WO2020215701A1 - 眼部图像处理方法、装置、设备及计算机可读存储介质 - Google Patents

眼部图像处理方法、装置、设备及计算机可读存储介质 Download PDF

Info

Publication number
WO2020215701A1
WO2020215701A1 PCT/CN2019/118548 CN2019118548W WO2020215701A1 WO 2020215701 A1 WO2020215701 A1 WO 2020215701A1 CN 2019118548 W CN2019118548 W CN 2019118548W WO 2020215701 A1 WO2020215701 A1 WO 2020215701A1
Authority
WO
WIPO (PCT)
Prior art keywords
eye image
instruction
type
abnormal
detection result
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2019/118548
Other languages
English (en)
French (fr)
Inventor
舒慧
杨光
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ping An Technology Shenzhen Co Ltd
Original Assignee
Ping An Technology Shenzhen Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ping An Technology Shenzhen Co Ltd filed Critical Ping An Technology Shenzhen Co Ltd
Priority to SG11202008484YA priority Critical patent/SG11202008484YA/en
Publication of WO2020215701A1 publication Critical patent/WO2020215701A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H15/00—ICT specially adapted for medical reports, e.g. generation or transmission thereof
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30041—Eye; Retina; Ophthalmic

Definitions

  • This application mainly relates to the field of image processing technology, and specifically to an eye image processing method, device, equipment, and computer-readable storage medium.
  • the eye image is usually obtained through fundus color photography, and then the optic disc area and macular area in the eye image are artificially determined by the medical staff, and then whether the macular area is diseased or abnormal Information etc.
  • This method of purely judging the abnormal information in the eye image by the medical staff will make the workload of the medical staff very heavy; especially for the medical staff who lack work experience, it will take more time, resulting in low work efficiency of the medical staff .
  • the main purpose of this application is to provide an eye image processing method, device, equipment, and computer readable storage medium, aiming to solve the problem that in the prior art, only medical personnel judge abnormal information in eye images, which leads to medical affairs.
  • the eye image processing method includes the following steps:
  • a detection instruction for an eye image When a detection instruction for an eye image is detected, a preset algorithm is called to detect the eye image, and a detection result is generated, where the detection result includes the eye type, macula position, and optic disc in the eye image Location and abnormal location;
  • the detection result, the risk-related factor, and the abnormal change probability are used to generate a reference detection report of the eye image, and the processing of the eye image is completed.
  • this application also proposes an eye image processing device, the eye image processing device comprising:
  • the detection module is used to call a preset algorithm to detect the eye image and generate a detection result when the detection instruction to the eye image is detected, wherein the detection result includes the eye type in the eye image , Macular position, optic disc position and abnormal position;
  • the prediction module is used to determine whether there is a risk-related factor in the detection result when a prediction instruction for an abnormal change trend is detected, and if the risk-related factor exists, according to the preset number and probability of the risk-related factor Correspondence between, predict the probability of abnormal change corresponding to the risk-related factor;
  • the generating module is used to generate a reference detection report of the eye image from the detection result, the risk-related factor, and the abnormal change probability when the trigger instruction to generate the detection report is detected, and complete the detection of the eye Image processing.
  • this application also proposes an eye image processing device, the eye image processing device comprising: a memory, a processor, a communication bus, and eye image processing readable instructions stored on the memory ;
  • the communication bus is used to realize connection and communication between the processor and the memory
  • the processor is configured to execute the eye image processing readable instructions to implement the following steps:
  • a detection instruction for an eye image When a detection instruction for an eye image is detected, a preset algorithm is called to detect the eye image, and a detection result is generated, where the detection result includes the eye type, macula position, and optic disc in the eye image Location and abnormal location;
  • the detection result, the risk-related factor, and the abnormal change probability are used to generate a reference detection report of the eye image, and the processing of the eye image is completed.
  • the present application also provides a computer-readable storage medium that stores one or more readable instructions, and the one or more readable instructions can be The above processor is executed for:
  • a detection instruction for an eye image When a detection instruction for an eye image is detected, a preset algorithm is called to detect the eye image, and a detection result is generated, where the detection result includes the eye type, macula position, and optic disc in the eye image Location and abnormal location;
  • the detection result, the risk-related factor, and the abnormal change probability are used to generate a reference detection report of the eye image, and the processing of the eye image is completed.
  • the eye image processing method of this embodiment obtains various detection results including eye type, macula position, optic disc position, and abnormal position in the eye image by performing image processing on the eye image; at the same time, according to the risk in the detection result Correlation factors, predict the abnormal change probability of abnormal change trends; and then use the detection results, risk-related factors and abnormal change probability to generate a reference test report for medical staff as a reference; it is convenient for medical staff to compare the eye image in the reference test report The abnormal information is quickly judged, avoiding the artificial judgment of the abnormal information in the eye image by the medical staff, which improves the work efficiency of the medical staff.
  • FIG. 1 is a schematic flowchart of a first embodiment of an eye image processing method of the present application
  • FIG. 2 is a schematic diagram of functional modules of the first embodiment of the eye image processing device of the present application.
  • FIG. 3 is a schematic diagram of the device structure of the hardware operating environment involved in the method of the embodiment of the present application.
  • This application provides an eye image processing method.
  • FIG. 1 is a schematic flowchart of a first embodiment of an eye image processing method according to the present application.
  • the eye image processing method includes:
  • Step S10 when a detection instruction for the eye image is detected, call a preset algorithm to detect the eye image, and generate a detection result, wherein the detection result includes the eye type and the macula in the eye image Location, optic disc location and abnormal location;
  • the eye image processing method of this application is suitable for processing eye images through a server to assist medical staff in judging abnormal information in the eye images.
  • the server is in communication connection with a preset terminal with a display device.
  • the preset terminal can be a tablet computer, desktop computer, smart phone, etc.; the display interface of the display device is used as the preset terminal interface, and the tool is set in the preset terminal interface Window, main image display area, thumbnail display area and function area.
  • tools in the tool window for editing eye images the main image display area is used to display the currently processed eye images, and the thumbnails are used for other eye images waiting to be processed Perform thumbnail display; multiple functional virtual buttons are set in the functional area to send various types of trigger instructions to the server to process eye images.
  • the functional area is provided with a virtual button for triggering the detection of the eye image, and the trigger operation of the virtual button is used as the detection instruction for the eye image; when the detection instruction is detected, it means that the eye image is detected
  • the requirements for detection include the type of eye, the position of the macula, the position of the optic disc, and the abnormal position in the eye image.
  • the eye types include left eye and right eye; the macula is 0.35cm and slightly below the temporal side of the optic disc of the fundus, in the optical center area of the human eye, which is the projection point of the vision axis; the optic disc is the retina from the macula There is a light red disc-like structure with a diameter of about 1.5mm and a clear boundary at about 3mm on the side of the nose; abnormalities are abnormal information that the eye image of the normal eye does not have, such as drusen; the position of the macula in the eye image It is the position of the macula, the position of the optic disc in the eye image is the position of the optic disc, and the position of the abnormal information in the eye image is the abnormal position.
  • Pre-set a preset algorithm for detecting the eye type, macula position, optic disc position and abnormal position in the eye image call the preset algorithm, and use the preset algorithm to detect the eye image to generate a characterization preset
  • calling the preset algorithm to detect the eye image includes:
  • Step S11 call a first preset algorithm, locate the center of the macula in the eye image, and determine the coordinates of the center of the macula;
  • Step S12 calling a second preset algorithm, segmenting the optic disc area in the eye image, and determining the optic disc position, optic disc center point coordinates, and optic disc diameter;
  • the preset algorithm includes a first preset algorithm and a second preset algorithm.
  • the first preset algorithm may be the yolo algorithm
  • the second preset algorithm may be the v-net algorithm; where yolo (You Only Look Once) algorithm is a target detection algorithm, V-Net algorithm is a 3D version of U-Net [1], and U-net is an algorithm for image classification and segmentation as a whole to reduce the amount of calculation.
  • the first preset algorithm is called first, and the center of the macula in the eye image is located through the first preset algorithm to obtain the coordinates of the center of the macula.
  • the second preset algorithm is called, the optic disc area in the eye image is segmented by the second preset algorithm, the optic disc position is obtained after segmentation, and the center of the optic disc position is positioned to obtain the coordinates of the optic disc center point; Detect the length of the optic disc in the vertical direction, and use this length as the optic disc diameter.
  • a virtual coordinate system can be established in the eye image; for example, the starting position of the lower left corner of the eye image is used as the coordinate circle, or the length and width of the eye image
  • the center position on the top is the coordinate circle point to establish a virtual coordinate system, and then the center point coordinates of the optic disc are determined according to the position point of the center of the optic disc position in the virtual coordinate system.
  • Step S13 determining the eye type according to the coordinates of the center point of the optic disc, and determining the position of the macula according to the coordinates of the center point of the macula and the diameter of the optic disc;
  • the eye type represented by the eye image is identified and the eye type is determined; at the same time, the center of the macula is the center coordinate, and the radius of the double optic disc is twice the diameter of the circle. Macular location.
  • Step S14 Determine whether there is abnormal information in the eye image according to the second preset algorithm. If the abnormal information exists, detect the position of the abnormal information, determine the abnormal position, and compare the eye The type, macula position, optic disc position, and abnormal position are generated as the detection result.
  • the image of the macular position can be optimized first, and the image style in it can be unified ⁇ . Thereafter, for the optimized macular position, it is determined whether there is abnormal information in the eye image based on the second preset algorithm, and if there is abnormal information, the position of the abnormal information is identified and detected to determine the abnormal position.
  • the detection result is displayed in the preset terminal interface functional area Among the corresponding detection items in, if the detected abnormal information is drusen, the area of drusen detected will be displayed in the detection items of drusen in the functional area; at the same time, the detection results will be displayed directly on the eye image , To intuitively reflect the position and area of each detection item in the eye image.
  • the same type of information in the eye image is selected and marked with the same color line; such as at the boundary of the macula Use red lines for selection, and use blue lines for the location of an abnormality A. Lines of different colors are used to characterize the type of abnormality in the eye image and the location of each abnormality.
  • the detection result obtained by the server detecting the eye image is inaccurate.
  • a modification mechanism is provided to modify the detection result; specifically, the step of generating the detection result includes :
  • Step a When a modification instruction to the detection result is detected, determine the modification type of the modification instruction according to the type identifier in the modification instruction;
  • the medical staff examines the eye image and its corresponding detection result, and determines whether the eye type, macula position, optic disc position, and abnormal position determined in the detection result are accurate based on experience; if If any item is inaccurate, you can change it through various editing tools in the tool window of the preset terminal interface.
  • the content of the change includes drawing various types of selection boxes such as rectangular boxes, circles, unclosed curves, polygons, etc., to change the inaccurate detection results; for example, the boundary of the macular position determined in the detection results is inaccurate, making the macula If the area of the location is larger, the circular tool can be used to select and adjust the location of the macula, so that the determined location of the macula is more accurate.
  • the tool window is also equipped with measuring tools, dragging tools, modification tools, translation tools, eraser tools, etc.; among them, various selection boxes that are changed and drawn can be measured by the measuring tools, and the area of each selection box can be determined and adjusted After that, the area of each position is updated; at the same time, you can add labels to the drawn selection box to name, modify the line width of the selection box, translate the selection box, and modify the outline of the selection box.
