A software and hardware cooperative cataract image standardization auxiliary acquisition method

Abstract

The application provides a software and hardware cooperative cataract image standardization auxiliary acquisition method and relates to the technical field of image processing. The application uses computer vision to analyze continuous frame displacement, predicts human physiological tremor trough to replace the traditional gyroscope hardware, automatically exposes at the best relative stationary point, solves the difficult single-hand operation and the blur problem caused by hand jitter. Through the closed-loop mechanism of "targeted offset calculation-UI instruction feedback-posture artificial correction", the application forces and standardizes the operation track of non-professionals, not only prevents jitter, but also completely eliminates the problems of shooting position offset and missing key lesions from the source, greatly improving the usability of screening data.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, specifically a standardized assisted acquisition method for cataract images using a hardware and software co-processing approach. Background Technology

[0002] Slit-lamp microscopy is an essential step in cataract diagnosis and preoperative evaluation. In grassroots cataract screening settings, due to a lack of specialized ophthalmologists, data collection is usually performed by primary care medical staff using portable slit lamps. The presence of cataracts is determined by observing the cloudiness of the lens under extremely narrow light bands.

[0003] Chinese patent application CN115530751A discloses a cataract screening system and its usage method. This method involves an operator manually taking photos using a mobile phone through the eyepiece of a handheld slit lamp, and then uploading the image containing the slit lamp light band to a server for processing. However, in practical grassroots cataract screening applications, this method has three main problems: First, existing portable devices have scattered control buttons and limited rotation adjustment mechanisms, making it impossible to simultaneously adjust the illumination angle, focus, and shoot with one hand, easily causing image shake. Second, relying entirely on the operator manually pointing the phone lens at the slit lamp eyepiece and manually pressing the shutter, the operator's unavoidable low-frequency breathing fluctuations and high-frequency physiological muscle tremors easily lead to blurred images and unclear slit lamp light bands. Third, this method lacks standardized guidance on the shooting path, making it easy for non-professionals at the grassroots level to miss key lens lesion areas or cause angle deviations. To improve image quality, existing technologies usually require adding expensive hardware image stabilization sensors or complex mechanical slip rings, which increases the cost and size of grassroots screening equipment. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to propose a hardware-software co-operated standardized assisted acquisition method for cataract images, comprising: Step 1: Set up a standardized timeline for cataract screening, which includes the auricular bulbar conjunctiva, the lens core area, and the nasal bulbar conjunctiva in sequence. Step 2: Real-time acquisition of eye video using a multi-degree-of-freedom intelligent handheld slit lamp device; Step 3: Process the acquired eye video. If the processing result meets the preset conditions for the auricular bulbar conjunctiva, acquire the auricular bulbar conjunctiva image using a multi-degree-of-freedom intelligent handheld slit lamp device. Step 4: Following Step 2 and Step 3, acquire images of the lens core area and the nasal bulbar conjunctiva in sequence.

[0005] Optionally, step 3 specifically includes: Step 3.1: Generate and output posture correction warning information based on each frame of the eye image in the eye video until the posture correction is passed, then proceed to step 3.2; Step 3.2: Calculate the instantaneous displacement magnitude based on two adjacent frames of eye images in the eye video. The instantaneous displacement modulus is calculated using a second-order difference algorithm. The first derivative of is used to obtain the acceleration. Instantaneous displacement modulus When the motion acceleration is less than the preset subpixel level stability threshold and the motion acceleration is less than the preset threshold, it indicates that the preset conditions for the auricular bulbar conjunctiva are met, and a prompt message for acquiring auricular bulbar conjunctiva images is output. Auricular bulbar conjunctiva images are acquired through a multi-degree-of-freedom intelligent handheld slit lamp device.

[0006] Optionally, step 3.1 specifically includes: For each frame of eye image in the eye video, the eye image is identified by an image recognition algorithm to obtain the region of interest, and the Euclidean distance, deflection angle and deflection direction between the center coordinates of the region of interest and the preset standard anatomical target frame of the auricular bulbar conjunctiva are calculated. Determine if the deflection angle is greater than the fault tolerance threshold. If it is, output the Euclidean distance, deflection angle, and deflection direction as warning information. If not, it indicates that the attitude correction has passed, output a prompt message indicating that the attitude correction has passed, and execute step 3.2.

