JP4827787B2 - Method for determining presence or absence of solution in drug discovery container - Google Patents

Description

  The present invention relates to a drug discovery sample storage system used for identifying, storing and observing a large number of samples in the field of drug discovery research and the like, and more specifically, in a container enclosing a drug discovery sample. The present invention relates to a method for determining the presence or absence of a solution in a drug discovery container for determining the presence or absence of a solution.

Conventionally, in fields such as drug discovery research, a solution in which a sample is dissolved is enclosed in a container called a microtube, and the container is stored in a grid-like storage plate, for example, SBS (Society for Biomolecular Screening) standard. A so-called 384 assay plate having a total number of 384 compartments in 16 rows and 24 columns, and a storage plate in which a solution in which a sample is dissolved is poured into a well called a well, and the well is partitioned in a grid pattern, For example, it is known to accommodate, store, and observe in a 1536 well plate having a total number of 1536 sections of 32 rows and 48 columns (see, for example, Patent Documents 1 and 2).
Japanese Patent No. 3421252 (2nd page, 5th paragraph, FIG. 1) JP 2007-33061 A (3rd page, 11th paragraph, FIG. 1)

However, the following problems are pointed out when determining whether or not a solution is contained in a plurality of containers accommodated in the storage plate as described above or in a plurality of wells formed in the storage plate. It had been.
(1) When visual inspection is performed, not only a large amount of labor is required for mass processing, but also the accuracy rate varies depending on the observer, and the accuracy rate varies depending on the date and time, resulting in poor inspection quality. .
(2) When using a CCD camera, the camera is moved to the observation position, and the presence or absence of the solution in the container or well is determined using the same threshold value at the end and the center of the storage rack. The frequency of recognition errors was high depending on the position of the container or well. In addition, when the threshold value is changed depending on the position of the container or the well, it takes a lot of time and effort. Furthermore, in order to capture a wide range with a high resolution, a plurality of high-resolution cameras are necessary, and in order to divide and capture, a mechanism for that is necessary, and the cost is high.
(3) When imaging with a scanner and performing conventional image processing, an illumination condition is set in advance corresponding to the matrix address, and after imaging, preprocessing such as mask processing and binarization processing is performed, and the matrix It was necessary to set a threshold value for each address, and operation, setting, and operation management were complicated.
(4) When visible light was used for discrimination, a transparent solution could not be detected.

  Therefore, an object of the present invention is to provide a simple method without interposing human eyes, and without setting a threshold value for determination in correspondence with the position of the container or well on the storage rack. It is an object of the present invention to provide a method for determining the presence or absence of a solution in a drug discovery container that can determine whether or not a solution is contained in the container.

The present invention is based on a captured image using a plurality of containers for storing a drug discovery sample, a storage plate for storing a plurality of the containers in a matrix, and an imaging means for imaging a state in the container. In the drug discovery container presence / absence determination method for determining the presence / absence of a solution in a container, a container in which a solution has been placed in advance and a container in which the solution has not been placed are alternately arranged in a checkered pattern on the storage plate. The respective images of the container containing the solution and the container not containing the solution are taken, accumulated and averaged to obtain a row average image with solution and a row average image without solution, and the row average image with solution and no solution A threshold calculation difference image that is a difference between row average images is calculated for each row, a threshold calculation feature value that is a feature value of a gray value of each pixel constituting the threshold calculation difference image is calculated, and the threshold calculation is performed. Multiply the feature value by a predetermined constant The value a solution existence discrimination threshold of each row. Then, a solution presence / absence determination difference image, which is a difference between an image of a container in which the presence / absence of a solution contained in an arbitrary column in an arbitrary row and a solution-free row average image in the row is determined, is calculated, and the solution presence / absence determination is calculated. Calculating a feature value for discrimination that is a feature value of a gray value of each pixel constituting the difference image, calculating a difference between the feature value for discrimination and the threshold value for determining the presence or absence of a solution, and when this difference is positive, It is determined that the solution is in the container, and in the negative case, it is determined that there is no solution in the container, thereby solving the above-described problem.