  • a new mechanism is also set up for the detection results, for the abnormal information in the eye image that is judged by the experience of the medical staff, but the server does not detect the abnormal information, such as the pigment abnormal information in the eye image; Through the new mechanism, the abnormal information is added to the detection result.
  • the same addition is also carried out through various editing tools in the tool window, such as drawing and measuring tools; the triggering of the change operation or the new operation is used as the modification instruction.
  • the modification instruction carries a type identifier that characterizes the operation type during the sending process. When the modification instruction is detected, the type identifier in it is read, and the type identifier read is used to determine the type identifier of the modification instruction. Modification type.
  • Step b when the modification type is a modification type, correct the eye type, macula position, optic disc position and abnormal position in the detection result according to the modification instruction;
  • the modification type includes two aspects: a modification type and a new type.
  • the modification type is used to change the detection result
  • the new type is used to add the detection result.
  • the correction instruction carries information that characterizes the need to change the detection result; if the area of the macular position determined in the detection result is a, the change information is to adjust the boundary of the macular position so that the boundary becomes larger.
  • the area occupied by the location of the macula after the change information is changed is a+x; then the functional area and the size of the macula displayed in the eye image are corrected based on this, and the value of the macula area is corrected from a to a+x.
  • Step c When the modification type is a newly-added type, determine the pigment abnormality information in the eye image according to the modification instruction, and add the pigment abnormality information to the detection result to compare The test results are updated.
  • the modification type of the modification instruction when determined to be a new type based on the type identifier, it means that there is pigment abnormality information in the eye image.
  • the pigment abnormality information includes pigment enhancement and pigment shedding, which is specifically determined by medical personnel .
  • the medical staff uses the editing tool in the tool window to mark the location of the abnormal pigment information.
  • the mark operation is the modification instruction whose modification type is a new type; the mark operation as the modification instruction is received, and the modification instruction carries The type information that characterizes the type of pigment abnormality information and the position information of the pigment abnormality information, the pigment abnormality information in the eye image can be determined according to the modification instruction, the pigment abnormality information is added to the detection result, and the detection result is updated to characterize Information and location of pigment abnormalities in the eye image.
  • Step S20 When the prediction instruction for the abnormal change trend is detected, it is determined whether there is a risk-related factor in the detection result. If the risk-related factor exists, according to the preset number of risk-related factors and the probability The corresponding relationship of, predict the probability of abnormal change corresponding to the risk-related factor;
  • the abnormal information in the detection result indicates that there is an abnormality in the eye image, that is, the patient from which the eye image is derived has a certain type of eye disease, and this type of eye disease may change over time .
  • This embodiment is provided with a mechanism for predicting the abnormal change trend based on the detection result.
  • the content of the prediction includes determining whether there is information in the eye image that characterizes the patient's risk of advanced AMD (age-related macular degeneration) in the next five years, and The size of the possibility of change in the amount of information; the predicted result is used as a reference basis for medical staff to judge the change trend of the eye disease.
  • the functional area is provided with a virtual button for predicting the abnormal change trend of the eye image, and the click operation of the virtual button is detected as a prediction instruction for the abnormal change trend.
  • the prediction instruction When the prediction instruction is detected, first determine whether there are risk-related factors in the test results; because the abnormal information in the test results includes drusen and pigment abnormalities, the risk-related factors are used to indicate that the maximum drusen diameter is greater than 125 ⁇ m, and abnormalities such as pigment loss or pigmentation enhancement.
  • the test result is judged to determine whether it carries any of the two risk-related factors. If any one of them is present, it can be determined that there is a risk-related factor in the test result, which means that There is information in the eye images that characterizes the patient's risk of having advanced AMD in the next five years.
  • the risk-related factors in the detection result are the risk information contained in the eye image, and the number of risk-related factors in the detection result is different, and the probability of the amount of information changes is different.
  • the probability of abnormal change the corresponding relationship between the number of risk-related factors and the probability is preset. For example, the probability corresponding to the absence of any risk factor is set to 0.5%, and the probability corresponding to any risk factor is 10%, the probability corresponding to two risk factors is 25%, etc.
  • the preset corresponding relationship is called, and the number of risk-related factors in the test result is compared with the corresponding relationship to determine the corresponding number of risk-related factors
  • the corresponding probability is the predicted probability of abnormal change corresponding to the risk-related factor.
  • the abnormal change probability output display is provided as reference information to the medical staff.
  • the server is inaccurate in predicting the risk information in the eye image and the probability of change in the amount of information, that is, the risk-related factors in the judged detection results and the predicted probability of abnormal changes are inaccurate ;
  • an adjustment mechanism is set up to adjust risk-related factors; specifically, the step of predicting the probability of abnormal changes corresponding to risk-related factors includes:
  • Step d when an adjustment instruction for the risk-related factor is detected, adjust the risk-related factor according to the adjustment information in the adjustment instruction to generate an adjusted risk-related factor;
  • Step e Update the abnormal change probability according to the adjusted risk-related factors.
  • the medical staff After the abnormal change probability is generated for the detection result of the eye image, the medical staff checks the eye image, the detection result, the risk-related factors, and the summary of abnormal changes, and based on experience, determines whether the risk-related factors determined in the detection result are accurate ; If it is judged that the risk-related factors determined by the detection result are not accurate, it can be adjusted through various editing tools in the tool window of the preset terminal interface.
  • the server detects the adjustment instruction, it reads the adjustment information carried in it, and adjusts the risk-related factors based on the adjustment information; after the adjustment operation, the detection items that were originally risk-related factors in the detection result may change to Non-risk-related factors that are not originally detected as risk-related factors may change to risk-related factors.
  • the number of risk-related factor items in the detection result of the adjustment operation will be changed, and the number of risk-related factor items for this change is the generated adjusted risk-related factor. Changes in the number of risk-related factors will cause changes in the probability of abnormal changes. Compare the adjusted risk-related factors with the preset risk factors and the corresponding relationship between the probability and determine the corresponding risk-related factors. The probability is the new probability of abnormal change, which realizes the update of the probability of abnormal change.
  • Step S30 When a trigger instruction to generate a detection report is detected, the detection result, the risk-related factor, and the abnormal change probability are used to generate a reference detection report of the eye image, and complete the evaluation of the eye image Processing.
  • the functional area is provided with a virtual button for generating a detection report for the eye image.
  • the virtual button is triggered, and the trigger operation is used as a trigger instruction for generating the detection report for detection.
  • the trigger instruction is detected, it indicates that there is a need to generate a detection report, so that the eye type, macula position, optic disc position, and abnormal position in the detection result, as well as risk-related factors, and the probability of abnormal changes are used to generate eye images.
  • the generated reference test report is used as reference information and provided to medical staff for viewing and reference.
  • the process of generating test results, judging the risk-related factors in the test results, and predicting the probability of abnormal changes involves the correction of the test results, risk-related factors and the probability of abnormal changes, if the correction operation is generating a reference test report After the steps; that is, after the reference detection report of the eye image is generated for the detection results, risk-related factors, and the probability of abnormal changes, any one of the detection results, risk-related factors, or the probability of abnormal changes is corrected to make the generated
  • the reference test report for is not accurate. At this time, the reference test report needs to be updated according to the correction items.
  • the position of the optic disc in the test result needs to be corrected accordingly to update the reference test report.
  • the correction operation is performed after the step of generating a reference test report, if any of the test results, risk-related factors or abnormal change probability is corrected, the corrected test results, risk-related factors and abnormal changes are directly used Probabilistically generate a reference detection report of the eye image.
  • buttons used to save and print the reference test report in the function area.
  • the save virtual button When the reference test report is generated, trigger the save virtual button to save the reference test report in PDF format in the preset Path. If the reference test report generated for the same eye image has been saved before, the subsequent save operation will overwrite the previously saved content. At the same time, by triggering the print virtual key, the latest saved reference test report can be printed. Through the saving and printing operation, it is convenient to provide the reference test report to the medical staff for viewing and reference.
  • the eye image processing method of this embodiment obtains various detection results including eye type, macula position, optic disc position, and abnormal position in the eye image by performing image processing on the eye image; at the same time, according to the risk in the detection result Correlation factors, predict the abnormal change probability of abnormal change trends; and then use the detection results, risk-related factors and abnormal change probability to generate a reference test report for medical staff as a reference; it is convenient for medical staff to compare the eye image in the reference test report The abnormal information is quickly judged, avoiding the artificial judgment of the abnormal information in the eye image by the medical staff, which improves the work efficiency of the medical staff.
  • the step when the detection instruction for the eye image is detected includes:
  • Step S40 when an opening instruction for the eye image is detected, the eye image is opened and displayed in a preset terminal interface, and it is detected whether a patient information input instruction corresponding to the eye image is received;
  • the eye image needs to be opened and displayed in the preset terminal interface; specifically, a tool for opening the eye image is provided in the tool window, and the tool is Select, and a dialog box for selecting the path of the eye image will pop up. Select the path of the eye image through this dialog box to trigger the opening instruction of the eye image.
  • the opening instruction is detected, the eye image in the selected path is opened and displayed in the main image display area of the preset terminal interface, so as to view and process the displayed eye image.
  • the patient information is the information of the patient corresponding to the eye image, including name, gender, age, examination number, and examination date.
  • the virtual button is provided with a corresponding identifier, and by detecting whether a trigger instruction carrying the identifier is received, it is determined whether a patient information input instruction corresponding to the eye image is received.
  • Step S50 If a patient information input instruction corresponding to the eye image is received, an information input box is displayed on the preset terminal interface, and it is detected whether the patient information input based on the information input box meets preset rules ;
  • the identifier determines whether the patient information of the patient from which the eye image is sourced needs to be input for the eye image, so that the information input box is displayed on the preset terminal interface.
  • the received patient information is tested to determine whether it meets the preset rules.
  • the preset rules are preset rules for judging whether the patient information is reasonable, such as setting the patient's name to be no more than 6 characters in length, the preset field of gender between "male” and "female", and the range of age Between 0 and 120 years old, the test date is within 30 years before the current date, etc. By comparing the received patient information with preset rules, it is determined whether the patient information meets the preset rules, so as to ensure the accuracy of the input patient information.
  • Step S60 If the patient information input based on the information input box meets a preset rule, display the patient information in the eye image, if the patient information input based on the information input box does not meet the preset rule , The prompt message for re-input is output.
  • the patient information will be saved; at the same time, the patient information will be displayed in the eye image.
