WO2021189771A1

WO2021189771A1 - Slide digitization information quality testing method and apparatus, and device and medium

Info

Publication number: WO2021189771A1
Application number: PCT/CN2020/112332
Authority: WO
Inventors: 初晓; 郭冰雪; 王季勇
Original assignee: 平安科技（深圳）有限公司
Priority date: 2020-07-30
Filing date: 2020-08-30
Publication date: 2021-09-30
Also published as: CN111899246B; CN111899246A

Abstract

A slide digitization information quality testing method and apparatus, and a device and a medium, which relate to the field of artificial intelligence and are applied to the field of smart medical treatment. The method comprises: extracting a region image, which includes a preset region of interest, from a slide digitization information image, and performing image cutting on the region image to acquire a plurality of first target image blocks after cutting (S20); performing contrast enhancement processing on each of the first target image blocks according to an image processing algorithm to obtain a plurality of second target image blocks (S30); performing quality classification on all the second target image blocks according to a preset convolutional neural network to obtain a plurality of third target image blocks which are successfully classified; and performing combined testing on quality classification results of all the third target image blocks to obtain a slide digitization information quality testing result of an entire slide (S50). The method and the apparatus are used for reducing the risk of missing detection of an abnormal lesion during a screening process of an artificial-intelligence-assisted screening system.

Description

Method, device, equipment and medium for detecting quality of slide digital information

This application claims the priority of a Chinese patent application filed with the Chinese Patent Office on July 30, 2020, the application number is 202010752434.2, and the invention title is "Methods, Apparatus, Equipment, and Medium for Quality Inspection of Digital Information on Glass Slides", all of which are approved The reference is incorporated in this application.

Technical field

This application relates to the field of artificial intelligence image processing, and in particular to a method, device, equipment, and medium for detecting the quality of slide digital information.

Background technique

At present, after scanning the glass slide with the test sample, analyzing the digitized information image of the scanned glass slide to determine the data to be analyzed of the test sample has become a common method. For example, the artificial intelligence-assisted screening system for cervical cancer currently on the market uses the digitized information image of the slide scanned by the scanner and detects the digitized information image of the slide to determine whether to screen for cervical cancer. The inventor realizes that the above-mentioned solutions in the prior art cannot guarantee the quality of the slide digital information. Among them, the quality of the slide digital information refers to whether the slide is blurred, whether the surface of the slide is contaminated, and whether the slide contains a lot of bubbles, etc. . In practical applications, if the quality of the digitized information of a digitized slide is unqualified, it will be impossible to observe the effective analysis data (for example, the unqualified quality of the digitized information of the slide will lead to the artificial intelligence-assisted screening of cervical cancer. In the system screening process, it is difficult to detect abnormal lesion information). It can be seen that this solution in the prior art will increase the risk of misdetection and missed detection of the digital information of the slide. Therefore, those skilled in the art urgently need to find a new technical solution to solve the above-mentioned problems.

Summary of the invention

This application provides a method for detecting the quality of slide digital information, including:

Obtain the slide digital information image of the entire slide;

Extracting a region image containing a preset region of interest from the slide digitized information image, and performing image cutting on the region image, to obtain a plurality of first target image blocks after the cutting;

Performing contrast enhancement processing on each of the first target image blocks according to an image processing algorithm to obtain a plurality of second target image blocks with enhanced image visual effects;

Perform quality classification on all the second target image blocks according to a preset convolutional neural network to obtain a plurality of third target image blocks successfully classified; each of the third target image blocks after successful classification corresponds to a quality classification result ；

After the combined detection of the quality classification results of all the third target image blocks, the quality detection result of the slide digital information of the entire slide is obtained.

This application also provides a device for detecting the quality of the digital information of the slide glass, which includes:

The first acquisition module is used to acquire the digitized information image of the entire slide;

The second acquisition module is used to extract a region image containing a preset region of interest from the slide digitized information image, and after performing image cutting on the region image, acquire a plurality of first target image blocks after cutting ；

An enhancement processing module, configured to perform contrast enhancement processing on each of the first target image blocks according to an image processing algorithm to obtain a plurality of second target image blocks with enhanced image visual effects;

The classification module is used to classify the quality of all the second target image blocks according to a preset convolutional neural network to obtain a plurality of successfully classified third target image blocks; each of the third target image blocks successfully classified Corresponding to a quality classification result;

The detection module is configured to perform a combined detection on the quality classification results of all the third target image blocks, and obtain the slide digitized information quality detection results of the entire slide.