また、単純に全ピクセルのグレイ値の平均を特徴値として用いることもできる。他にも画像の特徴を数値化するさまざまな公知の特徴値を用いることができる。 Here, the “feature value of the gray value of each pixel constituting the image” in the present invention is not particularly limited as long as the feature of the image can be quantified. For example, R is an image area. , P is a pixel in the region R having a gray value (gray value of 0 to 255 gradations) g (p), F is a surface (F = | R |), and M is an average of gray values of all pixels The standard deviation shown in Equation 1 can be used as the feature value.

Further, an average of gray values of all pixels can be simply used as a feature value. In addition, various known feature values for digitizing image features can be used.

  In addition, the “predetermined constant” in the case where the value obtained by multiplying the threshold calculation feature value by a predetermined constant is used as the solution presence / absence determination threshold is, for example, set to 1/2 when performing automatic calculation. A discrimination accuracy with no practical problem can be obtained. In addition, in the case of performing manually, the discrimination accuracy can be improved by taking into account the distribution of feature values for each row and setting the values to, for example, 1 / 1.8, 1 / 2.2, etc. it can.

  Here, the “container” in the present invention includes not only a cylindrical or square tube like a microtube, but also a depression called a “well” formed to store a solution on the surface of a storage plate. is doing.

According to the invention of claim 1, using a plurality of containers for storing a drug discovery sample, a storage plate for storing a plurality of the containers in a matrix, and an imaging means for imaging the state in the container, In the drug discovery container presence / absence determination method for determining presence / absence of a solution in a container based on a captured image, a container in which a solution has been put in advance and a container in which the solution has not been placed are alternately arranged in a checkered pattern on the storage plate. In addition, for each row, images of the container containing the solution and a container not containing the solution are taken, accumulated and averaged to obtain a row average image with solution and a row average image without solution, and with the solution A threshold calculation difference image that is a difference between a row average image and a solution-free row average image is calculated for each row, and a threshold calculation feature value that is a feature value of a gray value of each pixel constituting the threshold calculation difference image is calculated. Calculate the threshold value feature value A value multiplied by a predetermined constant is used as a threshold for determining the presence / absence of a solution for each row, and an image of a container for determining the presence / absence of a solution contained in an arbitrary column of an arbitrary row and the solution-free row average image in that row Calculating a difference image for determining presence / absence of a solution, which is a difference, calculating a feature value for determination as a feature value of a gray value of each pixel constituting the difference image for determining presence / absence of solution, and determining the feature value for determination and the presence / absence of the solution Calculate the difference with the threshold value for determination, if this difference is positive, it is determined that there is a solution in the container, and if it is negative, it is determined that there is no solution in the container. It is not necessary to individually set a threshold value for determination corresponding to the position of the container on the storage plate, and it is possible to determine whether or not the solution is contained in the container by a simple method.

  Next, a preferred embodiment of the method for determining the presence or absence of a solution in a drug discovery container according to the present invention will be described with reference to the drawings.

  As shown in FIG. 1, the hardware configuration used in the drug discovery container presence / absence determination method of the present invention accommodates a plurality of containers 1 for storing drug discovery samples and a plurality of these containers 1 in a matrix. The storage plate 2, the imaging means 3 for imaging the state in the container 1, and a personal computer 5 equipped with a program for determining the presence or absence of a solution in the container 1 based on the captured image. A positioning plate 4 for aligning the positions of the imaging region and the storage plate 2 is placed on the imaging means 3.

  As the image pickup means 3 used in the present invention, a general-purpose desktop flatbed scanner of a reading head moving type document fixed reading method using a CCD line sensor can be used as it is. In addition, it is preferable to use an infrared scanner because the presence or absence of a solution can be determined even if the solution in the container 1 is a transparent solution.

  Next, the flow for determining the presence / absence of a solution in a container by the drug discovery container presence / absence determination method of the present invention will be described based on the flowcharts shown in FIGS. The flow for determining the presence / absence of the solution in the container according to the present invention includes a threshold value calculation process (S0 to S19) for obtaining a solution presence / absence determination threshold value that is a threshold value for determining the presence / absence of the solution in the container, and the calculated solution. It comprises a discrimination process (S20 to S35) for discriminating the presence or absence of a solution in an arbitrary container using the presence / absence discrimination threshold.

