertical axes in FIG. 9 are the same as those in FIG.

[0092] The second judgment unit 42f judges each well image 13p as purple or yellow based on the acquired similarity distribution. For example, the well image 13p may be judged as purple or yellow using a threshold value for distinguishing between purple and yellow. The threshold value may be set by the evaluator by presenting the graph shown in FIG. 9 to the evaluator, or may be automatically set by the second judgment unit 42f using a hierarchical clustering method such as the Ward method in cluster analysis, a non-hierarchical clustering method such as K-means, a linear discriminant method, or the like.

[0093] The second reference creation unit 42g creates a second reference frequency distribution for each of the partition images Se1p to Se8p, which are images corresponding to the partitions Se1 to Se8 of the evaluation plate 10, based on the frequency distribution (frequency distribution created by the frequency distribution creation unit 42c) of multiple well images 13p judged to be purple and yellow, respectively, by the second judgment unit 42f.

[0094] Specifically, when any one of the section images Se1p to Se8p is referred to as a section image of interest, a second reference frequency distribution of purple is created based on an average frequency distribution obtained by averaging the frequency distributions of a plurality of well images 13p determined to be purple among a plurality of well images 13p (48 well images 13p in this embodiment) belonging to the section image of interest. Similarly, a second reference frequency distribution of yellow is created based on an average frequency distribution obtained by averaging the frequency distributions of a plurality of well images 13p determined to be yellow in the section image of interest. The section image of interest corresponds to the i-th section image (i is an integer between 2 and N) among the first to N-th section images which are images corresponding to the first to N-th sections when the plurality of sections set on the plate 11 are expressed as the first to N-th sections as described above.

[0095] The third judgment unit 42h performs a tertiary judgment on the colors of the multiple well images 13p belonging to each of the partition images Se1p to Se8p based on the second reference frequency distributions for purple and yellow created by the second reference creation unit 42g and the frequency distributions of the multiple well images 13p belonging to each of the partition images Se1p to Se8p. This results in a tertiary judgment on the colors of all wells 13 of the evaluation plate 10. In this embodiment, the color judgment result by the third judgment unit 42h is the final judgment result.

[0096] As described above, any one of the section images Se1p to Se8p is referred to as a section image of interest, and the determination method performed by the third determination unit 42h will be specifically described.

[0097] First, the similarity distribution in the well images 13p is obtained in the same manner as in the second judgment unit 42f, except that the frequency distribution of all well images 13p in the target section image is used instead of the frequency distribution of all well images 13p in the second judgment unit 42f, and the second frequency distribution is used instead of the first reference frequency distribution. FIG. 10 is a graph of the similarity distribution for each section image Se1p to Se8p. Se1p to Se8p in FIG. 10 indicate the section images Se1p to Se8p, respectively. The horizontal and vertical axes in FIG. 10 are the same as those in FIG. 8. The white circles in FIG. 10 correspond to well images 13p judged to be yellow, and the black circles correspond to well images 13p judged to be purple.

[0098] The third judgment unit 42h judges the color of all well images 13p in the partition images Se1p to Se8p based on the acquired similarity distribution. For example, the color of the well images 13p may be judged using a threshold value for separating purple and yellow. The threshold value may be set by the evaluator by presenting each graph shown in FIG. 10 to the evaluator, or may be automatically set by the third judgment unit 42h using a hierarchical clustering method such as Ward's method in cluster analysis, a non-hierarchical clustering method such as K-means, a linear discrimination method, or the like. As described above, the judgment result of the third judgment unit 42h is the final judgment result in this embodiment. Therefore, in the judgment of the third judgment unit 42h, for example, when the above threshold value is used, the color of the well image 13p (or well 13) located in the area on or near the threshold value may be judged to be an intermediate color between purple and yellow.

[0099] Since the colors of all well images 13p in the target section image are determined, the colors of all wells 13 in the evaluation plate 10 are determined.

[0100] The third determination unit 42h may create data showing the determination result. Specifically, it may create output image data in which each well image 13p (or well 13) is displayed in a color corresponding to the determined color (purple or yellow). In this case, the intermediate color may be displayed as, for example, green.

[0101] An example of an evaluation method (specimen evaluation method) for the evaluation plate 10 using the above-mentioned first color judgment process will be described. FIG. 11 is an example of a flowchart of the evaluation method using the first color judgment process. When the evaluation method using the first color judgment process is carried out, the calculation unit 42 functions as the calculation unit 42A. Here, a process after the process (input process) in which the image acquired by the imaging unit 33 of the observation device 30 is accepted by the input unit 41 will be described.

[0102] When an image from the imaging unit 33 is accepted by the input unit 41, the image processing device 40 generates a plate image P1 by synthesizing multiple images from the imaging unit 33 accepted by the input unit 41 using the plate image creation unit 42a (plate image creation process S11).

[0103] After the plate image creating step S11, the image processing device 40 converts the plate image P1 into an HSV image P2 by the HSV image creating section 42b (HSV image creating step S12).

[0104] After the HSV image creation step S12, the image processing device 40 creates a frequency distribution for each well image 13p by the frequency distribution creation unit 42c (frequency distribution creation step S13). In the frequency distribution creation step (statistical information acquisition step) S13, multiple frequency distributions corresponding to all well images 13p in the plate image P1 are created.

[0105] After the frequency distribution creation step S13, the image processing device 40 uses the first judgment unit 42d to calculate the similarity between the multiple frequency distributions created in the frequency distribution creation step S13, and makes a primary judgment as to whether the color of all of the well images 13p is purple or yellow (first judgment step S14). The method of calculating the similarity is as described above. By calculating the similarity between the multiple frequency distributions, a similarity distribution for purple and yellow is obtained, as described above. In the obtained similarity distribution, the color of all of the well images 13p is determined to be purple or yellow using a threshold value that separates purple and yellow.

[0106] After the first judgment step S14, the image processing device 40 creates a first reference frequency distribution (first reference statistical information) for each of purple and yellow by the first reference creation unit 42e (first reference creation step S15). The method of creating the first reference frequency distribution is the same as that described for the first reference creation unit 42e.

[0107] After the first reference creation step (first reference statistical information acquisition step) S15, the image processing device 40 uses the first reference frequency distribution to perform a second judgment of the colors of the multiple well images 13p by the second judgment unit 42f (second judgment step S16). In this embodiment, the similarity between each of the first reference frequency distributions for purple and yellow and the frequency distribution of all well images 13p is calculated. In the second judgment step S16, the similarity is calculated for the entire range of hue and saturation in the first reference frequency distribution. As a result, a set of similarities for purple and yellow is obtained for each well image 13p, so that a similarity distribution for purple and yellow is obtained. In the similarity distribution thus obtained, the color of all well images 13p is secondarily judged to be purple or yellow using a threshold value that separates purple and yellow. The threshold value used in the second judgment step S16 may be set by an evaluator as described above, or may be automatically set by the second judgment unit 42f.

[0108] After the second judgment step S16, the image processing device 40 creates a second reference frequency distribution for each of purple and yellow for each of the partition images Se1p to Se8p by the second reference creation unit 42g (second reference creation step S17). In this embodiment, for each of the partition images Se1p to Se8p, the second reference frequency distribution for purple (second reference statistical information) is created based on an average frequency distribution obtained by averaging the frequency distributions of the well images 13p determined to be purple in the second judgment step among the well images 13p (48 well images 13p in this embodiment) belonging to each of the partition images Se1p to Se8p. In the same manner, the second reference frequency distribution for yellow (second reference statistical information) is created.

[0109] After the second reference creation step (second reference statistical information acquisition step) S17, the image processing device 40 performs a tertiary judgment (final judgment) of the color of the well image 13p for each of the partition images Se1p to Se8p by the third judgment unit 42h (third judgment step S18). The third judgment step S18 in this embodiment will be specifically described.

[0110] First, for each of the partition images Se1p to Se8p, a similarity distribution is obtained for a plurality of well images 13p belonging to each of the partition images Se1p to Se8p in the same manner as in the second judgment step S16. Based on the obtained similarity distribution, a threshold value for dividing the well images 13p into purple and yellow is used to perform a tertiary judgment of the colors of all the well images 13p belonging to each of the partition images Se1p to Se8p. The threshold value may be set by the evaluator or may be automatically set by the third judgment unit 42h. When the above threshold value is used, the color of the well images 13p located in the area on or near the threshold value is judged to be an intermediate color between purple and yellow.

[0111] Since the colors of the multiple well images 13p are determined for each of the section images Se1p to Se8p, the colors of all of the wells 13 in the evaluation plate 10 are determined.

[0112] After the third determination step S18, the image processing device 40 may output the result of the third determination step S18 by the output unit 43 so as to be presented to the evaluator (output step).

[0113] [Second color judgment process] The calculation unit 42 when performing the second color judgment process is referred to as calculation unit 42B. Fig. 12 is a functional block diagram of calculation unit 42B. The calculation unit 42 has a plate image creation unit 42a, an HSV image creation unit 42b, a frequency distribution creation unit 42c, and a judgment unit 42i. The plate image creation unit 42a, the HSV image creation unit 42b, and the frequency distribution creation unit 42c are similar to those of calculation unit 42A, so their explanation will be omitted.

[0114] The determination unit 42i determines the color of the well image 13p by using the third reference frequency distribution (reference statistical information of color candidates) for each of purple and yellow, which are stored in advance in the image processing device 40. In this embodiment, the third reference frequency distribution used by the determination unit 42i is acquired for each of the section images Se1p to Se8p, similarly to the second reference frequency distribution. A method for creating the third reference frequency distribution will be described later. In this embodiment, the determination unit 42i determines the color of the well image 13p for each of the section images Se1p to Se8p in the same manner as the third determination unit 42h in the calculation unit 42, except that the determination unit 42i uses the third reference frequency distribution instead of the second reference frequency distribution. In this manner, the colors of all of the wells 13 in the evaluation plate 10 are determined.

[0115] Fig. 13 is a graph of the similarity distribution for each of the section images Se1p to Se8p acquired by the determination unit 42i. The horizontal and vertical axes in Fig. 13 are the same as those in Fig. 10. Fig. 13 shows the results when the same plate image P1 as in Fig. 10 is used. Se1p to Se8p in Fig. 13 respectively indicate the section images Se1p to Se8p.

[0116] A method for creating the third reference frequency distribution will be described. The calculation unit 42 when creating the third reference frequency distribution is referred to as the calculation unit 42C. Fig. 14 is a functional block diagram of the calculation unit 42C. The calculation unit 42C has a plate image creation unit 42a, an HSV image creation unit 42b, a frequency distribution creation unit 42c, and a reference creation unit (reference statistical information acquisition unit) 42j.

[0117] Here, a case will be described in which, in order to create the third reference frequency distribution, images of three evaluation plates 10 prepared under the same conditions are acquired by the observation device 30 and input to the image processing device 40 via the input unit 41, and the visual judgment results of the three evaluation plates 10 by an evaluator are input to the image processing device 40. At least the cells, indicator, culture medium, and S9mix in the plates under the same conditions are preferably the same as those used in the actual evaluation of the specimen.

[0118] Plate image creation section 42a creates plate images P1 of three evaluation plates 10 in the same manner as plate image creation section 42a in calculation section 42A.

[0119] The HSV image creation section 42b converts each of the three plate images P1 into an HSV image P2 in the same manner as the HSV image creation section 42b in the calculation section 42A.

[0120] The frequency distribution creating section 42c creates a frequency distribution for each well image 13p for each of the three HSV images P2 created by the HSV image creating section 42b, in the same manner as the frequency distribution creating section 42c in the calculation section 42A.

[0121] The reference creating unit 42j creates a third reference frequency distribution for purple and yellow for each of the section images Se1p to Se8p based on the frequency distribution created by the frequency distribution creating unit 42c and the visual judgment result of the evaluator. As described above, any one of the section images Se1p to Se8p will be referred to as a section image of interest.

[0122] All wells 13 belonging to the target section in the three evaluation plates 10 are divided into purple and yellow based on the visual judgment result. The reference creation unit 42j creates a third reference frequency distribution for purple by creating an average frequency distribution of the frequency distributions of all well images 13p divided into purple. Similarly, the reference creation unit 42j creates a third reference frequency distribution for yellow by creating an average frequency distribution of the frequency distributions of all well images 13p divided into yellow.

[0123] Here, the case where three evaluation plates 10 under the same conditions are used has been described, but the number of evaluation plates 10 is not limited. The more the number of evaluation plates 10, the more accurate the judgment. The third reference power distribution only needs to be created once. Alternatively, for example, the second reference power distribution previously created in the first color judgment method may be used as the third reference power distribution in the second color judgment process.

[0124] The third reference power distribution may be created for the entire evaluation plate 10, similarly to the first reference power distribution.

[0125] Next, an example of an evaluation method (sample evaluation method) of the evaluation plate 10 using the second color judgment process will be described. Fig. 15 is an example of a flowchart of the evaluation method using the second color judgment process. When the evaluation method using the second color judgment process is performed, the calculation unit 42 functions as the calculation unit 42B.

[0126] Here, a process after a process (input process) in which an image acquired by the imaging unit 33 of the observation device 30 is received by the input unit 41 will be described. A case will be described in which third reference frequency distributions for purple and yellow are generated in advance by the reference generating unit 42j shown in Fig. 14 and stored in the image processing device 40.

[0127] As in the case shown in FIG. 11, the plate image creating step S11, the HSV image creating step S12, and the frequency distribution creating step S13 are carried out.

[0128] After the frequency distribution creation step S13, the image processing device 40 judges the colors of all the well images 13p by the judgment unit 42i (judgment step S21). In this embodiment, the colors of the well images 13p are judged using the third reference frequency distributions for purple and yellow, which are stored in advance in the image processing device 40. In this embodiment, the third reference frequency distribution is acquired for each of the partition images Se1p to Se8p, similar to the second reference frequency distribution described in the second reference creation step S17 shown in FIG. 11. In this embodiment, in the judgment step S21, the colors of the well images 13p are judged for each of the partition images Se1p to Se8p in the same manner as in the third judgment step S18 of FIG. 11, except that the third reference frequency distribution is used instead of the second reference frequency distribution, thereby judging the colors of all the wells 13 of the evaluation plate 10.

[0129] After the determination step S21, the image processing device 40 may output the result of the determination step S21 by the output unit 43 so as to be presented to the evaluator (output step).

[0130] [Colony detection process] The calculation unit 42 when performing the colony detection process is referred to as the calculation unit 42D, and the calculation unit 42D will be described with reference to Figs. 16 and 17. Fig. 16 is a functional block diagram of the calculation unit 42D. Fig. 17 is a diagram for explaining the colony detection process by the calculation unit 42D. The calculation unit 42 has a plate image creation unit 42a, a well image identification unit 42k, a smoothing unit 42l, an inner region setting unit 42m, a first colony detection unit 42n, an outer region setting unit 42o, a second colony detection unit 42p, and a detection result creation unit 42q.

