JP2004248619A - System for automatically searching target cell - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To automatically carry out inspecting treatment of a target cell in good accuracy and remarkably shorten inspection time. <P>SOLUTION: An optical microscope 1 has an electric stage 17 having an XY stage mechanism and a objective lens control part 5 and a CCD camera 7. A PC for control has a stage (XY) moving part 53a for carrying out movement in the XY direction, an automatic focus-calculating part 53b for carrying out automatic focusing in the Z direction, an image input part 51a for inputting an image from a CCD camera 7, an image-recognizing part 51b and a result-storing database part 51c for storing image-treated result. The image photographed by the CCD camera 7 is input into an image input part 51a and displayed on a monitor 55 and fed to the image-recognizing part 51b and the automatic focus-calculating part 53b. The image recognized in the image-recognizing part 51b is memorized in the result-storing database 51c. The image input into the image input part 51a is also fed to the automatic focus-calculating part 53b and used for automatic focusing. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an automatic search system for target cells, and more particularly to an automatic search system for nucleated red blood cells derived from a fetus.
[0002]
[Prior art]
Conventional prenatal diagnostic techniques for performing prenatal diagnosis include screening techniques using serum markers, amniotic fluid testing techniques, and villus testing techniques. Recently, it has become clear that fetal blood cells are contained in a very small amount (for example, about 1 / 10th power) in the blood of pregnant women. By collecting the above-mentioned nucleated red blood cells derived from the fetus from maternal blood and performing DNA diagnosis, non-invasive fetal DNA diagnosis that is highly reliable and does not involve risk to the mother and baby is becoming possible.
[0003]
At present, research for practical use of a technique for recovering fetal nucleated red blood cells from maternal blood has been actively conducted worldwide. The main ones of these studies are Fluorescence-Activated Cell Sorting (FACS) technology by a research group in the United States, Magnetic-Activated Cell Sorting (MACS) technology by a research group in Europe, and a team of Dr. Takabayashi of Kanazawa Medical University. There are three well-known Percoll-micromanipulation technologies (Non-Patent Document 1, hereinafter, this technology is abbreviated as “Percoll technology”). In particular, in the United States, research and development of the FACS method has been underway since about 1994 under the control of the National Institutes of Health (NIH).
[0004]
Further, as an apparatus for detecting nucleated red blood cells derived from a fetus by image processing, an apparatus that uses a camera and a microscope to inspect a sample under computer control is disclosed (for example, see Patent Document 1).
[0005]
[Non-patent document 1]
Haruo Takabayashi, "Progress in prenatal genetic testing: fetal DNA diagnosis using maternal blood", Gene Medicine Vol. 5, no. 3 (2001) p410-416.
[Patent Document 1]
JP 2002-514762 A
[0006]
[Problems to be solved by the invention]
Among the above techniques, the serum marker screening technique has a problem in the reliability of collecting nucleated red blood cells derived from a fetus. Amniotic fluid tests and villus tests are not widely used because they involve the mother and fetus and pose risks to the mother.
[0007]
In addition, since the abundance of nucleated red blood cells derived from the fetus contained in maternal blood is extremely low, it is difficult to efficiently collect them.
[0008]
The FACS method by the U.S. research group and the MACS method by the European research group use a technique of marking fetal cells with an antibody specific to fetal cells and selecting fetal cells using these as indices. In addition, there are many contaminations with maternal cells, which is a serious problem for practical use.
[0009]
On the other hand, the Percoll technique in which blood is applied to a slide glass and fetal cells are identified by morphology and individually collected is used to reliably collect only fetal-derived red blood cells (nucleated) in the image 100 as shown in FIG. This is an effective and reliable technique because DNA diagnosis, which is a post-process, is easy to perform. However, since this technique involves a step in which a person in charge of inspection makes a visual judgment, it takes an enormous amount of time to perform a process of searching the entire surface of the slide glass with a microscope, which poses a serious problem for practical use. For example, it takes about 20 minutes to 1 hour to search for one nucleated red blood cell. In addition, among the test objects on the slide glass, there are many lymphocytes and the like that are similar in shape to nucleated red blood cells derived from the fetus, and it is difficult to discriminate from nucleated red blood cells even by a skilled person, This was a major bottleneck for practical use.
