CN-121998954-A - Identification method and system for oligoclonal gel electrophoresis image

CN121998954ACN 121998954 ACN121998954 ACN 121998954ACN-121998954-A

Abstract

The invention discloses a method and a system for identifying an oligoclonal gel electrophoresis image, wherein the method for identifying the oligoclonal gel electrophoresis image comprises the steps of obtaining an image to be identified, wherein the image to be identified is the oligoclonal gel electrophoresis image uploaded by a user based on a web browser; the method comprises the steps of extracting a lane region of an image to be identified by utilizing a multi-feature fusion identification algorithm, detecting a lane region to obtain a lane boundary, generating a lane feature parameter based on the lane boundary, parting the image to be identified according to the lane feature parameter by utilizing a preset parting rule to obtain a parting conclusion, generating a structured diagnosis report based on the parting conclusion, and displaying the diagnosis report through a web browser. The invention realizes the automation of the diagnostic report generation process by identifying and analyzing the oligoclonal gel electrophoresis image.

Inventors

Qin Xiyan
CUI YANWEI
LI WENHAN
ZHAN KE

Assignees

广州欧蒙未一医学检验实验室有限公司

Dates

Publication Date: 20260508
Application Date: 20260129

Claims (10)

1. A method for identifying an oligoclonal gel electrophoresis image, comprising: Acquiring an image to be identified, wherein the image to be identified is an oligoclonal gel electrophoresis image uploaded by a user based on a web browser; Extracting a lane region of the image to be identified by utilizing a multi-feature fusion identification algorithm, and carrying out strip detection on the lane region to obtain a strip boundary; Generating a strip characteristic parameter based on the strip boundary, and typing the image to be identified according to the strip characteristic parameter by utilizing a preset typing rule to obtain a typing conclusion; A structured diagnostic report is generated based on the typed conclusion and presented through a web browser.
2. The method for identifying an oligoclonal gel electrophoresis image of claim 1, further comprising, after the acquisition of the image to be identified: Performing format verification on the image to be identified, and generating a verification result of successful format verification under the condition that the image to be identified is in a preset format, wherein the preset format comprises PNG, JPG, JPEG and a TIFF format; Carrying out security verification on the image to be identified, and generating a verification result of successful security verification under the condition that the image to be identified passes the security verification, wherein the security verification comprises file integrity verification, session security verification and authority verification; If the verification result is that the format verification is successful and the security verification is successful, storing the image to be identified into an image database; if the verification result is format verification failure or security verification failure, a prompt message is returned to the user through the web browser, wherein the prompt message is used for prompting the user to upload the image to be identified again.
3. The method of claim 1, wherein extracting the lane region of the image to be identified using a multi-feature fusion recognition algorithm comprises: converting the image to be identified into an HSV color space to obtain an HSV image; Binarizing the HSV image based on a first hue threshold range and a second hue threshold range to obtain a mask image, wherein the mask image comprises at least one lane, the first hue threshold range is used for marking a purple area, and the second hue threshold range is used for marking a white interval; Performing morphological closing operation on the mask image to fill holes in the mask image and generating a lane mask; Calculating a vertical projection of the lane mask in a vertical direction, and locating a lane position list based on the vertical projection; labeling a cerebrospinal fluid lane and a serum lane based on the lane position list, the cerebrospinal fluid lane and the serum lane being spaced apart; the lane region is generated based on the cerebrospinal fluid lane and the serum lane.
4. The method for identifying an oligoclonal gel electrophoresis image of claim 3, wherein binarizing the HSV image based on a first hue threshold range and a second hue threshold range to obtain a mask image comprises: creating a purple region mask based on a first preset hue range, the first preset hue range being [120,40,40] to [160,255,255]; Creating a white area mask based on a second preset hue range, the second preset hue range being [0,0,200] to [180,50,255]; the mask image is generated based on the purple area mask and the white area mask.
