CN114983334A

CN114983334A - Slit lamp self-adjustment control method based on machine vision and slit lamp

Info

Publication number: CN114983334A
Application number: CN202210039862.XA
Authority: CN
Inventors: 刘奕志
Original assignee: Zhongshan Ophthalmic Center
Current assignee: Zhongshan Ophthalmic Center
Priority date: 2022-01-13
Filing date: 2022-01-13
Publication date: 2022-09-02

Abstract

The invention provides a slit lamp self-adjustment control method based on machine vision, which comprises the following steps of 1) collecting eye images by adopting a camera device; 2) comparing the collected eye image with standard data in a database to determine a focus area of the eye, and marking the focus area; 3) establishing a two-dimensional coordinate system with the pupil center as the origin of coordinates, and acquiring the coordinates of a focus area; 4) generating control data according to the coordinates of the focus area, controlling the slit lamp to obtain the optimal shooting angle, judging the type of the eye focus according to the focus image of the marked focus area, and generating data of a light filtering mode, the width of a projection light source and the irradiation angle required by the slit lamp when shooting the focus. The slit lamp eye disease detection system can quickly capture the region of the eye disease focus and identify the type of the disease focus, provides a basis for the operation of the slit lamp, and ensures the imaging effect of the slit lamp so as to facilitate the diagnosis of the disease by doctors.

Description

Slit lamp self-adjustment control method based on machine vision and slit lamp

Technical Field

The invention relates to a slit lamp self-adjusting control method based on machine vision and a slit lamp.

Background

When a patient suffering from an ophthalmic disease is treated in an ophthalmic manner, the patient needs to carry out comprehensive deep inspection on eyes to obtain accurate analysis, and the patient is treated in a manner which best meets the requirements of the patient, in the eye detection, a slit lamp microscope is most commonly used, and consists of an illumination system and a binocular microscope, so that superficial lesions can be observed clearly, focus and light source width can be adjusted to form an optical section, and lesions of deep tissues can also be clearly shown, but in the prior art, the following problems exist:

1) the doctor cannot be assisted in determining the focus position and adjusting the microscope to enable the slit lamp to be aligned with the focus position;

2) the slit lamp parameters can not be adjusted by judging the type of a focus according to the focus area image and combining the focus image.

Disclosure of Invention

Aiming at the defects of the prior art, the invention provides a slit lamp self-adjustment control method based on machine vision and a slit lamp, which can quickly capture an eye focus area and identify the focus category, provide a basis for the operation of the slit lamp, and ensure the imaging effect of the slit lamp so as to facilitate the diagnosis of diseases by doctors.

In order to achieve the above object, the present invention provides a slit lamp self-adjustment control method based on machine vision, which comprises the following steps:

1) collecting eye images by adopting a camera device;

2) comparing the collected eye image with standard data in a database to determine a focus area of the eye, and marking the focus area;

3) establishing a two-dimensional coordinate system with the pupil center as the origin of coordinates, and acquiring the coordinates of a focus area;

4) generating control data according to coordinates of a focus area, controlling a slit lamp to obtain an optimal shooting angle, judging the type of an eye focus according to a focus image of the marked focus area, and generating data of a light filtering mode, the width of a projection light source and an irradiation angle required by the slit lamp when the slit lamp shoots the focus;

the process of judging the eye focus type of the focus image of the marked focus area comprises the following steps:

the method comprises the steps of selecting eye images of multiple types of eye diseases, extracting specific information on focus images in each focus area through an algorithm to serve as identification information, further generating and storing a judgment matrix P (P1, P2 & cndot) corresponding to the focus images of the eye diseases, wherein P1 represents one type of disease identification information, P2 represents second type of disease identification information, and Pn represents nth type of disease identification information, and the focus images of the marked focus areas are compared with the judgment matrix one by one to determine the eye disease types corresponding to the focus images of the marked focus areas.