  • a predetermined position such as the upper left corner of the main image display area. If, after comparison, it is determined that any item of the input patient information does not meet the preset rules set in advance, the patient information and other patient information that meets the preset rules are displayed differently, and the output is Re-enter the patient information that does not meet its preset rules.
  • the eye image may be unclearly displayed due to various factors in the shooting process, which may affect the processing result of the eye image; and to avoid For the effect of the processing result, this embodiment is provided with a mechanism for enhancing the color in the eye image.
  • the step of displaying the patient information in the eye image includes:
  • Step S70 When an enhancement instruction for image enhancement is detected, determine the enhancement type of the image enhancement according to the type code in the enhancement instruction;
  • the image enhancement in this embodiment includes two types of three-color channel enhancement and preset algorithm enhancement.
  • the functional area is provided with different virtual buttons for the two enhancement types, and each virtual button carries a different type code.
  • the trigger operation performed on any one of the two virtual buttons will be detected as an enhancement command for image enhancement; when the enhancement command is detected, the type code carried in it will be read first, according to the The type code determines the enhancement type for image enhancement of the eye image.
  • Step S80 when the enhancement type is the first preset type, read the color information in the enhancement instruction, and enhance the color in the eye image corresponding to the color information;
  • the three-color channel enhancement type in the image enhancement is taken as the first preset type, where the three-color channels are red, green, and blue color channels, and the first preset type is the three color channels through red, green, and blue.
  • This type of image enhancement is for the entire image.
  • the enhancement instruction for image enhancement carries color information that characterizes the color that needs to be enhanced.
  • the color information in the instruction is read to determine the color channel that needs to be enhanced through the color information; then the color corresponding to the color information in the eye image is enhanced, that is, the color represented by the color information in the eye image The channel is enhanced so that the entire image of the eye image presents the color corresponding to the color information.
  • Step S90 When the enhancement type is the second preset type, the framed area in the eye image or the image in the preset default area is enhanced based on the enhancement instruction.
  • the preset algorithm enhancement in the image enhancement is used as the second preset type, where the preset algorithm may be the AREDS algorithm, and the second preset type is the type enhanced by the AREDS algorithm.
  • This type of image enhancement is aimed at the framed area or the preset default area in the eye image.
  • the type encoding determines that the enhancement type of the image enhancement is the second preset type, the eye image is detected and judged Whether there is a frame selection area; if it exists, the image in the frame selection area and the enhancement instructions are enhanced; if it does not exist, first determine the preset default area in the eye image, and set it in the preset default area
  • the image of is enhanced according to the enhancement instructions.
  • the preset default area is determined by a preset rule. For example, it is preset that the center point of the eye image is a dot and a circular area with a certain radius as the default area.
  • a virtual button for canceling the image enhancement is also set in the functional area.
  • the image enhancement part in the eye image Cancel and display the original image of the eye image to compare the eye image before and after enhancement.
  • the eye image processing method further includes the steps:
  • Step q1 When receiving a parameter configuration instruction, determine whether the eye image exists in the preset terminal interface, and if the eye image exists, output a prompt message to close the eye image;
  • the tool window is also provided with a virtual button for configuring the overall parameters of the eye image.
  • click the virtual button to receive the click operation as a parameter configuration instruction;
  • the parameters that support configuration include image size units, conversion between various size units, etc.
  • receiving the parameter configuration instruction first detect whether the eye image is displayed in the main image display area of the preset terminal interface. If the eye image is displayed, output a prompt message to close the displayed eye image; After closing the eye image, configure the parameters to prevent the configured parameters from affecting the displayed eye image.
  • Step q2 If the eye image does not exist, display a configuration input box in the preset terminal interface, and perform parameter configuration according to configuration information input based on the configuration input box.
  • a configuration input box will pop up and display in the preset terminal interface, so that various parameters can be input through the configuration input box for configuration.
  • configure the parameters according to the configuration information if the original image size unit is mm, and the size unit received by the input box is ⁇ m, the original The image size unit mm is modified to ⁇ m to realize the configuration of the image size unit.
  • this application provides an eye image processing device.
  • the eye image processing device includes:
  • the detection module 10 is used to call a preset algorithm to detect the eye image and generate a detection result when a detection instruction to the eye image is detected, wherein the detection result includes the eye in the eye image Type, macula location, optic disc location and abnormal location;
  • the prediction module 20 is used to determine whether there is a risk-related factor in the detection result when a prediction instruction for an abnormal change trend is detected, and if the risk-related factor exists, according to the preset number of risk-related factors and Correspondence between probabilities, predict the probability of abnormal change corresponding to the risk-related factor;
  • the generating module 30 is configured to generate a reference detection report of the eye image from the detection result, the risk-related factor, and the abnormal change probability when a trigger instruction to generate a detection report is detected, and complete the Processing of eye images.
  • the eye image processing device of this embodiment obtains various detection results including the eye type, macula position, optic disc position, and abnormal position in the eye image by performing image processing on the eye image; at the same time, according to the risk in the detection result Correlation factors, predict the abnormal change probability of abnormal change trends; and then use the detection results, risk-related factors and abnormal change probability to generate a reference test report for medical staff as a reference; it is convenient for medical staff to compare the eye image in the reference test report The abnormal information is quickly judged, avoiding the artificial judgment of the abnormal information in the eye image by the medical staff, which improves the work efficiency of the medical staff.
  • the detection module further includes:
  • a positioning unit configured to call a first preset algorithm, locate the center of the macula in the eye image, and determine the coordinates of the center of the macula;
  • a calling unit configured to call a second preset algorithm, segment the optic disc area in the eye image, and determine the position of the optic disc, the coordinates of the optic disc center point, and the optic disc diameter;
  • the determining unit is configured to determine the eye type according to the coordinates of the center point of the optic disc, and determine the position of the macula according to the coordinates of the center point of the macula and the diameter of the optic disc;
  • the detection unit is configured to determine whether abnormal information exists in the eye image according to the second preset algorithm, and if the abnormal information exists, detect the position of the abnormal information, determine the abnormal position, and compare the The eye type, macula position, optic disc position and abnormal position are generated as the detection result.
  • the eye image processing device further includes:
  • the first determining module is configured to determine the modification type of the modification instruction according to the type identifier in the modification instruction when the modification instruction to the detection result is detected;
  • the correction module is configured to correct the eye type, macula position, optic disc position, and abnormal position in the detection result when the modification type is a correction type, according to the modification instruction;
  • the first update module is configured to determine the pigment abnormality information in the eye image according to the modification instruction when the modification type is a newly added type, and add the pigment abnormality information to the detection result To update the detection result.
  • the eye image processing device further includes:
  • the adjustment module is configured to, when an adjustment instruction for the risk-related factor is detected, adjust the risk-related factor according to the adjustment information in the adjustment instruction to generate an adjusted risk-related factor;
  • the second update module is configured to update the abnormal change probability according to the adjusted risk-related factors.
  • the eye image processing device further includes:
  • the display module is used to open and display the eye image in the preset terminal interface when the opening instruction to the eye image is detected, and detect whether the patient information input corresponding to the eye image is received instruction;
  • the receiving module is configured to, if a patient information input instruction corresponding to the eye image is received, display an information input box on the preset terminal interface, and detect whether the patient information input based on the information input box meets the expected Set rules
  • the output module is configured to display the patient information in the eye image if the patient information input based on the information input box meets preset rules, and if the patient information input based on the information input box does not meet the expected Set the rules, then output the prompt information for re-input.
  • the eye image processing device further includes:
  • the second determining module is configured to, when an enhancement instruction for image enhancement is detected, determine the enhancement type of the image enhancement according to the type code in the enhancement instruction;
  • a reading module configured to read the color information in the enhancement instruction when the enhancement type is the first preset type, and enhance the color in the eye image corresponding to the color information;
  • the enhancement module is configured to, when the enhancement type is the second preset type, enhance the framed area in the eye image or the image in the preset default area based on the enhancement instruction.
  • the eye image processing device further includes:
  • the judging module is used for judging whether the eye image exists in the preset terminal interface when the parameter configuration instruction is received, and if the eye image exists, outputting a prompt message to close the eye image ;
  • the configuration module is configured to display a configuration input box in the preset terminal interface if the eye image does not exist, and perform parameter configuration according to the configuration information input based on the configuration input box.
  • each virtual function module of the above-mentioned eye image processing device is stored in the memory 1005 of the eye image processing device shown in FIG. 3, and when the processor 1001 executes the readable instructions for eye image processing, the implementation shown in the embodiment shown in FIG. The function of each module.
  • Fig. 3 is a schematic diagram of a device structure involving a hardware operating environment in a method of an embodiment of the present application.
  • the eye image processing device in the embodiment of the application may be a PC (personal computer, personal computer ), but also terminal devices such as smart phones, tablet computers, e-book readers, and portable computers.
  • PC personal computer, personal computer
  • terminal devices such as smart phones, tablet computers, e-book readers, and portable computers.
  • the eye image processing device may include: a processor 1001, such as a CPU (Central Processing Unit, central processing unit), memory 1005, communication bus 1002.
  • the communication bus 1002 is used to implement connection and communication between the processor 1001 and the memory 1005.
  • the memory 1005 may be a high-speed RAM (random access memory, random access memory), or stable memory (non-volatile memory), such as disk storage.
  • the memory 1005 may also be a storage device independent of the foregoing processor 1001.
  • the eye image processing equipment may also include a user interface, a network interface, a camera, RF (Radio Frequency, radio frequency) circuit, sensor, audio circuit, WiFi (Wireless Fidelity, wireless broadband) module and so on.
  • the user interface may include a display screen (Display) and an input unit such as a keyboard (Keyboard), and the optional user interface may also include a standard wired interface and a wireless interface.
  • the optional network interface can include standard wired interface and wireless interface (such as WI-FI interface).
  • the structure of the eye image processing device shown in FIG. 3 does not constitute a limitation on the eye image processing device, and may include more or less components than shown in the figure, or a combination of certain components, Or different component arrangements.
  • the memory 1005 which is a computer-readable storage medium, may include an operating system, a network communication module, and readable instructions for processing eye images.
  • the operating system is a readable instruction that manages and controls the hardware and software resources of the eye image processing equipment, and supports the operation of the eye image processing readable instructions and other software and/or readable instructions.
  • the network communication module is used to implement communication between various components in the memory 1005 and communication with other hardware and software in the eye image processing device.
  • the processor 1001 is configured to execute the eye image processing readable instructions stored in the memory 1005 to implement the steps of the foregoing eye image processing method embodiments.
  • the present application provides a computer-readable storage medium
  • the computer-readable storage medium may be a non-volatile storage medium
  • the computer-readable storage medium stores one or more readable instructions, the one or one
  • the above readable instructions may also be executed by one or more processors to implement the steps in each embodiment of the above eye image processing method.