5. The method of claim 3, wherein the detecting the lane area to obtain the lane boundary comprises: Acquiring a lane image corresponding to the lane region; graying treatment is carried out on the lane images to obtain lane gray level images, and average brightness values of the lane gray level images are calculated; Setting a dark color threshold value based on the average brightness value, and creating a composite dark color mask image corresponding to the HSV image according to the dark color threshold value, wherein the composite dark color mask image comprises a dark purple area corresponding to a third hue threshold value range and a black area corresponding to a fourth hue threshold value range; morphological enhancement processing is carried out on the composite dark mask image, and horizontal projection of the enhanced composite dark mask image is calculated to obtain projection data; A band position of the projection data is detected using a peak detection algorithm, and the band boundary is generated based on the band position.
6. The method for identifying an oligoclonal gel electrophoresis image of claim 5, wherein, the morphological enhancement processing of the composite dark mask image includes: Performing a close operation with the composite dark mask image with a horizontal check to enhance the lateral features and an open operation with the composite dark mask image with a vertical check to suppress longitudinal noise; the detecting the stripe position of the projection data using a peak detection algorithm and generating the stripe boundary based on the stripe position comprises: filtering the projection data to smooth the projection data; Identifying peak points of the smoothed projection data by using a peak detection algorithm, and determining a strip position from the peak points according to a preset peak height threshold; a band boundary is delineated in the lane image based on the band location.
7. The method of claim 5, wherein generating a band characteristic parameter based on the band boundary comprises: Marking a cerebrospinal fluid band and a serum band corresponding to the band boundary in the lane image based on the lane position list, wherein the cerebrospinal fluid band is positioned on the cerebrospinal fluid lane, and the serum band is positioned on the serum lane; Counting the number of cerebrospinal fluid bands and the number of serum bands; acquiring strip position and brightness information of the cerebrospinal fluid strip and the serum strip; the strip characteristic parameters are generated based on the cerebrospinal fluid strip number, the serum strip number, the strip position, and the brightness information.
8. The method for identifying an oligoclonal gel electrophoresis image of claim 7, wherein said typing said image to be identified according to said banding feature parameters using a preset typing rule comprises: if the cerebrospinal fluid strip number is smaller than a preset number threshold value and the serum strip number is smaller than a preset number threshold value, marking the image to be identified as a type one type, wherein the type one type represents negative, and the preset number threshold value is 2; If the number of the cerebrospinal fluid strips is greater than or equal to the preset number threshold and the number of the serum strips is 0, marking the image to be identified as a type II type, wherein the type II type characterizes the cerebrospinal fluid specificity; If the cerebrospinal fluid strip number and the serum strip number are both greater than or equal to the preset number threshold, and the cerebrospinal fluid strip number is greater than the serum strip number, marking the image to be identified as a three-type, wherein the three-type represents that the cerebrospinal fluid strip number is greater than the serum strip number; If the cerebrospinal fluid strip number and the serum strip number are both greater than or equal to the preset number threshold, the cerebrospinal fluid strip number is equal to the serum strip number, and the strip position is not at the edge of a lane, marking the image to be identified as a four-type category, wherein the four-type category characterizes that the cerebrospinal fluid strip number is equal to the serum strip number; And if the cerebrospinal fluid band number and the serum band number are both greater than or equal to the preset number threshold, the cerebrospinal fluid band number is equal to the serum band number, and the band position is positioned at the edge of a lane, marking the image to be identified as a five-type category, wherein the five-type category represents the monoclone.
9. The method of claim 1, wherein the diagnostic report comprises the image to be identified, an analysis result image and the typing conclusion, the analysis result image marking the lane area, the band border and the band characteristic parameter with different colors.
10. An identification system for an oligoclonal gel electrophoresis image, applied to the identification method for an oligoclonal gel electrophoresis image according to any of claims 1 to 9, comprising: the image acquisition module is configured to acquire an image to be identified, wherein the image to be identified is an oligoclonal gel electrophoresis image uploaded by a user based on a web browser; The strip detection module is configured to extract a lane region of the image to be identified by utilizing a multi-feature fusion recognition algorithm, and carry out strip detection on the lane region to obtain a strip boundary; The image typing module is configured to generate a strip characteristic parameter based on the strip boundary, and to type the image to be identified according to the strip characteristic parameter by utilizing a preset typing rule to obtain a typing conclusion; A report generation module configured to generate a structured diagnostic report based on the typing conclusions and to present the diagnostic report via a web browser.