  In this embodiment, a case where a 384 well plate having 384 wells of 16 rows and 24 columns is used as a storage plate will be described. However, 384 microtubes of 16 rows and 24 columns are stored. The invention is equally applicable to assay plates or 1536 well plates having 1536 wells in 32 rows and 48 columns.

  First, wells that have been pre-injected with a solution and non-injected wells are alternately arranged in a checkered pattern on a storage plate. FIG. 2 shows an example in which the solution is injected in a checkered pattern into a 384 well plate having 16 rows and 24 columns of wells in a grid pattern. The black compartments, that is, A row 1 column, A row 3 column, A row 5 column,... Are wells into which the solution has been injected, and the white compartments, that is, A row 2 columns, A row 4 columns, A row 6 Columns... Indicate wells not injected with solution. From this state, the threshold value calculation process shown in the flowchart of FIG. 3 starts (S0).

  The storage plate is placed on a scanner serving as an imaging unit, and an image is captured from the bottom side of the storage plate (S1). At this time, preprocessing such as enhancing the contrast as necessary is performed on the acquired image. Next, the value of the row loop counter N is initialized (S2), the row accumulated image (well without solution) and the row accumulated image (well with solution) are initialized (S3). The value is initialized (S4).

Next, the existence area of each well is calculated from the size of the storage plate, the size of the well, etc. given in advance, and in this embodiment, a well image of 16 rows and 24 columns is cut out (S5). Based on the even and odd numbers of the matrix, it is determined whether the well has no solution or the well has a solution (S6).
) Based on the result, the image is accumulated in a row accumulation image (none well) or a row accumulation image (well) (S7, S8). Thereafter, the value of the column loop counter M is incremented by 1 (S9), and the column loop of S5 to S10 is repeated until the accumulation of the image of the 24th column for the first row is completed (S10).

  When the accumulation of the images in the 24th column is completed for the first row, the image data of the first row accumulated image (none well) and the first row accumulated image (present well) are obtained and stored. Then, the value of the counter N of the row loop is incremented by 1 (S11) until image data of the row accumulated image (none well) on the 16th row and the row accumulated image (well present) on the 16th row are obtained (S12). , S3 to S12 are repeated.

  When image data of 16 rows of accumulated images (no wells) and row accumulated images (with wells) is obtained, each row accumulated image (without wells) and row accumulated image (with wells) has 24 columns. Since twelve images that are half are accumulated, after determining whether the image is a row accumulated image (none well) or a row accumulated image (with well) (S13), each row is divided by 12 to obtain a row average image. (No well), that is, a row average image without a solution and a row average image (well with a solution), that is, a row average image with a solution are obtained (S14, S15).

  Then, for each row, an image of the difference between the row average image (with well) and the row average image (without well), that is, a threshold calculation difference image is obtained (S16). In the case of the present embodiment, 16 lines of images are obtained. For each image, a feature value obtained by quantifying the feature of the image, that is, a threshold calculation feature value is obtained. In the case of the present embodiment, each image is obtained as a gradation value of 256 gradations with a resolution of 150 × 150 = 22,500 pixels. The standard deviation of the gray value of 22500 pixels is obtained and used as the feature value of each image (S17). The threshold value, that is, the solution presence / absence determination threshold value is obtained by multiplying the feature value obtained for each row in this way by a predetermined constant (S18), and the threshold value calculation process ends (S19).

  Next, a description will be given of a flow of determination processing for determining whether or not there is a solution in each well of an arbitrary storage plate using the solution presence / absence determination threshold for each row obtained by the threshold calculation processing.

  The discrimination process is started by placing the storage plate to be discriminated on the scanner as the imaging means (S20). First, an image is taken from the bottom side of the storage plate (S21). At this time, preprocessing such as enhancing the contrast as necessary is performed on the acquired image. Next, the value of the counter N of the row loop is initialized (S22), and the value of the counter M of the column loop is initialized (S23).