[0131] Plate image creation section 42a creates plate image P1 based on the image of evaluation plate 10 input to image processing device 40 in the same manner as plate image creation section 42a in calculation section 42A.

[0132] The well image specifying unit 42k specifies a well image (evaluation region) 13p, which is a region corresponding to the well 13, in the plate image P1. Specifically, a region of interest (ROI) may be set in the well image 13p. For illustrative purposes, (a) of FIG. 17 shows one well image 13p in which an ROI is set.

[0133] The smoothing unit 42l smoothes the well image 13p. The smoothing may be performed by a known method in image processing. (b) of Fig. 17 shows the result of smoothing the well image 13p of (a) of Fig. 17.

[0134] The internal region setting unit 42m identifies the internal region IA of the well image 13p. For example, as shown by the dashed line in FIG. 17(c), an ROI may be set in the internal region IA. FIG. 17(c) is a diagram showing a state in which an ROI is set in the internal region IA for the image in FIG. 17(b). The setting region of the ROI may be specified by the evaluator, or may be automatically set by the internal region setting unit 42m based on information (such as dimensions, arrangement state of the wells 13) of the plate 11 used in the evaluation plate 10 and brightness information of the well image 13p. The internal region IA of the well image 13p is a region corresponding to the bottom 13a of the well 13 when the evaluation plate 10 is viewed from its thickness direction. In other words, it is a region directly above the bottom 13a in the well 13 (see FIG. 2).

[0135] The first colony detection unit 42n detects the presence or absence of a colony in the internal area IA by identifying a change in shape of the evaluation object 14 in the internal area IA (a change in shape due to a colony). In this embodiment, when a colony is formed, the colony is detected by utilizing the fact that the colony is darker than its surroundings.

[0136] Specifically, the specified internal area IA is binarized based on a brightness threshold, and an area with a brightness value smaller (darker) than the set brightness threshold is detected as a colony candidate. The brightness threshold may be specified by an evaluator according to the brightness (brightness value) of the internal area IA, or may be set in advance based on past data or the like. In this embodiment, among the above-mentioned colony candidates, a candidate that further satisfies the shape (circularity) and size assumed as a colony is detected as a colony. The detection of a colony from a colony candidate based on the above-mentioned shape and size may be performed by an evaluator by looking at an image in which the colony candidate is presented, or may be determined by the first colony detection unit 42n based on the shape (circularity) and size set in advance.

[0137] In the section Se1, which is the solvent control section, a colony is not actually detected for wells 13 that are determined to be purple by color judgment. Therefore, for example, when a color judgment process is performed in advance and a colony candidate is identified in a well image 13p that is determined to be purple by color judgment of the section image Se1p (solvent control region) corresponding to the section Se1, the size of the largest colony candidate among the colony candidates in the section images Se2p to Se8p may be used as a reference. That is, in the section images Se2p to Se8p, a colony candidate that is larger than the largest colony candidate in the section image Se1p and meets the shape expected for the colony may be detected as a colony.

[0138] FIG. 17(d) is a diagram showing an example of the detection results of colonies in the internal region IA.

[0139] The external region setting unit 42o identifies the external region OA of the well image 13p. For example, an ROI may be set in the external region OA as shown by the dashed line in FIG. 17(e). FIG. 17(e) is a diagram showing a state in which an ROI has been set in the external region OA. The setting region of the ROI may be specified by the evaluator, or may be automatically set by the external region setting unit 42o based on data of the plate 11, etc. The external region OA of the well image 13p is the region of the wall of the well 13 (see FIG. 2).

[0140] The second colony detection unit 42p detects the presence or absence of a colony in the external area OA by identifying a change in shape (change due to a colony) of the evaluation object 14 in the external area OA. In this embodiment, similar to the first colony detection unit 42n, when a colony is formed, the inside of the colony is dark and the surrounding area is bright, and thus the colony is detected.

[0141] Specifically, the specified external area OA is binarized based on a brightness threshold, and an area with a brightness value smaller (darker) than the set brightness threshold is detected as a colony candidate. The brightness threshold may be specified by the evaluator according to the brightness (brightness value) of the external area OA, or may be set in advance based on past data, etc. Since the external area OA is darker than the internal area IA, the brightness threshold for detecting a colony in the external area OA is different from that in the internal area IA. In this embodiment, in the above-mentioned colony candidates, candidates that further satisfy the shape (circularity) and size assumed as a colony are detected as colonies, similar to the case of the first colony detection unit 42n. The conditions such as the shape and size may be different from those in the internal area IA. Furthermore, for the size, the result of the section image Se1p corresponding to the section Se1, which is the solvent control section, may be used, similar to the case of the first colony detection unit 42n.

[0142] FIG. 17(f) is a diagram showing an example of the detection results of colonies in the outer area OA.

[0143] The detection result creation unit 42q combines the results of the first colony detection unit 42n and the second colony detection unit 42p to create data indicating the result of colony detection. For example, the detection result creation unit 42q can create data in which the areas of the colonies detected by the first colony detection unit 42n and the second colony detection unit 42p are emphasized (for example, colored).

[0144] The calculation unit 42D outputs the detection result created by the detection result creation unit 42q as the final colony detection result to the output unit 43. (g) of FIG. 17 is a diagram showing an example of the final detection result by the calculation unit 42.

[0145] Next, an example of a method for evaluating the evaluation plate 10 using the colony detection process (method for evaluating a specimen) will be described. FIG. 18 is an example of a flowchart of the method for evaluating the evaluation plate 10 using the colony detection process. When the method for evaluating the evaluation plate 10 using the colony detection process is performed, the calculation unit 42 functions as the calculation unit 42D. Here, a process after the process (input process) in which the image acquired by the imaging unit 33 of the observation device 30 is accepted by the input unit 41 will be described.

[0146] As in the case shown in FIG. 11, the plate image creating step S11 is carried out.

[0147] After the plate image creation step S11, the image processing device 40 uses the well image specification unit 42k to specify each well image 13p in the plate image P1 (well image specification step S31). In this embodiment, an ROI is set for each well image 13p in the plate image. The ROI may be specified by the evaluator or may be automatically set by the well image specification unit 42k.

[0148] After the well image specifying step S31, the image processing device 40 smoothes the well image 13p by the smoothing unit 42l (smoothing step S32).

[0149] After the smoothing step S32, the image processing device 40 sets an internal area IA of each well image 13p by the internal area setting unit 42m (internal area setting step S33). In this embodiment, an ROI is set in the internal area IA of each well image 13p. The ROI may be specified by the evaluator or may be automatically set by the internal area setting unit 42m.

[0150] After the internal region setting step S33, the image processing device 40 detects the presence or absence of a colony in the internal region IA of each well image 13p by the first colony detection unit 42n (first colony detection step S34). In this embodiment, as described with respect to the first colony detection unit 42n, the presence or absence of a colony is detected by using the shape (circularity) and size as well as the brightness value.

[0151] After the first colony detection step S34, the image processing device 40 sets an external area OA of each well image 13p by the external area setting unit 42o (external area setting step S35). In this embodiment, an ROI is set in the external area OA of each well image 13p. The ROI may be specified by the evaluator or may be automatically set by the external area setting unit 42o.

[0152] After the external region setting step S35, the image processing device 40 detects the presence or absence of a colony in the external region OA of each well image 13p by the second colony detection unit 42p (second colony detection step S36). In this embodiment, as described with respect to the second colony detection unit 42p, the presence or absence of a colony is detected using the shape (circularity) and size as well as the brightness value.

[0153] After the second colony detection step S36, the image processing device 40 uses the detection result creation unit 42q to integrate the detection results of the first colony detection step S34 and the second colony detection step S36 to create a final detection result of the colony detection (detection result creation step S37).

[0154] After the detection result creation step S37, the image processing device 40 may output the detection result created in the detection result creation step S37 by the output unit 43 so as to present it to the evaluator (output step).

[0155] [Transparency change detection process] The calculation unit 42 when performing the transparency change detection process is referred to as calculation unit 42E, and the calculation unit 42E will be described. Fig. 19 is a functional block diagram of the calculation unit 42E. The calculation unit 42E has a plate image creation unit 42a, an HSV image creation unit 42b, a frequency distribution creation unit 42c, a section frequency distribution creation unit 42r, and a transparency change detection unit 42s. The plate image creation unit 42a, the HSV image creation unit 42b, and the frequency distribution creation unit 42c are the same as those in the calculation unit 42A, so their description will be omitted.

[0156] The section frequency distribution creating unit 42r creates a frequency distribution (hereinafter referred to as a "section frequency distribution") with the hue and saturation as variables for each of the section images Se1p to Se8p. When any one of the section images Se1p to Se8p is referred to as a section image of interest as described above, the section frequency distribution creating unit 42r creates a section frequency distribution of the section image of interest based on the average frequency distribution of the frequency distributions of all well images 13p belonging to the section image of interest. In addition, when the color judgment process has already been performed, the section frequency distribution of the section image of interest may be created based on the average frequency distribution of the frequency distributions of only well images 13p judged to be purple, for example (for example, only well images 13p judged to be purple in the third judgment step S18 (see FIG. 11) of the first color judgment process) among all well images 13p belonging to the section image of interest.

[0157] The transparency change detection unit 42s calculates the similarity between the section frequency distributions corresponding to the section images Se1p to Se8p. The transparency change detection unit 42s detects the change in transparency of the other section images Se2p to Se8p by using the transparency of the section image Se1p corresponding to the section Se1, which is the solvent control section, as a reference frequency distribution for detecting a change in transparency (reference statistical information for detecting a change in transparency). Specifically, the similarity between the section frequency distribution of the section image Se1p and the section frequency distributions of the other section images Se2p to Se8p is calculated. This similarity corresponds to the change in transparency of the other sections Se2 to Se8 relative to the transparency of the section Se1. FIG. 20 is a diagram showing the similarity of the section frequency distributions between the section images Se1p to Se8p, and (a) and (b) show different examples. (a) of FIG. 20 shows an example in which the transparency change is large, and (b) of FIG. 20 shows an example in which the transparency change is small enough to be practically unnoticeable. Se1p to Se8p in the chart respectively indicate the section images Se1p to Se8p. In Fig. 20, the row and column of Se1p indicate the similarity (change in transparency) between the transparency of the section image Se1p and the other section images Se2p to S8p and the section image Se1p itself.

[0158] Next, an example of an evaluation method (specimen evaluation method) of the evaluation plate 10 using a transparency change detection process will be described. FIG. 21 is an example of a flowchart of the evaluation method using a transparency change detection process. When the evaluation method using a transparency change detection process is performed, the calculation unit 42 functions as the calculation unit 42E. Here, a process after a process (input process) in which an image acquired by the imaging unit 33 of the observation device 30 is accepted by the input unit 41 will be described.

[0159] As in the case shown in FIG. 11, the plate image creating step S11, the HSV image creating step S12, and the frequency distribution creating step S13 are carried out.

[0160] After the frequency distribution creating step S13, the image processing device 40 creates a frequency distribution for each of the partition images Se1p to Se8p by the partition frequency distribution creating unit 42r (partition frequency distribution creating step S41). In this embodiment, if any one of the partition images Se1p to Se8p is referred to as a partition image of interest as described above, the partition frequency distribution of the partition image of interest is created based on the average frequency distribution of the frequency distributions of all wells 13 belonging to the partition image of interest in the partition frequency distribution step S41. As described above, if the color determination process has already been performed, the partition frequency distribution of the partition image of interest may be created based on the average frequency distribution of the frequency distributions of only well images 13p determined to be purple, for example, out of all well images 13p belonging to the partition image of interest.

[0161] After the section frequency distribution creation step S41, the image processing device 40 detects a transparency change of the other section images Se2p to Se8p relative to the section image Se1p (solvent control region) by the transparency change detection unit 42s (transparency change detection step S42). In this embodiment, the similarity between the section frequency distribution of the section image Se1p and the section frequency distribution of the other section images Se2p to Se8p is calculated. This similarity corresponds to the change in transparency of the other sections Se2 to Se8 relative to the transparency of the section Se1.

[0162] After the transparency change detection step S42, the image processing device 40 may output the detection result of the transparency change detection step S42 by the output unit 43 so as to be presented to the evaluator (output step).

[0163] [Foreign object detection processing] The calculation unit 42 when performing the foreign object detection process is referred to as the calculation unit 42F, and the calculation unit 42F will be described. FIG. 22 is a functional block diagram of the calculation unit 42F. FIG. 23 is a diagram for explaining the foreign object detection process by the calculation unit 42F. The calculation unit 42F has a plate image creation unit 42a, a well image identification unit 42k, a smoothing unit 42l, a first foreign object detection unit 42t, an external region setting unit 42o, a second foreign object detection unit 42u, and a detection result creation unit 42v. The plate image creation unit 42a, the well image identification unit 42k, and the smoothing unit 42l are the same as those in the calculation unit 42D, and therefore their description will be omitted. An example of a well image 13p identified by the well image identification unit 42k is shown in FIG. 23(a). FIG. 23(b) is a diagram after the well image 13p in FIG. 23(a) has been smoothed by the smoothing unit 42l.

[0164] The first foreign object detector 42t detects foreign objects using the first brightness value for the well image 13p smoothed by the smoother 42l. Usually, the inside of the foreign object is bright (high brightness) and the periphery of the foreign object is dark (low brightness). Therefore, a region where the inside is bright and the periphery is dark becomes a foreign object candidate. Therefore, a first brightness threshold value for detecting foreign objects is set and the well image 13p is binarized. In the binarized image, a region surrounded by a brightness value smaller than the first brightness threshold value and whose inside is equal to or greater than the first brightness value is detected as a foreign object candidate. The first brightness threshold value may be set to the brightness value of the internal region of the well image 13p (see FIG. 2). The brightness value (first brightness value) of the internal region may be specified by the evaluator based on the well image 13p. The internal region may be set in advance by the internal region setting unit 42m, for example. The first brightness value may be set automatically by the first foreign object detector 42t. When the first foreign object detector 42t automatically sets the first brightness value, it may be set automatically based on an average brightness value within a certain range with respect to the center of the well image 13p. Usually, foreign objects are circular and have a certain size. Therefore, in this embodiment, among the foreign object candidates, candidates that further satisfy the shape (circularity) and size assumed as a foreign object are detected as foreign objects. The detection of foreign objects based on the above-mentioned shape and size from the foreign object candidates may be performed by an evaluator looking at an image presenting the foreign object candidates, or may be determined by the first foreign object detector 42t based on the shape (circularity) and size that are set in advance. (c) of FIG. 23 is a diagram showing an example of a foreign object detection result by the first foreign object detector 42t.

