JPH0610676B2 - Automatic reticulocyte counter - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic blood cell counter using pattern recognition, and particularly to the influence of dust in an optical system when identifying a minute image such as reticulocyte count. The present invention relates to a reticulocyte automatic counting device suitable for performing automatic counting without being subjected to the treatment.

[Background of the Invention]

For the conventional reticulocyte automatic counting, refer to the 18th Japan ME
As described in the presentation entitled "Automatic measurement of reticulocytes from microscopic images" at the conference, 2-C-4, 594 (Sho 54), erythrocytes, injured erythrocytes,
Also known are methods of identifying the three members of reticulocytes.
However, in this method, the influence on dust due to dust of the optical system, defects of the image pickup device in the image pickup apparatus, etc. was not considered. If these dusts, defects, etc. are present in the image, those features are reticulated within reticulocytes (nuclear remnants).
It is difficult to distinguish between the two, and normal red blood cells may be misidentified as reticulocytes when red blood cells and obstacles overlap.

[Object of the Invention]

An object of the present invention is to preliminarily store the positions on the image such as dust on the optical path of the optical system and defects of the image pickup device, thereby eliminating the influence of these obstacles and performing reliable counting. An object of the present invention is to provide an automatic reticulocyte counting apparatus capable of performing the above.

[Outline of Invention]

In order to achieve such an object, the present invention relates to an imaging device that electrically scans a microscopic image of blood cells and outputs a density signal, and an image processing device that extracts characteristic parameters of blood cells from the density signal of the imaging device. And a reticulocyte automatic counting device comprising means for classifying blood cells based on the characteristic parameters of the image processing device, a storage means for storing the position of the blood cells on the image, and an optical system light between the microscope light source and the imaging device. The storage means is provided for storing the position on the image of the obstacle with respect to the blood cell count on the road and in the imaging device, and the means for determining whether or not the blood cell and the obstacle are overlapped from each of the storage means.

Example of Invention

An embodiment of the automatic reticulocyte counting apparatus according to the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing an automatic reticulocyte counting apparatus according to the present invention. In the figure, an autofocus control device 2 and a color TV camera 3 are attached to a microscope 1, and a microscope image of blood cells scattered on a blood sample (not shown) placed on the microscope 1 is input to the color TV camera 3. It is supposed to do. An electric signal corresponding to the density signal of the image output from the color TV camera 3 is input to the image processing device 4, and the image processing device 4 performs the feature amount extraction process of blood cells according to the instruction of the computer 5, and the feature amount is extracted. The characteristic parameters extracted by the extraction device are output to the computer 5. The computer classifies and counts normal red blood cells and reticulocytes based on the characteristic parameters.

In order to eliminate the influence of the obstacle in the automatic reticulocyte counting, the operator first inputs a command for registering the position of the obstacle into the computer 5 using the keyboard 6 before starting the blood cell counting. When this command is input, the computer 5 removes the blood sample placed on the microscope 1 by using the stage driving device 7. After that, the computer 5 instructs the image processing device 4 to start image processing. Then, the image processing device 4 is adapted to capture the image in a storage device in the device. The image thus captured contains obstacles in the optical path, as shown in FIG. The image processing device 4 and the computer 5 are adapted to perform a process for obtaining the position of an obstacle as shown in the flowchart of FIG. That is, the image is taken in at step 301, and the binarization processing of this image is performed at step 302. In this binarization process, the maximum and minimum values of the density in the image are detected, and based on the threshold value, 1 or more is set, and 0 is set below it. Next, in step 303, so-called labeling processing is performed in which the binarized image is numbered for each independent pattern. In this labeling process, for example, as shown in FIG. 4, the portions detected as 1 in the binarized image appear as one lump and are scattered, so that lump, that is, an independent pattern. It is done by numbering each item. At this point, the image processing device 4 determines the number of independent patterns (fourth pattern).
(3 in the case of the figure) is output to the computer 5, and step 3
At 04, the computer 5 determines whether this number is more or less than the preset number. If so, an alarm is displayed on the display device 8 in step 307 and the image processing is terminated. On the other hand, if the number is equal to or less than the set number, in step 305, the image processing apparatus 4 is instructed to continue image processing. The image processing apparatus 4 then finds the minimum and maximum values of the coordinates in both X and Y directions for each numbered pattern in step 306 and outputs these data to the computer 5, as shown in FIG. The computer 5 is adapted to store these data as the coordinates of the obstacle.

