Image registration method and device and electronic equipment
By dividing and processing images according to the parameters of the target image in the second-generation sequencing technology, the problem of image registration failure caused by base imbalance is solved, the reliability of image registration is improved, and the integrity of data is ensured.
Patent Information
- Application Number
- CN202310816195.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-07-04
- Publication Date
- 2025-01-07
AI Technical Summary
In second-generation sequencing technology, image registration failure caused by base imbalance samples leads to data loss and sequencing failure, and the prior art is difficult to effectively solve this problem.
By obtaining the target parameters of the target image, the images are divided into two categories: the presence and absence of base deletion, the image registration algorithm is used to register images without base deletion, and the images with base deletion are processed according to the results. Avoid the situation where images cannot be registered due to base deletion.
It improves the reliability of image registration, ensures that effective image registration can be achieved even in the absence of bases, and avoids data loss.
Smart Images
Figure CN119273722A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of image processing technology, and in particular to an image registration method, device and electronic equipment. Background Art
[0002] The second-generation sequencing technology uses high-resolution microscopic imaging to obtain the signals of fluorescent molecules and obtains the base sequence by decoding the image signals. Different sequencing platforms will have different optical systems and different fluorescent dyes to obtain different point light source signal images. If there are multiple images of the same scene during the sequencing process, it is first necessary to align the images of the same scene through positioning and registration algorithms, and then extract the point signals through the algorithm, and perform subsequent brightness information analysis and processing to obtain the base sequence. With the development of second-generation sequencing technology, sequencer products are equipped with real-time sequencing data processing and analysis software. First, different base signals need to be matched. A base signal corresponds to a specific coordinate position in the image. The corresponding position of the base in the image needs to be matched. Matching speed, accuracy and robustness are the basic requirements for image registration.
[0003] If the sequencing sample is a base-unbalanced sample, that is, in the same scene, the content of one or at most three bases is too small (less than 12.5%), the low data content leads to sparse fluorescent spots on the image, which is insufficient to extract image features for registration, and the registration of this scene fails, resulting in data loss. If similar situations occur in multiple scenes, the sequencing fails and no results are produced, resulting in irreparable data loss.
[0004] Therefore, there is an urgent need for a technical solution that can improve the reliability of image registration. Summary of the invention
[0005] In view of this, the present application provides an image registration method, device and electronic device to improve the reliability of image registration, as follows:
[0006] An image registration method, comprising:
[0007] Acquire multiple frames of target images, wherein the target images are images acquired by capturing images of a sample to be tested under laser irradiation, wherein the sample to be tested is dyed with a fluorescent dye;
[0008] Obtaining a target parameter corresponding to the target image, wherein the target parameter indicates whether there is a base deletion in the target image;
[0009] According to the target parameter, the target image is divided to obtain a first image and a second image, wherein the first image is an image without base deletion, and the second image is an image with base deletion;
[0010] Using an image registration algorithm, processing the first image to obtain a first registration result;
[0011] The second image is processed according to the first registration result to obtain a second registration result.
[0012] In the above method, preferably, obtaining the target parameters corresponding to the target image includes:
[0013] Selecting a target area in the target image;
[0014] Sort the signal values of the pixels in the target area from small to large;
[0015] According to the first signal value ranked at the first quantile and the second signal value ranked at the second quantile, a target parameter corresponding to the target image is obtained.
[0016] The above method preferably obtains the target parameter corresponding to the target image according to the first signal value ranked at the first quartile and the second signal value ranked at the second quartile, including:
[0017] When a ratio between a first signal value ranked at a first quartile and a second signal value ranked at a second quartile is less than a preset first threshold, obtaining a target parameter characterizing the presence of a base deletion in the target image;
[0018] When the ratio is greater than or equal to the first threshold, a target parameter characterizing that there is no base deletion in the target image is obtained.
[0019] In the above method, preferably, the first quantile is 98% and the second quantile is 30%.
