CN110689926A - An accurate detection method of high-throughput digital PCR image droplets - Google Patents

Abstract

The invention discloses a method for accurately detecting droplets of a high-throughput digital PCR image, which comprises the following steps of: acquiring a high-throughput PCR digital image; carrying out image preprocessing to obtain a high-quality image; positioning liquid drops; calculating to obtain the area of the liquid drop; the measuring range is wide, and the number of measuring objects is large; the complex state of the whole body can be directly measured and calculated, and a picture can store a very large amount of data; the number of the liquid drops can be accurately detected and the positions of the liquid drops can be positioned.

Description

An accurate detection method of high-throughput digital PCR image droplets

technical field

The invention relates to the technical field of gene sequencing, in particular to an accurate detection method for high-throughput digital PCR image droplets.

Background technique

High-throughput gene sequencing technology, also known as second-generation gene sequencing technology, is a very widely used gene sequencing method in recent years. Compared with the first-generation gene sequencing technology, it has the characteristics of high throughput and low cost. High-temperature polymerase chain reaction (polymerase chain reaction) is referred to as PCR technology, which can rapidly amplify DNA fragments and achieve exponential growth of DNA fragments in several cycles. It is currently the most commonly used DNA amplification technology. High-throughput PCR images contain a large number of fluorescent droplets to be detected, and have the characteristics of large amount of data, many targets to be detected, tight droplet intervals, frequent overlapping, and poor image quality, which are not easy for automatic image detection. How to automatically process these massive data has become an important topic of high-throughput sequencing technology.

In the actual high-throughput mass droplet digital PCR imaging process, the fluorescence collection efficiency of each point in the image field of the large-area fluorescence imaging system may be different, resulting in uneven brightness in the entire image field. Generally, droplets with the same brightness The brightness at the center of the image field is much larger than that at the edge of the image field, and this uneven brightness becomes more prominent as the image field increases, which directly leads to low contrast, unclear images, and inaccurate droplet extraction. . How to more effectively process massive droplet images and accurately detect droplets is an important part of a high-throughput massive droplet digital PCR system.

At present, digital PCR technology mainly uses the absolute quantitative fluorescence monitoring method to detect the number of droplets, and the detection method using digital image processing technology is still relatively small. Among several domestic papers discussing digital image processing, there are many discussions on micro-reaction/well plate digital PCR and large-scale integrated chip digital PCR. For these two types of digital PCR technologies, the droplet arrangement is relatively simple and regular, approximated as a matrix, and it is easy to predict the position of the droplet. Or use a relatively simple method to detect a large number of droplets, with high requirements.

SUMMARY OF THE INVENTION

In view of the above-mentioned shortcomings at present, the present invention provides an accurate detection method for high-throughput digital PCR image droplets, which can accurately detect the number of droplets and locate the position of the droplets.

To achieve the above object, the embodiments of the present invention adopt the following technical solutions:

An accurate detection method for high-throughput digital PCR image droplets, the accurate detection method for high-throughput digital PCR image droplets includes the following steps:

Obtain high-throughput PCR digital images;

Perform image preprocessing to obtain high-quality images;

perform droplet positioning;

Calculate the droplet area.

According to an aspect of the present invention, the specific steps of performing image preprocessing to obtain high-quality images include:

Using the file reading method, read the grayscale image of the specified format image as the source image;

Adaptive thresholding of the source image;

Direct thresholding of the source image;

ORing the adaptive thresholding image and the direct thresholding image, and save the processed image as a merged image;

The open operation operation is performed on the obtained combined image, the processed image is saved as an open operation image, and the preprocessing of the image is completed.

According to an aspect of the present invention, the adaptive thresholding of the source image includes:

For the source image, press the n*n (n>0) sliding frame to slide each pixel point;

Find the average value of all pixel gray values in the sliding box;

Take the average value minus a difference delta as the threshold for thresholding;

For each pixel in the sliding box, if its gray value is higher than the threshold, set its gray value to the maximum gray value, and if it is less than or equal to the threshold, set its gray value to 0.