Next, the operator sends a reticulocyte automatic classification command to the keyboard 6
Is input to the computer 5 through.
The computer 5 outputs the classification command to the image processing device 4 after moving the blood sample under the microscope 1 by using the stage driving device 7 and executing the autofocusing by using the autofocusing control device 2. The image processing apparatus 4 performs image capturing, binarization processing, labeling processing,
The position detection processing is performed, and the minimum value and the maximum value of the coordinates of each blood cell in the image are output to the computer 5. The minimum and maximum values of the coordinates for each blood cell are as shown in FIG.

If the condition (1) is satisfied for all the stored obstacles for each blood cell, the computer 5 does not overlap the blood and the obstacles, so that the image processing apparatus 4 continues the image processing. The image processing device 4 extracts the characteristic parameters, and the normal red blood cells and reticulocytes are discriminated based on the extracted characteristic parameters. If the formula (1) is not satisfied, the blood cells are not classified because the blood cells and the obstacles overlap each other.

The above equation (1) is an equation performed for each of the x and y coordinates.

In this way, the above processing is performed for all blood cells in one image, and when the number of red blood cells exceeds the set number, the counting for one sample is completed, and when the number does not exceed the set number, the stage driving device 7 After moving the sample by, the same processing is repeated.

A flowchart of the above process is shown in FIG.

As is clear from the above description, according to the present embodiment, it is possible to obtain an accurate count value without being affected by obstacles in reticulocyte automatic counting. Therefore, there is an effect that the reliability of the device can be remarkably improved, and even a defective image pickup device can be used, and the price of the image pickup device is about half that of a non-defective image pickup device. There is an effect that can be done.

〔The invention's effect〕

As described above, according to the reticulocyte automatic counting apparatus of the present invention, it is possible to perform counting without affecting the result of reticulocyte automatic counting even if an obstacle is present in the image. .

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of an automatic reticulocyte counting apparatus according to the present invention, FIG. 2 is an image pattern of an obstacle in automatic reticulocyte counting, and FIG. 3 is an obstacle in an automatic reticulocyte counting apparatus according to the present invention. FIG. 4 is a flowchart for registering the position of the object, FIG. 4 is an explanatory view showing the obstacle as a binary pattern, FIG. 5 is an explanatory view in which the obstacle in FIG. 4 is labeled, and FIG. 6 is a red blood cell image. FIG. 7 is a flow chart showing a process for eliminating the influence of obstacles in the reticulocyte automatic counting apparatus according to the present invention. 1 ... Microscope, 2 ... Automatic focus control device, 3 ... Imaging device, 4 ... Image processing device, 5 ... Computer, 6 ...
Keyboard, 7 ... Display device.

Claims (1)

[Claims] 1. An image pickup device for electrically scanning a microscopic image of blood cells to output a density signal, an image processing device for extracting characteristic parameters of blood cells from the density signal of the image pickup device, and features of the image processing device. In a reticulocyte automatic counting device consisting of means for classifying blood cells based on parameters,
Storage means for storing the position of the blood cell on the image, storage means for storing the position on the image of the harmful substance with respect to the blood cell count in the optical system optical path between the microscope light source and the imaging device and in the imaging device, and each of the storage An automatic reticulocyte counting device comprising: a means for determining whether or not blood cells and a harmful substance overlap with each other.