The present application also provides a computer device including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. The processor executes the computer program to implement the following slide Digital information quality inspection method:

Obtain the slide digital information image of the entire slide;

The present application also provides a computer-readable storage medium that stores a computer program, and when the computer program is executed by a processor, the method for detecting the quality of slide digital information as described below is realized:

Obtain the slide digital information image of the entire slide;

Description of the drawings

In order to explain the technical solutions of the embodiments of the present application more clearly, the following will briefly introduce the drawings that need to be used in the description of the embodiments of the present application. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative labor.

FIG. 1 is a schematic diagram of an application environment of a method for detecting the quality of slide digital information in an embodiment of the present application;

2 is a flowchart of a method for detecting the quality of slide digital information in an embodiment of the present application;

FIG. 3 is a schematic structural diagram of a device for detecting the quality of slide digital information in an embodiment of the present application;

Fig. 4 is a schematic diagram of a computer device in an embodiment of the present application.

Detailed ways

The technical solutions in the embodiments of the present application will be described clearly and completely in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of them. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of this application.

The method for detecting the quality of slide digital information provided in this application can be applied to the application environment as shown in Figure 1, where the client communicates with the server through the network. Among them, the client can be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices. The server can be implemented as an independent server or a server cluster composed of multiple servers.

In one embodiment, as shown in FIG. 2, a method for detecting the quality of slide digital information is provided. Taking the method applied to the server in FIG. 1 as an example, the method includes the following steps:

S10, acquiring a digital information image of the entire glass slide;

Understandably, the slide can be a slide of various pathological cells, such as a cervical cell pathology slide. Specifically, the slide digitized information image of the slide is scanned by different medical scanners, and the scanned glass A piece of digitized information image may include a variety of data formats, among which the data formats include but are not limited to svs, .kfb, .ndpi, .tif and .sdpc.

S20, extracting a region image containing a preset region of interest from the slide digitized information image, and after performing image cutting on the region image, obtain a plurality of cut first target image blocks;

Understandably, the preset area of interest refers to the area that is located in the digital information image of the slide and meets the requirements of the quality detection of the digital information of the slide; specifically, first determine the multiple of the objective lens used by the medical scanner during scanning (such as 20 Or 40 times objective lens), and then according to the multiple of the objective lens used by the medical scanner, the image magnification process is performed on the digitized information image of the slide, and then according to the preset selection rule (according to the preset target magnification required to determine the preset Resolution layer, for example, the digital information image of the glass slide is enlarged 20 times, and the target magnification is 15 times. At this time, the preset resolution layer is 15 times the digital information image of the glass slide) Select the preset resolution layer, And through the preset resolution layer to increase or decrease the resolution of the digitized information image of the slide, to realize the magnification or zoom processing of the digitized information image of the slide, and then use the Hough transform to find the cells in the digitized information image of the slide Finally, extract the image of the preset area of interest from the slide digitized information image of the determined foreground area, and use the sliding window to cut the image of the area to obtain multiple first targets after cutting Image block.

In this embodiment, the background area that does not contain effective information in the digitized information image of the slide can be efficiently removed. In the subsequent process of processing the digitized information image of the slide, the processing time for the digitized information image of the slide can be saved. In turn, the processing efficiency can be improved.

S30: Perform contrast enhancement processing on each of the first target image blocks according to the image processing algorithm to obtain multiple second target image blocks with enhanced image visual effects;

Understandably, the image processing algorithm refers to image contrast enhancement processing, where the image contrast enhancement processing stretches or compresses the brightness value range in the image into the brightness display range specified by the display system, thereby improving the overall or partial contrast of the image. This embodiment is mainly to improve the contrast of the first target image block, so as to achieve the purpose of enhancing the subjective visual effect of the image and enhancing the details of the image.

S40: Perform quality classification on all the second target image blocks according to a preset convolutional neural network to obtain a plurality of successfully classified third target image blocks; each of the third target image blocks after successful classification corresponds to a quality Classification result

Understandably, the preset convolutional neural network refers to the lightweight deep learning convolutional neural network MobileNetV3, where the MobileNetv3 version combines the deep separable convolution of MobileNetv1, the InvertedResiduals and LinearBottleneck of MobileNetv2, the SE module, and the use of NAS ( Neural structure search) to search for the configuration and parameters of the network; quality classification results include, but are not limited to, normal slides, blurred slides, contaminated slides, and bubbles on the slides. This embodiment uses MobileNetV3 to classify the quality of the second target image block. It mainly uses MobileNetV3 to predict the probability that the second target image block belongs to one of the quality classification results, and this embodiment uses MobileNetV3 to provide less The amount of calculations obtains a higher-precision quality classification result, which can achieve a better balance between real-time performance and accuracy.

S50: After the combined detection of the quality classification results of all the third target image blocks, the quality detection result of the slide digitized information of the entire slide is obtained.