[0007] Optionally, in step 3.2, the instantaneous displacement magnitude is calculated based on two adjacent frames of eye images in the eye video. ,include: Based on multiple consecutive instantaneous displacement moduli, the motion displacement of the region of interest in two adjacent eye images is calculated using the optical flow method. and ,according to and Calculate the instantaneous displacement modulus Specifically, this is achieved through the following formula: .

[0008] Optionally, step 4 specifically includes: Based on the preset standard anatomical target frame of the lens core area, repeat steps 3.1 to 3.2 to acquire images of the lens core area. Then, based on the preset standard anatomical target frame of the nasal bulbar conjunctiva, repeat steps 3.1 to 3.2 to acquire images of the nasal bulbar conjunctiva.

[0009] The beneficial effects of adopting the above technical solution are as follows: Zero-cost intelligent image stabilization driven by pure algorithms: It uses computer vision to analyze the displacement of continuous frames and predict the trough of human physiological tremor to replace the traditional gyroscope hardware. It automatically exposes at the best relative stationary point, solving the pain points of difficulty in one-handed operation and blur caused by hand tremors.

[0010] A human-computer interaction method for preventing missed images based on dynamic UI constraints: This innovative approach deeply integrates the medical screening path with the "step prompts" and "angle prompts" of the UI interface. Through a closed-loop mechanism of "target offset calculation - UI command feedback - posture correction," the operation trajectory of non-professionals is strictly standardized, which not only prevents image distortion but also eliminates the problems of shooting position deviation and missed images of key lesions from the source, greatly improving the usability of screening data. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating a hardware-software co-operated standardized assisted acquisition method for cataract images in an embodiment of the present invention. Figure 2 This is a flowchart illustrating another hardware-software co-operated cataract image standardization assisted acquisition method in an embodiment of the present invention. Figure 3 This is a schematic diagram of the overall appearance and magnetic disassembly structure of the multi-degree-of-freedom intelligent handheld slit lamp device in an embodiment of the present invention. Figure 4 This is a schematic diagram of the intelligent grip assembly and top slot distribution structure in an embodiment of the present invention; Among them, 1-intelligent grip assembly, 2-camera switch button, 3-one-click photo button, 4-light source switch button, 5-magnifying glass, 6-fan-shaped limiting groove, 7-L-shaped limiting rotating column, 8-brightness adjustment sliding rheostat. Detailed Implementation

[0012] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0013] To address the problems existing in the prior art, this invention provides a standardized assisted acquisition method for cataract images using a hardware and software collaborative approach, combining... Figure 1 and Figure 2 This may include the following steps: Step 1: Set up a standardized timeline for cataract screening, which includes the auricular bulbar conjunctiva, the lens core area, and the nasal bulbar conjunctiva in sequence. The present invention also includes a display interface that displays a step prompt bar, i.e., a standardized timing path.

[0014] Step 2: Real-time acquisition of eye video using a multi-degree-of-freedom intelligent handheld slit lamp device; Combination Figure 3 and Figure 4 In this configuration, 1 is the intelligent grip assembly, 2 is the camera switch button, 3 is the one-click photo button, 4 is the light source switch button, 5 is the magnifying glass, 6 is the fan-shaped limiting groove, 7 is the L-shaped limiting rotating column, and 8 is the brightness adjustment sliding rheostat. Therefore, the main structure of the multi-degree-of-freedom intelligent handheld slit lamp device includes a light source and its fixing module, a magnifying glass assembly, a camera and its components, and the intelligent grip assembly at the bottom. Single-handed full-function operation: The intelligent grip assembly serves as the device's supporting structure and control base, housing all control circuitry and the power management system. Its surface and sides integrate a camera switch button, a one-click photo button, a light source switch button, and a sliding rheostat for adjusting light source brightness. This layout allows the operator to perform all device functions with one hand, freeing up the other hand to stabilize the patient's head and reduce physiological shaking.