[0010]
In addition, the technique of Patent Document 1 is different in a method of determining a target cell, a method of presenting to a laboratory technician, exchanging a plurality of preparations, and an image processing function of positioning. Recovery is possible. However, in the above apparatus, a high-precision slide glass fixing mechanism and a microscope stage capable of high-precision positioning are required, and physical displacement errors in units of several μm to several tens of μm cannot be avoided, and the reproducibility is low. The workability was not good because the search was needed again because of the scarcity.
[0011]
An object of the present invention is to provide an automatic test apparatus that can accurately detect a test object such as a red blood cell derived from a fetus and can efficiently test a plurality of test objects.
[0012]
[Means for Solving the Problems]
According to one aspect of the present invention, a microscope provided with an objective lens for inspecting an object, a microscope associated with an imaging device, and a sample stage arranged in an inspectable area of the microscope, wherein an XY moving mechanism is provided. A sample table formed so that a sample plate on which a sample to be tested is mounted can be placed at a position observable by the optical microscope, a sample plate in which target cells to be tested have been searched, and a sample before searching. A stocker capable of accommodating a plurality of plates, a transport device for transporting the sample plate between the stocker and the sample table, an analysis unit for analyzing image data of the sample input from the imaging device, and An image processing unit having a storage unit for storing image data and its position of a target cell, and control for controlling focusing by the objective lens and the position of the sample plate. And knitting, the target cells automatic search system including a display device for displaying the observation image and the stored image is provided.
[0013]
Further, when reloading the sample plate to the sample stage after the search, position information when previously searched for any searched target cells, and position information obtained by performing image processing again on the spot, It is preferable to have a correction unit that compares the positional information of the target cells and corrects a positional deviation error relating to the sample plate or the sample table.
[0014]
According to the target cell automatic search system, even if the sample plate shifts during reloading, a plurality of searched target cells can be immediately moved to the center of the field of view of the microscope, and workability is improved. .
[0015]
According to another aspect of the present invention, there is provided a microscope provided with an objective lens for inspecting an object, associated with an imaging device, and a sample stage arranged in an inspectable area of the microscope, comprising: an XY moving mechanism A sample table formed so that a sample plate on which a sample to be inspected is placed can be placed at a position observable by the optical microscope, a stocker capable of accommodating a plurality of the sample plates before and after inspection, and the stocker A transport device that transports the sample plate between the sample and the sample table, an analysis unit that analyzes the image data of the sample input from the imaging device camera, and a storage unit that stores the image data and its position. Target cell including an image processing unit, a control unit for controlling focusing by the objective lens and a position of the sample plate, and a display device for displaying an observation image and a stored image A search method in a dynamic search system, wherein before performing a substantial search for a target cell, a position where cells are present on the sample plate and a position where the cell distribution density is appropriate are recognized by image processing at a low magnification. A method for automatically searching for target cells is provided. When the above method is used, the search area is limited, so that the search time can be significantly reduced.
[0016]
It is preferable that a search is performed at a low magnification while stepping the slide glass at a certain pitch in the search area, and when a candidate for a target cell considered to be a target cell is detected, the target cell candidate is precisely searched at a high magnification. Using this method, a target cell can be quickly searched and determined.
[0017]
BEST MODE FOR CARRYING OUT THE INVENTION
The target cell automatic search system according to the present invention is a system capable of searching for a target cell, loading the target cell again on the microscope after the search, and automatically moving the previously searched target cell to the center of the visual field of the microscope. Blood separated using the Percoll technique or the like that separates cells is applied to a slide glass, fetal cells are identified by morphology, and fetal-derived nucleated cells whose purpose is to reliably collect only fetal cells individually The configuration of the automatic target cell search system according to the present invention for searching for a target is briefly described below.
[0018]
The system according to the present invention includes a microscope, a high-precision digital camera, and a stage on which a slide glass is mounted, and can search for morphological characteristics of target cells. It also has a new search algorithm including image processing that can search for target cells in a short time. As a result, it is possible to search for a target cell with high reliability in a short time.
[0019]
In addition, processing can be performed on a plurality of slide glasses, and the search is stopped by storing an image of a predetermined number of target cells searched in each slide glass and position information (X, Y) on the microscope stage. Alternatively, after completion, the desired slide glass was reloaded on the stage, and the designated target cells were configured to be immediately moved to the center of the field of view of the microscope.
[0020]
At the time of reloading, any searched target cells were searched and stored in the past, and the stored position information was compared with the current and past target cells by performing image processing again on the spot, and the slide glass or microscope was used. By correcting the physical displacement error of the stage etc., even if there is a displacement of the slide glass during reloading, the target cells can be moved to the center of the field of view of the microscope any number of searched target cells immediately. . By displaying an image on a high-precision monitor or observing a microscope image, it is possible for an operator to directly determine cells, and to reliably sort out target cells. Blood was applied to a slide glass, fetal cells were discriminated by morphology, and an automated device for searching for nucleated cells for the purpose of individually collecting only fetal cells was devised.