According to an aspect of the present invention, the direct thresholding of the source image comprises:

Traverse all the pixels in the source image to obtain the average value of the gray values of all the pixels;

Take the average multiplied by a coefficient as the thresholding threshold;

For all pixels in the image, if the gray value is higher than the threshold, set the gray value to the maximum gray value; if it is less than or equal to the threshold, set the gray value to 0.

According to an aspect of the present invention, performing OR operation on the adaptive thresholding image and the direct thresholding image, and saving the processed image as a merged image includes:

Create a new blank image with the same size as the source image;

Comparing the adaptive thresholding and direct thresholding images, if for a point at the same position (x, y) in the two images, the gray value of the two images at this point is the maximum gray value, then in the newly created image In the blank image, the gray value of the pixel corresponding to the coordinates (x, y) is set to the maximum gray value; otherwise, the gray value of the point in the blank image is set to 0.

According to one aspect of the present invention, performing an open operation operation on the obtained combined image, the processed image is saved as an open operation image, and completing the preprocessing of the image includes:

Corrode the image;

Dilate the image after the erosion operation.

According to an aspect of the present invention, the etching operation on the image includes:

Perform sliding operation on each pixel point according to the sliding frame of m*m (m>0) on the image;

Compare the gray value of all pixels in the sliding box to find the minimum value;

Set the gray value of all pixels in the sliding box to the minimum value obtained.

According to an aspect of the present invention, the dilation operation on the image after the erosion operation includes:

Perform sliding operation on each pixel point according to the sliding frame of k*k (k>0) on the image;

Compare the grayscale values of all pixels in the sliding box to obtain the maximum value;

Set the gray value of all pixels in the sliding box to the maximum value obtained.

According to an aspect of the present invention, the specific steps of performing droplet positioning include:

Perform canny edge detection on the open operation image to obtain an edge detection image;

Contour detection is performed on the edge detection image, the contour information of all droplets is obtained, the position of the droplet is determined, and it is drawn on the source image.

According to an aspect of the present invention, the calculating to obtain the droplet area includes: using a contour area calculation function to calculate each droplet contour in the obtained droplet contour information to obtain all droplet areas.

Advantages of the implementation of the present invention:

1. This is a non-destructive, non-contact inspection.

2. Fast measurement and response; WYSIWYG measurement.

3. The measurement range is wide and there are many measurement objects; the complex state of the whole can be directly measured and calculated, and a photo can save a very large amount of data.

4. Images can be captured at any time, online collection, query and processing.

5. The focus can be adjusted, and it can be adjusted between the whole and the part at any time.

6. The equipment is simple. All that is needed is a set of photographic equipment, an analog-to-digital conversion module, and a data processing module, and the program can be copied to these devices.

Description of drawings

In order to illustrate the technical solutions in the embodiments of the present invention more clearly, the following briefly introduces the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained from these drawings without any creative effort.

Fig. 1 is the grayscale image to be processed of the source image in step S21;

Fig. 2 is step S22 adaptive thresholding image;

3 is a direct thresholding image in step S23;

Fig. 4 is step S24 merging images;

Fig. 5 is step S25 open operation image;

Fig. 6 is the edge detection image of step S31;

Fig. 7 is that step S32 draws an image;

FIG. 8 is a schematic diagram of the accurate detection method of high-throughput digital PCR image droplets according to the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

As shown in Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, Figure 6, Figure 7 and Figure 8, a high-throughput digital PCR image droplet accurate detection method, including:

S1. Acquisition of digital images: use CCD and other photosensitive elements to obtain digital images of the target area;

S2. Image preprocessing: use image processing methods such as histogram equalization and thresholding to improve image quality;

S3. Droplet positioning: determine the position of each droplet in the image, and remove the falsely detected droplets;

S4. Calculation of droplet area: the area of each droplet is calculated by using a digital image processing method.