[0165] The external region setting unit 42o sets an external region OA in the well image 13p in the same manner as the external region setting unit 42o in the calculation unit 42D (see FIG. 16). The external region setting unit 42o may set an ROI in the external region OA, for example, as shown by the dashed line in FIG. 23(d).

[0166] The second foreign object detection unit 42u detects a foreign object using the second brightness value. The second brightness value is set so that a foreign object can be detected according to the brightness of the external area OA. The second brightness value may be specified by an evaluator based on the brightness (luminance) of the external area OA. The second brightness value may be automatically set by the second foreign object detection unit 42u. When the second foreign object detection unit 42u automatically sets the second brightness value, for example, a value obtained by multiplying the average brightness of the external area OA by a constant value may be set. Since the external area OA is darker (lower in brightness) than the vicinity of the center of the well image 13p, the first brightness threshold value and the second brightness threshold value are usually different.

[0167] The second foreign object detection unit 42u performs binarization processing on the external area OA using the second brightness threshold, and detects an area in the binarized image having a brightness value greater or smaller than the second brightness value as a foreign object candidate. In the present embodiment, the second foreign object detection unit 42u further detects, as a foreign object, a candidate that further meets the shape (circularity) and size expected as a foreign object among the detected foreign object candidates. The detection of a foreign object from the foreign object candidates based on the shape and size may be performed, for example, by an evaluator looking at an image presenting the foreign object candidates, or may be determined by the second foreign object detection unit 42u based on a shape (circularity) and size that are set in advance. The shape and size may be the same as or different from those in the first foreign object detection unit 42t. FIG. 23(e) is a diagram showing an example of a foreign object detection result by the second foreign object detection unit 42u.

[0168] The detection result creation unit 42v combines the results of the first foreign object detection unit 42t and the second foreign object detection unit 42u to create data indicating the result of foreign object detection. For example, the detection result creation unit 42v can create data in which the areas of the foreign objects detected by the first foreign object detection unit 42t and the second foreign object detection unit 42u have been emphasized (e.g., colored).

[0169] The calculation unit 42 outputs the detection result created by the foreign object detection result creation unit 42v as the final foreign object detection result to the output unit 43. (f) of Figure 23 is a diagram showing an example of the final detection result by the calculation unit 42.

[0170] Next, an example of an evaluation method (sample evaluation method) for the evaluation plate 10 using a foreign matter detection process will be described. Fig. 24 is an example of a flowchart of the evaluation method using the foreign matter detection process. When the evaluation method using the foreign matter detection process is performed, the calculation unit 42 functions as the calculation unit 42F. Here, a process after the process (input process) in which the image acquired by the imaging unit 33 of the observation device 30 is accepted by the input unit 41 will be described.

[0171] As in the case shown in FIG. 18 (the case of colony detection processing), the plate image creating step S11, the well image specifying step S31, and the smoothing step S32 are performed.

[0172] After the smoothing step S32, the image processing device 40 detects the presence or absence of a foreign object in the well image 13p by the first foreign object detector 42t (first foreign object detection step S51). In this embodiment, as described with respect to the first foreign object detector 42t, a foreign object is detected based on the first brightness threshold, shape (circularity), and size.

[0173] After the first foreign matter detection step S51, the image processing device 40 sets an outer region OA of the well image 13p by the outer region setting unit 42o in the same manner as in the outer region setting step S35 shown in Fig. 18 (outer region setting step S52). In this embodiment, an ROI is set in the outer region OA. The setting region of the ROI may be specified by the evaluator, or may be automatically set by the outer region setting unit 42o based on the data of the plate 11.

[0174] After the external region setting step S52, the image processing device 40 detects the presence or absence of a foreign object in the well image 13p by the second foreign object detection unit 42u (second foreign object detection step S53). In this embodiment, as described with respect to the second foreign object detection unit 42u, a foreign object is detected based on the second brightness threshold, shape (circularity), and size.

[0175] After the second foreign object detection step S53, the image processing device 40 uses the detection result creation unit 42v to integrate the results of the first foreign object detection step S51 and the second foreign object detection step S53 to create a final result of foreign object detection (detection result creation step S54). In this embodiment, a process is performed to highlight the areas of foreign objects detected in the first foreign object detection step S51 and the second foreign object detection step S53, such as by coloring the areas of foreign objects.

[0176] After the detection result creation step S54, the image processing device 40 may output the detection result created in the detection result creation step S54 by the output unit 43 so as to present it to the evaluator (output step).

[0177] Next, an example of the procedure of the Ames test using the plate 11 will be specifically described.

[0178] (Step 1A) The test specimen to be used (bacteria having an amino acid auxotrophic mutation, for example, Salmonella typhimurium: TA100 strain, TA98 strain, TA1535 strain, TA1537 strain, Escherichia coli: WP2uvrA strain, etc.) is inoculated into 60 mL of nutrient broth medium and grown by shaking culture at 37°C for 9 to 10 hours.

[0179] (2nd A process) The required amount of sample (e.g., a compound to be evaluated for toxicity) is weighed out.

[0180] (3rd A process) The specimen prepared in step 2A is dissolved or suspended in an appropriate medium, and then serially diluted with the medium to prepare a concentration series of specimen liquid. For example, if the same specimen is to be evaluated at six different concentrations, six specimen liquids with different concentrations are prepared.

[0181] (4th A process) A solution of a positive control compound known to exhibit a given toxicity is prepared.

[0182] (Step 5A) Place 10 μL each of the medium used in preparing the sample solution, the concentration series of the sample solution, and the solution of the positive control compound (collectively referred to as sample solutions, etc.) into separate wells of a 24-well plate (a plate having 24 wells).

[0183] (Step 6A) 0.175-0.7 mL of the culture medium for the specimen (usually 1 × 10 9 The mixture is called the prepared bacterial solution under conditions without metabolic activation (a concentration exceeding 1000 cells / mL) and 6.83 to 6.3 mL of medium (a minimal medium supplemented with a carbon source and lacking amino acids).

[0184] (Step 7A) To a 24-well plate containing 10 μL of a sample solution, etc., 240 μL of the bacterial solution prepared in step 6A under conditions without metabolic activation is added.

[0185] (8th A process) The 24-well plate containing the mixture from step 7A is shaken at 37° C. and 250 rpm for 90 minutes (preincubation).

[0186] (Step 9A) After the preincubation, add 2.6 mL of indicator medium (minimal medium containing a pH indicator and a carbon source and lacking amino acids) to each well (250 µL) of the 24-well plate.

[0187] (Step 10A) 50 μL of the mixture from step 9A is transferred to each of 48 wells (one section) in each of the eight sections Se1 to Se8 of a 384-well plate, plate 11, so that the wells correspond to the sections Se1 to Se8. In the same manner, a total of three or more 384-well plates (plates 11) are seeded.

[0188] (Step 11A) The 384-well plate (plate 11) is incubated at 37° C. for 48 hours. This results in evaluation plate 10.

[0189] In the Ames test, after step 11A, the following steps 12A to 14A are performed. In this embodiment, step 12A and subsequent steps are performed by the evaluation method described above using the evaluation system 20.

[0190] (Step 12A) The color change, the presence or absence of colonies, the presence or absence of foreign matter, and the change in transparency of each well 13 of the evaluation plate 10, which is the plate 11 after the cultivation has been completed, are observed, and the number of wells that had color change, colonies, or foreign matter are counted, and the sections that had a change in transparency are recorded.

[0191] (Step 13A) The presence or absence of cytotoxicity is determined by combining the presence or absence of foreign matter and the results of the change in transparency.

[0192] (Step 14A) The presence or absence of genotoxicity is determined by combining the color change and the presence or absence of colonies. The order of Step 13A and Step 14A is not limited.

[0193] When detecting a specimen that shows toxicity after being metabolized by various enzymes in the living body, in step 6A, an animal liver extract with added coenzymes (S9 mixture) is prepared, and 0.175 to 0.7 mL of the culture solution of the test specimen, 5.78 to 5.25 mL of the medium, and 1.05 mL of the S9 mixture are mixed. The mixture is called the prepared bacterial solution under metabolic activation conditions. After that, the steps from step 7A onwards are carried out using the prepared bacterial solution under metabolic activation conditions instead of the prepared bacterial solution under metabolic activation conditions.

[0194] Next, the effects of the evaluation system 20, the image processing device 40, and the evaluation method described in this embodiment will be described.

[0195] When evaluating the evaluation plate 10, the evaluation plate 10 is imaged using the observation device 30 (imaging step). The image processing device 40 forms a plate image P1 based on the image of the evaluation plate 10 obtained in the imaging step.

[0196] The image processing device 40 performs a color judgment process using the plate image P1 to judge the color of the well image 13p. This judges the color of the well 13 corresponding to the well image 13p. In this embodiment, if the specimen has genotoxicity for the cells contained in the evaluation target 14, the color of the indicator changes from purple to yellow as a positive result. Therefore, the above color judgment process makes it possible to evaluate the genotoxicity of the specimen.

[0197] In the color judgment process described in this embodiment, the color is judged using a frequency distribution with the hue and saturation as variables created for each well image 13p. Therefore, even when it is difficult to judge the difference in color by visual observation, for example, the color can be judged reliably. Therefore, the color can be judged objectively and with high accuracy.

[0198] In the first color judgment process, judgment is performed three times according to the flowchart of the first color judgment process shown in Fig. 11. In this case, as shown in Figs. 8 to 10, as the judgment is repeated, the color of the well image 13p (i.e., the color of the well 13) can be more clearly judged to be purple or yellow.

[0199] As described above, Figures 10 and 13 show the results of the first and second color judgment processes performed on the same plate image P1. As can be seen from Figures 10 and 13, the first color judgment process can more clearly determine the color of the well image 13p (i.e., the color of the well 13) to be purple or yellow.

[0200] The image processing device 40 performs a colony detection process using the plate image P1 to determine the presence or absence of a colony in the well 13. In this embodiment, if the specimen has genotoxicity to the cells contained in the evaluation target 14, colonies may occur. In other words, the detection of a colony corresponds to a positive result. Therefore, the above-mentioned colony detection process makes it possible to evaluate the genotoxicity of the specimen.

[0201] In this embodiment, the well image 13p is divided into an inner area IA and an outer area OA, and a colony is detected in each area. Since the inner area IA and the outer area OA have different brightnesses, by detecting a colony in each of the inner area IA and the outer area OA, it is possible to detect the colony more reliably.

[0202] In the image processing device 40, it is possible to perform a colony detection process together with the color judgment process. In this case, if either the color judgment process or the colony detection process is positive, it is possible to more reliably evaluate the genotoxicity of the specimen by evaluating it as a positive well. In the case of colony detection, a colony (or a colony candidate) detected in the partition image Se1p corresponding to the partition Se1, which is the solvent control partition, is used to detect colonies in the partition images Se2p to Se8p, so that colonies can be detected more accurately. If the color judgment process has already been performed, a colony is not actually detected in the well image 13p determined to be purple by color judgment in the partition image Se1p, so that a colony detected in the well image 13p determined to be purple by color judgment of the partition image Se1p can be regarded as a detection error. Therefore, by detecting a colony larger than the colony detected in the partition image Se1p, it is easy to accurately detect the actual colony to be detected.

[0203] When the image processing device 40 is used to perform the colony detection process together with the color judgment process on the same evaluation plate 10, the data used in the earlier process may be used appropriately in the later process. For example, when the color judgment process is performed before the colony detection process, the plate image P1 created in the color judgment process may be used in the colony detection process. In this case, the plate image creation unit 42a is not required in the calculation unit 42D shown in FIG. 16. Similarly, in the evaluation method using the colony detection process shown in FIG. 18, the plate image creation step S11 can be omitted. The same applies to the case where the colony detection process is performed before the color judgment process. The order of the color judgment process and the colony detection process is not limited.

[0204] The image processing device 40 detects a change in transparency of the partition images Se1p-Se8p by performing a transparency detection process using the plate image P1. This detects a change in transparency of the partitions Se1-Se8. In this embodiment, if the specimen has cytotoxicity to the cells contained in the evaluation object 14, the transparency of the well 13 (specifically, the transparency of the evaluation object 14) changes due to the death of the cells, etc. Therefore, the above-mentioned transparency change detection process makes it possible to evaluate the cytotoxicity of the specimen.

[0205] If the specimen precipitates during culture, the transparency of the well 13 (specifically, the transparency of the evaluation object 14) also changes. Therefore, the transparency change detection process described above makes it possible to evaluate whether or not the specimen precipitates.

[0206] In the transparency change detection process described in this embodiment, the transparency change is determined using a frequency distribution with the hue and saturation as variables created for each well image 13p. Therefore, even when it is difficult to determine the difference in transparency by visual observation, for example, the transparency change can be determined reliably. Therefore, the transparency change can be detected objectively and with high accuracy.

[0207] In this embodiment, the transparency change is calculated based on the similarity between the section frequency distribution for the section image Se1p (solvent control region) in the same plate image P1 and the section frequency distributions of the other section images Se2p to Se8p. This reduces the influence of test conditions, etc., and allows for more accurate detection of transparency.

[0208] The image processing device 40 performs a foreign matter detection process using the plate image P1 to determine the presence or absence of a foreign matter in the well 13. In this embodiment, if the specimen has cytotoxicity to the cells contained in the evaluation target 14, the foreign matter occurs due to the death of the cells. Therefore, the cytotoxicity of the specimen can be evaluated by the above-mentioned foreign matter detection process.

[0209] In this embodiment, after foreign object detection is performed once on the well image 13p, an external area OA is set and foreign objects in the external area OA are detected again using a brightness threshold value (second brightness threshold value) corresponding to the external area OA. Since the external area OA is usually dark, foreign objects are difficult to detect. Therefore, by using a brightness threshold value (second brightness threshold value) corresponding to the external area OA, foreign objects can be detected more reliably.

[0210] In the image processing device 40, it is possible to perform a foreign body detection process in addition to a transparency change detection process. In this case, if a transparency change is detected by the transparency change detection process or if a foreign body is detected by the foreign body detection process, it is possible to more reliably evaluate the cytotoxicity of the specimen by evaluating that there is cytotoxicity.

[0211] The transparency detection process and the foreign matter detection process as cytotoxicity evaluation are used, for example, to set the concentration of the specimen for accurate genotoxicity evaluation. For example, the transparency detection process determines that the concentration of the specimen in the section that gives a positive result among sections Se1 to Se8 is inappropriate for genotoxicity evaluation. Therefore, it is sufficient to perform the transparency detection process or the foreign matter detection process at least once to determine the concentration range in which the cytotoxicity evaluation result is negative, and then perform the genotoxicity evaluation.