[0015] Magnetic Dual-Mode Switching: The connection between the camera box and the magnifying glass utilizes a Pogopin magnetic connector. Users can quickly switch between "direct vision mode" and "digital shooting mode" through simple adsorption and detachment, while the Pogopin spring pins ensure electrical continuity. This invention employs Pogopin magnetic connection technology and a fan-shaped limiting groove structure, enabling the device to switch between general screening and in-depth triage modes in seconds. Furthermore, it achieves omnidirectional, decoupled adjustment of the illumination beam and the field of view, alleviating the drawbacks of limited adjustment angles in traditional equipment.

[0016] Multi-degree-of-freedom decoupling structure: The L-shaped limiting rotating column at the bottom of the connecting arm cooperates with the fan-shaped limiting groove at the top of the intelligent grip assembly to realize the horizontal deflection degree of freedom of the head assembly; the connecting collar on the light source side and the connecting collar on the main body side are connected in a torsionable manner, so that the relative tilt angle of the light source optical axis can be continuously adjusted within the range of 30 degrees to 60 degrees.

[0017] Step 3: Process the acquired eye video. If the processing result meets the preset conditions for the auricular bulbar conjunctiva, acquire the auricular bulbar conjunctiva image using a multi-degree-of-freedom intelligent handheld slit lamp device. Step 3.1: Generate and output posture correction warning information based on each frame of the eye image in the eye video until the posture correction is passed, then proceed to step 3.2; For each frame of an eye image in an eye video, an image recognition algorithm is used to identify the region of interest (ROI). This invention dynamically anchors specific eye anatomical structures as ROIs. For example, when the prompt bar indicates "lens core area," the algorithm automatically anchors the pupil edge and corneal reflective points as reference features.

[0018] Calculate the Euclidean distance, deflection angle, and deflection direction between the center coordinates of the region of interest and the pre-defined standard anatomical target frame of the auricular bulbar conjunctiva; Determine if the deflection angle is greater than the fault tolerance threshold. If so, output the Euclidean distance, deflection angle, and deflection direction as warning information. Specifically, output the warning information to the angle prompt bar in the upper left corner of the display interface. At the same time, the angle prompt bar lights up with a high-contrast warning color (such as red) and displays "Angle: Please correct" and the specific correction direction (such as "fine adjustment to the left").

[0019] If not, the posture correction is successful, and a prompt message indicating successful posture correction is output. At this time, the angle prompt bar changes to a successful state (such as "Angle: Detecting" or disappears). Only then does the system unlock the subsequent image stabilization and capture logic. This ensures that frontline personnel strictly follow medical standards to collect accurate and complete images, and then proceeds to step 3.2; Step 3.2: Calculate the instantaneous displacement magnitude based on two adjacent frames of eye images in the eye video. Specifically, it includes: The motion displacement of the region of interest in two adjacent eye images was calculated using the Lucas-Kanade optical flow method. and ,according to and Calculate the instantaneous displacement modulus Specifically, this is achieved through the following formula: ; The instantaneous displacement modulus is calculated using a second-order difference algorithm based on multiple consecutive instantaneous displacement moduli. The first derivative is used to obtain the motion acceleration; specifically, the displacement magnitude of consecutive frames is stored in a sliding window, and the rate of change of discrete displacement data is analyzed using a second-order difference algorithm.

[0020] The physiological tremors of a person holding a device with one hand are similar to damped simple harmonic motion. The algorithm accurately predicts the "trough" (i.e., the relative rest point) of the hand tremor cycle by capturing feature points where the instantaneous acceleration passes zero and the velocity modulus is extremely small. At this point, the instantaneous displacement modulus is less than a preset subpixel level stability threshold. .

[0021] This invention designs a strict "AND" logic threshold: the system automatically interrupts the video stream buffer and extracts the current high-fidelity keyframe without motion blur only when both "spatial dimension: angle prompt bar verification passes (ROI is in the effective target area)" and "time dimension: tremor algorithm prediction reaches the physiological wave trough" are satisfied at the microsecond level.