Understandably, in this embodiment, various quality classification results need to be combined and tested to obtain the slide digitized information quality detection results of the slide. Among them, the proportion of the various quality classification results can be directly determined for the entire slide. If the proportion of the digital information quality of the slide is less than the preset proportion threshold, it can be determined that the quality of the digital information of the entire slide is normal. Otherwise, the digital information of the entire slide can be determined to be digitized. The information quality test result is abnormal.

Further, the extracting a region image containing a preset region of interest from the slide digitized information image, and performing image cutting on the region image includes:

Adjusting the resolution of the digitized information image of the slide glass through the selected preset resolution layer to obtain the adjusted digitized information image of the slide glass;

The Hough transform algorithm is used to determine the preset region of interest from the adjusted digitized information image of the slide, and the threshold image segmentation algorithm is used to extract the pre-contained information from the adjusted digitized information image of the slide. Set the region image of the region of interest;

Image cutting is performed on the region image containing the preset region of interest by using a wiped window of a preset size.

Understandably, the Hough transform algorithm is mainly used to separate geometric shapes (such as straight lines, circles, etc.) with the same characteristics from the adjusted digital information image of the slide. Specifically, the Hough transform algorithm can be used in this embodiment. The transformation algorithm is used to find the corresponding straight line in the opposite side of the parameter plane for each front spot on the adjusted slide digitized information image, and count the number of occurrences of all points on the straight line, and finally it is composed of the most frequent points Point positions are used to determine the preset area of interest (the foreground area where the cells in the slide are located); the threshold image segmentation algorithm is a region-based image segmentation technology, the principle is to divide the image pixel points into several categories, specifically, In this embodiment, the adjusted glass slide digital information image can be segmented by the global threshold method in the threshold segmentation image algorithm. The threshold method to segment the image mainly uses the maximum between-class variance method to divide the adjusted slide The histogram of the digitized information image is divided into two groups at a certain threshold. When the variance of the divided two groups is the largest, the threshold is obtained, and then the adjusted glass slide digitized information image pixels are divided into several categories through this threshold In order to obtain an area image containing a preset area of interest (the area image represents one type of pixels); the sliding window of the preset size can be a sliding window with a resolution of 5120*5120. In this embodiment, the sliding window can be designed To traverse the regional image and detect the regional image corresponding to the sliding window window can effectively overcome the problem of input heterogeneity caused by phenomena such as scale, position, and deformation, thereby improving the detection effect of the regional image.

Further, before determining the preset region of interest from the adjusted digitized information image of the slide glass by using the Hough transform algorithm, the method further includes:

When it is determined that the digitized information image of the slide glass is a black and white image, the image expansion and image erosion operations are sequentially performed on the adjusted digitized information image of the slide glass in sequence.

Understandably, image expansion and image erosion are two basic morphological operations, which are mainly used to find the maximum area and the minimum area in the image. The image expansion expands the highlighted area or white part of the image, and finally The running result image of is larger than the highlight area of the original image, and the image erosion reduces the highlight area or white part in the image, and the final running result image is smaller than the highlight area of the original image. This embodiment mainly uses two image algorithms to process the highlighted area in the black-and-white image of the slide digitization information image, so that the highlighted area is used as the foreground area where the cells in the slide are located, so as to facilitate the removal of the slide digitization. The background area of the information image.

Further, the performing contrast enhancement processing on each of the first target image blocks according to an image processing algorithm includes:

Acquiring channel data of each of the first target image blocks corresponding to three channels of R channel, G channel, and B channel;

Determine the maximum channel data and minimum channel data of each channel from all channel data;

Select one channel from the three channels in each of the first target image blocks as the current channel according to a preset selection rule, and compare the difference between the channel data of the current channel and the minimum channel data of the current channel After that, obtain the first difference result;

Under the current channel, under the current channel, after performing a difference comparison between the maximum channel data and the minimum channel data, obtaining a second difference result;

Determine the ratio between the first travel value result and the second difference result, and record the product of the ratio and a preset constant as the comparison value of the current channel;

After the three contrast values corresponding to the three channels of each of the first target image blocks are all recorded, perform contrast enhancement processing on the first target image block corresponding to the three contrast values .

Understandably, the preset selection rule can be selected in the order of R channel, G channel, and B channel, where the current channel is one of the R channel, G channel, and B channel, and the channel data refers to the first target image block In the gray value of the current channel [0-255], a first target image block corresponds to three channels, and one channel corresponds to one channel data, so a first target image block corresponds to three channel data; this implementation Example all steps can be converted into a formula, the formula is

Among them, V _in is the channel data of the current channel; V _min is the minimum channel data _{of the current channel; V max} is the maximum channel data of the current channel; V _out is the first ratio; the preset constant is 255; After the comparison value (using the maximum channel data and the minimum channel data to carry out the probability distribution of the channel data of each first target image block), the three comparison values of each first target image block are multiplied by the preset constant During calculation, the channel data of the R channel, G channel and B channel can be distributed more openly in the 0-255 gray value, which improves the contrast of the image, thereby improving the visual effect of the image, and can effectively reduce The style of the digitized information image of the digitized slides of different medical scanners has an impact on the image quality detection, thereby improving the accuracy of the preset convolutional neural network in identifying the second target image block.