Landscapes

  • Engineering & Computer Science (AREA)
  • Medical Informatics (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Primary Health Care (AREA)
  • Epidemiology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Biomedical Technology (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Pathology (AREA)
  • Quality & Reliability (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Measuring And Recording Apparatus For Diagnosis (AREA)
  • Eye Examination Apparatus (AREA)
  • Image Processing (AREA)

Abstract

一种眼部图像处理方法、装置、设备及计算机可读存储介质，所述方法包括：当侦测到对眼部图像的检测指令时，调用预设算法进行检测，生成包括眼睛类型、黄斑位置、视盘位置以及异常位置的检测结果；当侦测到对异常变化趋势的预测指令时，判断检测结果中是否存在风险相关因子，若存在则根据预设的风险相关因子的数量与概率之间的对应关系，预测与风险相关因子对应的异常变化概率；当检测到生成检测报告的触发指令时，则将检测结果、风险相关因子以及异常变化概率生成眼部图像的参考检测报告。通过图像检测技术，对眼部图像进行图像处理，生成用于供医务人员作为参考的参考检测报告；便于医务人员快速判断眼部图像中的异常信息。

Description

眼部图像处理方法、装置、设备及计算机可读存储介质
本申请要求于2019年06月05日提交中国专利局、申请号为201910494821.8、发明名称为“眼部图像处理方法、装置、设备及计算机可读存储介质”的中国专利申请的优先权，其全部内容通过引用结合在申请中。
技术领域
本申请主要涉及图像处理技术领域，具体地说，涉及一种眼部图像处理方法、装置、设备及计算机可读存储介质。
背景技术
目前，在眼部疾病检查的过程中，通常先通过眼底彩照的方式得到眼部图像，再由医务人员人为确定眼部图像中的视盘区域、黄斑区域，进而判断黄斑区域是否病变，是否存在异常信息等。该单纯由医务人员对眼部图像中的异常信息进行判断的方式，会使得医务人员的工作量非常大；尤其对于缺少工作经验的医务人员，花费的时间更多，导致医务人员的工作效率低。
发明内容
本申请的主要目的是提供一种眼部图像处理方法、装置、设备及计算机可读存储介质，旨在解决现有技术中仅由医务人员人为对眼部图像中的异常信息进行判断，导致医务人员工作效率低的问题。
为实现上述目的，本申请提供一种眼部图像处理方法，所述眼部图像处理方法包括以下步骤：
当侦测到对眼部图像的检测指令时，调用预设算法对所述眼部图像进行检测，生成检测结果，其中所述检测结果包括所述眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常位置；
当侦测到对异常变化趋势的预测指令时，判断所述检测结果中是否存在风险相关因子，若存在所述风险相关因子，则根据预设的风险相关因子的数量与概率之间的对应关系，预测与所述风险相关因子对应的异常变化概率；
当检测到生成检测报告的触发指令时，则将所述检测结果、所述风险相关因子以及所述异常变化概率生成所述眼部图像的参考检测报告，完成对所述眼部图像的处理。
此外，为实现上述目的，本申请还提出一种眼部图像处理装置，所述眼部图像处理装置包括：
检测模块，用于当侦测到对眼部图像的检测指令时，调用预设算法对所述眼部图像进行检测，生成检测结果，其中所述检测结果包括所述眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常位置；
预测模块，用于当侦测到对异常变化趋势的预测指令时，判断所述检测结果中是否存在风险相关因子，若存在所述风险相关因子，则根据预设的风险相关因子的数量与概率之间的对应关系，预测与所述风险相关因子对应的异常变化概率；
生成模块，用于当检测到生成检测报告的触发指令时，则将所述检测结果、所述风险相关因子以及所述异常变化概率生成所述眼部图像的参考检测报告，完成对所述眼部图像的处理。
此外，为实现上述目的，本申请还提出一种眼部图像处理设备，所述眼部图像处理设备包括：存储器、处理器、通信总线以及存储在所述存储器上的眼部图像处理可读指令；
所述通信总线用于实现处理器和存储器之间的连接通信；
所述处理器用于执行所述眼部图像处理可读指令，以实现以下步骤：
当侦测到对眼部图像的检测指令时，调用预设算法对所述眼部图像进行检测，生成检测结果，其中所述检测结果包括所述眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常位置；
当侦测到对异常变化趋势的预测指令时，判断所述检测结果中是否存在风险相关因子，若存在所述风险相关因子，则根据预设的风险相关因子的数量与概率之间的对应关系，预测与所述风险相关因子对应的异常变化概率；
当检测到生成检测报告的触发指令时，则将所述检测结果、所述风险相关因子以及所述异常变化概率生成所述眼部图像的参考检测报告，完成对所述眼部图像的处理。
此外，为实现上述目的，本申请还提供一种计算机可读存储介质，所述计算机可读存储介质存储有一个或者一个以上可读指令，所述一个或者一个以上可读指令可被一个或者一个以上的处理器执行以用于：
当侦测到对眼部图像的检测指令时，调用预设算法对所述眼部图像进行检测，生成检测结果，其中所述检测结果包括所述眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常位置；
当侦测到对异常变化趋势的预测指令时，判断所述检测结果中是否存在风险相关因子，若存在所述风险相关因子，则根据预设的风险相关因子的数量与概率之间的对应关系，预测与所述风险相关因子对应的异常变化概率；
当检测到生成检测报告的触发指令时，则将所述检测结果、所述风险相关因子以及所述异常变化概率生成所述眼部图像的参考检测报告，完成对所述眼部图像的处理。
本实施例的眼部图像处理方法，通过对眼部图像进行图像处理，得到包含眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常位置等各项检测结果；同时根据检测结果中的风险相关因子，预测异常变化趋势的异常变化概率；进而将检测结果、风险相关因子以及异常变化概率生成用于供医务人员作为参考的参考检测报告；便于医务人员依据参考检测报告对眼部图像中的异常信息快速判断，避免单纯由医务人员人为对眼部图像中的异常信息进行判断，提高了医务人员的工作效率。
附图说明
图1是本申请的眼部图像处理方法第一实施例的流程示意图；
图2是本申请的眼部图像处理装置第一实施例的功能模块示意图；
图3是本申请实施例方法涉及的硬件运行环境的设备结构示意图。
本申请目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。
具体实施方式
应当理解，此处所描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。
本申请提供一种眼部图像处理方法。
请参照图1，图1为本申请眼部图像处理方法第一实施例的流程示意图。在本实施例中，所述眼部图像处理方法包括：
步骤S10，当侦测到对眼部图像的检测指令时，调用预设算法对所述眼部图像进行检测，生成检测结果，其中所述检测结果包括所述眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常位置；
本申请的眼部图像处理方法适用于通过服务器对眼部图像进行处理，以辅助医务人员对眼部图像中的异常信息进行判断。具体地，服务器与具有显示装置的预设终端通信连接，预设终端可以是平板电脑、台式电脑、智能手机等终端；将显示装置的显示界面作为预设终端界面，预设终端界面中设置工具窗口、主图像显示区、缩略图显示区以及功能区。其中工具窗口中有各种类型的工具，以用于对眼部图像进行编辑；主图像显示区则用于对当前处理的眼部图像进行显示，缩略图用于对其他等待处理的眼部图像进行缩略显示；功能区中设置有多个功能虚拟按键，以用于向服务器发送各种类型的触发指令，对眼部图像进行处理。
功能区中设置有用于触发对眼部图像进行检测的虚拟按键，将对该虚拟按键的触发操作作为对眼部图像的检测指令；当侦测到该检测指令时，则说明有对眼部图像进行检测的需求，检测的内容包括眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常位置。其中眼睛类型包括左眼和右眼两种类型；黄斑是在眼底视神经盘的颞侧0.35cm处并稍下方，处于人眼的光学中心区，是视力轴线的投影点；视盘为视网膜由黄斑向鼻侧约3mm处有一直径约1.5mm，境界清楚的淡红色圆盘状结构；异常则为正常眼睛的眼部图像不具有的非正常信息，如玻璃膜疣；黄斑在眼部图像中的位置即为黄斑位置，视盘在眼部图像中的位置即为视盘位置，异常信息在眼部图像中的位置即为异常位置。预先设置有对眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常位置进行检测的预设算法，对该预设算法进行调用，并用预设算法对眼部图像进行检测，生成表征预设终端界面中所显示眼部图像的眼睛类型、黄斑位置、视盘位置以及异常位置的检测结果。
具体地，调用预设算法对眼部图像进行检测，生成检测结果的步骤包括：
步骤S11，调用第一预设算法，对所述眼部图像中的黄斑中心进行定位，确定黄斑中心点坐标；
步骤S12，调用第二预设算法，对所述眼部图像中的视盘区域进行分割，确定视盘位置、视盘中心点坐标以及视盘直径；
进一步地，预先设置的预设算法包括第一预设算法和第二预设算法，其中第一预设算法可以为yolo算法，第二预设算法可以为v-net算法；其中yolo（You Only Look Once）算法是目标检测算法，V-Net算法是U-Net[1]的3D版本，而U-net是以图片为整体来进行图片分类分割，以减少计算量的算法。在对眼部图像进行检测的过程中，先对第一预设算法进行调用，通过第一预设算法对眼部图像中的黄斑中心进行定位，得到黄斑中心点坐标。此后再对第二预设算法进行调用，通过第二预设算法对眼部图像中的视盘区域进行分割，经过分割得到视盘位置，并对视盘位置的中心进行定位，得到视盘中心点坐标；同时检测视盘在竖直方向的长度，将该长度作为视盘直径。
需要说明的是，为了便于定位得到中心点坐标，可在眼部图像中建立虚拟坐标系；如以眼部图像左下角的起始位置为坐标圆点，或者以眼部图像长度尺寸和宽度尺寸上的中心位置为坐标圆点建立虚拟坐标系，进而依据视盘位置的中心在该虚拟坐标系中的位置点，确定视盘中心点坐标。
步骤S13，根据所述视盘中心点坐标，确定眼睛类型，并根据所述黄斑中心点坐标和所述视盘直径，确定黄斑位置；
更进一步地，依据视盘中心点坐标，对眼部图像所表征的眼睛类型进行识别，确定眼睛类型；同时以黄斑中心坐标为圆心，两倍视盘直径为半径做圆，所得到的圆心区域即为黄斑位置。
步骤S14，根据所述第二预设算法判断所述眼部图像中是否存在异常信息，若存在所述异常信息，则对所述异常信息的位置进行检测，确定异常位置，并将所述眼睛类型、黄斑位置、视盘位置和异常位置生成为所述检测结果。
考虑到眼部图像中的异常信息主要针对黄斑区域中的异常，从而在确定黄斑位置后，为了对其中是否存在异常进行判断，可先对黄斑位置的图像进行优化，将其中的图像风格进行统一化处理。此后针对优化后的黄斑位置基于第二预设算法判断眼部图像中是否存在异常信息，若存在异常信息，则对异常信息的位置进行识别检测，确定异常位置。
在利用第一预设算法和第二预设算法对眼部图像进行检测，得到包含眼睛类型、黄斑位置、视盘位置以及异常位置的检测结果后，将该检测结果显示在预设终端界面功能区中的对应检测项中，如检测的异常信息为玻璃膜疣，则将检测得到的玻璃膜疣面积显示在功能区中玻璃膜疣的检测项中；同时还将检测结果直接显示在眼部图像中，以直观体现眼部图像中各检测项的位置及面积大小。在将检测结果显示在眼部图像中时，为了方便医务人员对检测结果中不同类型信息的查看，对眼部图像中同一类型的信息以相同颜色的线条进行选择标记；如在黄斑位置的边界用红色线条进行选择，某一异常A的位置边界用蓝色线条进行选择等；通过不同颜色的线条表征眼部图像中所具有的异常类型以及各异常的位置。