[0020] Preferably, the above method processes the second image according to the first registration result to obtain a second registration result, including:
[0021] Processing the offset and rotation value of the second image on the first channel according to the relative offset value and the relative rotation value to obtain the offset and rotation value of the second image on the second channel;
[0022] The second image is registered according to the offset and rotation value of the second image on each of the image channels to obtain a second registration result.
[0023] In the above method, preferably, the relative offset value is obtained by:
[0024] Obtaining a sample image, wherein the sample image is an image obtained by a base sequencing device to which the target image belongs, capturing an image of a target sample that is irradiated by a laser and dyed with a fluorescent dye;
[0025] Obtaining the offset of the sample image on each image channel;
[0026] Subtracting the offsets of the sample image on any two image channels to obtain a plurality of first difference values;
[0027] A peak value corresponding to the first difference in the normal distribution table is determined as the relative offset value.
[0028] In the above method, preferably, the rotation relative value is obtained by:
[0029] Obtaining a sample image, where the sample image is an image processed on a base sequencing device to which the target image belongs and does not have base deletions;
[0030] Obtaining a rotation value of the sample image on each image channel;
[0031] Subtracting the rotation values of the sample image on any two image channels to obtain a plurality of second difference values;
[0032] The peak value corresponding to the second difference in the normal distribution table is determined as the rotation relative value.
[0033] In the above method, preferably, the first difference is smaller than the second threshold, and the second difference is smaller than the third threshold.
[0034] An image registration device, comprising:
[0035] An image acquisition unit, used to acquire multiple frames of target images, wherein the target images are images acquired by acquiring images of a sample to be tested under laser irradiation, wherein the sample to be tested is dyed with a fluorescent dye;
[0036] A parameter obtaining unit, used to obtain a target parameter corresponding to the target image, wherein the target parameter indicates whether there is a base deletion in the target image;
[0037] an image division unit, configured to divide the target image according to the target parameter to obtain a first image and a second image, wherein the first image is an image without base deletion, and the second image is an image with base deletion;
[0038] A first registration unit, configured to process the first image using an image registration algorithm to obtain a first registration result;
[0039] The second registration unit is used to process the second image according to the first registration result to obtain a second registration result.
[0040] An electronic device, comprising:
[0041] A memory, used to store a computer program and data generated by the execution of the computer program;
[0042] A processor is used to execute the computer program to achieve: obtaining multiple frames of target images, the target images are images obtained by capturing images of a sample to be tested under laser irradiation, and the sample to be tested is stained with a fluorescent dye; obtaining target parameters corresponding to the target images, the target parameters characterizing whether there is a base deletion in the target image; dividing the target image according to the target parameters to obtain a first image and a second image, the first image is an image without base deletion, and the second image is an image with base deletion; using an image registration algorithm, processing the first image to obtain a first registration result; according to the first registration result, processing the second image to obtain a second registration result.
[0043] It can be seen from the above technical scheme that in an image registration method, device and electronic device disclosed in the present application, after obtaining multiple frames of target images of the sample to be tested under laser irradiation and dyed with fluorescent dye, a target parameter characterizing whether there is a base deletion in the target image is obtained, and then the target image is divided according to the target parameter, and then the first image without base deletion is registered using a preset image registration algorithm to obtain a first registration result, and for the second image with base deletion, a second registration result can be obtained according to the first registration result. Therefore, in the present application, image registration is performed in different ways according to whether there is a base deletion in the image, and for the image with base deletion, the registration result of the image without base deletion is used to achieve registration, thereby avoiding the situation where the image cannot be registered due to base deletion and the image is very dark, thereby improving the reliability of image registration. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0045] Figure 1 A flowchart of an image registration method according to the first embodiment of the present application;
[0046] Figure 2 is a flowchart of the image registration algorithm in this application;
[0047] Figure 3 A schematic diagram of the structure of an image registration device provided in Embodiment 2 of the present application;
[0048] Figure 4 Another structural schematic diagram of an image registration device provided in Embodiment 2 of the present application;
[0049] Figure 5 A schematic diagram of the structure of an electronic device provided in Embodiment 3 of the present application;
[0050] Figure 6 This is an example diagram of statistical relative values and other data in this application;
[0051] Figure 7 A flowchart for implementing image registration in this application;
[0052] Figure 8 This is an example diagram of the normal distribution of standard points in this application;
[0053] Fig. 9 This is an example diagram of the normal distribution of offset and rotation in this application;
[0054] Fig.10 This is an example diagram for image registration in this application. DETAILED DESCRIPTION
[0055] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0056] refer to Figure 1 FIG. 1 is a flowchart of an image registration method according to an embodiment of the present application, which can be applied to electronic devices capable of image processing, such as computers or servers. The technical solution in this embodiment is mainly used to improve the reliability of image registration.