Further, the quality classification of all the second target image blocks according to a preset convolutional neural network to obtain a plurality of third target image blocks successfully classified includes:

Input all the second target image blocks to the preset convolutional neural network;

Predictive classification is performed on each pixel in all the second target image blocks according to the weight coefficients set in the preset convolutional neural network to obtain a plurality of the third target image blocks successfully classified.

Specifically, first input the second target image block into the pre-trained convolutional neural network MobileNetv3, and then use the weight coefficient corresponding to each quality classification result in MobileNetv3 and each of the second target image blocks The pixels in the surrounding environment of the pixel detect each pixel in all the second target image blocks, and extract the image features composed of multiple pixels through the detection results, and then predict the second target image block based on the image features Classification, the probability that the second target image block belongs to a quality classification result is obtained, and finally the quality classification result of the third target image block is determined by the probability.

Further, the quality detection result of the digital information of the slide glass is stored in the blockchain, and after the combined detection of the quality classification results of all the third target image blocks, the slide digitization of the entire glass slide is obtained. Information quality test results, including:

Determine the abnormal quality detection result from all the quality detection results, and determine the proportion of the third target image block corresponding to the abnormal quality detection result in all the third target image blocks in the entire glass slide Ratio

When it is determined that the proportion reaches the preset proportion threshold, it is confirmed that the digital information image of the entire glass slide has a quality problem, and the quality inspection result corresponding to the entire glass slide is marked by the preset graphic structure The abnormal result of the quality inspection is sent and reminded to the preset recipient.

In this embodiment, the relationship between the proportion number and the preset proportion threshold is used to determine whether the digital information image of the entire glass slide has quality problems. In the case that the proportion number is too large, the entire glass slide can be determined The digitized information image of the slide glass has a quality problem. At this time, through this embodiment, the digitized information image of the slide glass that has a quality problem can be fed back in time, thereby speeding up the processing efficiency.

In addition, it should be emphasized that, in order to further ensure the privacy and security of the digital information quality inspection results of the slides, the digital information quality inspection results of the slides can also be stored in a blockchain node. Among them, the blockchain referred to in this application is a new application mode of computer technology such as distributed data storage, point-to-point transmission, consensus mechanism, and encryption algorithm. Blockchain, essentially a decentralized database, is a series of data blocks associated with cryptographic methods. Each data block contains a batch of network transaction information for verification. The validity of the information (anti-counterfeiting) and the generation of the next block. The blockchain can include the underlying platform of the blockchain, the platform product service layer, and the application service layer. The decentralized and fully distributed DNS service provided by the blockchain can realize the query and resolution of domain names through the point-to-point data transmission service between various nodes in the network, which can be used to ensure that the operating system and firmware of an important infrastructure are not available. If it is tampered with, it can monitor the status and integrity of the software, find bad tampering, and ensure that the transmitted data has not been tampered with, and store the digital information quality inspection results of the slides in the blockchain to ensure the slides The privacy and security of the digital information quality test results of the slides.

In summary, the above provides a method for detecting the quality of the digital information of the glass slide. The digital information image of the glass slide is processed in advance through image cutting and contrast enhancement processing, so that the image contrast is enhanced, the image visual effect is improved, and the image visual effect can be reduced. Different medical scanners have differences in the digitized information image style of the digitized slides; finally, the processed digitized slide digitized information images are input into the neural network MobileNetV3 for quality classification, and the accuracy and speed are considered comprehensively, that is, it can be effectively Improve the accuracy of quality classification and improve the efficiency of quality classification; it can be seen that this method can pre-classify the quality of slide digital information. For the screening process of the artificial intelligence-assisted screening system, the quality of the slide digital information of the slide is excluded. The impact of being qualified on the detection of abnormal lesions reduces the risk of misdetection and missed detection of abnormal lesions. This method can be applied in the field of smart medical care to promote the construction of smart cities.

It should be understood that the size of the sequence number of each step in the foregoing embodiment does not mean the order of execution. The execution sequence of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present application.