As another specific embodiment of the present invention, the eye image acquired in step 1) is subjected to data extraction to divide the acquired eye image into an eye corner image, an eye pupil image, an eye eyelid image, and an eye lid position image.

As another specific embodiment of the present invention, in step 2), a binarization process is performed on the acquired eye image, a contrast threshold is preset, and a region with a gray level variation exceeding the preset contrast threshold is marked as a lesion region.

As another specific embodiment of the present invention, the process of generating the manipulation data according to the coordinates of the lesion area in step 4) is:

establishing a focus area two-dimensional coordinate set f (x, y) according to the acquired coordinates of the focus area;

establishing a coordinate set adjustment matrix Si (Si1, Si2, Si3, Si4, i is 1, 2. cndot. n) according to the relation between the position and the shooting angle of the microscope and the lesion area, wherein Si1 represents a coordinate area, Si2 represents an X axial coordinate of microscope adjustment, Si3 represents a Y axial coordinate of microscope adjustment, and Si4 represents a microscope adjustment angle;

and judging whether the region represented by the two-dimensional coordinate set f (X, Y) of the lesion region exists in a coordinate region Si1, if so, calling a microscope to adjust an X-axis coordinate Si2, an Y-axis coordinate Si3 and a microscope adjusting angle Si4, and prompting that the position of the microscope should be adjusted to the X-axis coordinate Si1 and the Y-axis coordinate Si2 and the angle should be adjusted to the microscope adjusting angle Si 4.

As another embodiment of the present invention, the specific information on the lesion image in the lesion area in the step 4) includes a contour feature and a chromaticity feature.

As another embodiment of the present invention, when determining the type of the eye lesion based on the lesion image of the marked lesion area in step 4), the lesion image is divided into a lesion edge area and a lesion center area, the contour characteristics and the chromaticity characteristics of the lesion edge area and the lesion center area are compared with the i (i ═ 1, 2 · · n) th disease identification information in the determination matrix P, and the matching area and the non-matching area are recorded to determine the matching rate Sz of the lesion image:

wherein S1 represents the contour feature matching rate of the focus edge region, S2 represents the contour feature matching rate of the focus middle region, S3 represents the chroma matching rate of the focus edge region, S4 represents the chroma matching rate of the focus center region, alpha 1 is a weight parameter of S1 to Sz, alpha 2 is a weight parameter of S2 to Sz, alpha 3 is a weight parameter of S3 to Sz, and alpha 4 is a weight parameter of S4 to Sz;

comparing the coincidence rate Sz of the focus image with the set coincidence rate parameter S of the eye disease image:

when Sz is larger than S, judging that the eye disease image and the eye disease represented by the i-th class disease identification information are the same eye disease;

and when the Sz is less than or equal to S, judging that the eye disease image and the eye disease represented by the i-th class disease identification information are different eye diseases.

As another specific embodiment of the present invention, the process of generating the data of the filtering mode, the width of the projection light source and the irradiation angle required when the slit lamp shoots the lesion in step 4) is as follows:

a control matrix Ki (Ki1, Ki2, Ki3, Ki4, i ═ 1, 2 · · n) is established based on the relation between the type of the eye focus and an initial filtering mode, the width of an initial projection light source and an initial irradiation angle, wherein Ki1 represents the type of the eye focus, Ki2 represents the initial filtering mode, Ki3 represents the width of the initial projection light source, Ki4 represents the initial irradiation angle, and after the type Ki1 of the eye focus is determined, the initial filtering mode, the width of the initial projection light source and the initial irradiation angle can be determined.

As another specific embodiment of the present invention, the method further comprises the step 5) of performing the eye image acquisition again, and correcting the width and the illumination angle of the projection light source based on the definition and the chromaticity of the eye image acquired again.

As another specific embodiment of the present invention, the camera device in step 1) includes a wide-angle camera and an infrared camera.

The invention also provides a slit lamp for realizing the slit lamp self-adjustment control method based on the machine vision.