[0057] Specifically, the method in this embodiment may include the following steps:
[0058] Step 101: obtaining multiple frames of target images, where the target images are images acquired by capturing images of a sample to be tested under laser irradiation, and the sample to be tested is dyed with a fluorescent dye.
[0059] The sample to be tested is a sample that needs to be identified by base sequence. In this embodiment, the sample to be tested, which is irradiated by laser and dyed by fluorescent material, can be imaged by a base sequencing device to obtain a target image, which includes multiple image channels, such as A, C, G, T and other channels.
[0060] Step 102: Obtain target parameters corresponding to the target image, where the target parameters represent whether there is a base deletion in the target image.
[0061] In one implementation, step 102 may be implemented as follows:
[0062] First, select the target area in the target image.
[0063] Then, the signal values of the pixels in the target area are sorted from small to large;
[0064] Finally, the target parameter corresponding to the target image is obtained according to the first signal value ranked in the first quantile and the second signal value ranked in the second quantile.
[0065] The first quantile is 98% and the second quantile is 30%. For example, in this embodiment, the target parameter corresponding to the target image is obtained according to the first signal value ranked at 98% and the second signal value ranked at 30%.
[0066] Specifically, when the ratio between the first signal value ranked at the first quartile and the second signal value ranked at the second quartile is less than a preset first threshold, a target parameter characterizing the presence of base deletion in the target image is obtained in step 102;
[0067] When the ratio is greater than or equal to the first threshold, a target parameter characterizing that there is no base deletion in the target image is obtained in step 102 .
[0068] The first threshold may be set to 2. For example, when the ratio of the first signal value to the second signal value is less than 2, it indicates that the signal value is very close to the background value, and at this time, a target parameter representing the presence of base deletion in the target image is obtained.
[0069] Step 103: Divide the target image according to the target parameters to obtain a first image and a second image.
[0070] The first image is an image without base deletion, and the second image is an image with base deletion. For example, in this embodiment, according to the target parameter, the target image with the target parameter indicating the presence of base deletion is divided into the second image, and the target image with the target parameter indicating the absence of base deletion is divided into the first image.
[0071] Step 104: Process the first image using an image registration algorithm to obtain a first registration result.
[0072] The image registration algorithm in this embodiment may be an algorithm capable of realizing image registration, such as a registration algorithm based on image grayscale statistical characteristics, a registration algorithm based on image features, a registration algorithm based on image understanding, and the like.
[0073] For example, Figure 2 As shown in , in this embodiment, the first registration result can be obtained by the following method:
[0074] First, the first image is input into the image registration algorithm, and the feature template and the pixel size of the fluorescent molecular group are provided;
[0075] Then, the integral image is calculated, and the integral image features of the middle area are obtained;
[0076] Afterwards, the pixel-level position of the best feature point is corrected through local area search; and other feature positions are derived based on the best feature position and feature template; and the pixel position positioning correction of the local area of other feature positions is achieved;
[0077] Then, all points are evaluated according to the feature template as follows:
[0078] If the number of bad pixels exceeds 1 / 3, then all intersection point registration results and the evaluation value of each feature point registration result are output;
[0079] If the number of bad pixels is less than 1 / 3, first use the good neighborhood points to correct the bad pixels; then use the center method to obtain the sub-pixel position of all feature point positions; and deduce the boundary position points based on the feature template and the internal point position; then draw a grid based on the feature template and the sub-pixel position of the feature point; finally, output the sub-pixel position of all fluorescent molecular clusters and the evaluation value of the registration result of each feature point.