  Next, the existence area of each well is calculated from the size of the storage plate, the size of the well, and the like given in advance, and in this example, a well image of 16 rows and 24 columns is cut out (S24). First, a processed image that is a difference between the row in which the well image exists, that is, the row average image (none well) in the first row, that is, a solution presence / absence determination difference image is obtained from the well image in the first row and the first column ( S25). Then, a feature value obtained by digitizing the feature of the difference presence / absence discrimination image, that is, a discrimination feature value is obtained (S26). In the case of the present embodiment, each well image is obtained as a gradation value of 256 gradations with a resolution of 150 × 150 = 22,500 pixels. The standard deviation of the gray value for 22500 pixels is obtained and used as a distinguishing feature value.

  Then, the difference between the obtained discrimination feature value and the same row calculated in the above-described threshold value calculation process, here, one row of the solution presence / absence threshold value is calculated (S27), and when this difference is negative, It is determined that there is no solution in the well (S28), and if positive, it is determined that there is a solution in the well (S29).

  Next, 1 is added to the value of the column counter M (S30), the column loop from S24 to S31 is repeated, and the determination up to the well in the first row and the 24th column is performed. When the value of the column counter M exceeds 24 (S31), the value of the row counter N is incremented by 1 (S32), the row loop from S23 to S33 is repeated, and discrimination up to the well of 16 rows and 24 columns is performed. Is called. When the value of the row counter N exceeds 16 (S33), the presence / absence of the solution in all wells from the well in the first row and the first column to the well in the 16th row and the 24th column on the storage plate is determined. The data is displayed on the monitor of the personal computer or output from the printer (S34), and the discrimination process is terminated (S35).

  According to the method for determining the presence or absence of a solution in a container for drug discovery according to the present invention, the presence or absence of a solution for each row according to a flow determined using a storage plate in which a container containing a solution and a container not containing a solution are alternately arranged in a checkered pattern Since the determination threshold value is obtained to determine the presence or absence of the solution in each container, the determination result does not vary depending on the observer, the determination result can be obtained at high speed, and the effect is enormous.

The figure which showed the hardware constitutions used for the solution presence / absence discrimination | determination method in the drug discovery container of this invention. The figure explaining pouring a solution on a 384 well plate in a checkered pattern. The flowchart of the threshold value calculation process which calculates | requires the threshold value for solution presence determination of this invention. The flowchart of the discrimination | determination process which discriminate | determines the presence or absence of the solution in arbitrary containers using the threshold value for solution presence or absence discrimination.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Container 2 ... Storage plate 3 ... Imaging means 4 ... Positioning plate 5 ... Personal computer

Claims (1)

A plurality of containers for storing drug discovery samples, storage plates for storing a plurality of the containers in a matrix, and imaging means for imaging the state in the containers, and solutions in the containers based on the captured images In the method for determining the presence or absence of a solution in a drug container for determining the presence or absence of
In the storage plate, the container in which the solution is put in advance and the container in which the solution is not put are alternately arranged in a checkered pattern, and images of the container in which the solution is put and the container in which the solution is not put are taken for each row, Accumulated and averaged to obtain a row average image with solution and a row average image without solution, and calculate a threshold calculation difference image that is a difference between the row average image with solution and the row average image without solution for each row, A threshold calculation feature value that is a feature value of a gray value of each pixel constituting the threshold calculation difference image is calculated, and a value obtained by multiplying the threshold calculation feature value by a predetermined constant is used to determine whether or not there is a solution for each row. A threshold,
Calculating a difference image for determining the presence / absence of a solution, which is a difference between an image of a container for determining the presence / absence of a solution contained in an arbitrary column of an arbitrary row and the solution-free row average image in the row, and determining the presence / absence of the solution A feature value for discrimination, which is a feature value of the gray value of each pixel constituting the difference image, is calculated, and a difference between the feature value for discrimination and the threshold value for determining presence / absence of the solution is calculated. A method for determining the presence or absence of a solution in a drug discovery container, wherein it is determined that there is no solution in the container and, if negative, that there is no solution in the container.

Images