[0212] As described above, the transparency detection process or the foreign matter detection process is performed, for example, to evaluate the cytotoxicity of the specimen before the genotoxicity evaluation. Therefore, the conditions of the evaluation plate 10 when the transparency detection process or the foreign matter detection process is performed may be different from those of the evaluation plate 10 when at least one of the color judgment process and the colony detection process for the genotoxicity evaluation is performed (for example, the concentration setting of the specimen may be different).

[0213] However, for example, if the conditions of the evaluation plate 10 when performing the transparency detection process or the foreign matter detection process are the same as those of the evaluation plate 10 when performing at least one of the color judgment process and the colony detection process, data created during the transparency detection process or the foreign matter detection process may be used in the color judgment process and the colony detection process as appropriate.

[0214] For example, when a color judgment process is performed after a transparency detection process, the steps up to the frequency distribution creation step S13 are common, so in the color judgment process, the frequency distribution for each well image 13p created by the frequency distribution creation unit 42c (frequency distribution creation step S13) is used in the transparency detection process. Alternatively, when a colony detection process is performed after a foreign body detection process, the steps up to the well image identification step S31 are common, so in the colony detection process, the colony detection process is performed using the well image 13p identified by the well image identification unit 42k (well image identification step S31). In the process flow when the calculation unit 42D performs the colony detection process or the process flow when the calculation unit 42F performs the foreign body detection process, the calculation unit 42D and the calculation unit 42F do not need to have functions corresponding to the steps omitted as described above.

[0215] For example, as shown in Fig. 1, when judging the colors of 384 wells 13, the burden on the evaluator is large and the evaluator's judgment ability is likely to decrease, so that objectivity and accuracy tend to decrease. In contrast, in the evaluation method using the image processing device 40, the color is judged by the image processing device 40, so that the color can be judged more objectively and accurately than when the evaluator judges the color visually. The same applies to the detection of the presence or absence of colonies, the detection of a change in transparency, and the detection of foreign matter.

[0216] Next, a description will be given of an experimental example 1 in which the evaluation system 20 and the evaluation method described in this embodiment are applied to an Ames test. Hereinafter, unless otherwise specified, the same reference numerals will be used to designate elements similar to those described so far, and duplicated descriptions will be omitted.

[0217] In Experimental Example 1, a View-Plate-384 manufactured by Perkinelmer was used as the plate 11. As shown in Fig. 1, the View-Plate-384 had 384 wells 13. In this Experimental Example 1, as in Fig. 1, sections Se1 to Se8 were virtually set.

[0218] In this Experimental Example 1, the Ames test kit ("Ames MPF") manufactured by Xenometrix, Inc., which includes an indicator, a culture medium, etc. TM The Penta I Microplate Format Mutagenicity Assay was used.

[0219] Nine types of evaluation plates 10 with different test conditions were prepared. The evaluation plates 10 used in this Experimental Example 1 are evaluation plates 10 in a state where toxicity testing is possible, in which cells, specimens, etc. are administered into the wells 13, and cultured according to the manual of the Ames test kit. In this Experimental Example 1, the nine types of evaluation plates 10 with different test conditions are referred to as evaluation plates 10_1 to 10_9.

[0220] The test conditions for the evaluation plates 10_1 to 10_9 were as shown in the table in FIG. 25. In the figure, condition No. 1 to No. 9 indicate the evaluation plates 10_1 to 10_9. In the column "Metabolic activation present or absent," "absent" means that the evaluation target 14 does not contain S9mix, and "present" means that the evaluation target 14 contains S9mix. The column "Concentration (ng / ml)" indicates the concentration of the specimen in the evaluation target 14 (specifically, evaluation targets 14b to 14g) held by the wells 13 in the sections Se2 to Se7. The six concentrations shown in the column "Concentration (ng / mL)" for No. 1 to No. 9 (evaluation plates 10_1 to 10_9) are the concentrations of specimens contained in the evaluation targets 14b to 14g. The concentrations of specimens in the evaluation targets 14b to 14g increase from the evaluation target 14b to the evaluation target 14g. The "concentration" in "positive control compound / solvent (concentration (ng / mL))" indicates the concentration of the positive control compound in compartment Se8 (positive control compartment). DMSO was used as the solvent for the sample solution and the positive control solution, as shown in the figure. The solvent control solution was also DMSO.

[0221] In Experimental Example 1, N sheets of each of the evaluation plates 10_1 to 10_9 were prepared. N was 3 for the evaluation plates 10_1, 10_3, 10_4, and 10_7 to 10_9, and 1 for the rest.

[0222] In this experimental example 1, first, an imaging process was performed in which the evaluation plates 10_1 to 10_9 were imaged using the observation device 30. Then, a color judgment process, a colony detection process, a transparency change detection process, and a foreign matter detection process were performed using the image processing device 40. In addition, the evaluation plates 10_1 to 10_9 were visually judged for color.

[0223] A description will be given of the imaging steps in Experimental Example 1. In Experimental Example 1, two imaging steps (hereinafter referred to as a "first imaging step" and a "second imaging step") were carried out.

[0224] The first imaging step was performed on all of the evaluation plates 10_1 to 10_9. Since the first imaging step is the same for the evaluation plates 10_1 to 10_9, the first imaging step for the evaluation plate 10_1 will be described as a representative example.

[0225] In the first imaging step, the observation device 30 shown in Fig. 3 was used. An LED light that outputs visible light was used for the illumination device 32 of the observation device 30. A lens and a line sensor (CCD camera) were used for the light collecting section and the imager 33a of the imaging section 33, respectively.

[0226] In this experimental example 1, since the imaging range of the imager 33a was smaller than the size of the evaluation plate 10_1, half of the area of ​​the evaluation plate 10_1 was imaged using the imager 33a (and the corresponding light collecting unit 33b) while the evaluation plate 10_1 was transported in one direction by the stage 31. After that, the evaluation plate was shifted by half in a direction perpendicular to the transport direction, and the remaining area of ​​the evaluation plate 10_1 was imaged while moving the evaluation plate 10_1 in one direction again.

[0227] In this experimental example 1, the first imaging step was performed on N evaluation plates 10_1.

[0228] The second imaging step was similarly performed on the evaluation plates 10_1 and 10_4. Since the second imaging steps for the evaluation plates 10_1 and 10_4 are the same, the second imaging step for the evaluation plate 10_1 will be described as a representative example.

[0229] In the second imaging process, in the observation device 30 shown in Figure 3, the positions of the illumination device 32 and the imaging unit 33 relative to the evaluation plate 10_1 were inverted (i.e., in Figure 3, the imaging unit 33 was placed on the plate and the illumination device 32 was placed below the stage 31), and the evaluation plate 10_1 was imaged in the same manner as in the first imaging process.

[0230] In this experimental example 1, the second imaging step was performed on N evaluation plates 10_1.

[0231] Next, in this experimental example 1, a color judgment process was performed on the images acquired in the imaging process using the image processing device 40. In the color judgment process, the images acquired in the first imaging process were used. In addition, a foreign matter detection process was performed on the evaluation plate 10_1. Furthermore, a colony detection process was performed on the evaluation plate 10_4. As described later, the colony detection process and the foreign matter detection process were performed on the images acquired in the first imaging process, and also on the images acquired in the second imaging process.

[0232] [Color Judgment Processing] The color judgment process was performed on all of the evaluation plates 10_1 to 10_9. The method of color judgment for the evaluation plates 10_1 to 10_9 was the same. Therefore, the color judgment process performed on the evaluation plate 10_1 will be described as a representative example.

[0233] The color judgment process was performed using the image processing device 40, following the flow chart of the color judgment process (first color judgment process) shown in Fig. 11. In the first judgment step S14, the calculated similarity distribution was presented to the evaluator in the form of a graph as shown in Fig. 8, and the evaluator set a threshold value for separating purple from yellow. In the second judgment step S16 and the third judgment step S18, threshold values ​​were set in the same manner.

[0234] The results of the color judgment process were as shown in the tables of Figures 26 to 29. Figures 26 to 29 also show the color camera image, the HSV image, and the visual judgment results in addition to the detection results, which are the color judgment results.

[0235] No. 1 to No. 9 in Fig. 26 to Fig. 29 correspond to the evaluation plates 10_1 to 10_9. The result of No. 1 in Fig. 26 is one result out of N evaluation plates 10_1. The same applies to the results of No. 2 to No. 9 in Fig. 26 to Fig. 29.

[0236] The "color camera image" in Figs. 26 to 29 is the plate image P1 created in the plate image creation step S11 shown in Fig. 11. The "HSV image" in Figs. 26 to 29 is the HSV image P2 created in the HSV image creation step S12 shown in Fig. 11. The "visual judgment result" in Figs. 26 to 29 is the result of an evaluator visually judging the evaluation plates 10_1 to 10_9, inputting the results into the image processing device 40 and manually color-coding them into purple and yellow, and displaying the results as an image. The "image processing result" in Figs. 26 to 29 is a diagram showing the result of the color judgment process (first color judgment process) of the image processing device 40.

[0237] The numbers "A to P" and "1 to 24" around each image shown in FIGS. 26 to 29 correspond to the numbers "A to P" and "1 to 24" in FIG.

[0238] In the images of the visual judgment results and the image processing results, the light-colored well images 13p show yellow, and the dark-colored well images 13p show purple. Also, the well images 13p with intermediate colors show neutral colors. In the visual judgment results of the evaluation plates 10_1 to 10_9, the well images 13p marked with "m" show that they were intermediate colors between purple and yellow.

[0239] From the results shown in Figs. 26 to 29, it can be understood that the color judgment process using the image processing device 40 yields substantially the same results as the visual judgment. In fact, the rates at which the detection results using the image processing device 40 differed from the visual judgment results for the evaluation plates 10_1, 10_2, 10_5, and 10_6 (hereinafter, "difference rates") were 0.0%, 0.2%, 0.1%, and 0.0%, respectively. The difference rate for the evaluation plate 10_1 is a value (%) obtained by dividing the total number of wells 13 that differed from the visual judgment results in the detection results by the image processing device 40 for the three evaluation plates 10_1 under the same conditions by the total number of wells 13 (384 x N), and the difference rates for the evaluation plates 10_2, 10_5, and 10_6 were also values ​​calculated in the same manner.

[0240] [Colony detection process] The colony detection process was performed on the evaluation plate 10_4. As described above, the colony detection process was performed on the images acquired in the first imaging step and the second imaging step.

[0241] First, the colony detection process using the image acquired in the first imaging step will be described.

[0242] The colony detection process was performed using an image processing device 40, following the flowchart of the colony detection process shown in Fig. 18. In a well image specification step S31, the plate 11 with nothing in the well 13 was photographed as a mask image, and the corresponding well area was set as an ROI. Furthermore, in a first colony detection step S34 and a second colony detection step S36, the brightness thresholds used in each step were specified by the evaluator. Similarly, the shape and size for colony detection were also specified by the evaluator.

[0243] The results of the colony detection process were as shown in the table in FIG. 30. No. 1 to No. 7 in FIG. 30 are excerpts of a portion of a plurality of well images 13p ("original images" in FIG. 30) in plate image P1 corresponding to evaluation plate 10_4, and show the corresponding colony detection results. In the detection results, the area surrounded by a dashed dotted line is the area detected as a colony. No. 2 shows a well image 13p in which no colony has occurred as shown in the original image. In this case, no colony was detected in the colony detection process by image processing device 40. Therefore, the detection result of No. 2 is omitted.

[0244] Comparing the original image in Fig. 30 with the detection results, it can be seen that colonies are appropriately detected by the colony detection process in the image processing device 40. Furthermore, it can be seen that the colony detection process using the image processing device 40 can reliably detect small colonies that are difficult to distinguish in the well image 13p. As in No. 2 above, when no colony is present in the well image 13p, no colony is detected in the colony detection process by the image processing device 40. In other words, it can be seen that no false detection occurs.

[0245] Next, a colony detection process using the image acquired in the second imaging process will be described. In this case, colonies were detected in the same manner as in the case of using the image acquired in the first imaging process, except that the image of the evaluation plate 10_4 acquired in the second imaging process was used instead of the first imaging process.

[0246] The results of the colony detection process were as shown in the table in FIG. 31. No. 1 to No. 7 in FIG. 31 are excerpts of a portion of a plurality of well images 13p ("original images" in FIG. 31) in the plate image P1 corresponding to the evaluation plate 10_4, and show the corresponding colony detection results. The well image 13p in FIG. 31 was an image of the same well 13 as in the well image 13p in FIG. 30. In the detection results, the area surrounded by the dashed line is the area detected as a colony. The reason for omitting the detection results in No. 2, No. 3, and No. 5 is the same as in the case of No. 2 in FIG. 30. That is, since no colony was present in the original image, no colony was detected even in the results using the image processing device 40.

[0247] Comparing the original image in FIG. 31 with the detection results, it can be seen that some colonies have been properly detected by the colony detection process in the image processing device 40.

[0248] As described above, Fig. 30 and Fig. 31 are the results for the well image 13p corresponding to the same well 13. Comparing the results of Fig. 30 and Fig. 31, it can be understood that the colony can be detected more clearly when the image of 10_4 is acquired by the first imaging step (i.e., when 10_4 is imaged from the bottom 13a side of the well 13) than when the image of 10_4 is acquired by the second imaging step (i.e., when 10_4 is imaged from the side opposite to the bottom 13a of the well 13) (see, for example, the results of No. 3 and 5 in Fig. 30 and Fig. 31). This is considered to be because the colony usually sinks to the bottom 13a side of the well 13.

[0249] [Foreign object detection processing] The foreign matter detection process was performed on the evaluation plate 10_1. Here, the foreign matter detection was performed using the images acquired in the first imaging step and the second imaging step described in the colony detection process.

[0250] First, foreign object detection using the image acquired in the first imaging process will be described.

[0251] The foreign object detection process was performed using the image processing device 40 according to the flowchart of the foreign object detection process shown in Fig. 24. In the well image specification step S31, similar to the colony detection process, the plate 11 with nothing in the well 13 was photographed as a mask image, and the corresponding well area was set as the ROI. In the foreign object detection process, the first brightness threshold and the second brightness threshold were specified by the evaluator in the first foreign object detection step S51 and the second foreign object detection step S53. Similarly, the shape and size for foreign object detection were also specified by the evaluator.