可理解地，服务器对眼部图像进行检测所得到的检测结果具有不准确性，为了提高检测结果的准确性，设置有对检测结果进行修改的修改机制；具体地，生成检测结果的步骤之后包括：
步骤a，当侦测到对所述检测结果的修改指令时，根据所述修改指令中的类型标识符，确定所述修改指令的修改类型；
在针对眼部图像生成检测结果后，医务人员对眼部图像及其对应的检测结果进行查看，并基于经验判定检测结果中所确定的眼睛类型、黄斑位置、视盘位置、异常位置是否准确；若其中存在任意一项不准确，则可通过预设终端界面的工具窗口中的各种编辑工具进行更改。更改的内容包括绘制矩形框、圆形、不封闭曲线、多边形等各种类型的选择框，以对不准确的检测结果进行更改；如检测结果中所确定的黄斑位置的边界不准确，使得黄斑位置的面积较大，则可通过圆形工具对黄斑位置进行选择调节，使得所确定的黄斑位置更为准确。同时工具窗口中还设置有测量工具、拖拽工具，修改工具、平移工具、橡皮擦工具等；其中可通过测量工具对更改绘制的各种选择框进行测量，确定各个选择框的面积，对调整后各位置的面积进行更新；同时还可对绘制的选择框添加标签以进行命名、修改选择框的线宽、平移选择框、修改选择框轮廓等操作。
此外，对检测结果还设置有新增机制，对于经医务人员的经验判断存在于眼部图像中的异常信息，而服务器没有检测出此异常信息，如眼部图像中的色素异常信息；则可通过新增机制，将该异常信息新增到检测结果中。同样的新增也通过工具窗口中的各种编辑工具进行，如绘制和测量工具等；将对更改操作或新增操作的触发作为修改指令，当侦测到修改指令时，因针对更改操作和新增操作的处理方式不一样，而需要对修改指令所针对的操作类型进行确定。具体地，修改指令在发送的过程中携带有表征操作类型的类型标识符，当侦测到修改指令时，读取其中的类型标识符，并依据该读取的类型标识符，确定修改指令的修改类型。
步骤b，当所述修改类型为修正类型时，则根据所述修改指令对所述检测结果中的眼睛类型、黄斑位置、视盘位置和异常位置进行修正；
进一步地的，修改类型包括修正类型和新增类型两方面，其中修正类型用于对检测结果进行更改，而新增类型用于对检测结果进行新增。当通过类型标识符确定修改指令的修改类型为修正类型时，则根据修改指令对检测结果中的眼睛类型、黄斑位置、视盘位置和异常位置进行修正。具体地，修正指令中携带有表征对检测结果所需要更改的信息；如检测结果中确定的黄斑位置所占的面积为a，更改信息为对黄斑位置的边界进行调整，使得边界变大，经该更改信息更改后黄斑位置所占的面积为a+x；则依据此对功能区以及眼部图像中所显示的黄斑面积大小进行修正，将黄斑面积的数值由a修正为a+x。
步骤c，当所述修改类型为新增类型时，则根据所述修改指令确定所述眼部图像中的色素异常信息，并将所述色素异常信息添加到所述检测结果中，以对所述检测结果进行更新。
更进一步地，当依据类型标识符确定修改指令的修改类型为新增类型时，则说明眼部图像中存在色素异常信息，该色素异常信息包括色素增强和色素脱落两方面，具体由医务人员确定。医务人员通过工具窗口中的编辑工具对色素异常信息所在的位置进行标注，该标注操作即为修改类型为新增类型的修改指令；对该作为修改指令的标注操作进行接收，修改指令中携带有表征色素异常信息类型的类型信息以及色素异常信息的位置信息，根据修改指令即可确定眼部图像中的色素异常信息，将该色素异常信息添加到检测结果中，对检测结果进行更新，以表征眼部图像中所存在的色素异常信息及位置。
步骤S20，当侦测到对异常变化趋势的预测指令时，判断所述检测结果中是否存在风险相关因子，若存在所述风险相关因子，则根据预设的风险相关因子的数量与概率之间的对应关系，预测与所述风险相关因子对应的异常变化概率；
可理解地，检测结果中的异常信息，表征了眼部图像中存在异常，即眼部图像所来源的患者患有某一类眼部疾病，而该类眼部疾病可能随着时间而发生变化。本实施例设置有依据检测结果对异常变化趋势进行预测的机制，预测的内容包括判断眼部图像中是否存在表征患者在未来五年内患有进展期AMD（老年性黄斑变性）风险的信息，以及该信息数量的变化可能性大小；将预测的结果作为医务人员判断该眼部疾病的变化趋势的参考依据。具体地，功能区中设置有用于对眼部图像进行异常变化趋势预测的虚拟按键，将对该虚拟按键的点击操作作为对异常变化趋势的预测指令进行侦测。
当侦测到该预测指令时，先判断检测结果中是否存在风险相关因子；因检测结果中的异常信息包括玻璃膜疣和色素异常两类，从而风险相关因子用来表征最大玻璃膜疣直径大于125μm、以及存在色素脱失或色素增强的色素异常等异常事项。在侦测到预测指令时，则对检测结果进行判断，确定其中是否携带有两项风险相关因子中的任意一项，若存在任意一项，则可判定检测结果中存在风险相关因子，即说明眼部图像中存在表征患者在未来五年内患有进展期AMD风险的信息。该存在于检测结果中的风险相关因子即为眼部图像中所具有的风险信息，且检测结果中所具有风险相关因子的数量不同，信息数量变化的可能性大小不一样，将该可能性大小作为异常变化概率；预先设置有表征风险相关因子的数量与概率之间的对应关系，如设置不存在任意一项风险因子所对应的概率为0.5%，存在任意一项风险因子所对应的概率为10%，存在两项风险因子所对应的概率为25%等。在判断出检测结果中存在风险相关因子后，则对该预先设定的对应关系进行调用，并将检测结果中所具有的风险相关因子数量和对应关系进行对比，确定与该风险相关因子数量对应的概率，该对应的概率即为预测的与风险相关因子对应的异常变化概率。将该异常变化概率输出显示，作为参考信息提供给医务人员。
可理解地，服务器对眼部图像中所存在的风险信息，以及该信息数量的变化可能性大小的预测具有不准确性，即判断的检测结果中的风险相关因子以及预测的异常变化概率不准确；为了提高两者的准确性，设置有对风险相关因子进行调整的调整机制；具体地，预测与风险相关因子对应的异常变化概率的步骤之后包括：
步骤d，当侦测到对所述风险相关因子的调整指令时，根据所述调整指令中的调整信息对所述风险相关因子进行调整，生成调整后的风险相关因子；
步骤e，根据调整后的所述风险相关因子，对所述异常变化概率进行更新。
在针对眼部图像的检测结果生成异常变化概率后，医务人员对眼部图像、检测结果、风险相关因子以及异常变化概括进行查看，并基于经验判定由检测结果中所确定的风险相关因子是否准确；若判断由检测结果所确定的风险相关因子不准确，则可通过预设终端界面的工具窗口中的各种编辑工具进行调整。如选择修改选择框轮廓的工具，对玻璃膜疣的区域进行更改，使得玻璃膜疣的面积变大或变小；变大或变小后的面积，即为对风险相关因子进行调整的调整信息，而对玻璃膜疣的更改操作为对风险相关因子进行调整的调整指令；将调整信息添加到调整指令中一并发送到服务器，以对风险相关因子进行调整。服务器在侦测到该调整指令时，读取其中所携带的调整信息，并依据该调整信息对风险相关因子进行调整；经调整操作后检测结果中原本为风险相关因子的检测项，可能变化为非风险相关因子，而原本不是风险相关因子的检测项，可能变化为风险相关因子。通过调整操作检测结果中的风险相关因子的项数会发送变化，该变化的风险相关因子项数即为所生成的调整后的风险相关因子。因风险相关因子项数的变化，会引起异常变化概率的变化，将调整后的风险相关因子和预设的风险因子与概率之间的对应关系进行对比，确定与调整后的风险相关因子所对应的概率，该概率即为新的异常变化概率，实现对异常变化概率的更新。
步骤S30，当检测到生成检测报告的触发指令时，则将所述检测结果、所述风险相关因子以及所述异常变化概率生成所述眼部图像的参考检测报告，完成对所述眼部图像的处理。
进一步地，为了使检测结果、风险相关因子以及异常变化概率更好的给医务人员提供参考，功能区中设置有针对眼部图像生成检测报告的虚拟按键。当有生成检测报告的需求时，对该虚拟按键进行触发，将该触发操作作为生成检测报告的触发指令进行检测。当检测到该触发指令时，则说明有生成检测报告的需求，从而将检测结果中的眼睛类型、黄斑位置、视盘位置和异常位置等信息，以及风险相关因子，异常变化概率生成眼部图像的参考检测报告，完成对眼部图像的处理。此后将该所生成的参考检测报告作为参考信息，提供给医务人员进行查看参考。
考虑到在生成检测结果、判断检测结果中的风险相关因子，以及预测异常变化概率的过程中，涉及到对检测结果、风险相关因子以及异常变化概率的修正，若修正操作在生成参考检测报告的步骤之后；即在针对检测结果、风险相关因子以及异常变化概率生成眼部图像的参考检测报告后，对检测结果、风险相关因子或异常变化概率中的任意一项进行修正，则会使得所生成的参考检测报告不准确，此时需要依据修正项对参考检测报告进行更新。如在生成参考检测报告后，对检测结果中的视盘位置进行修正，则需要对参考检测报告的检测结果中的视盘位置进行相应的修正，以对参考检测报告进行更新。而当修正操作在生成参考检测报告的步骤之后，存在对检测结果、风险相关因子或异常变化概率中的任意一项进行修正的情况，则直接用修正后的检测结果、风险相关因子和异常变化概率生成眼部图像的参考检测报告。
此外，功能区中还设置有用于对参考检测报告进行保存和打印操作的虚拟按键，当生成参考检测报告后，触发保存虚拟按键则可将参考检测报告以PDF的格式保存在预先所设定的路径中。对于同一眼部图像所生成的参考检测报告若此前进行了保存操作，后续的保存操作会把之前保存的内容覆盖掉。同时通过对打印虚拟按键的触发，可对最新所保存的参考检测报告进行打印操作。通过该保存和打印操作，便于将参考检测报告提供给医务人员进行查看和参考。
本实施例的眼部图像处理方法，通过对眼部图像进行图像处理，得到包含眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常位置等各项检测结果；同时根据检测结果中的风险相关因子，预测异常变化趋势的异常变化概率；进而将检测结果、风险相关因子以及异常变化概率生成用于供医务人员作为参考的参考检测报告；便于医务人员依据参考检测报告对眼部图像中的异常信息快速判断，避免单纯由医务人员人为对眼部图像中的异常信息进行判断，提高了医务人员的工作效率。
进一步地，在本申请眼部图像处理方法另一实施例中，所述当侦测到对眼部图像的检测指令时的步骤之前包括：
步骤S40，当侦测到对眼部图像的打开指令时，将所述眼部图像打开并显示在预设终端界面中，并检测是否接收到与所述眼部图像对应的患者信息输入指令；
可理解地，在对眼部图像进行检测之前，需要先将眼部图像打开显示在预设终端界面中；具体地，工具窗口中设置有用于对眼部图像进行打开的工具，对该工具进行选择，弹出选择眼部图像所在路径的对话框，通过该对话框选择眼部图像所在的路径，即可触发对眼部图像的打开指令。当侦测到该打开指令时，将所选择路径中的眼部图像打开并显示在预设终端界面的主图像显示区，以便对该显示的眼部图像进行查看和处理。同时功能区中设置有用于对患者信息进行输入的虚拟按键，其中患者信息为与眼部图像所对应患者具有的信息，包括姓名、性别、年龄、检查号和检查日期等。该虚拟按键设置有对应的标识符，通过检测是否接收到携带该标识符的触发指令，来确定是否接收到与眼部图像对应的患者信息输入指令。
步骤S50，若接收到与所述眼部图像对应的患者信息输入指令，则在所述预设终端界面中显示信息输入框，并检测基于所述信息输入框输入的患者信息是否满足预设规则；
进一步地，若通过标识符确定接收到与眼部图像对应的患者信息输入指令，则说明需要针对该眼部图像输入其所来源患者的患者信息，从而在预设终端界面中显示信息输入框，以便于通过信息输入框输入患者信息。在接收到基于信息输入框输入的患者信息后，对该接收的患者信息进行检测，判断其是否满足预设规则。其中预设规则为预先设定用于判断患者信息是否合理的规则，如设定患者的姓名长度不超过6个字、性别的预设字段在“男”和“女”之间、年龄的范围在0到120岁之间、检测日期在当前日期往前30年以内等。通过将接收到的患者信息和预设规则对比，判断患者信息是否满足预设规则，以确保所输入患者信息的准确性。
步骤S60，若基于所述信息输入框输入的患者信息满足预设规则，则将所述患者信息显示在所述眼部图像中，若基于所述信息输入框输入的患者信息不满足预设规则，则输出重新输入的提示信息。
更进一步地，若经对比确定所输入的各项患者信息均满足各自预先所设定的预设规则，则对各项患者信息进行保存操作；同时将各项患者信息显示在眼部图像中的预定位置，如主图像显示区的左上角。而若经对比确定输入的患者信息中存在任意一项不满足其预先所设定的预设规则，则对该项患者信息和其他满足预设规则的患者信息进行区别显示，并输出对该项不满足其预设规则的患者信息进行重新输入的提示信息。