In one embodiment, a device for detecting the quality of digital information on a slide is provided, and the device for detecting the quality of digital information on a slide corresponds one-to-one with the method for detecting the quality of digital information on a slide in the foregoing embodiment. As shown in FIG. 3, the device for detecting the quality of slide digital information includes a first acquisition module 11, a second acquisition module 12, an enhancement processing module 13, a classification module 14, and a detection module 15. The detailed description of each functional module is as follows:

The first acquisition module 11 is used to acquire the digitized information image of the entire glass slide;

The second acquisition module 12 is configured to extract a region image containing a preset region of interest from the slide digitized information image, and after performing image cutting on the region image, acquire multiple first target images after cutting Piece;

The enhancement processing module 13 is configured to perform contrast enhancement processing on each of the first target image blocks according to an image processing algorithm to obtain a plurality of second target image blocks with enhanced image visual effects;

The classification module 14 is configured to perform quality classification on all the second target image blocks according to a preset convolutional neural network to obtain a plurality of successfully classified third target image blocks; each of the third target images after successful classification The block corresponds to a quality classification result;

The detection module 15 is configured to perform a combined detection on the quality classification results of all the third target image blocks to obtain a slide digitized information quality detection result of the entire slide.

Further, the second acquisition module includes:

The adjustment sub-module includes adjusting the resolution of the digitized information image of the slide glass through the selected preset resolution layer to obtain the digitized information image of the slide glass after adjustment;

The extraction sub-module includes using the Hough transform algorithm to determine the preset region of interest from the adjusted digitized information image of the slide, and using a threshold image segmentation algorithm to extract the adjusted digitized information image of the slide Output the region image containing the preset region of interest;

The cutting sub-module is used to perform image cutting on the region image containing the predetermined region of interest by using a scribe window of a preset size.

Further, the second acquisition module includes:

The image expansion and image erosion sub-module is used to sequentially perform image expansion and image erosion operations on the adjusted slide digitized information image when it is determined that the slide digitized information image is a black and white image.

Further, the enhanced processing module includes:

The first acquisition sub-module is configured to acquire the channel data of each of the first target image blocks corresponding to the R channel, the G channel and the B channel respectively;

The first determining sub-module is used to determine the maximum channel data and the minimum channel data of each channel from all the channel data;

The second acquisition sub-module is used to select one channel from the three channels in each of the first target image blocks as the current channel according to a preset selection rule, and compare the channel data of the current channel and the data of the current channel. After the difference value comparison is performed on the minimum channel data, the first difference value result is obtained;

The third obtaining submodule is configured to obtain a second difference result after performing a difference comparison between the maximum channel data and the minimum channel data under the current channel;

A recording sub-module, configured to determine the ratio between the first travel value result and the second difference result, and record the product of the ratio and a preset constant as the comparison value of the current channel;

The enhancement processing sub-module is configured to, after the three contrast values corresponding to the three channels of each of the first target image blocks are all recorded, pass the three contrast value pairs to the corresponding first The target image block undergoes contrast enhancement processing.

Further, the classification module includes:

An input sub-module for inputting all the second target image blocks to the preset convolutional neural network;

The classification sub-module is used to predict and classify each pixel in all the second target image blocks according to the weight coefficients set in the preset convolutional neural network to obtain a plurality of successfully classified third targets Image block.

Further, the detection module includes:

The second determination sub-module is used to determine the abnormal quality detection result from all the quality detection results, and determine all the third target image blocks in the entire glass slide corresponding to the abnormal quality detection result The proportion of the third target image block;

The confirmation sub-module is used to confirm that there is a quality problem in the digitalized information image of the entire glass slide when it is determined that the accounted ratio reaches the preset accounted ratio threshold. The preset graphic structure is used to correspond to the entire glass slide. Mark the abnormal quality inspection result in the quality inspection result, and send and prompt the preset recipient.

For the specific limitation of the device for detecting the quality of the digital information of the slide, please refer to the above limitation on the method of detecting the quality of the digitized information of the slide, which will not be repeated here. Each module in the above-mentioned glass slide digital information quality inspection device can be implemented in whole or in part by software, hardware, and a combination thereof. The above-mentioned modules may be embedded in the form of hardware or independent of the processor in the computer equipment, or may be stored in the memory of the computer equipment in the form of software, so that the processor can call and execute the operations corresponding to the above-mentioned modules.

In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. 4. The computer equipment includes a processor, a memory, a network interface, and a database connected through a system bus. Among them, the processor of the computer device is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer equipment is used to store the data involved in the method for detecting the quality of the digital information of the slide. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for detecting the quality of the slide digital information is realized.

In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the computer program, the quality of the slide digital information in the above-mentioned embodiment is realized. The steps of the detection method are, for example, step S10 to step S50 shown in FIG. 2. Or, when the processor executes the computer program, the function of each module/unit of the apparatus for detecting the quality of slide digital information in the foregoing embodiment is realized, for example, the functions of the modules 11 to 15 shown in FIG. 3. To avoid repetition, I won’t repeat them here.