[0252] The results of the foreign body detection process were as shown in the table in Fig. 32. No. 1 to No. 7 in Fig. 32 are excerpts of a portion of a plurality of well images 13p ("original images" in Fig. 32) in plate image P1 corresponding to evaluation plate 10_1, and show the corresponding detection results. In the detection results, the area surrounded by a dashed dotted line is the area detected as a foreign body.

[0253] 32 with the detection results, it can be seen that foreign objects are appropriately detected by the foreign object detection process using the image processing device 40. Furthermore, it can be seen that the foreign object detection process using the image processing device 40 can reliably detect even small foreign objects that are difficult to distinguish in the original image.

[0254] Next, a foreign object detection process using the image acquired in the second imaging process will be described. In this case, a foreign object was detected in the same manner as in the case of using the image acquired in the first imaging process, except that the image of the evaluation plate 10_1 acquired in the second imaging process was used instead of the first imaging process.

[0255] The results of the foreign body detection process were as shown in the table in FIG. 33. No. 1 to No. 7 in FIG. 33 are excerpts of a portion of a plurality of well images 13p ("original images" in FIG. 33) in the plate image P1 corresponding to the evaluation plate 10_1, and show the corresponding detection results. The well image 13p in FIG. 33 was an image of the same well 13 as in the well image 13p in FIG. 32. In the detection results, the area surrounded by the dashed line is the area detected as a foreign body. The reason for omitting the detection results in No. 1 and No. 2 is the same as in the case of No. 2 in FIG. 30. That is, since there was no foreign body in the original image, no foreign body was detected even in the results using the image processing device 40.

[0256] Comparing the original image in FIG. 33 with the detection results, it is apparent that some foreign objects have been properly detected as foreign objects by the foreign object detection process performed by image processing device 40.

[0257] As described above, Fig. 32 and Fig. 33 are the results for the well image 13p corresponding to the same well 13. Comparing the results of Fig. 32 and Fig. 33, it can be understood that the foreign matter can be detected more clearly when the image of the evaluation plate 10_1 is acquired by the first imaging step (i.e., when the evaluation plate 10_1 is imaged from the bottom 13a side of the well 13) than when the image of the evaluation plate 10_1 is acquired by the second imaging step (i.e., when the evaluation plate 10_1 is imaged from the side opposite the bottom 13a of the well 13) (see, for example, the results of No. 1 and 2 in Fig. 32 and Fig. 33). This is considered to be because the foreign matter usually sinks to the bottom 13a side of the well 13.

[0258] Next, a modification of the above embodiment will be described.

[0259] (Variation 1) In the first modification, in the irradiating step for capturing an image of the evaluation plate 10, the evaluation plate is irradiated with light L having a wavelength within at least one of the wavelength ranges of 430 nm to 460 nm and 520 nm to 620 nm.

[0260] The irradiation step can be performed, for example, by adopting an LED illumination including at least one of an LED outputting light having a wavelength in the wavelength range of 430 nm to 460 nm and an LED outputting light having a wavelength in the wavelength range of 520 nm to 620 nm as the illumination device 32 shown in Fig. 3. Alternatively, the irradiation step can be performed by irradiating the evaluation plate 10 with the light output from the illumination device 32 via a wavelength filter that selectively passes at least one of an LED outputting light having a wavelength in the wavelength range of 430 nm to 460 nm and an LED outputting light having a wavelength in the wavelength range of 520 nm to 620 nm. Such a wavelength filter can be incorporated in the illumination device 32.

[0261] In the irradiation step, light L having a first wavelength selected from the wavelength range of 430 nm to 460 nm and a second wavelength selected from the wavelength range of 520 nm to 620 nm, that is, a mixture of light of the first wavelength and light of the second wavelength, may be irradiated onto the evaluation plate 10. As an example of a combination of the first wavelength and the second wavelength, the first wavelength is 440 nm and the second wavelength is 590 nm.

[0262] Except for the above condition regarding the wavelength of the irradiated light L, the present embodiment is similar to the above embodiment. Therefore, the first modification has the same effects as the embodiment.

[0263] By using light L within the above wavelength range, the luminance difference between wells 13 (e.g., between sections Se1 to Se8) with different states of the evaluation object 14 becomes larger, so that, for example, the color difference becomes clearer. As a result, the safety evaluation can be performed more accurately.

[0264] For example, when the color of the wells 13 (the color of the evaluation object 14) changes due to a change in color of the pH indicator, the evaluation plate 10 includes wells 13 whose colors are close to the colors represented by wavelengths in the wavelength range of 430 nm to 460 nm, and wells 13 whose colors are close to the colors represented by wavelengths in the wavelength range of 520 nm to 620 nm. When the light L has the above-mentioned first and second wavelengths, it is easy to ensure the brightness of the wells 13 for both the colors on the 430 nm to 460 nm side of the wavelength range and the colors on the 520 nm to 620 nm side of the wavelength range. In this case, the color difference between wells 13 (for example, between sections Se1 to Se8) that differ in the state of the evaluation object 14, etc., becomes clearer. As a result, the safety evaluation can be performed more accurately.

[0265] Experimental Example 2 describes a verification experiment on the fact that the color differences between sections Se1 to Se8 become clearer by irradiating the evaluation plate with light L having a wavelength within at least one of the wavelength ranges of 430 nm to 460 nm and 520 nm to 620 nm.

[0266] In Experimental Example 2, five evaluation plates, Sample No. 1 to Sample No. 5, were prepared as follows as models of evaluation plate 10 in a state for color judgment, i.e., evaluation plate 10 obtained after completion of step 11A of the above-mentioned Ames test procedure (steps 1A to 14A). Sample No. 1 to Sample No. 5 mimic the range of color tones that may occur on an actual evaluation plate in the Ames test. For ease of explanation, the evaluation plates, Sample No. 1 to Sample No. 5, prepared in Experimental Example 2 will also be referred to as evaluation plate 10.

[0267] (Sample No. 1) Bromocresol purple (pH indicator) was injected into 384 wells 13 of the plate 11 shown in FIG. 1. Bromocresol purple was a pH indicator that can change from pH 5.2 (yellow) to pH 6.8 (purple). The long side of the plate 11 was 127 mm long and the short side was 87 mm long. Different amounts of hydrochloric acid were further added to the sections Se1 to Se8 to adjust the color (pH indicator) of the wells 13 in the sections Se1 to Se8 so as to obtain a state (color state) that imitates an example of the evaluation result in the color evaluation of the evaluation plate 10. The amount of hydrochloric acid added, pH, and visual color judgment in the sections Se1 to Se8 of sample No. 1 were as shown in Table 1. The intermediate colors in Table 1 are between purple and yellow. The color of well 13 in section Se1 (negative control section) was considered as purple, and the color of well 13 in section Se8 (positive control section) was considered as yellow, and the color of well 13 in section Se1 and the color of well 13 in section Se8 were used as the standard for judgment. [Table 1]

[0268] (Sample No. 2 to Sample No. 5) The colors (pH indicators) of the wells 13 in the sections Se1 to Se8 of each of Sample No. 2 to Sample No. 5 were adjusted by adding hydrochloric acid so that Sample No. 2 to Sample No. 5 were in states (color states) that imitated different evaluation results in the color evaluation of the evaluation plate 10. The amount of hydrochloric acid added, pH, and visual color judgments in the sections Se1 to Se8 of Sample No. 2 to No. 5 were as shown in Tables 2, 3, 4, and 5. The intermediate colors in Tables 2 and 3 are colors between purple and yellow. In the judgment of the colors of Sample No. 2 to Sample No. 3, the colors of the wells 13 in the sections Se1 and the wells 13 in the sections Se8 were used as the standards in each of Sample No. 2 to No. 3, as in the case of Sample No. 1. In judging the color of Sample No. 4 to Sample No. 5, the color of well 13 in section Se1 (negative control section) in each of Sample No. 4 to No. 5 was regarded as purple, and the color of well 13 in section Se1 was used as the standard for judgment. [Table 2] [Table 3] [Table 4] [Table 5]

[0269] 34 and an observation device including an imaging section 331 arranged opposite the illumination device 321. Specifically, the illumination device 321 was turned on, and the evaluation plate 10 of sample No. 1 was placed on the illumination device 321 and was imaged by the imaging section 331 while being transported together.

[0270] The illumination device 321 was a surface light source device using LEDs, and was a surface light source device (KDB-100 manufactured by KOS21) that outputs light L with a wavelength of 400 to 700 nm. The imaging unit 331 had an imager 331a and a light collecting unit 331b attached to the imager 331a. The imager 331a was a hyperspectral camera (Pika-L manufactured by Resonon). The frame rate of the imager 331a was 249.2 Hz, and the exposure time was 3.93 ms. For the light collecting unit 331b, a Xenoplan1.4 / 23-0902 (aperture: F5.6) manufactured by Shneider was used. The conveying speed of the evaluation plate 10 was 22 mm / s.

[0271] FIG. 35 is a drawing showing the spectral distribution by section based on the imaging results of the evaluation plate 10 of sample No. 1. FIG. 35 shows an average spectrum obtained by averaging the spectra obtained by the imaging device 331a (hyperspectral camera) for each of sections Se1 to Se8. In FIG. 35, the horizontal axis is wavelength, and the vertical axis is brightness value. Lines of different densities represented by "1" to "8" represent the lines of average spectra corresponding to sections Se1 to Se8, and "1" to "8" correspond to sections Se1 to Se8, respectively. That is, the color of the line showing the average spectrum becomes lighter in the order of section Se1, section Se2, ..., and section Se8.

[0272] Images of the evaluation plates 10 of samples No. 2 to No. 5 were taken in the same manner as for the evaluation plate 10 of sample No. 1, except that the evaluation plates 10 to be imaged were changed to the evaluation plates 10 of samples No. 2 to No. 5.

[0273] FIG. 36 is a diagram showing the spectral distribution by section based on the imaging results of the evaluation plate 10 of sample No. 2. FIG. 37 is a diagram showing the spectral distribution by section based on the imaging results of the evaluation plate 10 of sample No. 3. FIG. 38 is a diagram showing the spectral distribution by section based on the imaging results of the evaluation plate 10 of sample No. 4. FIG. 39 is a diagram showing the spectral distribution by section based on the imaging results of the evaluation plate 10 of sample No. 5. The horizontal axis, vertical axis, and lines showing the average spectra corresponding to sections Se1 to Se8 in FIGS. 36 to 39 are represented in the same way as in FIG. 35.

[0274] Fig. 40 is a graph showing the luminance difference between sections for each wavelength. Specifically, the maximum luminance difference between sections for each wavelength is plotted for the results of Figs. 35 to 39. The difference in the lines showing the results of Sample No. 1 to Sample No. 5 shown in Fig. 40 is represented by shading, with the line colors becoming darker in the order of Sample No. 4, Sample No. 5, Sample No. 2, Sample No. 3, and Sample No. 1.

[0275] As shown in FIG. 40, in each of the wavelength ranges of 430 nm to 460 nm and 520 nm to 620 nm, the luminance difference between the sections is large in all of Samples No. 1 to No. 5.

[0276] Therefore, by irradiating the evaluation plate with light L having a wavelength within at least one of the wavelength ranges of 430 nm to 460 nm and 520 nm to 620 nm, it becomes easier to clearly distinguish color changes between the sections, and as a result, safety evaluation can be performed more accurately.

[0277] (Variation 2) In the irradiation step in capturing an image of the evaluation plate 10 in the second modification, the evaluation plate is irradiated with light L having a wavelength of 520 nm or more. The rest of the process is the same as in the above embodiment. Therefore, the first modification has the same effects as the embodiment.

[0278] 40 showing the results of Experimental Example 2 described in Modification 1, the luminance difference between the sections can be increased by using light L having a wavelength of 520 nm or more. Therefore, by using light L having a wavelength of 520 nm or more, the color difference between wells 13 (for example, sections Se1 to Se8) having different states of the evaluation object 14 becomes clearer. As a result, the safety evaluation can be performed more accurately.

[0279] (Variation 3) In the third modification, in the imaging step of imaging the evaluation plate, light having a wavelength within at least one of the wavelength ranges of 430 nm to 460 nm and 520 nm to 620 nm is selectively received. Other than this, the third modification is the same as the above embodiment. Therefore, the first modification has the same effects as the embodiment.

[0280] Such an imaging process can be carried out in the observation device 30 shown in FIG. 3 by receiving light from the evaluation plate 10 in the imaging unit 33 via a filter that selectively transmits light having wavelengths within at least one of the wavelength ranges of 430 nm to 460 nm and 520 nm to 620 nm.

[0281] In the third modification, light having a wavelength within at least one of the wavelength ranges of 430 nm to 460 nm and 520 nm to 620 nm is selectively received. In this case, the wavelength is selected on the light receiving side, and the other points are substantially the same as those of the first modification. Therefore, the third modification has the same effects as the first modification.

[0282] (Variation 4) The plate 11 may be a plate that has been determined to be suitable by the following plate determination method. The plate determination method will be described below by taking as an example a case where the plate 11 shown in Fig. 1 is determined by the plate determination method. The foreign matter described later in Modification 4 is a substance (or structure) that is generated when cells die, as described in the above embodiment.

[0283] (Plate Judgment Method) An evaluation sample containing a test substance and specimen selected and adjusted to produce the above-mentioned foreign matter is injected into multiple wells (corresponding to wells to be evaluated (evaluation wells) in the evaluation method) 13 of the plate 11 to be evaluated, and after a certain period of time has elapsed, if there is at least one well 13 that retains the foreign matter, the plate is judged to be proper.

[0284] The timing of counting after the certain time (hereinafter referred to as "counting timing") is a time at which it can be ensured that a foreign matter has certainly been generated, and for example, when evaluating the presence or absence of a foreign matter in an Ames test, it may be the timing at which the evaluation is performed. The counting timing may differ, for example, depending on the test. For example, the counting timing may be a first counting timing in one test, and a second counting timing a predetermined time after the first counting timing in another test.

[0285] An example of the first counting timing is after injecting the specimen and test specimen into plate 11 and culturing for 48 hours as in step 11A. For example, the counting timing is within 12 hours, and preferably within 3 hours, after the end of 48 hours of culturing after injecting the evaluation samples into the multiple wells 13. The second counting timing may be, for example, after storing the specimen and test specimen at a constant temperature (for example, 10°C or less) after a predetermined time has elapsed after injecting the specimen and test specimen into plate 11 as in step 11A. The predetermined time is, for example, 1 day, 2 days, 3 days, 4 days, or 5 days after the start of storage.

[0286] In the plate evaluation method, the test specimen is a cell, and the evaluation sample is a combination of a cell and a specimen that are known in advance to contain foreign matter, and the specimen concentration is a concentration that causes the foreign matter.