The invention has the following beneficial effects:

the image information is shot by the camera device, so that the focus area of the eyes can be conveniently captured, meanwhile, the moving coordinate and the irradiation angle of the microscope when the microscope is aligned with the focus area can be calculated, a doctor can conveniently and quickly check the focus area, the disease type can be judged according to the preset algorithm, the filtering mode of the slit lamp, the width of the projection light source and the irradiation angle can be correspondingly adjusted according to the disease type, high image quality can be obtained, and the doctor can conveniently and quickly align and correct the slit lamp and align the focus area.

In addition, after the focus area is identified, the operating data of the slit lamp is correspondingly adjusted according to the type of the focus, so that the imaging effect of the slit lamp is ensured, and a doctor can conveniently diagnose diseases.

The present invention will be described in further detail with reference to the accompanying drawings.

Drawings

FIG. 1 is a schematic flow diagram of the present invention;

FIG. 2 is a schematic diagram of a slit lamp that may be used in the present invention;

FIG. 3 is a photograph taken with a wide angle camera according to the present invention;

FIG. 4 is a photograph taken by an infrared camera in accordance with the present invention;

fig. 5 is a photograph showing a lesion area according to the present invention.

Detailed Description

In order that the above objects, features and advantages of the present invention can be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflict.

In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those specifically described herein, and therefore the scope of the present invention is not limited to the specific embodiments disclosed below.

Example 1

The embodiment provides a slit lamp self-adjustment control method based on machine vision, as shown in fig. 1-2, the slit lamp is provided with an illumination system and a microscope system, and the embodiment does not limit the specific structure of the slit lamp, and aims to control multiple parameters (an illumination angle, a light source width, a filtering mode and the like) of the slit lamp to obtain the optimal use effect.

Which comprises the following steps:

1) collecting eye images by adopting a camera device;

wherein camera device includes wide-angle camera and infrared camera, gathers eye image through wide-angle camera and infrared camera, as shown in fig. 3-4 (the color that will show in fig. 3 is handled by grey scale), wherein wide-angle camera and infrared camera simultaneous working, the wide-angle camera receives the visible light, judges the pathological change that easily observes under the visible light, and infrared camera receives the infrared light, judges the pathological change that easily observes under the infrared light.

2) Comparing the collected eye image with standard data in a database to determine an eye focus area, and marking the focus area;

wherein a standard database is established based on the healthy eye database, and an eye lesion area is obtained by image contrast analysis, as shown in fig. 5, wherein the framed area is an observed lesion area of a suspected lesion.

Specifically, after the acquired eye image is acquired, binarization processing is performed on the acquired eye image, and a lesion area is determined by presetting a contrast threshold value in a processor, specifically, an area with gray level variation exceeding the preset contrast threshold value is marked as the lesion area.

as shown in fig. 5, the (X, Y) coordinates of the lesion area may be obtained according to the position of the suspected lesion in the photograph, and the (Z) coordinates of the lesion area may be determined according to the position difference between the images captured by the wide-angle camera and the infrared camera on the same portion.

the process of generating the control data according to the coordinates of the lesion area comprises the following steps:

That is, after obtaining the coordinates of the lesion area, the X-axis coordinates, the Y-axis coordinates and the adjustment angle data of the microscope system, which correspond to the coordinates of the lesion area one by one, may be obtained through the related data (i.e., the matrix Si) preset in the processor.

Wherein the process of determining the type of the ocular lesion of the lesion image of the marked lesion region comprises:

the method comprises the steps of selecting eye images of multiple types of eye diseases, extracting specific information on focus images in each focus area through an Ai algorithm to serve as identification information, further generating and storing a judgment matrix P (P1, P2 & cndot) corresponding to the focus images of the eye diseases, wherein P1 represents one type of disease identification information, P2 represents second type of disease identification information, and Pn represents nth type of disease identification information, and the focus images of the marked focus areas are compared with the judgment matrix one by one to determine the eye disease types corresponding to the focus images of the marked focus areas.