可理解地，在将眼部图像显示在预设终端界面之后，该眼部图像可能因拍摄过程中的各项因素，而存在显示不清楚的问题，影响眼部图像的处理结果；而为了避免对处理结果的影响，本实施例设置有对眼部图像中的颜色进行增强的机制。具体地，将患者信息显示在眼部图像中的步骤之后包括：
步骤S70，当侦测到对图像增强的增强指令时，根据所述增强指令中的类型编码，确定所述图像增强的增强类型；
进一步地，本实施例中的图像增强包括三色道增强和预设算法增强两种类型，功能区针对该两种增强类型设置有不同的虚拟按键，且各虚拟按键携带有不同的类型编码。将对两个虚拟按键中的任意一个所进行的触发操作，作为对图像增强的增强指令进行侦测；当侦测到该增强指令时，则先读取其中所携带的类型编码，以根据该类型编码确定对眼部图像进行图像增强的增强类型。
步骤S80，当所述增强类型为第一预设类型时，读取所述增强指令中的颜色信息，并对所述眼部图像中与所述颜色信息对应的颜色进行增强；
更进一步地，将图像增强中的三色道增强类型作为第一预设类型，其中三色道为红、绿、蓝三种颜色通道，第一预设类型即为通过红、绿、蓝三种颜色通道进行增强的类型。该类型的图像增强所针对的是整个图像，当经类型编码确定图像增强的增强类型为第一预设类型时，图像增强的增强指令中携带有表征所需要增强颜色的颜色信息，先对增强指令中的颜色信息进行读取，以通过颜色信息确定所需要增强的颜色通道；进而对眼部图像中与该颜色信息所对应的颜色进行增强，即对眼部图像中颜色信息所表征的颜色通道进行增强，使得眼部图像的整个图像呈现与颜色信息对应的颜色。
步骤S90，当所述增强类型为第二预设类型时，则对所述眼部图像中的框选区域或预设默认区域内的图像基于所述增强指令进行增强。
进一步地，将图像增强中的预设算法增强作为第二预设类型，其中预设算法可以是AREDS算法，第二预设类型即为通过AREDS算法进行增强的类型。该类型的图像增强所针对的是眼部图像中的框选区域或预设默认区域，当经类型编码确定图像增强的增强类型为第二预设类型时，则对眼部图像进行检测，判断其中是否存在框选区域；若存在，则对该框选区域中的图像以及增强指令进行增强；若不存在，则先确定眼部图像中的预设默认区域，并对该预设默认区域中的图像依据增强指令进行增强。其中预设默认区域由预先设定的规则所确定，如预先设定将眼部图像的中心点为圆点，半径为某一值的圆形区域，作为默认区域。
此外为了便于对眼部图像的原图进行查看，功能区中还设置有对图像增强进行取消的虚拟按键，当检测到对该虚拟按键的触发指令时，则对眼部图像中的图像增强部分进行取消，显示眼部图像的原图，以对增强前后的眼部图像进行对比。
进一步地，在本申请眼部图像处理方法另一实施例中，所述眼部图像处理方法还包括步骤：
步骤q1，当接收到参数配置指令时，判断所述预设终端界面中是否存在所述眼部图像，若存在所述眼部图像，则输出对所述眼部图像进行关闭的提示信息；
更进一步地，工具窗口中还设置有用于对眼部图像的整体参数进行配置的虚拟按键，当有对参数进行配置的需求时，点击该虚拟按键，将该点击操作作为参数配置指令进行接收；其中支持配置的参数包括图像尺寸单位，各尺寸单位之间的转换等。当接收到该参数配置指令时，先检测预设终端界面的主图像显示区中是否显示有眼部图像，若显示有眼部图像，则输出将显示的眼部图像进行关闭的提示信息；以在将眼部图像关闭后，再对参数进行配置，避免配置的参数对显示的眼部图像产生影响。
步骤q2，若不存在所述眼部图像，则在所述预设终端界面中显示配置输入框，并根据基于所述配置输入框输入的配置信息进行参数配置。
而若经检测预设终端界面的主图像显示区中不存在显示的眼部图像，则在预设终端界面中弹出并显示配置输入框，以便于通过配置输入框输入各项参数进行配置。在接收到基于该配置输入框所输入的配置信息后，则依据配置信息对各项参数进行配置；如原图像尺寸单位为mm，而输入框所接收到的尺寸单位为μm，则将原来的图像尺寸单位mm修改为μm，实现对图像尺寸单位的配置。通过参数的配置，确保眼部图像中各参数的一致性，且可针对不同的处理需求，配置不同的参数，以使得处理的精度与处理需求相一致。
此外，请参照图2，本申请提供一种眼部图像处理装置，在本申请眼部图像处理装置第一实施例中，所述眼部图像处理装置包括：
检测模块10，用于当侦测到对眼部图像的检测指令时，调用预设算法对所述眼部图像进行检测，生成检测结果，其中所述检测结果包括所述眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常位置；
预测模块20，用于当侦测到对异常变化趋势的预测指令时，判断所述检测结果中是否存在风险相关因子，若存在所述风险相关因子，则根据预设的风险相关因子的数量与概率之间的对应关系，预测与所述风险相关因子对应的异常变化概率；
生成模块30，用于当检测到生成检测报告的触发指令时，则将所述检测结果、所述风险相关因子以及所述异常变化概率生成所述眼部图像的参考检测报告，完成对所述眼部图像的处理。
本实施例的眼部图像处理装置，通过对眼部图像进行图像处理，得到包含眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常位置等各项检测结果；同时根据检测结果中的风险相关因子，预测异常变化趋势的异常变化概率；进而将检测结果、风险相关因子以及异常变化概率生成用于供医务人员作为参考的参考检测报告；便于医务人员依据参考检测报告对眼部图像中的异常信息快速判断，避免单纯由医务人员人为对眼部图像中的异常信息进行判断，提高了医务人员的工作效率。
进一步地，在本申请眼部图像处理装置另一实施例中，所述检测模块还包括：
定位单元，用于调用第一预设算法，对所述眼部图像中的黄斑中心进行定位，确定黄斑中心点坐标；
调用单元，用于调用第二预设算法，对所述眼部图像中的视盘区域进行分割，确定视盘位置、视盘中心点坐标以及视盘直径；
确定单元，用于根据所述视盘中心点坐标，确定眼睛类型，并根据所述黄斑中心点坐标和所述视盘直径，确定黄斑位置；
检测单元，用于根据所述第二预设算法判断所述眼部图像中是否存在异常信息，若存在所述异常信息，则对所述异常信息的位置进行检测，确定异常位置，并将所述眼睛类型、黄斑位置、视盘位置和异常位置生成为所述检测结果。
进一步地，在本申请眼部图像处理装置另一实施例中，所述眼部图像处理装置还包括：
第一确定模块，用于当侦测到对所述检测结果的修改指令时，根据所述修改指令中的类型标识符，确定所述修改指令的修改类型；
修正模块，用于当所述修改类型为修正类型时，则根据所述修改指令对所述检测结果中的眼睛类型、黄斑位置、视盘位置和异常位置进行修正；
第一更新模块，用于当所述修改类型为新增类型时，则根据所述修改指令确定所述眼部图像中的色素异常信息，并将所述色素异常信息添加到所述检测结果中，以对所述检测结果进行更新。
进一步地，在本申请眼部图像处理装置另一实施例中，所述眼部图像处理装置还包括：
调整模块，用于当侦测到对所述风险相关因子的调整指令时，根据所述调整指令中的调整信息对所述风险相关因子进行调整，生成调整后的风险相关因子；
第二更新模块，用于根据调整后的所述风险相关因子，对所述异常变化概率进行更新。
进一步地，在本申请眼部图像处理装置另一实施例中，所述眼部图像处理装置还包括：
显示模块，用于当侦测到对眼部图像的打开指令时，将所述眼部图像打开并显示在预设终端界面中，并检测是否接收到与所述眼部图像对应的患者信息输入指令；
接收模块，用于若接收到与所述眼部图像对应的患者信息输入指令，则在所述预设终端界面中显示信息输入框，并检测基于所述信息输入框输入的患者信息是否满足预设规则；
输出模块，用于若基于所述信息输入框输入的患者信息满足预设规则，则将所述患者信息显示在所述眼部图像中，若基于所述信息输入框输入的患者信息不满足预设规则，则输出重新输入的提示信息。
进一步地，在本申请眼部图像处理装置另一实施例中，所述眼部图像处理装置还包括：
第二确定模块，用于当侦测到对图像增强的增强指令时，根据所述增强指令中的类型编码，确定所述图像增强的增强类型；
读取模块，用于当所述增强类型为第一预设类型时，读取所述增强指令中的颜色信息，并对所述眼部图像中与所述颜色信息对应的颜色进行增强；
增强模块，用于当所述增强类型为第二预设类型时，则对所述眼部图像中的框选区域或预设默认区域内的图像基于所述增强指令进行增强。
进一步地，在本申请眼部图像处理装置另一实施例中，所述眼部图像处理装置还包括：
判断模块，用于当接收到参数配置指令时，判断所述预设终端界面中是否存在所述眼部图像，若存在所述眼部图像，则输出对所述眼部图像进行关闭的提示信息；
配置模块，用于若不存在所述眼部图像，则在所述预设终端界面中显示配置输入框，并根据基于所述配置输入框输入的配置信息进行参数配置。
其中，上述眼部图像处理装置的各虚拟功能模块存储于图3所示眼部图像处理设备的存储器1005中，处理器1001执行眼部图像处理可读指令时，实现图2所示实施例中各个模块的功能。
参照图3，图3是本申请实施例方法涉及硬件运行环境的设备结构示意图。
本申请实施例眼部图像处理设备可是PC( personal computer，个人计算机 )，也可是智能手机、平板电脑、电子书阅读器、便携计算机等终端设备。
如图3所示，该眼部图像处理设备可以包括：处理器1001，例如CPU（Central Processing Unit，中央处理器），存储器1005，通信总线1002。其中，通信总线1002用于实现处理器1001和存储器1005之间的连接通信。存储器1005可以是高速RAM（random access memory，随机存取存储器），也可以是稳定的存储器（non-volatile memory），例如磁盘存储器。存储器1005可选的还可以是独立于前述处理器1001的存储装置。
可选地，该眼部图像处理设备还可以包括用户接口、网络接口、摄像头、RF（Radio Frequency，射频）电路，传感器、音频电路、WiFi（Wireless Fidelity，无线宽带）模块等等。用户接口可以包括显示屏（Display）、输入单元比如键盘（Keyboard），可选用户接口还可以包括标准的有线接口、无线接口。网络接口可选的可以包括标准的有线接口、无线接口（如WI-FI接口）。
本领域技术人员可以理解，图3中示出的眼部图像处理设备结构并不构成对眼部图像处理设备的限定，可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件布置。
如图3所示，作为一种计算机可读存储介质的存储器1005中可以包括操作系统、网络通信模块以及眼部图像处理可读指令。操作系统是管理和控制眼部图像处理设备硬件和软件资源的可读指令，支持眼部图像处理可读指令以及其它软件和/或可读指令的运行。网络通信模块用于实现存储器1005内部各组件之间的通信，以及与眼部图像处理设备中其它硬件和软件之间通信。
图3所示的眼部图像处理设备中，处理器1001用于执行存储器1005中存储的眼部图像处理可读指令，实现上述眼部图像处理方法各实施例的步骤。
本申请提供了一种计算机可读存储介质，所述计算机可读存储介质可为非易失性存储介质，所述计算机可读存储介质存储有一个或者一个以上可读指令，所述一个或者一个以上可读指令还可被一个或者一个以上的处理器执行以用于实现上述眼部图像处理方法各实施例中的步骤。
还需要说明的是，在本文中，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者装置不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者装置所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括该要素的过程、方法、物品或者装置中还存在另外的相同要素。
上述本申请实施例序号仅仅为了描述，不代表实施例的优劣。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在如上所述的一个计算机可读存储介质（如ROM/RAM、磁碟、光盘）中，包括若干指令用以使得一台终端设备（可是手机，计算机，服务器，或者网络设备等）执行本申请各个实施例所述的方法。
以上所述仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是在本申请的构思下，利用本申请说明书及附图内容所作的等效结构变换，或直接/间接运用在其他相关的技术领域均包括在本申请的专利保护范围内。