[0287] An example of the cell employed in the evaluation sample is bacteria. Examples of the bacteria employed in the evaluation sample may include Escherichia coli (e.g., Escherichia coli WP2uvrA strain) or Salmonella typhimurium (e.g., Salmonella typhimurium TA100 strain). Examples of the analyte employed in the evaluation sample are tert-butylhydroquinone, 1-decanol, 2-ethoxynaphthalene, 4-(1,1,3,3-tetramethylbutyl)phenol, diethyl chlorothiophosphate, or 4-chlorobenzyl chloride. For example, the analyte may be 1-decanol having a concentration of 62.5 μg / mL or more, or tert-butylhydroquinone having a concentration of 667 μg / mL or more.

[0288] Specific examples of the evaluation sample include the following first to fourth evaluation samples. (First evaluation sample) The Escherichia coli WP2uvrA strain was used as the cell, and 1-decanol with a concentration of 62.5 μg / mL was used as the specimen, which was an evaluation sample. (2nd evaluation sample) The Escherichia coli WP2uvrA strain was used as the cells, and 1-decanol at a concentration of 125 μg / mL was used as the specimen, which was an evaluation sample. (Third evaluation sample) The evaluation sample used was the Salmonella typhimurium TA100 strain as the cells and 1-decanol with a concentration of 667 μg / mL as the specimen. (4th evaluation sample) The evaluation sample used was Salmonella typhimurium TA100 strain as the above-mentioned cells, and 1-decanol with a concentration of 2000 μg / mL as the above-mentioned specimen.

[0289] In the above plate determination method, the evaluation sample is cultured, for example, at 37° C. for 48 hours.

[0290] In the plate evaluation method, in order to appropriately evaluate a plate, the number of wells 13 into which evaluation samples are injected is preferably at least 12, and more preferably 48.

[0291] An example of a specific procedure for the plate determination method will be described. (1B process) The same process as in step 2A is carried out, except that the test specimen used is the cells used for the evaluation sample on plate 11. Specifically, the cells for the evaluation sample are inoculated into 60 mL of nutrient broth medium and grown by shaking culture at 37°C for 9 to 10 hours.

[0292] (2nd B process) A step similar to step 1B is carried out, except that the specimen used is the specimen used for the evaluation sample on plate 11. Specifically, a required amount of specimen for the evaluation sample is weighed out.

[0293] (3B process) The specimen from step 2B is dissolved or suspended in an appropriate medium to prepare a specimen solution at a concentration 25 times higher than the concentration at which the specimen exerts cytotoxicity.

[0294] (4th B process) 10 μL of the sample solution obtained in step 3B is placed into each well of a 24-well plate.

[0295] (Step 5B) Similarly to step 6A, prepare a bacterial solution under conditions without metabolic activation. Specifically, prepare 0.175 to 0.7 mL of cell culture solution (usually 1 × 10 9 The culture medium is mixed with 6.83 to 6.3 mL of medium (a minimal medium supplemented with a carbon source and lacking amino acids) to prepare a bacterial solution in the absence of metabolic activation.

[0296] (Step 6B) To a 24-well plate containing 10 μL of a sample solution, etc., 240 μL of the prepared bacterial solution prepared in step 5B under conditions without metabolic activation is added.

[0297] (Step 7B) The 24-well plate containing the mixture prepared in step 6B is shaken at 37° C. and 250 rpm for 90 minutes (preincubation).

[0298] (Step 8B) After the preincubation, add 2.6 mL of indicator medium (minimal medium containing a pH indicator and a carbon source and lacking amino acids) to each well (250 µL) of the 24-well plate.

[0299] (Step 9B) The mixture obtained in step 8B is transferred in 50 μL portions to 48 wells (one section) of a 384-well plate (plate 11).

[0300] (Step 10B) The 384-well plate (plate 11) is incubated at 37° C. for 48 hours.

[0301] (Step 11B) In the plate 11 after the incubation, the 48 wells 13 into which the mixed liquid was injected in step 9B are observed for the presence or absence of foreign matter, the number of wells 13 containing foreign matter is counted, and it is confirmed whether there is at least one well 13 containing foreign matter.

[0302] The plate 11 that is evaluated as proper by carrying out the above-mentioned plate evaluation method can retain foreign matter until the counting timing after the fixed time has elapsed. Therefore, when the evaluation method in the Ames test described above is carried out using the plate 11 that is evaluated as proper, the toxicity of the specimen to cells can be accurately evaluated by evaluating the evaluation plate 10 at the counting timing after the fixed time has elapsed.

[0303] Next, experimental examples 3 to 8 relating to the modified example 3 will be described. In experimental examples 3 to 8, the following six plates A to H were evaluated using the plate evaluation method. Plates A to H had the same configuration except for the materials and coating conditions described below, and had 384 wells, similar to plate 11 shown in FIG. 1.

[0304] (Plate A) The material of plate A was polystyrene. The surface of plate A (including the inner surface and bottom surface of well 13) was subjected to TC (tissue culture) treatment.

[0305] (Plate B) The material of plate B was polystyrene. No surface coating was applied to plate B. That is, plate B was a polystyrene plate with no surface coating applied.

[0306] (Plate C) Plate C had a glass plate, and was constructed by bonding a structure having 384 wells 13 onto the glass plate. The material of the structure was cyclic olefin copolymer (COC). Thus, the bottom surface of each well 13 in Plate C was made of glass, and the wall surface was made of COC. In Plate C, the bottom surface of the wells 13 was coated with fibronectin.

[0307] (Plate D) The material of Plate D was polystyrene. The surface of Plate D was coated with Poly-D-Lysine.

[0308] (Plate E) Plate E had a glass plate, and was configured by bonding a structure having 384 wells 13 onto the glass plate. The material of the structure was COC. Therefore, the bottom material of each well 13 in Plate E was glass, and the wall material was COC. No surface coating was applied to Plate E.

[0309] (Plate F) The material of plate F was cycloolefin polymer. No surface coating was applied to plate F. That is, plate F was a cycloolefin polymer plate not applied with a surface coating.

[0310] (Plate G) Plate G had a glass plate, and was configured by bonding a structure having 384 wells 13 onto the glass plate. The material of the structure was polystyrene. Therefore, the bottom surface of each well 13 in plate G was made of glass, and the wall surface was made of polystyrene. No surface coating was applied to plate G.

[0311] (Plate H) The material of plate H was COC. The surface of plate H was treated with TC (tissue culture).

[0312] The cells used in Experimental Examples 3 to 8 were prepared according to the National Institute of Health Sciences, Biotechnology Center of the National Institute of Technology and Evaluation, American Type Culture Collection, and the Ames test kit ("Ames MPF" manufactured by Xenometrix). TM The samples used in Experimental Examples 3 to 8 were available from Tokyo Chemical Industry Co., Ltd.

[0313] (Experimental Example 3) In Experimental Example 3, the following cells and specimens for evaluation samples were used, and the above steps 1B to 8B were carried out. Then, in step 9B, the mixture prepared in step 8B was transferred to plate A, plate B, and plate C, respectively, as explained in step 9B, and then steps 10B to 11B were carried out.

[0314] <Cells and specimens for evaluation samples in Experimental Example 3> Cells: E. coli WP2uvrA strain Sample: 1-decanol

[0315] The vehicle used in step 3B of Experimental Example 3 was DMSO. In step 5B, 0.7 mL of cells were mixed with 6.3 mL of medium. The concentration in step 6B was 62.5 μg / mL.

[0316] In Experimental Example 3, in step 11B, the presence or absence of foreign matter in each well 13 was visually observed and the number of wells containing foreign matter was counted, and then the plate 11 was stored at 4°C, and the presence or absence of foreign matter in each well 13 was visually observed 1 day, 2 days, 3 days, 4 days, and 5 days later, and the number of wells containing foreign matter was counted. The results of Experimental Example 3 are shown in Table 6. "Immediately after completion of culture" in Table 6 means that visual observation for foreign matter was performed at the end of 48 hours of culture. [Table 6]

[0317] (Experimental Example 4) In Experimental Example 4, the following cells and specimens for evaluation samples were used, and the above steps 1B to 8B were carried out. Then, in step 9B, the mixture prepared in step 8B was transferred to plate A, plate B, and plate C, respectively, as explained in step 9B, and then steps 10B to 11B were carried out.

[0318] <Cells and specimens for evaluation samples in Experimental Example 4> Cells: E. coli WP2uvrA strain Sample: 1-decanol

[0319] The vehicle used in step 3B of Experimental Example 4 was DMSO. In step 5B, 0.7 mL of cells were mixed with 6.3 mL of medium. The concentration in step 6B was 125 μg / mL.

[0320] In Experimental Example 4, in step 11B, the presence or absence of foreign matter in each well 13 was visually observed and the number of wells containing foreign matter was counted, and then the plate 11 was stored at 4°C, and the presence or absence of foreign matter in each well 13 was visually observed 1 day, 2 days, 3 days, 4 days, and 5 days later, and the number of wells containing foreign matter was counted. The results of Experimental Example 4 are shown in Table 7. The meaning of "immediately after completion of culture" in Table 7 is the same as in Table 6. [Table 7]

[0321] (Experimental Example 5) In Experimental Example 5, the following cells and specimens for evaluation samples were used, and the above steps 1B to 8B were carried out. Then, in step 9B, the mixture prepared in step 8B was transferred to plate A, plate B, and plate C, respectively, as explained in step 9B, and then steps 10B to 11B were carried out.

[0322] <Cells and specimens for evaluation samples in Experimental Example 5> Cells: Salmonella typhimurium strain TA100 Sample: 1-decanol

[0323] The vehicle used in step 3B of Experimental Example 5 was DMSO. In step 5B, 0.35 mL of cells were mixed with 6.65 mL of medium. In step 6B, the concentration was 667 μg / mL.

[0324] In Experimental Example 5, in step 11B, the presence or absence of foreign matter in each well 13 was visually observed and the number of wells containing foreign matter was counted, and then the plate 11 was stored at 4°C, and the presence or absence of foreign matter in each well 13 was visually observed 1 day, 2 days, 3 days, 4 days, and 5 days later, and the number of wells containing foreign matter was counted. The results of Experimental Example 5 are shown in Table 8. The meaning of "immediately after completion of culture" in Table 8 is the same as in Table 6. [Table 8]

[0325] (Experimental Example 6) In Experimental Example 6, the following cells and specimens for evaluation samples were used, and the above steps 1B to 8B were carried out. Then, in step 9B, the mixture prepared in step 8B was transferred to plate A, plate B, and plate C, respectively, as explained in step 9B, and then steps 10B to 11B were carried out.

[0326] <Cells and specimens for evaluation samples in Experimental Example 6> Cells: Salmonella typhimurium strain TA100 Sample: 1-decanol

[0327] The vehicle used in step 3B of Experimental Example 6 was DMSO. In step 5B, 0.35 mL of cells were mixed with 6.65 mL of medium. In step 6B, the concentration was 2000 μg / mL.

[0328] In Experimental Example 6, in step 11B, the presence or absence of foreign matter in each well 13 was visually observed and the number of wells containing foreign matter was counted, and then the plate 11 was stored at 4°C, and the presence or absence of foreign matter in each well 13 was visually observed 1 day, 2 days, 3 days, 4 days, and 5 days later, and the number of wells containing foreign matter was counted. The results of Experimental Example 6 are shown in Table 9. The meaning of "immediately after completion of culture" in Table 9 is the same as in Table 6. [Table 9]

[0329] (Experimental Example 7) In Experimental Example 7, the following cells and specimens for evaluation samples were used, and the above steps 1B to 8B were carried out. Then, in step 9B, the mixture prepared in step 8B was transferred to plates A, D, and E, respectively, as explained in step 9B, and then steps 10B to 11B were carried out.

[0330] <Cells and specimens for evaluation samples in Experimental Example 7> Cells: Salmonella typhimurium strain TA98 Sample: tert-butylhydroquinone

[0331] The medium used in step 3B of Experimental Example 7 was DMSO. Furthermore, the concentration in step 3B was 2500 μg / mL. In step 5B, 0.7 mL of cells was mixed with 6.3 mL of medium. In step 6B, the concentration was 2500 μg / mL.

[0332] In Experimental Example 7, the presence or absence of foreign matter in each well 13 was visually observed in step 11B, and the number of wells containing foreign matter was counted. The results of Experimental Example 7 were as shown in Table 10. Table 10 shows the number of wells in which foreign matter was confirmed out of 48 wells 13. The meaning of "immediately after completion of culture" in Table 10 is the same as in Table 6. [Table 10]

[0333] (Experimental Example 8) In Experimental Example 8, the following cells and specimens for evaluation samples were used, and the above steps 1B to 8B were carried out. Then, in step 9B, the mixture prepared in step 8B was transferred to plates A, F, G, and H, respectively, as explained in step 9B, and then steps 10B to 11B were carried out.

[0334] <Cells and specimens for evaluation samples in Experimental Example 8> Cells: E. coli WP2uvrA strain Sample: 1-decanol

[0335] The vehicle used in step 3B of Experimental Example 8 was DMSO. In step 5B, 0.7 mL of cells were mixed with 6.3 mL of medium. In step 6B, the concentration was 125 μg / mL.

[0336] In Experimental Example 8, the presence or absence of foreign matter in each well 13 was visually observed in step 11B, and the number of wells containing foreign matter was counted. The results of Experimental Example 8 were as shown in Table 11. Table 11 shows the number of wells in which foreign matter was confirmed out of 48 wells 13. The meaning of "immediately after completion of culture" in Table 11 is the same as in Table 6. [Table 11]

[0337] (Evaluation of Experimental Examples 3 to 8) In Experimental Examples 3 to 8, when counted immediately after the end of the culture, wells holding foreign matter were present in Plates B, C, D, E, F, G and H. Therefore, it can be seen that the combinations of cells and specimens and the concentrations of the specimens used in Experimental Examples 3 to 8 were conditions that would produce foreign matter, and that Plates B, C, D, E, F, G and H were plates capable of holding foreign matter immediately after the end of the culture. On the other hand, in Plate A, there were no wells 13 holding foreign matter immediately after the end of the culture in Experimental Examples 3 to 8. Therefore, Plate A is a plate that cannot hold foreign matter.

[0338] Therefore, for a cytotoxicity evaluation test to determine the presence or absence of foreign matter immediately after culture, Plates B to H are appropriate plates (i.e., plates that can be used to properly evaluate the presence or absence of foreign matter), while Plate A can be evaluated as a plate that cannot be used to properly evaluate the presence or absence of foreign matter.

