JP2007017345A - Method and apparatus for automatically detecting observation target - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform the automatic sorting of a metaphase cell capable of observing a chromosome with higher precision. <P>SOLUTION: When the metaphase cell being the aggregate of chromosomes is sorted from an image to be displayed, small particles having a size almost same to that of the chromosomes in the image are extracted to be binarized and the aggregate formed of the small particles gathered in the density of the degree of the chromosomes is extracted to be sorted based on feature quantity. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an observation object automatic detection method and apparatus, and in particular, an observation object automatic detection method suitable for use as a metaphase finder for detecting metaphase cells (metaphase) when performing chromosome analysis, and Relates to the device.

  By observing the occurrence frequency of chromosomal abnormalities observed when a living body is exposed to radiation, the exposure dose can be estimated. In particular, the phenomenon of abnormal chromosomes appearing specifically at the time of radiation exposure, and can be easily observed with an ordinary staining method using an optical microscope. It is a useful indicator. However, when the exposure dose is low, the frequency of occurrence is low, and therefore, it is necessary to observe a large amount of specimens. Therefore, observation by human power is difficult, and automation has been expected.

  A device with the metaphase finder function that automatically selects and displays metaphase cells that can observe chromosomes from a sample under a microscope is commercially available by foreign companies, but its image processing function is not disclosed. It is unknown, and at least the structural elements of mathematical morphological operations (also called morphological operations) are not used, and the processing results are inaccurate.

  Thus, as described in Non-Patent Document 1, the inventors made a prototype of an apparatus that uses a structural element of a morphological operation for metaphase finder image processing.

Akira Furukawa et al. “Automation of chromosome aberration analysis” Cytometry Research 7 (1), pp1-8, 1997

  However, in Non-Patent Document 1, image processing using structural elements is only performed after binarization of the original image, and structural elements are used for image processing when binarizing the input image. Because it was not known, cell nuclei that were stained as light as the threshold could not be separated from the chromosomes, often resulting in false results.

  The present invention has been made to solve the above-mentioned conventional problems, and enables automatic collection of an aggregate to be observed, such as metaphase cells containing many chromosomes, with higher accuracy. Is an issue.

  The present invention relates to an automatic detection method of an observation object for selecting and displaying a collection of observation objects from an image, and a procedure for extracting small particles having the same size as an observation object in an image and binarization thereof. And a procedure for extracting an aggregate in which the small particles are gathered at a density of an observation target, and a procedure for selecting the aggregate based on a feature amount. is there.

  Further, the procedure for extracting the small particles removes small particles having the same size as the observation target by applying a connection process to the entire image using a structural element having the same size as the observation target. A procedure, and a procedure of adding a contraction to the result and subtracting it from the original image.

  Further, the procedure for extracting the aggregate is a procedure for connecting the extracted small particles by a connection process using a structural element having a size that can fill a gap of an observation object scattered in one aggregate. And a procedure for performing a cutting process on the obtained image to smooth the outline shape of the aggregate in which the interior is filled.

  Further, the present invention provides an automatic detection method for an observation object, wherein the observation object is a chromosome and the aggregate is a metaphase cell.

  The present invention also provides an automatic detection device for an observation target for selecting and displaying a collection of observation targets from an image, and means for extracting small particles having a size about the observation target in the image. The problem has been solved by comprising means for pricing, means for extracting an aggregate in which the small particles are gathered at a density about the observation target, and means for selecting the aggregate based on a feature amount. Is.

  Further, the means for extracting small particles removes small particles having the same size as the observation object by applying a connection process to the entire image using a structural element having the same size as the observation object. Means and means for applying a contraction to the result and subtracting it from the original image.

  Further, the means for extracting the aggregate is a means for combining the extracted small particles by a connection process using a structural element having a size that can fill the gaps of the observation object scattered in one aggregate. And means for performing a cutting process on the obtained image and smoothing the outline shape of the aggregate in which the interior is filled.