Claims (20)

  1. 一种眼部图像处理方法，其特征在于，所述眼部图像处理方法包括以下步骤：
    当侦测到对眼部图像的检测指令时，调用预设算法对所述眼部图像进行检测，生成检测结果，其中所述检测结果包括所述眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常信息的异常位置；
    当侦测到对异常变化趋势的预测指令时，判断所述检测结果中是否存在风险相关因子，若存在所述风险相关因子，则根据预设的风险相关因子的数量与概率之间的对应关系，预测与所述风险相关因子对应的异常变化概率，其中所述风险相关因子与所述异常信息中的玻璃膜疣和色素异常相关；
    当检测到生成检测报告的触发指令时，则将所述检测结果、所述风险相关因子以及所述异常变化概率生成所述眼部图像的参考检测报告，完成对所述眼部图像的处理。
  2. 如权利要求1所述的眼部图像处理方法，其特征在于，所述调用预设算法对所述眼部图像进行检测，生成检测结果的步骤包括：
    调用第一预设算法，对所述眼部图像中的黄斑中心进行定位，确定黄斑中心点坐标；
    调用第二预设算法，对所述眼部图像中的视盘区域进行分割，确定视盘位置、视盘中心点坐标以及视盘直径；
    根据所述视盘中心点坐标，确定眼睛类型，并根据所述黄斑中心点坐标和所述视盘直径，确定黄斑位置；
    根据所述第二预设算法判断所述眼部图像中是否存在异常信息，若存在所述异常信息，则对所述异常信息的位置进行检测，确定异常位置，并将所述眼睛类型、黄斑位置、视盘位置和异常位置生成为所述检测结果。
  3. 如权利要求1所述的眼部图像处理方法，其特征在于，所述生成检测结果的步骤之后包括：
    当侦测到对所述检测结果的修改指令时，根据所述修改指令中的类型标识符，确定所述修改指令的修改类型；
    当所述修改类型为修正类型时，则根据所述修改指令对所述检测结果中的眼睛类型、黄斑位置、视盘位置和异常位置进行修正；
    当所述修改类型为新增类型时，则根据所述修改指令确定所述眼部图像中的色素异常信息，并将所述色素异常信息添加到所述检测结果中，以对所述检测结果进行更新。
  4. 如权利要求1所述的眼部图像处理方法，其特征在于，所述预测与所述风险相关因子对应的异常变化概率的步骤之后包括：
    当侦测到对所述风险相关因子的调整指令时，根据所述调整指令中的调整信息对所述风险相关因子进行调整，生成调整后的风险相关因子；
    根据调整后的所述风险相关因子，对所述异常变化概率进行更新。
  5. 如权利要求1所述的眼部图像处理方法，其特征在于，所述当侦测到对眼部图像的检测指令时的步骤之前包括：
    当侦测到对眼部图像的打开指令时，将所述眼部图像打开并显示在预设终端界面中，并检测是否接收到与所述眼部图像对应的患者信息输入指令；
    若接收到与所述眼部图像对应的患者信息输入指令，则在所述预设终端界面中显示信息输入框，并检测基于所述信息输入框输入的患者信息是否满足预设规则；
    若基于所述信息输入框输入的患者信息满足预设规则，则将所述患者信息显示在所述眼部图像中，若基于所述信息输入框输入的患者信息不满足预设规则，则输出重新输入的提示信息。
  6. 如权利要求5所述的眼部图像处理方法，其特征在于，所述将所述患者信息显示在所述眼部图像中的步骤之后包括：
    当侦测到对图像增强的增强指令时，根据所述增强指令中的类型编码，确定所述图像增强的增强类型；
    当所述增强类型为第一预设类型时，读取所述增强指令中的颜色信息，并对所述眼部图像中与所述颜色信息对应的颜色进行增强；
    当所述增强类型为第二预设类型时，则对所述眼部图像中的框选区域或预设默认区域内的图像基于所述增强指令进行增强。
  7. 如权利要求6所述的眼部图像处理方法，其特征在于，所述眼部图像处理方法还包括步骤：
    当接收到参数配置指令时，判断所述预设终端界面中是否存在所述眼部图像，若存在所述眼部图像，则输出对所述眼部图像进行关闭的提示信息；
    若不存在所述眼部图像，则在所述预设终端界面中显示配置输入框，并根据基于所述配置输入框输入的配置信息进行参数配置。
  8. 一种眼部图像处理装置，其特征在于，所述眼部图像处理装置包括：
    检测模块，用于当侦测到对眼部图像的检测指令时，调用预设算法对所述眼部图像进行检测，生成检测结果，其中所述检测结果包括所述眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常位置；
    预测模块，用于当侦测到对异常变化趋势的预测指令时，判断所述检测结果中是否存在风险相关因子，若存在所述风险相关因子，则根据预设的风险相关因子的数量与概率之间的对应关系，预测与所述风险相关因子对应的异常变化概率；
    生成模块，用于当检测到生成检测报告的触发指令时，则将所述检测结果、所述风险相关因子以及所述异常变化概率生成所述眼部图像的参考检测报告，完成对所述眼部图像的处理。
  9. 如权利要求8所述的眼部图像处理装置，其特征在于，所述检测模块还包括：
    定位单元，用于调用第一预设算法，对所述眼部图像中的黄斑中心进行定位，确定黄斑中心点坐标；
    调用单元，用于调用第二预设算法，对所述眼部图像中的视盘区域进行分割，确定视盘位置、视盘中心点坐标以及视盘直径；
    确定单元，用于根据所述视盘中心点坐标，确定眼睛类型，并根据所述黄斑中心点坐标和所述视盘直径，确定黄斑位置；
    检测单元，用于根据所述第二预设算法判断所述眼部图像中是否存在异常信息，若存在所述异常信息，则对所述异常信息的位置进行检测，确定异常位置，并将所述眼睛类型、黄斑位置、视盘位置和异常位置生成为所述检测结果。
  10. 如权利要求8所述的眼部图像处理装置，其特征在于，所述眼部图像处理装置还包括：
    第一确定模块，用于当侦测到对所述检测结果的修改指令时，根据所述修改指令中的类型标识符，确定所述修改指令的修改类型；
    修正模块，用于当所述修改类型为修正类型时，则根据所述修改指令对所述检测结果中的眼睛类型、黄斑位置、视盘位置和异常位置进行修正；
    第一更新模块，用于当所述修改类型为新增类型时，则根据所述修改指令确定所述眼部图像中的色素异常信息，并将所述色素异常信息添加到所述检测结果中，以对所述检测结果进行更新。
  11. 如权利要求8所述的眼部图像处理装置，其特征在于，所述眼部图像处理装置还包括：
    调整模块，用于当侦测到对所述风险相关因子的调整指令时，根据所述调整指令中的调整信息对所述风险相关因子进行调整，生成调整后的风险相关因子；
    第二更新模块，用于根据调整后的所述风险相关因子，对所述异常变化概率进行更新。
  12. 如权利要求8所述的眼部图像处理装置，其特征在于，所述眼部图像处理装置还包括：
    显示模块，用于当侦测到对眼部图像的打开指令时，将所述眼部图像打开并显示在预设终端界面中，并检测是否接收到与所述眼部图像对应的患者信息输入指令；
    接收模块，用于若接收到与所述眼部图像对应的患者信息输入指令，则在所述预设终端界面中显示信息输入框，并检测基于所述信息输入框输入的患者信息是否满足预设规则；
    输出模块，用于若基于所述信息输入框输入的患者信息满足预设规则，则将所述患者信息显示在所述眼部图像中，若基于所述信息输入框输入的患者信息不满足预设规则，则输出重新输入的提示信息。
  13. 如权利要求12所述的眼部图像处理装置，其特征在于，所述眼部图像处理装置还包括：
    第二确定模块，用于当侦测到对图像增强的增强指令时，根据所述增强指令中的类型编码，确定所述图像增强的增强类型；
    读取模块，用于当所述增强类型为第一预设类型时，读取所述增强指令中的颜色信息，并对所述眼部图像中与所述颜色信息对应的颜色进行增强；
    增强模块，用于当所述增强类型为第二预设类型时，则对所述眼部图像中的框选区域或预设默认区域内的图像基于所述增强指令进行增强。
  14. 如权利要求13所述的眼部图像处理装置，其特征在于，所述眼部图像处理装置还包括：
    判断模块，用于当接收到参数配置指令时，判断所述预设终端界面中是否存在所述眼部图像，若存在所述眼部图像，则输出对所述眼部图像进行关闭的提示信息；
    配置模块，用于若不存在所述眼部图像，则在所述预设终端界面中显示配置输入框，并根据基于所述配置输入框输入的配置信息进行参数配置。
  15. 一种眼部图像处理设备，其特征在于，所述眼部图像处理设备包括：存储器、处理器、通信总线以及存储在所述存储器上的眼部图像处理可读指令；
    所述通信总线用于实现处理器和存储器之间的连接通信；
    所述处理器用于执行所述眼部图像处理可读指令，以实现以下步骤：
    当侦测到对眼部图像的检测指令时，调用预设算法对所述眼部图像进行检测，生成检测结果，其中所述检测结果包括所述眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常信息的异常位置；
    当侦测到对异常变化趋势的预测指令时，判断所述检测结果中是否存在风险相关因子，若存在所述风险相关因子，则根据预设的风险相关因子的数量与概率之间的对应关系，预测与所述风险相关因子对应的异常变化概率，其中所述风险相关因子与所述异常信息中的玻璃膜疣和色素异常相关；
    当检测到生成检测报告的触发指令时，则将所述检测结果、所述风险相关因子以及所述异常变化概率生成所述眼部图像的参考检测报告，完成对所述眼部图像的处理。
  16. 如权利要求15所述的眼部图像处理设备，其特征在于，所述调用预设算法对所述眼部图像进行检测，生成检测结果的步骤包括：
    调用第一预设算法，对所述眼部图像中的黄斑中心进行定位，确定黄斑中心点坐标；
    调用第二预设算法，对所述眼部图像中的视盘区域进行分割，确定视盘位置、视盘中心点坐标以及视盘直径；
    根据所述视盘中心点坐标，确定眼睛类型，并根据所述黄斑中心点坐标和所述视盘直径，确定黄斑位置；
    根据所述第二预设算法判断所述眼部图像中是否存在异常信息，若存在所述异常信息，则对所述异常信息的位置进行检测，确定异常位置，并将所述眼睛类型、黄斑位置、视盘位置和异常位置生成为所述检测结果。
  17. 如权利要求15所述的眼部图像处理设备，其特征在于，所述生成检测结果的步骤之后，所述处理器用于执行所述眼部图像处理可读指令，以实现以下步骤：
    当侦测到对所述检测结果的修改指令时，根据所述修改指令中的类型标识符，确定所述修改指令的修改类型；
    当所述修改类型为修正类型时，则根据所述修改指令对所述检测结果中的眼睛类型、黄斑位置、视盘位置和异常位置进行修正；
    当所述修改类型为新增类型时，则根据所述修改指令确定所述眼部图像中的色素异常信息，并将所述色素异常信息添加到所述检测结果中，以对所述检测结果进行更新。
  18. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质上存储有眼部图像处理可读指令，所述眼部图像处理可读指令被处理器执行时实现以下步骤：
    当侦测到对眼部图像的检测指令时，调用预设算法对所述眼部图像进行检测，生成检测结果，其中所述检测结果包括所述眼部图像中的眼睛类型、黄斑位置、视盘位置以及异常信息的异常位置；
    当侦测到对异常变化趋势的预测指令时，判断所述检测结果中是否存在风险相关因子，若存在所述风险相关因子，则根据预设的风险相关因子的数量与概率之间的对应关系，预测与所述风险相关因子对应的异常变化概率，其中所述风险相关因子与所述异常信息中的玻璃膜疣和色素异常相关；
    当检测到生成检测报告的触发指令时，则将所述检测结果、所述风险相关因子以及所述异常变化概率生成所述眼部图像的参考检测报告，完成对所述眼部图像的处理。
  19. 如权利要求18所述的计算机可读存储介质，其特征在于，所述调用预设算法对所述眼部图像进行检测，生成检测结果的步骤包括：
    调用第一预设算法，对所述眼部图像中的黄斑中心进行定位，确定黄斑中心点坐标；
    调用第二预设算法，对所述眼部图像中的视盘区域进行分割，确定视盘位置、视盘中心点坐标以及视盘直径；
    根据所述视盘中心点坐标，确定眼睛类型，并根据所述黄斑中心点坐标和所述视盘直径，确定黄斑位置；
    根据所述第二预设算法判断所述眼部图像中是否存在异常信息，若存在所述异常信息，则对所述异常信息的位置进行检测，确定异常位置，并将所述眼睛类型、黄斑位置、视盘位置和异常位置生成为所述检测结果。
  20. 如权利要求18所述的计算机可读存储介质，其特征在于，所述生成检测结果的步骤之后，所述眼部图像处理可读指令被处理器执行时实现以下步骤：
    当侦测到对所述检测结果的修改指令时，根据所述修改指令中的类型标识符，确定所述修改指令的修改类型；
    当所述修改类型为修正类型时，则根据所述修改指令对所述检测结果中的眼睛类型、黄斑位置、视盘位置和异常位置进行修正；
    当所述修改类型为新增类型时，则根据所述修改指令确定所述眼部图像中的色素异常信息，并将所述色素异常信息添加到所述检测结果中，以对所述检测结果进行更新。
PCT/CN2019/118548 2019-06-05 2019-11-14 眼部图像处理方法、装置、设备及计算机可读存储介质 Ceased WO2020215701A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
SG11202008484YA SG11202008484YA (en) 2019-06-05 2019-11-14 Eye image processing method, apparatus and device, and computer readable storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910494821.8A CN110335253B (zh) 2019-06-05 2019-06-05 眼部图像处理方法、装置、设备及计算机可读存储介质
CN201910494821.8 2019-06-05