The step S2 includes:

S21, using the file reading method, read in the grayscale image of the image in the specified format, this image is called the source image, and the source image is shown in Figure 1;

S22. Perform adaptive thresholding processing on the source image in S21. The step S22 includes:

S221, perform a sliding operation on each pixel point according to the sliding frame of n*n on the image, and the value of n is, for example, 9;

S222. Find the average value of the grayscale values of all the pixels in the sliding frame;

S223, take the average value minus a difference delta as the threshold value taken by thresholding;

S224. For each pixel in the sliding frame, if its gray value is higher than the threshold, set its gray value to the maximum gray value, and if it is less than or equal to the threshold, set its gray value to 0;

The image after adaptive thresholding is shown in Figure 2;

S23. Perform direct thresholding processing on the source image in S21, and the step S23 includes:

S231, traversing all the pixels in the image to obtain the average value of the gray values of all the pixels;

S232, taking the average value and multiplying it by a coefficient as the thresholding threshold, and the value of the coefficient is, for example, 0.75;

S233. For all pixels in the image, if the gray value is higher than the threshold, set the gray value to the maximum gray value, and if it is less than or equal to the threshold, set the gray value to 0;

The image after direct thresholding is shown in Figure 3;

S24, performing an OR operation on the adaptive thresholding image described in S22 and the direct thresholding image described in S23, and step S24 includes:

S241. Create a new blank image whose size is the same as that of the source image;

S242. Comparing the adaptive thresholding and direct thresholding images, if for a point at the same position (x, y) in the two images, the gray value of the two images at this point is the maximum gray value, then In the newly created blank image, the gray value of the pixel corresponding to the coordinates (x, y) is set as the maximum gray value. Otherwise, set the gray value of the point in the blank image to 0;

The combined image is shown in Figure 4;

S25, perform an opening operation on the merged image obtained in S24, and the step S25 includes:

S251, perform an erosion operation on the image, and the step S251 includes:

S2511, perform a sliding operation on each pixel point according to the sliding frame of n*n on the image, and the value of n is, for example, 15;

S2512, compare the gray values of all the pixels in the sliding frame, and obtain the minimum value;

S2513, set the gray value of all pixels in the sliding frame as the obtained minimum value;

S252, performing an expansion operation on the image after the corrosion operation, and the step S252 includes:

S2511, perform a sliding operation on each pixel point according to the sliding frame of n*n on the image, and the value of n is, for example, 15;

S2512, compare the gray values of all pixels in the sliding frame, and obtain the maximum value;

S2513, setting the gray value of all pixels in the sliding frame as the obtained maximum value;

The image obtained by the open operation operation is shown in Figure 5;

S31, performing canny edge detection on the open operation image described in S25;

The resulting image is shown in Figure 6;

S32, perform contour detection on the edge detection image described in S31, obtain contour information of all droplets, determine the positions of the droplets, and draw them on the source image;

The resulting image is shown in Figure 7;

Use the contour area calculation function to calculate each droplet contour in the droplet contour information obtained in S32 to obtain all droplet areas, and store the results in a local file.

Advantages of the implementation of the present invention:

1. This is a non-destructive, non-contact inspection.

2. Fast measurement and response; WYSIWYG measurement.

3. The measurement range is wide and there are many measurement objects; the complex state of the whole can be directly measured and calculated, and a photo can save a very large amount of data.

4. Images can be captured at any time, online collection, query and processing.

5. The focus can be adjusted, and it can be adjusted between the whole and the part at any time.

6. The equipment is simple. All that is needed is a set of photographic equipment, an analog-to-digital conversion module, and a data processing module, and the program can be copied to these devices.

The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited to this. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, All should be included within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (10)

1. The accurate detection method of the high-throughput digital PCR image droplets is characterized by comprising the following steps of:

acquiring a high-throughput PCR digital image;

carrying out image preprocessing to obtain a high-quality image;

positioning liquid drops;

and calculating to obtain the area of the liquid drop.