[0339] From the counting results after 1 day, 2 days, 3 days, 4 days, and 5 days in Experimental Examples 3 to 6, Plate C is suitable for the cytotoxicity evaluation test to determine the presence or absence of foreign matter after 1 day, 2 days, 3 days, 4 days, and 5 days, respectively, just like the case immediately after culture, that is, it is a plate that can properly evaluate the presence or absence of foreign matter. On the other hand, Plates A and B did not have wells 13 holding foreign matter after 1 day. Therefore, Plates A and B can be evaluated as plates that cannot properly evaluate the presence or absence of foreign matter when evaluating the presence or absence of foreign matter after 1 day.

[0340] Immediately after culturing, similar results were obtained in Experimental Examples 3 to 8. Therefore, for example, if the evaluation sample of Experimental Example 3 (the cells, specimen, and specimen concentration used in Experimental Example 3) is used as a reference sample for plate evaluation, it can be seen that by using this reference sample, plates can be appropriately evaluated even in the case of other combinations of cells, specimens, and specimen concentrations, as in Experimental Examples 4 to 8.

[0341] The same is true when the evaluation samples of Experimental Examples 4 to 8 (the cells, specimens, and specimen concentrations used in Experimental Examples 4 to 8) are used as the reference samples for plate evaluation. That is, when the evaluation samples of Experimental Examples 4 to 8 are used as the reference samples, it can be seen that the plates evaluated as appropriate by the reference samples can also be appropriately evaluated in the case of combinations of cells, specimens, and specimen concentrations used in Experimental Examples 3 to 8 other than the Experimental Examples in which the reference samples were used.

[0342] Therefore, it was verified that the evaluation samples employed in Experimental Examples 3 to 8 can be used as evaluation samples for evaluating plates immediately after culture.

[0343] The same applies when the counting timing is 1 day, 2 days, 3 days, 4 days, or 5 days later. That is, it was verified that the evaluation samples used in each of Experimental Examples 3 to 6 can be used as evaluation samples for evaluating plates when the counting timing is 1 day, 2 days, 3 days, 4 days, or 5 days later.

[0344] The above describes various embodiments and modifications of the present invention as well as experimental examples. However, the present invention is not limited to the exemplified various embodiments, modifications, and experimental examples, and is intended to include the scope indicated by the claims, as well as all modifications within the meaning and scope equivalent to the claims.

[0345] For example, in the flowchart shown in Fig. 11, the color judgment is repeated three times, but the color judgment only needs to be performed at least once. For example, the result of the first judgment step in the flowchart shown in Fig. 11 may be the final judgment. The colors of all well images (corresponding to the colors of the wells) may be finally judged based on the similarity between the first reference statistical information (e.g., frequency distribution) for each of purple and yellow created in the first reference creation step shown in the flowchart shown in Fig. 11 and the statistical information (e.g., frequency distribution) of all well images of the plate image.

[0346] The color candidates used in color judgment and the like are not limited to purple and yellow. Other two colors may be used, or three or more colors may be used as color candidates depending on the test contents. In addition, even if the color of, for example, a well (or the color to be evaluated) of an evaluation plate is colorless to the naked eye, the present invention is applicable as long as the color feature amount (e.g., hue and saturation) in the image acquired by the imaging unit changes. For example, if the color feature amount (e.g., hue and saturation) in the image acquired by the imaging unit changes by irradiating visible light or a wavelength other than visible light, the color candidate may be selected from the false color provisionally assigned in the image acquired by the imaging unit. In this case, the image corresponding to the evaluation plate obtained by assigning the false color functions as a plate image.

[0347] In addition to hue and saturation, lightness, luminance, tone value, etc. may be used alone or in combination as at least one color feature value used in color judgment. Also, the R tone value, G tone value, and B tone value in an RGB image may be treated as separate color feature values. Furthermore, three or more color features may be selected from the above instead of two.

[0348] The statistical information used in the color judgment includes at least one type of statistical quantity. The statistical quantity may be the average value or median value of the color feature quantity of the entire evaluation area other than the frequency distribution. The statistical quantity may also be the maximum value, minimum value, deviation (for example, standard deviation) of the color feature quantity in the evaluation area.

[0349] For example, a gradation value can be used as the color feature amount, and an average value and a standard deviation (two types of statistics) can be used as the statistical information.

[0350] An example of the color judgment method in this case will be described. First, the average value and standard deviation of each RGB gradation value of all pixels included in each of the multiple well images (evaluation area) are obtained. As a result, a set of the average value and standard deviation is obtained as statistical information for each of the multiple well images. The statistical information (set of the average value and standard deviation) for each of the multiple well images thus obtained is used instead of the frequency distribution, and color judgment may be performed in the same manner as when the frequency distribution is used. For example, the similarity between the statistical information for the multiple well images and the reference statistical information of yellow and purple is calculated. The colors of the multiple wells are judged based on the calculated similarity. The reference statistical information of yellow and purple can be obtained in the same manner as when creating the third reference frequency distribution described in relation to the second color judgment method using the frequency distribution. For example, the average value and standard deviation of each RGB gradation value of all pixels included in all well images corresponding to all wells visually judged to be yellow may be used as the reference statistical information of yellow. The same applies to the reference statistical information of purple. Here, the case where the statistical information includes two statistics (average value and standard deviation) has been described, but for example, one of the average value and the standard deviation may be used as the statistical information. In the above explanation, the average value and standard deviation of the gradation values ​​are used, but other color feature amounts such as hue, saturation, etc. can also be used instead of the gradation values.

[0351] In the colony detection process and the foreign object detection process, the smoothing process does not have to be performed. In the colony detection process and the foreign object detection process, the detection result creation process does not have to be performed. Alternatively, in the colony detection process, the process performed in the detection result creation process may be performed (for example, performed by the second colony detection unit) in the second colony detection process shown in FIG. 18. Similarly, in the foreign object detection process, the process performed in the detection result creation process may be performed (for example, performed by the second foreign object detection unit) in the second foreign object detection process shown in FIG. 24.

[0352] In the foreign object detection process, it is not necessary to perform the external area setting step and the second foreign object detection step shown in Fig. 24. However, as described above, by performing the external area setting step and the second foreign object detection step, it is possible to more reliably detect a foreign object in the external area.

[0353] In the color judgment process and transparency detection process using statistical information (e.g., frequency distribution), for example, a well image corresponding to each well is used as an evaluation area, and statistical information is obtained for each of a plurality of well images (a plurality of evaluation areas). However, an evaluation area may have at least one well image. Therefore, a well image group including a plurality of well images may be used as an evaluation area, and statistical information may be obtained for the well image group. For example, in each of the sections Se1 to Se8 shown in FIG. 1, the conditions of the evaluation target in each of the sections Se1 to Se8 are the same. Therefore, one piece of statistical information may be created with the section images Se1p to Se8p corresponding to each of the sections Se1 to Se8 as evaluation areas, and the color judgment process and transparency detection process may be performed using the similarity between the statistical information for the plurality of section images.

[0354] In the transparency change detection process, the transparency change is detected based on statistical information (e.g., a zone frequency distribution) of the zone image (solvent control region) corresponding to the solvent control zone. However, the transparency change may be detected using statistical information (e.g., a frequency distribution) of a standard for transparency. When the multiple wells include at least one solvent control well, the solvent control region is a region including at least one solvent control well image corresponding to the image of the at least one solvent control well.

[0355] The specimen is not limited to a cell. The specimen is not limited to a chemical substance. The specimen and specimen may be any combination that produces at least a color change (including a color change when an indicator is used) when combined. Another example of a specimen is an antibody, and another example of an specimen is an antigen.

[0356] In the above embodiment, colonies are detected as a change in shape of the test object to be evaluated, but the evaluation object is not limited to colonies when the shape of the test object changes due to the test results. A foreign object is a structure different from the change in shape of the test object, and may be, for example, a substance (or structure) that was not originally included, and may have a clear boundary between the outline and the inside. For example, a substance (or structure) that is separated from the surroundings of the foreign object by an interface with a different refractive index. Examples of foreign objects include oil lumps and air bubbles.

[0357] In the above description, the evaluation method at least performs a color judgment process. However, the present invention also relates to an evaluation method that at least performs a transparency change detection process. Even in this case, when detecting a transparency change, the transparency change may be detected using statistical information (e.g., a frequency distribution) of a standard for transparency.

[0358] The present disclosure may include an evaluation method including a step of detecting the presence or absence of foreign matter in a plurality of wells (hereinafter, for convenience of explanation, referred to as an "evaluation method based on foreign matter detection"). In this evaluation method based on foreign matter detection, a color judgment process does not need to be performed. In the "step of detecting the presence or absence of foreign matter in a plurality of wells" of the evaluation method based on foreign matter detection, the presence or absence of foreign matter may be detected visually, or may be detected based on a well image as described in the above embodiment.

[0359] In the case where the presence or absence of foreign matter in a plurality of wells is detected based on well images, the evaluation method based on foreign matter detection includes: detecting the presence or absence of foreign matter in a plurality of wells based on the well images; The step of detecting the presence or absence of foreign matter in the plurality of wells includes: detecting the presence or absence of the foreign matter based on a first brightness threshold set for each of the multiple well images; detecting the presence or absence of the foreign matter in an outer region surrounding an inner region corresponding to a bottom region of the well when viewed from a depth direction of the well, based on a second brightness threshold set in the outer region in each of the plurality of well images; having the first luminance threshold and the second luminance threshold are different; In each of the plurality of wells, the area of ​​the bottom is smaller than the area of ​​the opening opposite the bottom. This is an evaluation method.

[0360] As one form of the evaluation method based on foreign body detection, which has a step of detecting the presence or absence of foreign bodies in the multiple wells based on the well images, there is a form in the above embodiment in which a foreign body detection process is performed on an image obtained by imaging the evaluation plate and the evaluation object is evaluated based on the processing result. In this evaluation method based on foreign body detection, a color judgment process does not need to be performed. Furthermore, when detecting the presence or absence of foreign bodies in the multiple wells based on the well images, the well images (images obtained by imaging the evaluation plate) may be visually inspected to detect the presence or absence of foreign bodies.

[0361] The plate may be a plate in which the bottom and wall surfaces of the wells are made of the same material, or a plate in which the bottom and wall surfaces of the wells are made of different materials.

[0362] In the above evaluation method based on foreign matter detection, it is preferable to use a plate that has been determined to be proper by the plate determination method described in Modification 4. This allows foreign matter to be detected more accurately.

[0363] Plates that are judged to be appropriate by the plate judgment method include polystyrene plates with a surface coating of Poly-D-Lysine, cyclic olefin copolymer plates with a surface treatment for TC (tissue culture), and cycloolefin polymer plates without a surface coating.

[0364] The plate judged to be proper by the plate judgment method may be the plate 11A shown in FIG. 41. The plate 11A has a glass plate 111 and a resin layer 112 provided on a surface 111a of the glass plate 111. A plurality of through holes 113 corresponding to a plurality of wells 13 are formed in the resin layer 112. The wells 13 are formed by the through holes 113 and the areas of the glass plate 111 exposed from the through holes 113. The resin layer 112 is a cyclic olefin copolymer or polystyrene. In the plate 11A, a coating layer 114 is formed at least in an area corresponding to the bottom surface of the well 13 (an area of ​​the glass plate 111 exposed from the through holes 113). The material of the coating layer 114 is fibronectin. The plate 11A shown in FIG. 41 corresponds to the plate C used in the experimental example described in the modification 4. Therefore, the plate 11A is a plate suitable for the evaluation method based on the above-mentioned foreign matter detection. The material of the coating layer 114 may be fibronectin, collagen, laminin, etc. Fig. 41 shows an example of a plate having the coating layer 114. However, in the plate 11A, the region of the glass plate 111 exposed at the through-hole 113 may not be coated.

[0365] Although the Ames test has been used in the above explanation, the present invention can also be applied to other biological safety evaluations. For example, the present invention can be suitably applied to cases where a change in color (including a change in transparency) of an evaluation target or a change in shape of a test object to be evaluated occurs in a safety test. As described above, the change in color of an evaluation target also includes a color change occurring in an image captured by an imaging unit due to the influence of illumination light for capturing an image of an evaluation plate.

[0366] The illustrated embodiments and modified examples may be combined as appropriate without departing from the spirit of the present invention. [Explanation of symbols]

[0367] 10...evaluation plate, 20...evaluation system, 13...well, 13a...bottom, 14, 14a to 14h...evaluation target, 33...imaging unit, 40...image processing device, IA...internal area, OA...external area, P1...plate image, 13p...well image, Se1 to Se8...sections (1st to Nth sections), Se1p...section image (solvent control area).

Claims

1. acquiring statistical information based on at least one color feature for each of a plurality of evaluation regions in a plate image corresponding to an image of an evaluation plate in which evaluation targets are held in a plurality of wells of the plate; determining colors of the plurality of evaluation regions using the statistical information; Equipped with The evaluation object includes a test specimen, The plurality of wells include a specimen well that holds the evaluation target further including a specimen, the plate image includes a plurality of well images corresponding to the plurality of wells; Each of the plurality of evaluation regions has at least one well image corresponding to at least one well; the at least one color feature is two or more color features, the statistical information is a frequency distribution of the two or more color feature quantities as variables; How the subject of evaluation is assessed.

2. The two or more color feature amounts are hue and saturation. The method for evaluating an evaluation object according to claim 1.

3. In the step of determining the colors of the plurality of evaluation regions, the colors of the plurality of evaluation regions are determined based on a similarity of the statistical information. A method for evaluating an evaluation object according to claim 1 or 2.

4. The step of determining the colors of the plurality of evaluation areas includes: a step of making a primary judgment on the colors of the plurality of evaluation regions from among a plurality of color candidates based on the similarity of the statistical information; obtaining first reference statistical information for the at least one color feature amount for the plurality of color candidates based on a result of the primary judgment; a step of making a final judgment on colors of the plurality of evaluation areas using a similarity between the first reference statistical information and the statistical information; Equipped with In the step of acquiring the first reference statistical information, the first reference statistical information is acquired based on statistical information of at least one evaluation area that has been determined to be a color in a result of the primary determination, among the statistical information for the plurality of evaluation areas. The method for evaluating an evaluation object according to claim 3.

