Ophthalmic disease examination system based on remote slit lamp

Abstract

The invention belongs to the field of intelligent medical treatment, and particularly relates to an ophthalmic disease examination system based on a remote slit lamp. The system comprises a remote control end and a near-end execution end, and the remote control end and the near-end execution end are connected through a network. The system executes the following method: S1, a remote control end acquisition unit acquires a control instruction of an electric signal, inputs the control instruction into a remote signal conversion unit, encodes the control instruction into a corresponding audio signal, and sends the audio signal to a near-end execution end through an audio channel of a remote interface unit; s2, a near-end execution end receives the audio signal through an audio channel of a near-end interface unit, a near-end signal conversion unit decodes the audio signal into a control instruction of an electric signal, and after a control instruction execution unit drives a slit lamp to make an action corresponding to the control instruction, a camera shooting unit shoots an ophthalmology examination image. According to the application, plug-and-play remote medical slit lamp examination is realized, the inconvenience of installing a driving program is avoided, and the compatibility of the system is higher.

Description

Technical Field

[0001] The present invention relates to the field of intelligent medical treatment, and more specifically, to an ophthalmic disease diagnosis system based on a remote slit lamp. Background Art

[0002] Medical and health resources are skewed in distribution, with scattered and numerous primary health resources and a relatively insufficient supply of high-quality medical and health resources. With the help of advanced technology, the establishment of telemedicine systems and telemedicine management can effectively solve the problem of uneven distribution of medical resources. The current telemedicine system mainly includes remote surgery, remote consultation and remote outpatient clinics. The clinical application of remote surgery has been reported at home and abroad. Remote surgery requires the use of dedicated networks, remote consultations and remote outpatient clinics. Through remote guidance from consulting experts, experts make diagnoses based on the uploaded information. Among them, the main purpose of the dedicated network is to ensure the quality of communication during remote diagnosis and treatment, but the maintenance cost of the dedicated network is high, which is unacceptable for systems aimed at popularizing remote diagnosis and treatment. Therefore, the signal transmission speed during remote diagnosis and treatment in the Internet network is a severe challenge.

[0003] The remote virtual diagnosis and treatment platform of the hospital's ophthalmology department adopts the above-mentioned remote consultation and remote outpatient model. The consultation and outpatient clinic realize diagnosis and examination through remote film reading and remote viewing of test reports. There are many examination instruments in the ophthalmology outpatient clinic. The slit lamp microscope, as the most basic medical device in the ophthalmology outpatient clinic, needs to be operated by an ophthalmologist, and its examination results are closely related to the doctor's operation. In order to further improve the accessibility of medical resources for people in special positions and improve the accuracy of eye disease diagnosis, the slit lamp diagnosis system of the remote virtual diagnosis and treatment platform using a dedicated network for people in special positions is urgently needed to be developed. Que Tianxing et al. designed a remote slit lamp diagnosis platform based on the Internet of Things technology, and used a doctor-side / server-side / device-side distributed model to implement the platform, but the platform can only run under dedicated remote consultation software, operation-side driver and execution-side driver. Chen Junfa et al. designed a Web-based remote slit lamp microscope diagnosis system, which is implemented using a B / S architecture, but a dedicated driver needs to be installed on the handle end. G. LahaieLuna et al. proposed a stereo robot remote control unmanned slit lamp, and the system uses a dedicated line network for communication data and command data transmission. N. Tanabe et al. proposed a remote-operated slit lamp microscope system, which is based on a C / S architecture and installs supporting remote consultation software to realize system functions. D. Nankivil et al. designed a robot remote-controlled slit lamp system, which uses a remote desktop to realize remote control of the slit lamp and requires the installation of corresponding remote desktop control software.

[0004] It can be seen that the remote control system in the prior art often involves the installation of drivers and desktop operating programs, and the drivers, desktop programs, etc. are written in a specific programming language, and their operation process often depends on a specific language and language dependency packages. Since the system platform and system environment of the remote computer used are different, it is very easy to encounter installation failure problems in actual application scenarios, which also makes it difficult to deploy, implement and promote remote diagnosis and treatment systems. Summary of the invention

[0005] In view of the above problems, the present invention provides an ophthalmic disease diagnosis system based on a remote slit lamp, which realizes driverless remote diagnosis by converting the control electrical signal of the slit lamp into an audio frequency signal for transmission and decoding. At the same time, encoding is used to compress the space during signal transmission to speed up data transmission, thereby constructing a lightweight remote slit lamp ophthalmic disease diagnosis system.