[0080] Step 105: Process the second image according to the first registration result to obtain a second registration result.
[0081] Specifically, step 105 can be implemented in the following manner:
[0082] Firstly, according to the relative offset value and the relative rotation value, the offset and the rotation value of the second image on the first channel are processed to obtain the offset and the rotation value of the second image on the second channel;
[0083] Then, the second image is registered according to the offset and rotation value of each image channel of the second image to obtain a second registration result.
[0084] The first channel is an image channel whose pixel brightness value meets the high brightness condition, and the second channel is an image channel whose pixel brightness value does not meet the high brightness condition. The high brightness condition can be: the number of pixels whose brightness value is greater than or equal to the brightness threshold exceeds a preset number threshold.
[0085] Specifically, the relative offset value can be obtained by:
[0086] First, a sample image is obtained. The sample image is an image obtained by capturing an image of a target sample under laser irradiation and stained with a fluorescent dye by a base sequencing device to which the target image belongs. The target sample here can be the sample to be tested as mentioned above, or other samples.
[0087] Then, the offset of the sample image on each image channel is obtained, which is recorded as offset. For example, in this embodiment, the track intersection point on each image channel, that is, the trackcross point, can be first obtained, and then the trackcross point is counted and the offset is removed to obtain the offset.
[0088] Afterwards, the offsets of the sample image on any two image channels are subtracted to obtain a plurality of first difference values, such as the offset difference between image channel A and image channel C, image channel A and image channel G, and image channel A and image channel T;
[0089] Finally, the peak value corresponding to the first difference in the normal distribution table is determined as the relative offset value. For example, the first difference is normally distributed, and the first difference at the peak value in the normal distribution table is selected as the relative offset value.
[0090] Alternatively, the rotation relative value can be obtained by:
[0091] First, a sample image is obtained, where the sample image is an image processed on a base sequencing device to which the target image belongs and does not have base deletions;
[0092] Then, the rotation value of the sample image on each image channel is obtained, which is recorded as rotation; for example, in this embodiment, the track intersection point on each image channel, that is, the trackcross point, can be first obtained, and then the rotation is counted and removed for the trackcross point to obtain the rotation;
[0093] Afterwards, the rotation values of the sample image on any two image channels are subtracted to obtain a plurality of second difference values; such as the rotation difference between image channel A and image channel C, image channel A and image channel G, and image channel A and image channel T;
[0094] Finally, the peak value corresponding to the second difference in the normal distribution table is determined as the rotation relative value. For example, the second difference is normally distributed, and the second difference at the peak value in the normal distribution table is selected as the rotation relative value.
[0095] It should be noted that the base sequencing device in this embodiment is a device in which the first difference and the second difference obtained meet the base balance condition. If the first difference and the second difference do not meet the base balance condition, then the base sequencing device cannot apply the method in this application. The base balance condition is: the first difference is less than the second threshold value, and the second difference is less than the third threshold value.
[0096] For example, when the offset difference obtained in this embodiment is less than 0.2 pixel (second threshold) and the rotation difference is less than one ten-thousandth of a radian (third threshold), the base sequencing device can implement the method of performing registration processing on the second image in this application.