[0021]
Hereinafter, the target cell automatic search system according to the present embodiment will be described with reference to the drawings. FIG. 1 and FIG. 2 are diagrams showing a configuration example of a target cell automatic search system according to the present embodiment.
[0022]
The apparatus shown in FIGS. 1 and 2 is an automated apparatus that applies blood to a slide glass, identifies fetal cells by morphology, and searches for nucleated cells for the purpose of reliably collecting only fetal cells individually. is there.
[0023]
As shown in FIG. 1, the automatic target cell search system according to the present embodiment includes an optical transport system A, an image processing system (image processing PC) B, a control system (control PC) C, and an output system (high precision). Monitor display) D.
[0024]
The optical transport system A is an optical microscope 1 which includes an objective lens 5 for inspecting an object above the concave groove 3 formed in the upper portion, and an optical microscope 1 having a CCD camera (imaging device) 7. The microscope 1, a sample stage 11 disposed above the optical microscope 1 and having an XY stage mechanism 17, and a sample stage formed so that, for example, a slide glass 23 can be placed on its upper surface 21. The sample on the slide glass 23 provided in the sample placement unit 15 can be inspected by the optical microscope 1 having the objective lens 5 and the CCD camera 7.
[0025]
On the sample stand 11, a stocker 41 having a storage part 45 capable of accommodating a large number of slide glasses 23 before or after the inspection, for example, about 50, is provided. A transport device 31 provided with an operation arm 33 for performing an operation of making it possible to perform an inspection or to be stored in the storage portion 45 of the stocker 41 is provided.
[0026]
The image processing system (image processing PC) B includes an image processing unit 51 that performs data analysis, storage (data storage), and the like. The control PC system C includes a control unit 53 that controls the position of the XY stage 17 and the like and controls the entire processing sequence. The output system D includes, for example, a high-precision monitor display (display device) 55.
[0027]
By including the stocker 41 capable of storing a plurality of slide glasses 23 and the automatic transport device 31, a large amount of search can be continuously performed such as at night. The XY stage 17 provided in the optical microscope 1 is an electric stage capable of performing a high-speed and high-accuracy positioning operation, and is capable of performing a high-speed step feed in a short time within the sample surface. The high-precision digital camera (CCD camera) 7 provided in the optical microscope 1 and image processing software (to be described later) corresponding to the morphological characteristics of the target cells enable a highly accurate search for the target cells in a short time. Become.
[0028]
In the vicinity of the center of the sample stage 11, an opening 15 whose inner diameter is different in two steps (the upper inner diameter is large) is formed. The size of the slide glass 23 is larger than the smaller (lower) opening diameter of the openings 15, and smaller than the larger opening diameter of the openings 15. Thus, for example, the peripheral portion of the slide glass 23 can be placed on the larger bottom surface 21 of the opening.
[0029]
By recording the images of the designated number of target cells searched in each slide glass 23 and the position coordinates (X, Y) on the stage in a storage device, any slide glass can be relocated on the stage after the search. Once loaded, the designated target cells can be immediately moved to the center of the microscope field of view. After the search, when reloading the slide glass 23 onto the sample stage 11, the position information obtained by searching for any searched target cells in the past and the position information obtained by performing image processing again on the spot, By comparing the position information of the target cell between the present and the past and correcting the physical displacement error of the slide glass 23 or the sample stage 11 having the XY stage mechanism 17, the slide glass 23 at the time of reloading is corrected. Even if displacement occurs, a plurality of searched target cells can be immediately moved to the center of the field of view of the microscope, and workability is improved. With the use of the above system, even when a high-precision and expensive slide glass or microscope stage is not used, a mechanism (including an algorithm) for correcting the positional shift of the slide glass 23 is provided, so that low cost and high precision can be achieved. It is possible to search for an inspection object. Further, by displaying an image on the high-accuracy monitor 55 or observing the microscope image, the operator can directly discriminate the cells, and the target cells can be reliably selected.