2. The method for accurately detecting the droplets in the high-throughput digital PCR image according to claim 1, wherein the specific step of performing image preprocessing to obtain a high-quality image comprises:

reading in a gray image of an image in a specified format as a source image by using a file reading mode;

carrying out self-adaptive thresholding treatment on a source image;

performing direct thresholding on the source image;

performing OR operation on the self-adaptive thresholding image and the direct thresholding image, and storing the processed images as a combined image;

and performing opening operation on the obtained combined image, and storing the processed image as an opening operation image to finish the image preprocessing.

3. The method for accurately detecting the high-throughput digital PCR image droplets as set forth in claim 2, wherein the adaptively thresholding the source image comprises:

carrying out sliding operation on each pixel point on a source image according to a sliding frame of n x n (n is more than 0);

calculating the average value of the gray values of all pixel points in the sliding frame;

subtracting a difference value delta from the average value to be used as a threshold value for thresholding;

for each pixel in the sliding frame, if the gray value is higher than the threshold value, the gray value is set as the maximum gray value, and if the gray value is smaller than or equal to the threshold value, the gray value is set as 0.

4. The method for accurately detecting droplets in high-throughput digital PCR images according to claim 2, wherein the directly thresholding the source image comprises:

traversing all pixel points in the source image to obtain an average value of gray values of all the pixel points;

taking the average value multiplied by a coefficient as a threshold value;

and setting the gray value of all pixel points in the image as the maximum gray value if the gray value is higher than the threshold value, and setting the gray value as 0 if the gray value is less than or equal to the threshold value.

5. The method for accurately detecting droplets in a high-throughput digital PCR image according to claim 2, wherein the performing or operation on the adaptive thresholded image and the direct thresholded image, and the saving of the processed images as a combined image comprises:

newly building a blank image, wherein the size of the blank image is consistent with that of the source image;

comparing the self-adaptive thresholding image with the direct thresholding image, and setting the gray value of a pixel point with the corresponding coordinate (x, y) as the maximum gray value in a newly-built blank image if the gray values of the two images at the point with the same position (x, y) in the two images are the maximum gray value; otherwise, setting the gray value of the point in the blank image as 0.

6. The method for accurately detecting the droplets in the high-throughput digital PCR image according to claim 2, wherein the merging images obtained by the method are subjected to an on operation, the processed images are saved as on operation images, and the preprocessing of the images comprises:

carrying out corrosion operation on the image;

and performing expansion operation on the image subjected to the corrosion operation.

7. The method for accurate detection of droplets in high-throughput digital PCR images according to claim 6, wherein said performing erosion operations on the images comprises:

performing sliding operation on each pixel point of the image according to a sliding frame of m by m (m is more than 0);

comparing the gray values of all pixel points in the sliding frame to obtain the minimum value;

and setting the gray values of all pixel points in the sliding frame as the obtained minimum value.

8. The method for accurately detecting the high-throughput digital PCR image droplets as claimed in claim 7, wherein the expanding the image after the erosion operation comprises:

performing sliding operation on each pixel point of the image according to a sliding frame of k x k (k is more than 0);

comparing the gray values of all pixel points in the sliding frame to obtain the maximum value;

and setting the gray values of all pixel points in the sliding frame as the maximum value.

9. The method for accurately detecting high-throughput digital PCR image droplets according to any one of claims 2 to 8, wherein the specific step of performing droplet positioning comprises:

carrying out canny edge detection on the opening operation image to obtain an edge detection image;

and carrying out contour detection on the edge detection image to obtain contour information of all liquid drops, determining the positions of the liquid drops and drawing the liquid drops on the source image.

10. The method for accurately detecting droplets in high-throughput digital PCR images according to claim 9, wherein the calculating the droplet area comprises: and calculating the outline of each droplet in the obtained droplet outline information by utilizing an outline area calculation function to obtain the area of all droplets.

Images