  The present invention also provides an automatic detection apparatus for an observation object, wherein the observation object is a chromosome and the aggregate is a metaphase cell.

  According to the present invention, it is possible to perform automatic collection of observation target aggregates with higher accuracy. In particular, the processing prior to binarization leaves only small particles and removes the effects of non-uniformity of light sources in the image, uneven distribution of aggregates, uneven color and density of aggregates, etc. Can do. In addition, cell nuclei stained as thin as the threshold, which could not be obtained by the method described in Non-Patent Document 1, can be separated from chromosomes.

  Hereinafter, an embodiment of the present invention applied to a metaphase finder of a chromosome abnormality analysis apparatus will be described in detail with reference to the drawings.

  As shown in FIG. 1, the chromosomal abnormality analysis apparatus for carrying out the present invention includes an XY stage 12 driven by, for example, a stepping motor (not shown) for mounting a slide glass 16 mount (not shown) and an automatic A small microscope 10 having a focusing device 14, an imaging device 20 such as a normal video camera or a high-vision camera, and an image processing computer 30 such as a workstation are provided.

  The XY stage 12 has sufficient position resolution and position reproducibility compared with the size of the chromosome.

  As the automatic focusing device 14, a passive type that does not emit light for focusing is used. For example, an active type that irradiates infrared light for focusing can also be used.

  The image processing computer 30 includes software for performing the following processing.

  (1) According to the present invention, metaphase cells on the slide glass 16 are detected and their positions are recorded.

  (2) Taking an enlarged view of metaphase cells and selecting those suitable for subsequent image processing.

  (3) Separate individual chromosomes in the cell and detect the location of the centromere on the chromosome.

  (4) Display abnormal chromosome candidates on the screen and, if necessary, correct them with human hands.

  (5) Aggregate the results and determine the incidence of abnormal chromosomes.

  Hereinafter, the metaphase finder processing according to the present invention will be described with reference to FIG.

  First, in step 100, as shown in FIG. 3, the XY stage 12 is moved only by the visual field 22 of the image pickup device 20, the slide glass 16 is moved, and after focusing is performed by the autofocus device 14, the image is taken. A photographed image (also referred to as an original image) as shown is obtained.

  Next, in step 110, small particles having a size about the chromosome that is the observation target in the original image are extracted. Specifically, first, a small particle having the same size as a chromosome is removed as shown in FIG. At this time, a structural element having the same size as the chromosome image is used.

  Next, a slight erosion process is added to the connection processing result shown in FIG. 5 to obtain an image as shown in FIG. 6 and subtracted from the original image, as shown in FIG. Extract the small particles that disappeared in step 5. The structural elements at this time are sized so that the peripheral portions of the object other than the small particles to be extracted do not remain when subtracting.

  Next, the routine proceeds to step 120 where binarization is performed with an appropriate threshold value to obtain a binarized image as shown in FIG.

  Next, at step 130, a portion where small particles are gathered at a density of about the chromosome in the cell is extracted. Specifically, the extracted small particles are combined by a connection process to obtain an outline image of the aggregate as shown in FIG. As a structural element at this time, a structural element having a size that can fill a space between chromosomes scattered inside one metaphase cell is used. If the structuring element is too small, the gap cannot be filled, and conversely, if the structuring element is too large, it is likely to join adjacent separate metaphase cells. It is necessary to use it.

  Next, a cutting process (Opening) is performed on the obtained contour image to smooth the contour shape of the metaphase cells filled in the interior, thereby obtaining an image of an aggregate as shown in FIG.

  Next, the routine proceeds to step 140, where the extracted aggregates are selected based on the feature values for the shape (area, aspect ratio, circularity, etc.).

  From the measurement value obtained as a result, it is determined whether the image includes a chromosome, and the position on the slide glass 16 is obtained and recorded.