[0030]
FIG. 2 is a functional block diagram showing a schematic configuration example of the automatic target cell search system according to the present embodiment shown in FIG. 1, and the description will be given with particular emphasis on the image processing system. As shown in FIG. 2, the optical microscope 1 has an electric stage having an XY stage mechanism 17, an objective lens control unit 5, and a CCD camera 7. The control PC includes a stage (XY) moving unit 53a that moves in the XY directions, an automatic focus calculating unit 53b that performs automatic focusing in the Z direction, an image input unit 51a that inputs an image from the CCD camera 7, It has an image recognition unit 51b and a result storage database unit 51c for storing image processing results.
[0031]
The image captured by the CCD camera 7 is input to the image input unit 51a, displayed on the monitor 55, and sent to the image recognition unit 51b and the automatic focus calculation unit 53b. The image recognized by the image recognition unit 51b is stored in the result storage database 51c. Automatic focusing is performed based on the image sent to the automatic focus calculation unit 53b.
[0032]
In the following description, FIG. 1 and FIG.
First, an example of a target cell search procedure will be described with reference to FIG. As shown in step S1, a large number of slide glasses (for example, 50) are stored in a stocker (cassette holder) 41. The stocker 41 is attached to an elevator (not shown) (step S2). As shown in step S3, target search conditions are input on the high-precision monitor display 55, and the slide glass 23 is automatically loaded on the sample stage (XY stage) 17 on the optical microscope 1 using the transport device 31. (Step S4). The sample on the slide glass 23 is focused by the automatic focusing function (step S5). A digital image obtained from the optical microscope 1 is captured (step S6), and a target cell is searched for by image processing (step S7). Image processing is performed according to the shape, color, size, and the like, which are characteristics of the target cells, to select target cells from other cells. If the target cell search is successful in step S7, the process proceeds to step S9, and information on the searched target cell image and position coordinates is recorded in the storage device (database) 51c (FIG. 2).
[0033]
The motorized stage is moved by a small step, the slide glass is moved to a position where another search area can be inspected, and the next search is started. When the search for the specified number of target cells is completed (step S10), the slide glass is unloaded from the stage (S11), and another slide glass 23 is loaded from the stocker 41 (all the slide glasses 23 in the stocker 41 or designated). This is repeated until the search for the plurality of (or single) slide glasses 23 is completed (loop returning from step S11 to step S4). In step S7, if the target cell cannot be searched for by the image processing (NG), the process proceeds to step S8, where the stage 17 is moved, and the slide glass 23 is moved together with the stage 17 to another search area. Return to When the search for the entire amount of the designated slide glass is completed in step S12, the actual search is completed (step S13), and the search result is confirmed on the display screen of the monitor display 55 (step S15). And the target cell on it is specified on the display screen (step S17), and in step S18, the slide glass 23 related to the target cell specified in step S17 is loaded on the XY stage 17 (step S18). Next, in step S19, the target cell is moved by the XY stage 17 to the center of the visual field of the optical microscope 1 or its vicinity based on the position information of the specified target cell. It is confirmed by the optical microscope 1 that the cell at the center of the visual field is the target cell (step S20). If it is determined in step S20 that this is the final confirmation of the target cell, the process proceeds to step S21. If it is determined in step S20 that the confirmation is not final, the process returns to step S17, and the next target cell is designated.
[0034]
If the final confirmation is made in step S20, the target cells are taken out of the slide glass 23 using a micromanipulator (not shown) or the like (step S21). Using the nuclei of the collected target cells, the process proceeds to a chromosome test, a genetic test, and the like (step S22).
[0035]
9A to 9D, after acquiring an image (reference numeral 103 in FIG. 9A) from a CCD camera connected to an optical microscope, as shown in FIGS. A nuclear region candidate is extracted (reference numeral 107 in FIG. 9 (B), reference numeral 115 in FIG. 9 (C)), and it is determined whether the target cell is a target cell by extracting a cell membrane region and a cytoplasmic region (FIG. 9 (D)). ), Reference numeral 118).
[0036]
First, a description will be given of the process of extracting a nucleus region with reference to FIG. In step S31, an image is obtained from the CCD camera 7 connected to the optical microscope 1. This input image is a digital image having RGB components. When an analog input device is used, the digital image is converted into a digital image (RGB digital signal) and the subsequent processing is performed.