Publications (1)

Publication Number Publication Date
WO2020215701A1 true WO2020215701A1 (zh) 2020-10-29

Family

ID=68140861

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/118548 Ceased WO2020215701A1 (zh) 2019-06-05 2019-11-14 眼部图像处理方法、装置、设备及计算机可读存储介质

Country Status (3)

Country Link
CN (1) CN110335253B (zh)
SG (1) SG11202008484YA (zh)
WO (1) WO2020215701A1 (zh)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110335253B (zh) * 2019-06-05 2023-12-15 平安科技（深圳）有限公司 眼部图像处理方法、装置、设备及计算机可读存储介质
CN111192679B (zh) * 2019-12-25 2024-04-19 上海联影智能医疗科技有限公司 一种影像数据异常的处理方法、装置及存储介质
CN113610746B (zh) * 2021-02-23 2026-02-06 腾讯科技（深圳）有限公司 一种图像处理方法、装置、计算机设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105786385A (zh) * 2016-02-29 2016-07-20 深圳康桥网络科技股份有限公司 基于人体图像检测疾病症状的方法及装置
CN108803994A (zh) * 2018-06-14 2018-11-13 四川和生视界医药技术开发有限公司 视网膜血管的管理方法及视网膜血管的管理装置
US20190164308A1 (en) * 2017-11-27 2019-05-30 Huntercraft Limited Intelligent shooting training management system
CN110335253A (zh) * 2019-06-05 2019-10-15 平安科技（深圳）有限公司 眼部图像处理方法、装置、设备及计算机可读存储介质

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2010071898A2 (en) * 2008-12-19 2010-06-24 The Johns Hopkins Univeristy A system and method for automated detection of age related macular degeneration and other retinal abnormalities
CN105513077B (zh) * 2015-12-11 2019-01-04 北京大恒图像视觉有限公司 一种用于糖尿病性视网膜病变筛查的系统
CN107169257A (zh) * 2016-03-07 2017-09-15 刘江 基于视觉影像的多眼疾“慧查”筛查方法和系统
US20180060535A1 (en) * 2016-08-31 2018-03-01 International Business Machines Corporation Updating probabilities of conditions based on annotations on medical images
CN109191451B (zh) * 2018-09-11 2020-10-23 百度在线网络技术（北京）有限公司 异常检测方法、装置、设备和介质
CN109584152A (zh) * 2018-11-30 2019-04-05 深圳市脸萌科技有限公司 图像处理方法、装置、电子设备及计算机可读存储介质

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105786385A (zh) * 2016-02-29 2016-07-20 深圳康桥网络科技股份有限公司 基于人体图像检测疾病症状的方法及装置
US20190164308A1 (en) * 2017-11-27 2019-05-30 Huntercraft Limited Intelligent shooting training management system
CN108803994A (zh) * 2018-06-14 2018-11-13 四川和生视界医药技术开发有限公司 视网膜血管的管理方法及视网膜血管的管理装置
CN110335253A (zh) * 2019-06-05 2019-10-15 平安科技（深圳）有限公司 眼部图像处理方法、装置、设备及计算机可读存储介质

Also Published As

Publication number Publication date
CN110335253B (zh) 2023-12-15
CN110335253A (zh) 2019-10-15
SG11202008484YA (en) 2020-11-27

Similar Documents

Publication Publication Date Title
WO2015167160A1 (en) Command displaying method and command displaying device
WO2020085666A1 (en) Electronic device and control method thereof
WO2020215701A1 (zh) 眼部图像处理方法、装置、设备及计算机可读存储介质
WO2020251217A1 (ko) 사용자 맞춤형 피부 진단 시스템 및 방법
WO2019024485A1 (zh) 数据共享方法、装置及计算机可读存储介质
WO2020015067A1 (zh) 数据采集方法、装置、设备及存储介质
WO2020073688A1 (zh) 预测网络设备异常的方法、装置、设备及存储介质
WO2020107765A1 (zh) 语句分析处理方法、装置、设备以及计算机可读存储介质
WO2020087981A1 (zh) 风控审核模型生成方法、装置、设备及可读存储介质
WO2018236058A1 (en) ELECTRONIC DEVICE FOR PROVIDING PROPERTY INFORMATION OF AN EXTERNAL LIGHT SOURCE FOR AN OBJECT OF INTEREST
WO2020107895A1 (zh) 慢病检查报告生成方法、装置、设备及存储介质
WO2020253048A1 (zh) 基于深度学习的影像识别方法、装置、设备及存储介质
WO2019174090A1 (zh) 截屏文件分享的控制方法、装置、设备和计算机存储介质
WO2016204466A1 (en) User authentication method and electronic device supporting the same
WO2021012508A1 (zh) Ai影像识别方法、装置、设备及存储介质
WO2016024835A1 (en) Apparatus and method for processing drag and drop
WO2020119115A1 (zh) 数据审核方法、装置、设备及存储介质
WO2020207038A1 (zh) 基于人脸识别的人数统计方法、装置、设备及存储介质
WO2019000967A1 (zh) 企业年金交易方法、装置及计算机可读存储介质
WO2009145596A2 (ko) 얼굴 분석 서비스 제공 방법 및 장치
EP3360322A1 (en) Electronic apparatus and controlling method thereof
WO2023282371A1 (ko) 인공지능 학습모델을 이용하여 다국어 자막 서비스를 제공하는 서버 및 방법과, 서버의 제어 방법
WO2023043032A1 (ko) 전자 장치 및 그 제어 방법
WO2018145597A1 (zh) 基于移动终端的屏幕补光拍照方法及系统、移动终端
WO2019045499A1 (en) IMAGE PROCESSING APPARATUS, IMAGE PROCESSING METHOD, AND COMPUTER-READABLE RECORDING MEDIUM

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19925595

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19925595

Country of ref document: EP

Kind code of ref document: A1