  Next, it is determined in step 150 whether or not it is the end point. If it is not the end point, the process proceeds to step 160 where the slide glass 16 is moved and the next image is taken as shown in FIG.

  By repeating the above, the portion on the slide glass 16 where the sample is placed is scanned.

  By reproducing the scan result record and bringing the detected cells to the center of the field of view of the microscope, subsequent processing such as chromosome detection can be easily performed.

  In the above embodiment, the present invention has been applied to a metaphase finder for detecting metaphase cells in which chromosomes can be observed. However, the application target of the present invention is not limited to this, and for example, obtained by a telescope The same applies to selecting galaxy groups and globular clusters from images, detecting fish schools from ultrasound images, and finding groups of people and ships from aerial photographs. In addition, when the size of the object to be observed is different, it is possible to know the distribution for each size by performing the processing according to the present invention for each size.

The block diagram which shows the structure of the chromosome abnormality analysis apparatus which is embodiment of this invention. Flow chart showing the processing procedure of the embodiment The top view which similarly shows the scanning position on the slide glass The figure which shows the example of the same photographed image (original image) The figure which shows the image which performed connection (Closing) processing in FIG. FIG. 5 is a diagram showing an image subjected to the erosion process. The figure which shows the image which subtracted FIG. 6 from the original image The figure which shows the image which binarized FIG. 7 and extracted the small particle The figure which shows the outline image of the aggregate | assembly obtained by giving a connection process to FIG. The figure which shows the image which performed the cutting (Opening) processing with respect to FIG. 9, and smoothed and extracted the aggregate | assembly

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Optical microscope 12 ... XY stage 14 ... Automatic focus apparatus 16 ... Slide glass 20 ... Imaging device 30 ... Computer for image processing

Claims (8)

In the automatic detection method of the observation target for selecting and displaying the aggregate of the observation target from the image,
A procedure for extracting small particles about the size of the observation object in the image;
A procedure to binarize this,
A procedure for extracting an aggregate in which the small particles are gathered at a density of an observation target;
A procedure for selecting the aggregate based on the feature amount;
A method for automatically detecting an observation target, comprising: The procedure for extracting the small particles comprises:
A procedure for removing small particles of the same size as the observation target by applying a connection process to the entire image using a structural element of the same size as the observation target;
A procedure to shrink the result and subtract it from the original image;
The method for automatically detecting an observation target according to claim 1, wherein: The procedure for extracting the aggregate is:
A procedure of combining the extracted small particles by a connection process using a structural element having a size that can fill the gaps of the observation object scattered in one aggregate,
A procedure for cutting the obtained image and smoothing the outline shape of the assembly filled in the interior,
The method for automatically detecting an observation target according to claim 1, wherein:   The method for automatically detecting an observation target according to any one of claims 1 to 3, wherein the observation target is a chromosome and the aggregate is a metaphase cell. In the automatic detection device for the observation target for selecting and displaying the aggregate of the observation target from the image,
Means for extracting small particles about the size of the observation object in the image;
Means for binarizing this,
Means for extracting an aggregate in which the small particles are gathered at a density of an observation target;
Means for selecting the aggregate based on the feature amount;
A device for automatically detecting an observation target. Means for extracting the small particles;
Means for removing small particles of the same size as the observation target by applying a connection process to the entire image using a structural element of the same size as the observation target;
Means to add shrinkage to the result and subtract it from the original image;
The observation object automatic detection apparatus according to claim 5, comprising: Means for extracting said aggregate;
Means for connecting the extracted small particles by a connection process using a structural element having a size sufficient to fill the gaps in the observation object scattered in one aggregate;
Means for performing a cutting process on the obtained image and smoothing the outline shape of the assembly filled in the interior;
The observation object automatic detection apparatus according to claim 5, comprising:   8. The observation object automatic detection apparatus according to claim 5, wherein the observation object is a chromosome and the aggregate is a metaphase cell.

Images