[0037]
In step S32, a density level image 1 (L1 = 1-R + 1-G) is created from the input image (RGB). In step S33, a density histogram 1 is created from the density level image 1. From the density histogram 1, a density reference value 1 is determined. In the density level image 1, a binarized image in which pixels higher than the density reference value 1 are set to “1” (area including a nucleus) and the other pixels are set to “0” (background) is created as a mask image 1 ( Step S34). The R component of the input image is masked by the mask image 1 to create an R image 1 (step S35). In step S36, the average value of the R values is calculated from the R image 1. In the R image 1, a pixel smaller than the average value is set to “1” (area including a nucleus), and the other pixels are set to “0” (background). The image is set as a mask image 2 (step S37). The input image (RGB) is masked by the mask image 2 to generate a nucleus candidate region image 1 (RGB) (step S38).
[0038]
In step S39, a density level image 2 (L2 = (R + G + B) / 3) is created from the created nucleus candidate area image 1. In the density level image 2, a connected component in an area other than “0” is searched to perform a labeling process (step S40).
[0039]
In step S41, it is determined whether or not the area of the connected component is smaller than the reference value 2, for example. If the area of the connected component is smaller than the reference value 2, it is determined that the target cell is not included in the region. A determination is made (step S43), and the pixels in the area are set to 0 (background). Here, the reference value 2 is a minimum area that is considered to be occupied by the nucleus of the target cell in the image.
[0040]
For the connected component larger than the reference value 2, in step S44, a density level image 2 including a connected portion is created. Next, a peripheral region including each of the connected components is cut out (this is referred to as a partial image 1. This is shown in FIG. 10A. In FIG. 10A, reference numeral 151 denotes a kernel candidate, and reference numeral 153 denotes other components. is there). In the cut-out partial image 1, the minimum value and the maximum value of the pixels other than “0” (background) are obtained, and the minimum value obtained here is set as the threshold 1 (step S51). For the partial image 1, a binarized image is created according to the set threshold value (step S52). In the binarized partial image, a pixel having a value of "1" is searched for, and connected components of the partial image are labeled (step S53). Here, “labeling a connected component” means assigning a label of the same value to a cluster of connected white (or black) pixels to identify connected pixels (connected components), The connected components indicate a process of assigning different labels.
[0041]
In step S54, it is determined whether or not the area of the connected component is within the range of the size considered to be the nucleus of the target cell, and whether the circularity is, for example, 0.7 or more. If so, consider that area as the nucleus of the target cell (the circularity is 4πS / P, where P is the perimeter and S is the area). ² (The maximum value is 1). In step S54, if no region considered as the nucleus of the target cell is found, in step S56, the threshold value 1 is changed within the range between the minimum value and the maximum value obtained in step S51, and the process returns to step S52. , The processing from step S52 to step S54 is repeated. If an area that is considered to be the nucleus of the target cell is not found even after changing to the maximum value (YES in step S55), it is determined that there is no target cell in this range, and another search area is searched (step S57). .
[0042]
In step S54, when a region considered to be a nucleus of the target cell is found, the process proceeds to step S58, in which a peripheral image is cut out centering on the barycentric coordinates, and a cell peripheral image 1 is created. With respect to the cell peripheral image 1, a mask image 3 (nuclear region mask, FIG. 10D) in which the region considered to be a nucleus is “1” and the other regions are “0” is created (step S59). Next, the process proceeds to the cell membrane region extraction process (FIG. 7) in step S60.
[0043]
As shown in FIG. 7, in step S61, a binary image in which the edge portion is "1" and the other portions are "0" is created in the cell peripheral image 1 by using an edge (boundary line) detection filter. (FIG. 10B). Edge connection processing is performed in this binary image, and the portions where the film is missing are connected on the image (the film image in FIG. 10C). In step S62, a labeling process is performed by searching for a connected component of an edge portion in the binary image. It is determined whether or not the outer periphery of the connected component is within a range that is considered to be the cell membrane of the target cell (step S63). If the outer periphery is within the range, an image in which the region is filled is created and a mask image is created. 4 (area mask in cell membrane) (step S65, FIG. 10E). If it is not within the range, the process proceeds to step S64, where it is determined that the target cell is not the cell membrane, and another search area is searched.
[0044]
Next, AND processing is performed on the created mask image 4 (the intra-cell membrane region mask in FIG. 10E) and the mask image 3 (the nucleus extraction image in FIG. 10D) created in step S59 (FIG. 6). This is performed (FIG. 10F), and it is determined in step S67 whether a nucleus exists in the cell membrane. If it is determined in step S67 that a nucleus (indicated by reference numeral 155 in FIG. 10F) exists in the cell membrane, the process proceeds to the cytoplasm extraction process in step S69. If it is determined in step S67 that no nucleus exists in the cell membrane, the process proceeds to step S68, where it is determined that the cell is not a target cell, and the process proceeds to another search area.