Description

Identification method and system for oligoclonal gel electrophoresis image Technical Field The invention relates to the technical field of medical image processing, in particular to a method and a system for identifying an oligoclonal gel electrophoresis image. Background The oligonucleotide gel electrophoresis technology is a key detection means for clinical diagnosis of central nervous system diseases, and the core of the technology is to judge whether the central nervous system has intrathecal synthesized immunoglobulin or not by analyzing protein band distribution difference in cerebrospinal fluid (CSF) and serum samples, thereby providing an important basis for diagnosis of diseases such as multiple sclerosis and the like. Along with the improvement of clinical examination flux, higher requirements are put on the analysis efficiency, accuracy and standardization degree of the oligoclonal gel electrophoresis image. At present, the existing oligoclonal gel electrophoresis image analysis technology is realized by multi-dependent single-version software, and has the remarkable technical limitations that on one hand, the single-version software is complex in deployment and maintenance flow, cannot support multi-mechanism remote collaboration and data sharing, is complex in user interface operation and high in professional threshold, common inspection personnel can work only through professional training, on the other hand, the existing technology lacks a standardized automatic parting function, mainly relies on manpower to judge and parting lanes and strips in an electrophoresis image, so that the analysis efficiency is low, single-sample analysis takes up to 5-10 minutes, high-throughput detection requirements are difficult to meet, and the reliability of clinical diagnosis is seriously affected due to poor result consistency and insufficient accuracy caused by subjectivity of manual judgment. In addition, the existing system has single result display form, lacks visual support and structural clinical interpretation, and is not beneficial to the rapid interpretation of results and clinician reference by inspectors. Therefore, a technical scheme capable of realizing automatic, efficient and accurate identification of the oligoclonal gel electrophoresis image is needed, so as to solve the problems of dependence on manual interpretation, poor result consistency, insufficient visualization and the like in the prior art. Disclosure of Invention Aiming at the defects existing in the prior art, the invention provides a method and a system for identifying an oligoclonal gel electrophoresis image, which aim to identify and analyze the oligoclonal gel electrophoresis image and realize the automation of the diagnostic report generation flow. The invention discloses an identification method of an oligoclonal gel electrophoresis image, which comprises the following steps: acquiring an image to be identified, wherein the image to be identified is an oligoclonal gel electrophoresis image uploaded by a user based on a web browser; extracting a lane region of an image to be identified by utilizing a multi-feature fusion identification algorithm, and carrying out strip detection on the lane region to obtain a strip boundary; Generating a strip characteristic parameter based on a strip boundary, and typing an image to be identified according to the strip characteristic parameter by utilizing a preset typing rule to obtain a typing conclusion; A structured diagnostic report is generated based on the typing conclusions and presented through a web browser. Preferably, after acquiring the image to be identified, the method further comprises: Performing format verification on the image to be identified, and generating a verification result of successful format verification under the condition that the image to be identified is in a preset format, wherein the preset format comprises PNG, JPG, JPEG and a TIFF format; Carrying out security verification on the image to be identified, and generating a verification result of successful security verification under the condition that the image to be identified passes the security verification, wherein the security verification comprises file integrity verification, session security verification and authority verification; if the verification result is that the format verification is successful and the security verification is successful, storing the image to be identified into an image database; if the verification result is that the format verification fails or the security verification fails, a prompt message is returned to the user through the web browser, wherein the prompt message is used for prompting the user to upload the image to be identified again. Preferably, extracting the lane region of the image to be identified using a multi-feature fusion recognition algorithm includes: converting the image to be identified into an HSV color space to obtain an HSV image; Performing binar