[0097] It can be seen from the above technical scheme that in an image registration method provided in Example 1 of the present application, after obtaining multiple frames of target images of the sample to be tested under laser irradiation and stained with fluorescent dye, a target parameter characterizing whether there is a base deletion in the target image is obtained, and then the target image is divided according to the target parameter, and then the first image without base deletion is registered using a preset image registration algorithm to obtain a first registration result, and for the second image with base deletion, a second registration result can be obtained according to the first registration result. Therefore, in the present application, image registration is performed in different ways according to whether there is a base deletion in the image, and for the image with base deletion, the registration result of the image without base deletion is used to achieve registration, thereby avoiding the situation where the image cannot be registered due to base deletion and the image is very dim, thereby improving the reliability of image registration.
[0098] refer to Figure 3 , is a schematic diagram of the structure of an image registration device provided in the second embodiment of the present application, the device can be configured in an electronic device capable of image processing, such as a computer or a server. The technical solution in this embodiment is mainly used to improve the reliability of image registration.
[0099] Specifically, the device in this embodiment may include the following units:
[0100] An image acquisition unit 301 is used to acquire multiple frames of target images, where the target images are images acquired by capturing images of a sample to be tested under laser irradiation, where the sample to be tested is dyed with a fluorescent dye;
[0101] A parameter obtaining unit 302 is used to obtain a target parameter corresponding to the target image, wherein the target parameter indicates whether there is a base deletion in the target image;
[0102] An image division unit 303 is used to divide the target image according to the target parameter to obtain a first image and a second image, wherein the first image is an image without base deletion, and the second image is an image with base deletion;
[0103] A first registration unit 304, configured to process the first image using an image registration algorithm to obtain a first registration result;
[0104] The second registration unit 305 is configured to process the second image according to the first registration result to obtain a second registration result.
[0105] It can be seen from the above technical scheme that in an image registration device of the second embodiment of the present application, after obtaining multiple frames of target images of the sample to be tested under laser irradiation and dyed with fluorescent dye, a target parameter characterizing whether there is a base deletion in the target image is obtained, and then the target image is divided according to the target parameter, and then the first image without base deletion is registered using a preset image registration algorithm to obtain a first registration result, and for the second image with base deletion, a second registration result can be obtained according to the first registration result. Therefore, in this embodiment, image registration is performed in different ways according to whether there is a base deletion in the image, and for the image with base deletion, the registration result of the image without base deletion is used to achieve registration, thereby avoiding the situation where the image cannot be registered due to base deletion and the image is very dark, thereby improving the reliability of image registration.
[0106] In one implementation, the parameter acquisition unit 302 is specifically used to: select a target area in the target image; sort the signal values of the pixels in the target area from small to large; and obtain the target parameters corresponding to the target image based on the first signal value sorted in the first quartile and the second signal value sorted in the second quartile.
[0107] Preferably, when the parameter acquisition unit 302 obtains the target parameter corresponding to the target image based on the first signal value ranked at the first quartile and the second signal value ranked at the second quartile, it is specifically used to: obtain the target parameter characterizing the presence of base deletions in the target image when the ratio between the first signal value ranked at the first quartile and the second signal value ranked at the second quartile is less than a preset first threshold; and obtain the target parameter characterizing the absence of base deletions in the target image when the ratio is greater than or equal to the first threshold.
[0108] Among them, the first quartile is 98% and the second quartile is 30%.
[0109] In one implementation, the second registration unit 305 is specifically used to: process the offset and rotation values of the second image on the first channel according to the relative offset value and the relative rotation value to obtain the offset and rotation values of the second image on the second channel; and perform registration processing on the second image according to the offset and rotation values of the second image on each of the image channels to obtain a second registration result.
[0110] In one implementation, the device in this embodiment further includes the following units: Figure 4 As shown in:
[0111] A relative value obtaining unit 306, used to obtain an offset relative value and a rotation relative value;
[0112] Among them, when obtaining the relative value of the offset, the relative value obtaining unit 306 is specifically used to: obtain a sample image, where the sample image is an image obtained by the base sequencing device to which the target image belongs to acquire an image of a target sample under laser irradiation and stained with a fluorescent dye; obtain the offset of the sample image on each image channel; subtract the offset of the sample image on any two image channels to obtain multiple first difference values; and determine the peak value corresponding to the first difference value in the normal distribution table as the relative value of the offset.