[0045]
FIG. 8 is a flow relating to a cytoplasm extraction step. As shown in FIG. 8, when a nucleus is included in the cell membrane in step S67, a difference image is created by subtracting the mask image 3 (nuclear mask image) from the mask image 4 (mask image in the cell membrane), and the mask image 5 (cytoplasmic mask image) (step S71). Based on the mask image 5 (cytoplasmic mask image) and the mask image 3 (nuclear region mask) created in step S59 (FIG. 6), the ratio of the area between the nuclear region and the cytoplasmic region is calculated (step S72). . In step S73, it is determined whether the calculated ratio is within the range of the ratio indicated by the target cell. If it is determined in step S73 that the ratio is not within the ratio, it is determined that the cell is not a target cell (step S80). Move to another search area.
[0046]
If it is determined in step S73 that the ratio is within the ratio, the input image (RGB) is masked with the mask image 5 (cytoplasm mask image) to create a cytoplasm image (RGB) (step S74). Next, in step S75, an RG histogram is created for the cytoplasm image. It is determined whether the peak of the RG histogram is larger than the reference value 3 (step S76). If the peak of the RG histogram is larger than the reference value 3, the cytoplasm is reddish, it is determined that the cell is not a target cell (NO in step S76), and the cell moves to another search area. If the peak of the RG histogram is smaller than the reference value 3, the process proceeds to step S77 to create a gray level histogram (L3 = (R + G + B) / 3) from the cytoplasm image.
[0047]
In step S78, it is determined whether the peak of the gray level histogram is larger than the reference value 4. If the peak of the gray level histogram is smaller than the reference value 4, the lightness of the cytoplasm is low. The cell is determined not to be a target cell and moves to another search area. If the peak of the gray level histogram is larger than the reference value 4, the process proceeds to step S79, where the finally remaining cells are considered to be target cells, and the process proceeds to step S9 (FIG. 3). The degree of circularity, the lightness of the nucleus, the outer peripheral value of the cell membrane, the area ratio between the cytoplasm and the nucleus, the degree of redness of the cytoplasm, the lightness of the cytoplasm, and the like are stored as probability elements indicating whether the cell is a target cell. When the target cell is found, the barycentric coordinates of the target cell with respect to the input image (RGB) are stored. When the target cell is found, the input image (RGB) is stored in the database 51c (FIG. 2). The stored data on the target cells can be retrieved from the database 51c at any time in response to a request from the user. The processes from S6 to S9 are repeated until a specified number of target cells are found.
[0048]
FIG. 11 shows an example of a monitor display screen relating to the test results obtained by using the target cell automatic search technology according to the present embodiment. As shown in FIG. 11, the target cell automatic search result display screen 200 includes an operation start button 201, a search mode button 203, a search result display button 205, a maintenance button 207, and an end button 209. . At present, the search result display button 205 is selected, and the search result display 210 is displayed on the screen. The search result display 210 is a display related to the sample number 0005, and displays target cell candidates 1 to 5 211a to 211e. The currently selected target cell candidate 211a is displayed in a large size, and various information related to the microscope image display is also displayed.
[0049]
In addition, the search result list 217 displays a sample 221, a target cell number 223, a search start date and time 225, and a search end date and time 227 for each sample No. Further, a read button 231, a save button 233, a delete button 235, and an Exit button 237 are provided, and the above-described read, save, delete, and Exit processes can be performed on each sample. When each button on such a display screen is selected, the selected processing is performed and a simple user interface is formed.
[0050]
As described above, according to the automatic target cell search system according to the present embodiment, target cells contained in blood collected from the mother, for example, using nucleated cells from the fetus, highly reliable to the mother and child Prenatal testing and diagnosis is possible without any risk of illness. According to the automatic target cell search system using nucleated cells derived from the fetus according to the present embodiment, a manual operation that conventionally required about 20 minutes to 1 hour as a search time per nucleated red blood cell is reduced to one slide. Five nucleated red blood cells could be searched for in about 10 minutes per glass. That is, automation of sample processing and long-time unmanned operation are made possible by the above-described apparatus and image processing technology, and conventionally, the burden of a laboratory technician for long-term microscopic observation has been greatly reduced. By using an automated high-precision search device, if the device is set before the end of work and unmanned operation is performed as it is at night, the search result can be displayed on the display the next morning.