5. the plurality of evaluation regions are the plurality of well images, The step of determining the colors of the plurality of evaluation areas includes: A step of initially determining the colors of the plurality of well images from among a plurality of color candidates based on the similarity of the statistical information; obtaining first reference statistical information for the at least one color feature amount for the plurality of color candidates based on a result of the primary judgment; A step of secondarily determining the colors of the plurality of well images from among the plurality of color candidates using the similarity between the statistical information and the first reference statistical information; obtaining second reference statistical information for the at least one color feature amount for the plurality of color candidates based on a result of the secondary judgment; A step of tertiarily determining the colors of the plurality of well images using the second reference statistical information; having The plate includes first to Nth sections (N is an integer of 2 or more) each having at least one well among the plurality of wells; In the step of acquiring the second reference statistical information, in an i-th partition image among the first to N-th partition images corresponding to the first to N-th partitions in the plate image, the second reference statistical information for the plurality of color candidates is acquired based on statistical information of at least one well image determined to be the same color candidate among the statistical information for the plurality of well images; In the step of tertiary determining the colors of the plurality of well images, the colors of the well images belonging to the i-th section among the plurality of well images are determined based on the similarity between the second reference statistical information acquired for the i-th section image and statistical information of at least one well belonging to the i-th section image among the statistical information for the plurality of well images, thereby finally determining the colors of the plurality of well images. The method for evaluating an evaluation object according to claim 3.

6. In the step of determining the colors of the plurality of evaluation regions, the colors of the plurality of evaluation regions are determined based on a similarity between reference statistical information for each of a plurality of color candidates, the reference statistical information being based on the at least one color feature amount, and the statistical information. A method for evaluating an evaluation object according to claim 1 or 2.

7. detecting a transparency change in the plurality of evaluation regions based on a similarity between the statistical information and reference statistical information for detecting a transparency change based on the at least one color feature amount; A method for evaluating an evaluation object according to any one of claims 1 to 6.

8. the plurality of wells further comprising at least one solvent control well for a solvent control; one of the plurality of evaluation regions is a solvent control region including at least one solvent control well image corresponding to the at least one solvent control well; In the step of detecting a change in transparency of the plurality of evaluation regions, and detecting a transparency change in the plurality of evaluation regions based on a similarity between the statistical information of the solvent contrast region and the remaining statistical information of the statistical information of the plurality of evaluation regions, using the statistical information of the solvent contrast region among the statistical information of the plurality of evaluation regions as reference statistical information for detecting a transparency change. The method for evaluating an evaluation object according to claim 7.

9. The method further includes detecting the presence or absence of a change in shape of the test object to be evaluated held in the wells based on the well images. A method for evaluating an evaluation object according to any one of claims 1 to 8.

10. Further comprising a step of detecting the presence or absence of foreign matter in the plurality of wells based on the plurality of well images. A method for evaluating an evaluation object according to any one of claims 1 to 9.

11. The step of detecting the presence or absence of a foreign object in the plurality of wells includes: detecting the presence or absence of the foreign matter based on a first brightness threshold set for each of the plurality of well images; detecting the presence or absence of the foreign matter in an outer region surrounding an inner region corresponding to a bottom region of the well when viewed from a depth direction of the well, based on a second brightness threshold value set in the outer region in each of the plurality of well images; having the first luminance threshold and the second luminance threshold are different; In each of the plurality of wells, the area of ​​the bottom is smaller than the area of ​​the opening opposite the bottom. The method for evaluating an evaluation object according to claim 10.

12. The plate is determined to be proper by a plate determination method in which, after a certain time has elapsed since an evaluation sample containing the test substance and the specimen selected and adjusted so as to cause the foreign matter to occur in a plurality of evaluation wells of the plate to be evaluated, the plate is determined to be proper if at least one of the plurality of evaluation wells holds the foreign matter. The method for evaluating an evaluation target according to claim 10 or 11.

13. The plate is Uncoated polystyrene plates, Polystyrene plate with surface coating of Poly-D-Lysine, A plate made of a cyclic olefin copolymer having a tissue culture-treated surface, or A cycloolefin polymer plate having no surface coating. The method for evaluating an evaluation object according to claim 12.

14. The plate is A glass plate and A resin layer provided on a surface of the glass plate; having the resin layer is formed of a cyclic olefin copolymer or polystyrene, A plurality of through holes corresponding to the plurality of wells are formed in the resin layer, The plate has no coating on the surface of the plate exposed through the through holes. The method for evaluating an evaluation object according to claim 12.

15. The plate is A glass plate and A resin layer provided on a surface of the glass plate; having the resin layer is formed of a cyclic olefin copolymer or polystyrene, A plurality of through holes corresponding to the plurality of wells are formed in the resin layer, a coating layer is formed on an area of ​​the surface exposed through the plurality of through holes; The material of the coating layer is fibronectin, collagen or laminin; The method for evaluating an evaluation object according to claim 12.

16. acquiring statistical information based on at least one color feature for each of a plurality of evaluation regions in a plate image corresponding to an image of an evaluation plate in which evaluation targets are held in a plurality of wells of the plate; detecting a transparency change in the plurality of evaluation regions based on a similarity between the plurality of evaluation regions and reference statistical information for detecting a transparency change based on the at least one color feature amount; Equipped with The evaluation object includes a test specimen, The plurality of wells include a specimen well that holds the evaluation target further including a specimen, Each of the plurality of evaluation regions has at least one well image corresponding to at least one well; the at least one color feature is two or more color features, the statistical information is a frequency distribution of the two or more color feature quantities as variables; How the subject of evaluation is assessed.

17. acquiring statistical information based on at least one color feature for each of a plurality of evaluation regions in a plate image corresponding to an image of an evaluation plate in which evaluation targets are held in a plurality of wells of the plate; determining colors of the plurality of evaluation regions using the statistical information; Equipped with The evaluation object includes a test specimen, The plurality of wells include a specimen well that holds the evaluation target further including a specimen, the plate image includes a plurality of well images corresponding to the plurality of wells; Each of the plurality of evaluation regions has at least one well image corresponding to at least one well; In the step of determining the colors of the plurality of evaluation regions, the colors of the plurality of evaluation regions are determined based on a similarity of the statistical information; The step of determining the colors of the plurality of evaluation areas includes: a step of making a primary judgment on the colors of the plurality of evaluation regions from among a plurality of color candidates based on the similarity of the statistical information; obtaining first reference statistical information for the at least one color feature amount for the plurality of color candidates based on a result of the primary judgment; a step of making a final judgment on colors of the plurality of evaluation areas using a similarity between the first reference statistical information and the statistical information; Equipped with In the step of acquiring the first reference statistical information, the first reference statistical information is acquired based on statistical information of at least one evaluation area that has been determined to be a color in a result of the primary determination, among the statistical information for the plurality of evaluation areas. How the subject of evaluation is assessed.

18. acquiring statistical information based on at least one color feature for each of a plurality of evaluation regions in a plate image corresponding to an image of an evaluation plate in which evaluation targets are held in a plurality of wells of the plate; determining colors of the plurality of evaluation regions using the statistical information; Equipped with The evaluation object includes a test specimen, The plurality of wells include a specimen well that holds the evaluation target further including a specimen, the plate image includes a plurality of well images corresponding to the plurality of wells; Each of the plurality of evaluation regions has at least one well image corresponding to at least one well; In the step of determining the colors of the plurality of evaluation regions, the colors of the plurality of evaluation regions are determined based on a similarity of the statistical information; the plurality of evaluation regions are the plurality of well images, The step of determining the colors of the plurality of evaluation areas includes: A step of initially determining the colors of the plurality of well images from among a plurality of color candidates based on the similarity of the statistical information; obtaining first reference statistical information for the at least one color feature amount for the plurality of color candidates based on a result of the primary judgment; A step of secondarily determining the colors of the plurality of well images from among the plurality of color candidates using the similarity between the statistical information and the first reference statistical information; obtaining second reference statistical information for the at least one color feature amount for the plurality of color candidates based on a result of the secondary judgment; A step of tertiarily determining the colors of the plurality of well images using the second reference statistical information; having The plate includes first to Nth sections (N is an integer of 2 or more) each having at least one well among the plurality of wells; In the step of acquiring the second reference statistical information, in an i-th partition image among the first to N-th partition images corresponding to the first to N-th partitions in the plate image, the second reference statistical information for the plurality of color candidates is acquired based on statistical information of at least one well image determined to be the same color candidate among the statistical information for the plurality of well images; In the step of tertiary determining the colors of the plurality of well images, the colors of the well images belonging to the i-th section among the plurality of well images are determined based on the similarity between the second reference statistical information acquired for the i-th section image and statistical information of at least one well belonging to the i-th section image among the statistical information for the plurality of well images, thereby finally determining the colors of the plurality of well images. How the subject of evaluation is assessed.

19. acquiring statistical information based on at least one color feature for each of a plurality of evaluation regions in a plate image corresponding to an image of an evaluation plate in which evaluation targets are held in a plurality of wells of the plate; determining colors of the plurality of evaluation regions using the statistical information; detecting a transparency change in the plurality of evaluation regions based on a similarity between the reference statistical information for detecting a transparency change based on the at least one color feature amount and the statistical information; Equipped with The evaluation object includes a test specimen, The plurality of wells include a specimen well that holds the evaluation target further including a specimen, the plate image includes a plurality of well images corresponding to the plurality of wells; Each of the plurality of evaluation regions has at least one well image corresponding to at least one well; the plurality of wells further comprising at least one solvent control well for a solvent control; one of the plurality of evaluation regions is a solvent control region including at least one solvent control well image corresponding to the at least one solvent control well; the statistical information based on the at least one color feature is a frequency distribution, In the step of detecting a change in transparency of the plurality of evaluation regions, and detecting a transparency change in the plurality of evaluation regions based on a similarity between the statistical information of the solvent contrast region and the remaining statistical information of the statistical information of the plurality of evaluation regions, using the statistical information of the solvent contrast region among the statistical information of the plurality of evaluation regions as reference statistical information for detecting a transparency change. How the subject of evaluation is assessed.

20. acquiring statistical information based on at least one color feature for each of a plurality of evaluation regions in a plate image corresponding to an image of an evaluation plate in which evaluation targets are held in a plurality of wells of the plate; determining colors of the plurality of evaluation regions using the statistical information; Equipped with The evaluation object includes a test specimen, The plurality of wells include a specimen well that holds the evaluation target further including a specimen, the plate image includes a plurality of well images corresponding to the plurality of wells; Each of the plurality of evaluation regions has at least one well image corresponding to at least one well; Further comprising a step of detecting the presence or absence of foreign matter in the plurality of wells based on the plurality of well images; the statistical information based on the at least one color feature is a frequency distribution, The step of detecting the presence or absence of a foreign object in the plurality of wells includes: detecting the presence or absence of the foreign matter based on a first brightness threshold set for each of the plurality of well images; detecting the presence or absence of the foreign matter in an outer region surrounding an inner region corresponding to a bottom region of the well when viewed from a depth direction of the well, based on a second brightness threshold value set in the outer region in each of the plurality of well images; having the first luminance threshold and the second luminance threshold are different; In each of the plurality of wells, the area of ​​the bottom is smaller than the area of ​​the opening opposite the bottom. How the subject of evaluation is assessed.

21. The plate is determined to be proper by a plate determination method in which, after a certain time has elapsed since an evaluation sample containing the test substance and the specimen selected and adjusted so as to cause the foreign matter to occur in a plurality of evaluation wells of the plate to be evaluated, the plate is determined to be proper if at least one of the plurality of evaluation wells holds the foreign matter. The method for evaluating an evaluation object according to claim 20.

22. acquiring statistical information based on at least one color feature for each of a plurality of evaluation regions in a plate image corresponding to an image of an evaluation plate in which evaluation targets are held in a plurality of wells of the plate; determining colors of the plurality of evaluation regions using the statistical information; Equipped with The evaluation object includes a test specimen, The plurality of wells include a specimen well that holds the evaluation target further including a specimen, the plate image includes a plurality of well images corresponding to the plurality of wells; Each of the plurality of evaluation regions has at least one well image corresponding to at least one well; Further comprising a step of detecting the presence or absence of foreign matter in the plurality of wells based on the plurality of well images; the statistical information based on the at least one color feature is a frequency distribution, The plate is determined to be proper by a plate determination method in which, after a certain time has elapsed since an evaluation sample containing the test substance and the specimen selected and adjusted so as to cause the foreign matter to occur in a plurality of evaluation wells of the plate to be evaluated, the plate is determined to be proper if at least one of the plurality of evaluation wells holds the foreign matter. How the subject of evaluation is assessed.

23. The plate is Uncoated polystyrene plates, Polystyrene plate with surface coating of Poly-D-Lysine, A plate made of a cyclic olefin copolymer having a tissue culture-treated surface, or A cycloolefin polymer plate having no surface coating.

23. A method for evaluating an evaluation target according to claim 21 or 22.

24. The plate is A glass plate and A resin layer provided on a surface of the glass plate; having the resin layer is formed of a cyclic olefin copolymer or polystyrene, A plurality of through holes corresponding to the plurality of wells are formed in the resin layer, The plate has no coating on the surface of the plate exposed through the through holes.

23. A method for evaluating an evaluation target according to claim 21 or 22.

25. The plate is A glass plate and A resin layer provided on a surface of the glass plate; having the resin layer is formed of a cyclic olefin copolymer or polystyrene, A plurality of through holes corresponding to the plurality of wells are formed in the resin layer, a coating layer is formed on an area of ​​the surface exposed through the plurality of through holes; The material of the coating layer is fibronectin, collagen or laminin; 23. A method for evaluating an evaluation target according to claim 21 or 22.

26. The method further includes an irradiation step of irradiating the evaluation plate with light, In the irradiation step, the evaluation plate is irradiated with light having a wavelength within at least one of a wavelength range of 430 nm to 460 nm and a wavelength range of 520 nm to 620 nm. A method for evaluating an evaluation object according to any one of claims 1 to 25.

27. The light irradiated onto the evaluation plate in the irradiation step is light having a first wavelength selected from a wavelength range of 430 nm to 460 nm and a second wavelength selected from a wavelength range of 520 nm to 620 nm. The method for evaluating an evaluation object according to claim 26.

28. The method further includes an imaging step of imaging the evaluation plate. A method for evaluating an evaluation object according to any one of claims 1 to 27.

29. In the imaging step, light having a wavelength within at least one of a wavelength range of 430 nm to 460 nm and a wavelength range of 520 nm to 620 nm is selectively received. The method for evaluating an evaluation object according to claim 28.

30. In the imaging step, an image of the evaluation plate is taken from a bottom side of the plurality of wells.

30. A method for evaluating an evaluation target according to claim 28 or 29.

31. An image processing device configured to execute the method for evaluating an evaluation object according to any one of claims 1 to 30.

32. a plate holder for holding an evaluation plate having evaluation targets held in a plurality of wells provided in the plate; an imaging unit that images the evaluation plate; an image processing device according to claim 31 , which evaluates a plate image corresponding to the evaluation plate obtained based on an imaging result of the imaging unit; Equipped with The rating system being assessed.

33. The imaging unit is disposed on the bottom side of the plurality of wells. The evaluation system for an evaluation target according to claim 32.

Images