In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method for detecting the quality of the slide digital information in the above-mentioned embodiment are implemented, as shown in FIG. 2的Steps S10 to S50. Or, when the computer program is executed by the processor, the functions of the modules/units of the apparatus for detecting the quality of slide digital information in the foregoing embodiment are realized, for example, the functions of the modules 11 to 15 shown in FIG. 3. To avoid repetition, I won’t repeat them here.

A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage. In the medium, when the computer program is executed, it may include the processes of the above-mentioned method embodiments. Wherein, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and/or volatile memory. Non-volatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not a limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Channel (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

The blockchain referred to in this application is a new application mode of computer technology such as distributed data storage, point-to-point transmission, consensus mechanism, and encryption algorithm. Blockchain, essentially a decentralized database, is a series of data blocks associated with cryptographic methods. Each data block contains a batch of network transaction information for verification. The validity of the information (anti-counterfeiting) and the generation of the next block. The blockchain can include the underlying platform of the blockchain, the platform product service layer, and the application service layer.

Those skilled in the art can clearly understand that, for the convenience and conciseness of description, only the division of the above functional units and modules is used as an example. In practical applications, the above functions can be allocated to different functional units and modules as needed. Module completion, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above.

The above-mentioned embodiments are only used to illustrate the technical solutions of the present application, not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, a person of ordinary skill in the art should understand that it can still implement the foregoing The technical solutions recorded in the examples are modified, or some of the technical features are equivalently replaced; these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the application, and should be included in Within the scope of protection of this application.

Claims

A method for detecting the quality of digital information on slides, which includes:

Obtain the slide digital information image of the entire slide;

Extracting a region image containing a preset region of interest from the slide digitized information image, and performing image cutting on the region image, to obtain a plurality of first target image blocks after the cutting;

Performing contrast enhancement processing on each of the first target image blocks according to an image processing algorithm to obtain a plurality of second target image blocks with enhanced image visual effects;

Perform quality classification on all the second target image blocks according to a preset convolutional neural network to obtain a plurality of third target image blocks successfully classified; each of the third target image blocks after successful classification corresponds to a quality classification result ；

After the combined detection of the quality classification results of all the third target image blocks, the quality detection result of the slide digital information of the entire slide is obtained.
The method for detecting the quality of digital information of a slide glass according to claim 1, wherein extracting a region image containing a preset region of interest from the digital information image of the slide glass and performing image cutting on the region image comprises:

Adjusting the resolution of the digitized information image of the slide glass through the selected preset resolution layer to obtain the adjusted digitized information image of the slide glass;

The Hough transform algorithm is used to determine the preset region of interest from the adjusted digitized information image of the slide, and the threshold image segmentation algorithm is used to extract the pre-contained information from the adjusted digitized information image of the slide. Set the region image of the region of interest;

Image cutting is performed on the region image containing the preset region of interest by using a wiped window of a preset size.
The method for detecting the quality of the slide digital information according to claim 2, wherein before the determining the preset region of interest from the adjusted digital information image of the slide using the Hough transform algorithm, the method further comprises:

When it is determined that the digitized information image of the slide glass is a black and white image, the image expansion and image erosion operations are sequentially performed on the adjusted digitized information image of the slide glass in sequence.
The method for detecting the quality of digital information of a slide glass according to claim 1, wherein said performing contrast enhancement processing on each of said first target image blocks according to an image processing algorithm comprises:

Acquiring channel data of each of the first target image blocks corresponding to three channels of R channel, G channel, and B channel;

Determine the maximum channel data and minimum channel data of each channel from all channel data;

Select one channel from the three channels in each of the first target image blocks as the current channel according to a preset selection rule, and compare the difference between the channel data of the current channel and the minimum channel data of the current channel After that, obtain the first difference result;

Under the current channel, after performing a difference comparison between the maximum channel data and the minimum channel data, obtain a second difference result;

Determine the ratio between the first travel value result and the second difference result, and record the product of the ratio and a preset constant as the comparison value of the current channel;

After the three contrast values corresponding to the three channels of each of the first target image blocks are all recorded, perform contrast enhancement processing on the first target image block corresponding to the three contrast values .
The method for detecting the quality of slide digital information according to claim 1, wherein the quality classification is performed on all the second target image blocks according to a preset convolutional neural network to obtain a plurality of successfully classified third target image blocks ,include:

Input all the second target image blocks to the preset convolutional neural network;

Predictive classification is performed on each pixel in all the second target image blocks according to the weight coefficients set in the preset convolutional neural network to obtain a plurality of the third target image blocks successfully classified.
The method for detecting the quality of digital information of the slides according to claim 1, wherein the results of the quality detection of the digital information of the slides are stored in a blockchain;