[0051]
In addition, using the target cell automatic search device according to the present embodiment, even if the external accuracy of the slide glass, the slide glass fixing mechanism on the microscope stage, or the operation accuracy of the microscope stage is not particularly precise, It can operate normally. If the position of the microscope stage is corrected by image processing, a plurality of arbitrarily searched target cells on the same slide glass can be immediately moved to the center of the field of view of the microscope, so that work can be performed in a short time. In addition, it is not necessary to use a high-precision and expensive slide glass, and the inspection cost can be reduced. Regarding the determination of nucleated red blood cells from countless blood cells that were difficult to determine even by an experienced person, it is sufficient to make a determination from several cells searched using the apparatus according to the present embodiment. Can be greatly reduced.
[0052]
Although the present invention has been described with reference to each embodiment, those skilled in the art can easily understand that various other modifications and changes are possible. For example, it goes without saying that an apparatus storing the above search method as a program, a program thereof, or a recording medium storing the program are also included in the scope of the present invention. In addition, even if the target cell is not a nucleated red blood cell derived from a fetus, it can be tested as long as it is another characteristic search target. For example, various objects including living organisms and physiological substances including cancer cells and lymphocytes.
[0053]
【The invention's effect】
ADVANTAGE OF THE INVENTION According to the target cell automatic search system by this invention, test processing of a target cell can be performed automatically and accurately, and a test time can be shortened significantly.
[0054]
In addition, regarding the external accuracy of the slide glass, the operation accuracy of the slide glass fixing mechanism on the microscope stage or the operation accuracy of the microscope stage, etc., no special accuracy is required due to the position correction of the microscope stage by image processing. For example, it is possible to move the target cells to the center of the visual field of the microscope with a plurality of arbitrary searched target cells immediately, so that the work efficiency can be improved and the inspection cost can be reduced. In addition, even if the user is not skilled in the inspection, the inspection can be performed with high accuracy, and the load on the operator can be greatly reduced.
[Brief description of the drawings]
FIG. 1 is a diagram showing a configuration example of a target cell automatic search system according to an embodiment of the present invention.
FIG. 2 is a functional block diagram showing a configuration example of a target cell automatic search system shown in FIG.
FIG. 3 is a flowchart showing a processing flow in the target cell search system according to one embodiment of the present invention.
FIG. 4 is a flowchart illustrating a flow of a process in the target cell search system according to the embodiment of the present invention, and is a view subsequent to FIG. 3;
FIG. 5 is a flowchart showing a processing flow in the target cell search system according to one embodiment of the present invention, and is a drawing showing a flow of a nuclear region extraction process.
FIG. 6 is a flowchart showing a processing flow in the target cell search system according to one embodiment of the present invention, and is a view showing a processing flow following FIG. 5;
FIG. 7 is a flowchart illustrating a flow of a process in the target cell search system according to the embodiment of the present invention, and is a diagram illustrating a flow of a process of extracting a membrane region.
FIG. 8 is a flowchart showing a process flow in the target cell search system according to the embodiment of the present invention, and is a diagram showing a flow of a cytoplasm extraction process.
FIGS. 9 (A) to 9 (D) are examples of images in each step after performing processing in the target cell search system according to the present embodiment.
FIGS. 10A to 10F are examples of images in each step after processing in the target cell search system according to the present embodiment, from partial image extraction to final nuclear image; 5 is an image after each step up to the extraction of.
FIG. 11 is a diagram showing an example of a display screen in the target cell search system according to the present embodiment.
FIG. 12 is an image of fetal-derived red blood cells (nucleated).
[Explanation of symbols]
A: optical transport system, B: image processing system (image processing PC), C: control system (control PC), D: output system (high-precision monitor display), 1: optical microscope, 3: concave groove portion, 5: objective Lens, 7: CCD camera (imaging device), 11: sample stage, 15: sample placement unit, 17: XY stage mechanism, 21: upper surface, 23: slide glass, 31: transport device, 33: operating arm, 41: stocker , 45 ... accommodation unit, 51 ... image processing unit, 53 ... control unit, 55 ... monitor display (display device).