[0113] Among them, when obtaining the rotation relative value, the relative value obtaining unit 306 is specifically used to: obtain a sample image, wherein the sample image is an image without base deletion processed on the base sequencing device to which the target image belongs; obtain the rotation value of the sample image on each image channel; subtract the rotation values of the sample image on any two image channels to obtain multiple second difference values; and determine the peak value corresponding to the second difference value in the normal distribution table as the rotation relative value.
[0114] The first difference is smaller than a second threshold, and the second difference is smaller than a third threshold.
[0115] It should be noted that the specific implementation of each unit in this embodiment can refer to the corresponding content in the previous text and will not be described in detail here.
[0116] refer to Figure 5 , is a schematic diagram of the structure of an electronic device provided in Embodiment 3 of the present application, and the electronic device may include the following structure:
[0117] Memory 501, used to store computer programs and data generated by the execution of the computer programs;
[0118] The processor 502 is used to execute the computer program to achieve: obtaining multiple frames of target images, the target images are images obtained by capturing images of a sample to be tested under laser irradiation, and the sample to be tested is stained with a fluorescent dye; obtaining target parameters corresponding to the target images, the target parameters characterizing whether there is a base deletion in the target image; dividing the target image according to the target parameters to obtain a first image and a second image, the first image is an image without base deletion, and the second image is an image with base deletion; using an image registration algorithm, processing the first image to obtain a first registration result; according to the first registration result, processing the second image to obtain a second registration result.
[0119] It can be seen from the above technical scheme that in an electronic device provided in the third embodiment of the present application, after obtaining multiple frames of target images of the sample to be tested under laser irradiation and dyed with fluorescent dye, a target parameter characterizing whether there is a base deletion in the target image is obtained, and then the target image is divided according to the target parameter, and then the first image without base deletion is registered using a preset image registration algorithm to obtain a first registration result, and for the second image with base deletion, a second registration result can be obtained according to the first registration result. Therefore, in this embodiment, image registration is performed in different ways according to whether there is a base deletion in the image, and for the image with base deletion, the registration result of the image without base deletion is used to achieve registration, thereby avoiding the situation where the image with very dark light due to base deletion cannot be registered, thereby improving the reliability of image registration.
[0120] Taking the scenario of base sequence recognition of a sample as an example, the existing image registration algorithm will register each channel of the base signal image based on fluorescent molecules, such as Figure 2 As shown, when the base content of the current channel is low, the sparse luminous points may cause the channel registration to fail. If one channel registration fails, all data of the entire scene will not be processed. The algorithm has low robustness against this situation.
[0121] In response to the above situation, this application designs a method for positioning and alignment based on stable features between channels. The method is simple and efficient, has high accuracy, good applicability and robustness, and can be compatible with the efficient acquisition of high-precision fluorescent molecule luminescence positions when fluorescent molecules are extremely sparse.
[0122] The specific steps of this application are as follows:
[0123] (1) For each machine (i.e., the base sequencing device mentioned above), use the original image registration algorithm to count the image offset and rotation differences between different image channels of the base-balanced sample. The obtained data should conform to the normal distribution. The peak value of each distribution is selected to represent the theoretical value of the image offset and rotation difference between different channels. The results are recorded. The specific process is as follows: Figure 6 As shown. For example, for the trackcross points of the standard data, the rotation is counted and removed, and then the relative values between the rotations of each image channel are obtained; then the offset is counted and removed, and then the relative values between the offsets of each image channel are obtained; finally, the standard point and its channel templates are obtained. The output results are used to measure whether this machine meets the requirements of base imbalance registration. The sigma of the offset distribution of the registration difference between the two channels is less than 0.2pixel, and the rotation difference between the two channels is less than one ten-thousandth of a radian. If the statistical results do not meet the requirements, this machine cannot be used for base imbalance experiments.