Specifically, the process of generating the decision matrix P may be:

firstly, selecting an eye focus picture of a type of eye diseases, extracting characteristic information of a focus region in the eye focus picture through an Ai algorithm, generating and storing identification information capable of judging the type of the diseases;

repeating the above process to generate identification information of multiple types of focus regions, and further generating a determination matrix P corresponding to focus images of multiple eye diseases.

The process of generating the data of the filtering mode, the width of the projection light source and the irradiation angle required by the slit lamp to shoot the focus is as follows:

For example, when the disease is determined to be a type 1 disease, the processor can determine that the initial filtering mode is K12, the initial projection light source width is K13, and the initial projection angle is K14;

one specific process of the Ai algorithm is as follows:

integrating the Transformer and the Unet into a model to realize the segmentation of the image at the anterior segment of the eye and the identification of the focus; wherein, the encoder module of the U-net uses a vision transform to encode; a Transformer encodes a marked image block from a Convolutional Neural Network (CNN) feature map into an input sequence for extracting a global context; the decoder upsamples the encoded features and then combines them with the high resolution CNN feature map to achieve accurate positioning.

In the embodiment, the specific information on the lesion image in the lesion area is a contour feature and a chromaticity feature, and for convenience of image comparison processing, the eye image acquired in step 1) is subjected to data extraction and is divided into an eye corner image, an eye pupil image, an eye eyelid image and an eyelid position image, and the image information of the lesion area in each type of image is subjected to data processing through the controller, so that the directivity in the image processing process is stronger.

Further, when determining the type of the eye lesion according to the lesion image of the marked lesion area in step 4), dividing the lesion image into a lesion edge area and a lesion center area, respectively comparing the contour characteristics and the chromaticity characteristics of the lesion edge area and the lesion center area with the i (i ═ 1, 2 · · n) th class disease identification information in the determination matrix P, and recording an anastomotic area and a non-anastomotic area to determine the anastomotic rate Sz of the lesion image:

when Sz > S, determining that the eye disease image is the same eye disease as the eye disease represented by the i-th type of disease identification information;

In this embodiment, after determining the initial filtering manner, the initial width of the projection light source, and the initial illumination angle, the method further includes:

step 5): collecting the eye image again, and correcting the width and the illumination angle of the projection light source based on the definition and the chromaticity of the eye image collected again, wherein the specific correction process comprises the following steps:

establishing a brightness matrix C0 and a light source width adjusting parameter matrix D0;

for the luminance matrices C0, C0(C1, C2, C3, C4), where C1 is a first preset luminance parameter, C2 is a second preset luminance parameter, C3 is a third preset luminance parameter, and C4 is a fourth preset luminance parameter, the preset luminance parameters are sequentially increased;

for the light source width adjustment parameter matrix D0, D0(D1, D2, D3, D4), wherein D1 is a first preset light source width adjustment parameter, D2 is a second preset light source width adjustment parameter, D3 is a third preset light source width adjustment parameter, and D4 is a fourth preset light source width adjustment parameter;

detecting the brightness Ci of the ith area (the ith area can be any area, and the operation really most indicates a focus area or a suspected focus area) of the eye image by a processor and comparing the Ci with parameters in a C0 matrix:

when Ci is less than or equal to C1, judging that the i-th area is insufficient in brightness, and selecting D1 from the D0 matrix as a light source width adjusting parameter;

when Ci is more than C1 and less than or equal to C2, judging that the i-th area is insufficient in brightness, and selecting D2 from a D0 matrix as a light source width adjusting parameter;

when C2 < Ci is less than or equal to C3, judging that the i-th area brightness is in a standard state;

when Ci is more than C3 and less than or equal to C4, judging that the i-th area has overhigh brightness, and selecting D3 from a D0 matrix as a light source width adjusting parameter;