Claims (16)

With an objective lens for inspecting the object, and a microscope associated with the imaging device,
A sample stage arranged in an inspectable area of the microscope, the sample stage having an XY moving mechanism, and formed so that a sample plate on which a sample to be inspected is placed can be placed at a position observable by the optical microscope. A sample stage,
A stocker capable of accommodating a plurality of sample plates in which target cells to be examined have been searched and a sample plate before search,
A transfer device for transferring the sample plate between the stocker and the sample table,
An analysis unit that analyzes the image data of the sample input from the imaging device, and a storage unit that stores the image data and the position of the searched target cells, and an image processing unit that has
A control unit for controlling the focusing by the objective lens and the position of the sample plate,
An automatic target cell search system including a display device that displays an observation image and a stored image. Further, when reloading the sample plate to the sample stage after the search, position information when previously searched for any searched target cells and position information obtained by performing image processing again on the spot, 2. The automatic target cell search system according to claim 1, further comprising a correction unit configured to compare positional information of the target cells and correct a positional error with respect to the sample plate or the sample table. 3. A target cell automatic search system that performs image processing of an inspection image by a microscope and automatically searches for a target cell placed on a sample plate,
An automatic target image search system, wherein the morphological characteristics of the target cells are divided into a nucleus region, a cell membrane region, and a cytoplasm region, and the image processing of the inspection image is performed. 4. The automatic target cell search system according to claim 3, wherein the target cell is distinguished from other cells based on at least one of a color, a shape, a positional relationship, and an area ratio. Select from a group that includes the area of the nucleus area of the cell, the degree of circularity of the nucleus, the color of the nucleus, the peripheral value of the cell membrane, the area ratio of the cytoplasm and the nucleus, the degree of redness of the cytoplasm, and the brightness of the cytoplasm The target cell automatic search system according to claim 3, wherein a probability value regarding whether or not the cell is the target cell is calculated based on the characteristic to be performed. The target cell automatic search according to any one of claims 3 to 5, further comprising: a display control unit configured to permutate the probability values and display the probability values in descending order. system. The threshold according to the image processing for determining whether or not the target cell is the target cell is set as an adjustable parameter value, wherein the threshold value is determined as an adjustable parameter value. Automatic target cell search system. Furthermore, the position information of the plurality of cells determined to be the target cells during the search processing is stored, and the target cells specified when the sample plate specified after the search processing is reloaded are located at the center of the visual field of the microscope. The target cell automatic search system according to any one of claims 3 to 7, further comprising an automatic positioning mechanism that is automatically moved. The automatic positioning mechanism reloads the designated slide glass after the search, and moves the target cell to the center of the field of view of the microscope from the position information obtained when the specified target cell was searched in the past. By performing image processing again, by comparing positional information of the target cell between the present and the past, it is possible to have a correction means for correcting a physical displacement error between the sample plate and the microscope. The target cell automatic search system according to any one of claims 3 to 7, characterized in that: 10. The automatic target cell search system according to claim 8, wherein the automatic positioning is performed based on a target cell. 11. The image forming apparatus according to claim 3, further comprising an adjusting unit configured to adjust at least one of a threshold value and an importance level of a morphological feature serving as a criterion of a target image based on a previous inspection result. Item 6. The automatic target cell search system according to Item 1. A microscope provided with an objective lens for inspecting an object, a microscope associated with an imaging device, and a sample table arranged in an inspectable area of the microscope, having an XY moving mechanism, and on which a sample to be inspected is placed. A sample table formed so that a sample plate can be placed at a position observable by the optical microscope, a stocker capable of accommodating a plurality of the sample plates before and after inspection, and the sample plate between the stocker and the sample table. An image processing unit having a transport device that transports the image data, an analysis unit that analyzes image data of the sample input from the imaging device, and a storage unit that stores the image data and its position, and focusing by the objective lens. And a control unit for controlling the position of the sample plate and a display device for displaying an observation image and a stored image. There is,
A method for automatically searching for target cells, comprising the step of recognizing at a low magnification a position where cells are present on the sample plate and a position where the cell distribution density is appropriate by image processing before substantially searching for target cells. 13. A step of performing a search at a low magnification while stepping a slide glass at a certain pitch in a search area, and precisely detecting the target cell candidate at a high magnification when detecting a target cell candidate considered to be a target cell. 3. The method for automatically searching for target cells according to 1. A search is performed at a low magnification while stepping the slide glass in the search area at a first pitch, and when a target cell candidate is not detected in the searched area, a second different from the first pitch is used. The target cell automatic search method according to claim 13, wherein the search is performed at a second pitch that does not overlap the search position at the first pitch. The method for automatically searching for target cells according to claim 13 or 14, further comprising cutting out the searched target cells and performing at least one of chromosome and gene tests. The method according to any one of claims 3 to 15, wherein the target cells are nucleated red blood cells derived from a fetus contained in pregnant woman's blood.

Images