[0124] (2) Select a certain area, extract the original image signal, sort the original image signal in ascending order, select the data of the 98% percentile as the signal (i.e., the first signal value), select the data of the 30% percentile as the background (i.e., the second signal value), and use the ratio of the signal to the background to measure whether this image belongs to a sparse image of fluorescent molecules, that is, an image with base deletions.
[0125] (3) If the ratio of signal to background is less than 2, it means that the signal and background are very close, and it is judged as a base deletion; otherwise, it is an image with normal base content.
[0126] (4) If the image is judged to be base-deficient, registration is not performed first, and the registration of the image with normal base content is waited for. The image with normal base content is registered using the existing registration algorithm to obtain the registration result.
[0127] (5) Using the registration result of the normal base content plus offset and rotation, the registration result of the base-deficient image is obtained, such as Figure 7 For example, after obtaining an image for base sequencing, the data of each channel in the image is obtained, the offset and rotation of the high-brightness channel data are calculated, and the offset of the low-brightness channel is obtained by the relative value between the offsets of each channel, and the rotation of the low-brightness channel is obtained by the relative value between the rotations of each channel. For the data of the low-brightness channel, the derived trackcross is obtained by adding the template, the offset and the rotation.
[0128] It can be seen that the present application can register images with very dim light due to base deletion, quickly and accurately extract the intersection coordinates to achieve image positioning and registration, and it is not necessary to use the original registration algorithm for each image using the original image information. The result of one channel can be used to obtain the registration results of other channels through simple matrix operations. It will not increase the time consumption on the original basis, and it is fast and has strong anti-interference ability.
[0129] The following is an example of this application:
[0130] 1. Use data: V2-30 PE2+10 data;
[0131] 2. Count the trackcross points of each image channel (such as A, C, G, T) and remove the rotation and offset to get the standard point;
[0132] 3. Count the distribution of each standard point, such as Figure 8 As shown, the distribution of each trackcross point is relatively concentrated, and it can be observed that it basically conforms to the normal distribution. The median of each standard point is calculated as the template coordinate point of the trackcross point;
[0133] 4. Count the relationship between the offset and rotation of different channels (that is, confirm that the relative values of the offset and rotation between different channels are relatively fixed). It can be observed that the offset and rotation between channels conform to the normal distribution, such as Fig. 9 As shown in , the peak value is then used as a fixed difference between the channels;
[0134] 5. Determine the offset and rotation of the brighter channel (e.g. A);
[0135] 6. Obtain the offset and rotation of the C, G, and T channels through the fixed differences in offset and rotation between AC, AG, and AT;
[0136] 7. By adding the corresponding offset and rotation to the templates of the C, G, and T channels, the trackcross points of the C, G, and T channels are obtained, such as Fig.10 As shown in . For example, image A is the brightest. Based on the coordinates of image A, offsetA and rotationA are obtained by registration. Then, offsetC and rotationC are obtained by offsetA-C and rotationA-C. Finally, based on the template of image C plus offsetC and rotationC, the coordinates of image C are obtained and trackcross is derived.
[0137] In summary, this application proposes a high-precision image registration algorithm based on point features, which can achieve high accuracy in positioning and registering base information graphs of different sequencing platforms. The algorithm is real-time, efficient and has strong resistance to impurity interference. In addition, this application proposes an acceleration method based on integral image extraction of line features. It can cope with base balance and imbalance, and has good compatibility with images with large bright spots and impurities.
[0138] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.
[0139] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.
[0140] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
[0141] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. An image registration method, characterized in that: include: Acquire multiple frames of target images, wherein the target images are images acquired by capturing images of a sample to be tested under laser irradiation, wherein the sample to be tested is dyed with a fluorescent dye; Obtaining a target parameter corresponding to the target image, wherein the target parameter indicates whether there is a base deletion in the target image; According to the target parameter, the target image is divided to obtain a first image and a second image, wherein the first image is an image without base deletion, and the second image is an image with base deletion; Using an image registration algorithm, processing the first image to obtain a first registration result; The second image is processed according to the first registration result to obtain a second registration result.