when Ci is larger than C4, judging that the ith area of the brightness is too high, and selecting D4 from the D0 matrix as a light source width adjusting parameter;

when the luminance of the ith area is judged not to be in the standard state (the standard state is the state that C2 is larger than Ci and is less than or equal to C3), the width of the light source of the ith area is adjusted to Ci';

when the width of the light source in the ith area is judged to be insufficient, Ci ═ Ci + (C3-Ci). times.dj, j ═ 1, 2;

when the light source width of the i-th area is judged to be too narrow and too high, Ci ═ Ci- (Ci-C2). times Dr, r ═ 3, 4;

when the width of the light source of the ith area is adjusted to Ci ', comparing Ci' with the parameters in the C0 matrix:

when C2 is less than Ci' and less than C3, judging that the luminance of the ith area is in a standard state;

when Ci 'is not in C2-C3, the operation is repeated until C2 is less than Ci' and less than or equal to C3.

Although the present invention has been described with reference to the preferred embodiments, it is not intended to limit the scope of the invention. It will be appreciated by those skilled in the art that changes may be made without departing from the scope of the invention, and it is intended that all matter contained in the above description or shown in the accompanying drawings shall be interpreted as illustrative and not in a limiting sense.

Claims

1. A slit lamp self-adjustment control method based on machine vision comprises the following steps:

1) collecting eye images by adopting a camera device;

2. The machine vision-based slit lamp self-adjustment control method as claimed in claim 1, wherein the eye image acquired in step 1) is subjected to data extraction to divide the acquired eye image into an eye corner image, an eye pupil image, an eye eyelid image and an eye eyelid image.

3. The slit lamp self-adjustment control method based on machine vision as claimed in claim 1, wherein in step 2), the collected eye image is binarized, a contrast threshold value is preset, and a region with a gray scale change exceeding the preset contrast threshold value is marked as a lesion region.

4. The machine vision-based slit lamp self-adjustment control method as claimed in claim 1, wherein the process of generating the manipulation data according to the coordinates of the lesion area in step 4) is as follows:

establishing a coordinate set adjustment matrix Si (Si1, Si2, Si3, Si4, i is 1, 2. cndot. n) according to the relation between the microscope position and the shooting angle and the lesion area, wherein Si1 represents a coordinate area, Si2 represents a microscope adjustment X axial coordinate, Si3 represents a microscope adjustment Y axial coordinate, and Si4 represents a microscope adjustment angle;

5. The machine vision-based slit lamp self-adjustment control method as claimed in claim 1, wherein the specific information on the lesion image in the lesion area in step 4) comprises contour features and chromaticity features.

6. The slit lamp self-adjustment control method based on machine vision as claimed in claim 5, wherein, when determining the type of the eye focus according to the focus image of the marked focus area in step 4), the focus image is divided into a focus edge area and a focus center area, the contour characteristics and the chromaticity characteristics of the focus edge area and the focus center area are respectively compared with the i (i ═ 1, 2 · · n) class disease identification information in the determination matrix P, and the anastomotic area and the non-anastomotic area are recorded to determine the anastomotic ratio Sz of the focus image:

7. The slit lamp self-adjustment control method based on machine vision as claimed in claim 6, wherein the process of generating the data of the filtering mode, the width of the projection light source and the irradiation angle required by the slit lamp to shoot the lesion in step 4) is as follows:

8. The slit lamp self-adjustment control method based on machine vision as claimed in claim 7, further comprising step 5) of re-acquiring the eye image, and correcting the width and the illumination angle of the projection light source based on the definition and chromaticity of the re-acquired eye image.

9. The machine vision-based slit lamp self-adjustment control method as claimed in claim 1, wherein the camera in step 1) comprises a wide-angle camera and an infrared camera.

10. A slit lamp implementing the machine vision based slit lamp self-adjustment control method of any one of claims 1-9.