After the combined detection of the quality classification results of all the third target image blocks, the digital information quality detection result of the entire glass slide is obtained, including:

Determine the abnormal quality detection result from all the quality detection results, and determine the proportion of the third target image block corresponding to the abnormal quality detection result in all the third target image blocks in the entire glass slide Ratio

When it is determined that the proportion reaches the preset proportion threshold, it is confirmed that the digital information image of the entire glass slide has a quality problem, and the quality inspection result corresponding to the entire glass slide is marked by the preset graphic structure The abnormal result of the quality inspection is sent and reminded to the preset recipient.
A glass slide digital information quality detection device, which includes:

The first acquisition module is used to acquire the digitized information image of the entire slide;

The second acquisition module is used to extract a region image containing a preset region of interest from the slide digitized information image, and after performing image cutting on the region image, acquire a plurality of first target image blocks after cutting ；

An enhancement processing module, configured to perform contrast enhancement processing on each of the first target image blocks according to an image processing algorithm to obtain a plurality of second target image blocks with enhanced image visual effects;

The classification module is used to classify the quality of all the second target image blocks according to a preset convolutional neural network to obtain a plurality of successfully classified third target image blocks; each of the third target image blocks successfully classified Corresponding to a quality classification result;

The detection module is configured to perform combined detection on the quality classification results of all the third target image blocks to obtain the quality detection result of the slide digital information of the entire slide.
8. The device for detecting quality of slide digital information according to claim 7, wherein the second acquisition module comprises:

The adjustment sub-module includes adjusting the resolution of the digitized information image of the slide glass through the selected preset resolution layer to obtain the digitized information image of the slide glass after adjustment;

The extraction sub-module includes using the Hough transform algorithm to determine the preset region of interest from the adjusted digitized information image of the slide, and using a threshold image segmentation algorithm to extract the adjusted digitized information image of the slide Output the region image containing the preset region of interest;

The cutting sub-module is used to perform image cutting on the region image containing the predetermined region of interest by using a scribe window of a preset size.
A computer device including a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor implements the following slide digital information when the computer program is executed Quality inspection method:

Obtain the slide digital information image of the entire slide;

Extracting a region image containing a preset region of interest from the slide digitized information image, and performing image cutting on the region image, to obtain a plurality of first target image blocks after the cutting;

Performing contrast enhancement processing on each of the first target image blocks according to an image processing algorithm to obtain a plurality of second target image blocks with enhanced image visual effects;

Perform quality classification on all the second target image blocks according to a preset convolutional neural network to obtain a plurality of third target image blocks successfully classified; each of the third target image blocks after successful classification corresponds to a quality classification result ；

After the combined detection of the quality classification results of all the third target image blocks, the quality detection result of the slide digital information of the entire slide is obtained.
9. The computer device according to claim 9, wherein extracting a region image containing a preset region of interest from the slide digitized information image, and performing image cutting on the region image, comprises:

Adjusting the resolution of the digitized information image of the slide glass through the selected preset resolution layer to obtain the adjusted digitized information image of the slide glass;

The Hough transform algorithm is used to determine the preset region of interest from the adjusted digitized information image of the slide, and the threshold image segmentation algorithm is used to extract the pre-contained information from the adjusted digitized information image of the slide. Set the region image of the region of interest;

Image cutting is performed on the region image containing the preset region of interest by using a wiped window of a preset size.
10. The computer device according to claim 10, wherein before the determining the preset region of interest from the adjusted digitized information image of the slide using the Hough transform algorithm, the method further comprises:

When it is determined that the digitized information image of the slide glass is a black and white image, the image expansion and image erosion operations are sequentially performed on the adjusted digitized information image of the slide glass in sequence.
9. The computer device according to claim 9, wherein said performing contrast enhancement processing on each of said first target image blocks according to an image processing algorithm comprises:

Acquiring channel data of each of the first target image blocks corresponding to three channels of R channel, G channel, and B channel;

Determine the maximum channel data and minimum channel data of each channel from all channel data;

Select one channel from the three channels in each of the first target image blocks as the current channel according to a preset selection rule, and compare the difference between the channel data of the current channel and the minimum channel data of the current channel After that, obtain the first difference result;

Under the current channel, after performing a difference comparison between the maximum channel data and the minimum channel data, obtain a second difference result;

Determine the ratio between the first travel value result and the second difference result, and record the product of the ratio and a preset constant as the comparison value of the current channel;

After the three contrast values corresponding to the three channels of each of the first target image blocks are all recorded, perform contrast enhancement processing on the first target image block corresponding to the three contrast values .
9. The computer device according to claim 9, wherein the quality classification of all the second target image blocks according to a preset convolutional neural network to obtain a plurality of third target image blocks successfully classified comprises:

Input all the second target image blocks to the preset convolutional neural network;

Predictive classification is performed on each pixel in all the second target image blocks according to the weight coefficients set in the preset convolutional neural network to obtain a plurality of the third target image blocks successfully classified.
9. The computer device according to claim 9, wherein the digital information quality inspection result of the slide is stored in a blockchain;

After the combined detection of the quality classification results of all the third target image blocks, the digital information quality detection result of the entire glass slide is obtained, including:

Determine the abnormal quality detection result from all the quality detection results, and determine the proportion of the third target image block corresponding to the abnormal quality detection result in all the third target image blocks in the entire glass slide Ratio

When it is determined that the proportion reaches the preset proportion threshold, it is confirmed that the digital information image of the entire glass slide has a quality problem, and the quality inspection result corresponding to the entire glass slide is marked by the preset graphic structure The abnormal result of the quality inspection is sent and reminded to the preset recipient.
A computer-readable storage medium, the computer-readable storage medium stores a computer program, wherein the computer program is executed by a processor to implement the following method for detecting the quality of slide digital information:

Obtain the slide digital information image of the entire slide;

Extracting a region image containing a preset region of interest from the slide digitized information image, and performing image cutting on the region image, to obtain a plurality of first target image blocks after the cutting;

Performing contrast enhancement processing on each of the first target image blocks according to an image processing algorithm to obtain a plurality of second target image blocks with enhanced image visual effects;

Perform quality classification on all the second target image blocks according to a preset convolutional neural network to obtain a plurality of third target image blocks successfully classified; each of the third target image blocks after successful classification corresponds to a quality classification result ；

After the combined detection of the quality classification results of all the third target image blocks, the quality detection result of the slide digital information of the entire slide is obtained.
15. The computer-readable storage medium according to claim 15, wherein extracting a region image containing a preset region of interest from the slide digitized information image, and performing image cutting on the region image, comprises:

Adjusting the resolution of the digitized information image of the slide glass through the selected preset resolution layer to obtain the adjusted digitized information image of the slide glass;

The Hough transform algorithm is used to determine the preset region of interest from the adjusted digitized information image of the slide, and the threshold image segmentation algorithm is used to extract the pre-contained information from the adjusted digitized information image of the slide. Set the region image of the region of interest;

Image cutting is performed on the region image containing the preset region of interest by using a wiped window of a preset size.
15. The computer-readable storage medium according to claim 16, wherein before the determining the preset region of interest from the adjusted digitized information image of the slide using the Hough transform algorithm, the method further comprises:

When it is determined that the digitized information image of the slide glass is a black and white image, the image expansion and image erosion operations are sequentially performed on the adjusted digitized information image of the slide glass in sequence.
15. The computer-readable storage medium according to claim 15, wherein the performing contrast enhancement processing on each of the first target image blocks according to an image processing algorithm comprises:

Acquiring channel data of each of the first target image blocks corresponding to three channels of R channel, G channel, and B channel;

Determine the maximum channel data and minimum channel data of each channel from all channel data;

Select one channel from the three channels in each of the first target image blocks as the current channel according to a preset selection rule, and compare the difference between the channel data of the current channel and the minimum channel data of the current channel After that, obtain the first difference result;

Under the current channel, after performing a difference comparison between the maximum channel data and the minimum channel data, obtain a second difference result;

Determine the ratio between the first travel value result and the second difference result, and record the product of the ratio and a preset constant as the comparison value of the current channel;

After the three contrast values corresponding to the three channels of each of the first target image blocks are all recorded, perform contrast enhancement processing on the first target image block corresponding to the three contrast values .
The computer-readable storage medium according to claim 15, wherein the quality classification of all the second target image blocks according to a preset convolutional neural network to obtain a plurality of third target image blocks successfully classified includes :

Input all the second target image blocks to the preset convolutional neural network;

Predictive classification is performed on each pixel in all the second target image blocks according to the weight coefficients set in the preset convolutional neural network to obtain a plurality of the third target image blocks successfully classified.
15. The computer-readable storage medium according to claim 15, wherein the result of the quality detection of the digital information of the slide is stored in a blockchain;

After the combined detection of the quality classification results of all the third target image blocks, the digital information quality detection result of the entire glass slide is obtained, including:

Determine the abnormal quality detection result from all the quality detection results, and determine the proportion of the third target image block corresponding to the abnormal quality detection result in all the third target image blocks in the entire glass slide Ratio

When it is determined that the proportion reaches the preset proportion threshold, it is confirmed that the digital information image of the entire glass slide has a quality problem, and the quality inspection result corresponding to the entire glass slide is marked by the preset graphic structure The abnormal result of the quality inspection is sent and reminded to the preset recipient.