[0022] Specifically, in the instantaneous displacement modulus When the motion acceleration is less than the preset subpixel level stability threshold and the preset threshold is less than the preset threshold, it indicates that the preset conditions for the auricular bulbar conjunctiva are met, and a prompt message for acquiring auricular bulbar conjunctiva images is output. Auricular bulbar conjunctiva images are acquired through a multi-degree-of-freedom intelligent handheld slit lamp device. Step 4: Following steps 2 and 3, acquire images of the lens core area and the nasal bulbar conjunctiva in sequence; Specifically, based on the preset standard anatomical target frame of the lens core area, repeat steps 3.1 to 3.2 to acquire images of the lens core area, and then based on the preset standard anatomical target frame of the nasal bulbar conjunctiva, repeat steps 3.1 to 3.2 to acquire images of the nasal bulbar conjunctiva. The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A standardized assisted acquisition method for cataract images using hardware and software collaboration, characterized in that, include: Step 1: Set up a standardized timeline for cataract screening, which includes the auricular bulbar conjunctiva, the lens core area, and the nasal bulbar conjunctiva in sequence. Step 2: Real-time acquisition of eye video using a multi-degree-of-freedom intelligent handheld slit lamp device; Step 3: Process the acquired eye video. If the processing result meets the preset conditions for the auricular bulbar conjunctiva, acquire the auricular bulbar conjunctiva image using a multi-degree-of-freedom intelligent handheld slit lamp device. Step 4: Following Step 2 and Step 3, acquire images of the lens core area and the nasal bulbar conjunctiva in sequence.

2. The hardware-software co-processing cataract image standardization assisted acquisition method according to claim 1, characterized in that, Step 3 specifically includes: Step 3.1: Generate and output posture correction warning information based on each frame of the eye image in the eye video until the posture correction is passed, then proceed to step 3.2; Step 3.2: Calculate the instantaneous displacement magnitude based on two adjacent frames of eye images in the eye video. The instantaneous displacement modulus is calculated using a second-order difference algorithm. The first derivative of is used to obtain the acceleration. Instantaneous displacement modulus When the motion acceleration is less than the preset subpixel level stability threshold and the motion acceleration is less than the preset threshold, it indicates that the preset conditions for the auricular bulbar conjunctiva are met, and a prompt message for acquiring auricular bulbar conjunctiva images is output. Auricular bulbar conjunctiva images are acquired through a multi-degree-of-freedom intelligent handheld slit lamp device.

3. The hardware-software co-processing standardized assisted acquisition method for cataract images according to claim 2, characterized in that, Step 3.1 specifically includes: For each frame of eye image in the eye video, the eye image is identified by an image recognition algorithm to obtain the region of interest, and the Euclidean distance, deflection angle and deflection direction between the center coordinates of the region of interest and the preset standard anatomical target frame of the auricular bulbar conjunctiva are calculated. Determine if the deflection angle is greater than the fault tolerance threshold. If it is, output the Euclidean distance, deflection angle, and deflection direction as warning information. If not, it indicates that the attitude correction has passed, output a prompt message indicating that the attitude correction has passed, and execute step 3.

2.

4. The hardware-software co-processing standardized assisted acquisition method for cataract images according to claim 3, characterized in that, In step 3.2, the instantaneous displacement magnitude is calculated based on two adjacent frames of eye images in the eye video. ,include: Based on multiple consecutive instantaneous displacement moduli, the motion displacement of the region of interest in two adjacent eye images is calculated using the optical flow method. and ,according to and Calculate the instantaneous displacement modulus Specifically, this is achieved through the following formula: 。 5. The hardware-software co-processing standardized assisted acquisition method for cataract images according to claim 4, characterized in that, Step 4 specifically includes: Based on the preset standard anatomical target frame of the lens core area, repeat steps 3.1 to 3.2 to acquire images of the lens core area. Then, based on the preset standard anatomical target frame of the nasal bulbar conjunctiva, repeat steps 3.1 to 3.2 to acquire images of the nasal bulbar conjunctiva.

Images