2. The method according to claim 1, characterized in that Obtaining target parameters corresponding to the target image includes: Selecting a target area in the target image; Sort the signal values of the pixels in the target area from small to large; According to the first signal value ranked at the first quantile and the second signal value ranked at the second quantile, a target parameter corresponding to the target image is obtained.
3. The method according to claim 2, characterized in that Obtaining a target parameter corresponding to the target image according to a first signal value ranked at a first quartile and a second signal value ranked at a second quartile includes: When a ratio between a first signal value ranked at a first quartile and a second signal value ranked at a second quartile is less than a preset first threshold, obtaining a target parameter characterizing the presence of a base deletion in the target image; When the ratio is greater than or equal to the first threshold, a target parameter characterizing that there is no base deletion in the target image is obtained. The method according to claim 2 or 3, wherein the first quantile is 98% and the second quantile is 30%.
5. The method according to claim 1 or 2, characterized in that: Processing the second image according to the first registration result to obtain a second registration result includes: Processing the offset and rotation value of the second image on the first channel according to the relative offset value and the relative rotation value to obtain the offset and rotation value of the second image on the second channel; The second image is registered according to the offset and rotation value of the second image on each of the image channels to obtain a second registration result.
6. The method according to claim 5, characterized in that The relative offset value is obtained by: Obtaining a sample image, wherein the sample image is an image obtained by a base sequencing device to which the target image belongs, capturing an image of a target sample that is irradiated by a laser and dyed with a fluorescent dye; Obtaining the offset of the sample image on each image channel; Subtracting the offsets of the sample image on any two image channels to obtain a plurality of first difference values; A peak value corresponding to the first difference in the normal distribution table is determined as the relative offset value.
7. The method according to claim 5, characterized in that The rotation relative value is obtained by: Obtaining a sample image, where the sample image is an image processed on a base sequencing device to which the target image belongs and does not have base deletions; Obtaining a rotation value of the sample image on each image channel; Subtracting the rotation values of the sample image on any two image channels to obtain a plurality of second difference values; The peak value corresponding to the second difference in the normal distribution table is determined as the rotation relative value.
8. The method according to claim 6 or 7, characterized in that: The first difference is smaller than a second threshold, and the second difference is smaller than a third threshold.
9. An image registration device, characterized in that: include: An image acquisition unit, used to acquire multiple frames of target images, wherein the target images are images acquired by acquiring images of a sample to be tested under laser irradiation, wherein the sample to be tested is dyed with a fluorescent dye; A parameter obtaining unit, used to obtain a target parameter corresponding to the target image, wherein the target parameter indicates whether there is a base deletion in the target image; an image division unit, configured to divide the target image according to the target parameter to obtain a first image and a second image, wherein the first image is an image without base deletion, and the second image is an image with base deletion; A first registration unit, configured to process the first image using an image registration algorithm to obtain a first registration result; The second registration unit is used to process the second image according to the first registration result to obtain a second registration result.
10. An electronic device, characterized in that: include: A memory, used to store a computer program and data generated by the execution of the computer program; A processor is used to execute the computer program to achieve: obtaining multiple frames of target images, wherein the target images are images obtained by collecting images of a sample to be tested under laser irradiation, and the sample to be tested is dyed with a fluorescent dye; obtaining target parameters corresponding to the target images, wherein the target parameters represent whether there is a base deletion in the target image; According to the target parameter, the target image is divided to obtain a first image and a second image, wherein the first image is an image without base deletion, and the second image is an image with base deletion; Using an image registration algorithm, processing the first image to obtain a first registration result; The second image is processed according to the first registration result to obtain a second registration result.
Citation Information
Cited By
Multi-slice registration method and device, electronic equipment and storage medium
CN119446255A