[0006] The present application (first aspect) discloses an ophthalmic disease diagnosis system based on a remote slit lamp, comprising:

[0007] A remote control end and a near-end execution end, wherein the remote control end and the near-end execution end are connected via a network, wherein the remote control end includes: an acquisition unit, a far-end signal conversion unit, and a far-end interface unit; wherein the near-end execution end includes: a near-end interface unit, a near-end signal conversion unit, an execution unit, and a camera unit; wherein the system executes the following method:

[0008] S1: The remote control end acquisition unit acquires the control command of the electrical signal and inputs it into the remote signal conversion unit, which encodes the control command into a corresponding audio signal and then sends it to the near-end execution end through the audio channel of the remote interface unit;

[0009] S2: The proximal execution end receives the audio signal through the audio channel of the proximal interface unit, and the proximal signal conversion unit decodes the audio signal into a control instruction of an electrical signal. Based on the control instruction execution unit, the slit lamp is driven to perform an action corresponding to the control instruction, and then the camera unit captures the ophthalmic examination image.

[0010] Furthermore, the method performed by the system also includes:

[0011] S3: sending the ophthalmic examination image to the remote control end through the video channel of the proximal interface unit,

[0012] S4: The remote control end receives the diagnosis image through the video channel of the remote interface unit, and obtains an updated control instruction from the acquisition unit according to the diagnosis image;

[0013] S1-S4 are repeated until the remote control terminal receives the ophthalmic examination image required for ophthalmic disease examination, and the ophthalmic disease diagnosis is performed based on the ophthalmic examination image.

[0014] Further, the method executed by the system further includes: the proximal execution end further includes an adaptive control unit, and the adaptive control unit automatically executes a set of control instructions based on an adaptive control strategy according to the received control instructions to adjust the slit lamp to the next inspection state. The method for obtaining the adaptive control strategy includes:

[0015] Step 1: Obtain the initial robotic arm movement and the initial state corresponding to the initial movement, where the initial state is the current inspection setting of the slit lamp;

[0016] Step 2: Reinforcement learning takes the next action in the action space, where the action space includes adjusting the slit lamp intensity, slit width, slit color, slit lamp diffusion, lamp irradiation angle, imaging system angle, left - right movement of the microscope, front - back movement of the microscope, up - down movement of the microscope, and imaging magnification;

[0017] Step 3: After executing the action, the slit lamp is converted to the next inspection setting, and an updated inspection image is obtained based on the current inspection setting;

[0018] Step 4: Obtain the reward function of the action based on the updated inspection image, and obtain the adaptive control strategy of the robotic arm through the training iteration of the reward function.

[0019] Further, the obtaining unit uses a controller handle to obtain the control instructions of the electrical signals of the handle operation;

[0020] Optionally, the distal signal conversion unit is built into the controller handle;

[0021] Optionally, the distal signal conversion unit is connected to the controller handle and accesses the computer through the audio interface of the computer at the remote control end.

[0022] Further, the control instructions or the actions corresponding to the control instructions include adjusting any one or more of the following: slit intensity, slit width, slit color, slit angle, slit light diffusion, slit lamp irradiation angle, imaging system angle, left - right movement of the microscope, front - back movement of the microscope, up - down movement of the microscope, and imaging system magnification.

[0023] Further, encoding the control instructions as audio signals means using different audio frequencies to represent different control instructions, and different control instructions correspond to corresponding actions at the instrument end;

[0024] Optionally, the distal interface unit and the proximal interface unit are two computers connected through a network.

[0025] Furthermore, the imaging unit of the proximal execution end includes a slit lamp microscope inspection camera and a monitoring camera. The slit lamp microscope inspection camera is used to capture the inspection images of the microscope, and the monitoring camera is used to capture the instrument operation status images during the inspection by the slit lamp microscope. The proximal execution end transmits the inspection images and the instrument operation status images to the remote control end through the proximal interface unit for display.

[0026] Furthermore, the slit lamp microscope inspection camera is an electronic eyepiece.

[0027] Optionally, the proximal execution end further includes a video unit. The video unit combines the inspection images of the microscope and the instrument operation status images as the main picture and the secondary picture of the picture-in-picture respectively into one image and then transmits it to the instrument end interface unit. The diagnostic images are transmitted to the remote control end through the video interface. The remote control end acquisition unit also acquires the switching instructions for the main picture and the secondary picture of the picture-in-picture and then transmits them to the distal interface unit. The distal interface unit transmits the switching instructions and the audio signal to the proximal interface unit and then separates the audio signal and the switching instructions. The video unit sets the original main picture of the picture-in-picture as the secondary picture of the new picture-in-picture according to the received switching instructions, sets the original primary and secondary pictures of the picture-in-picture as the main picture of the new picture-in-picture, and then transmits the new picture-in-picture to the proximal interface unit and sends it to the remote control end.

[0028] Furthermore, the acquisition unit further includes a microphone module for acquiring the voice communication signals between the doctor and the patient, transmitting the voice communication signals to the distal interface unit. The control end interface unit transmits the reverberation signal obtained by reverberating the audio signal and the voice communication signals to the proximal interface unit. The proximal interface unit separates the reverberation signal to obtain the audio signal and the voice communication signals. The voice communication signals are played through the voice playback device.

[0029] Furthermore, the remote control end further includes an imaging unit for acquiring the communication picture between the doctor and the patient, transmitting the communication picture to the control end interface unit, and transmitting the communication picture to the display on the instrument end through the data interface. The instrument end interface unit transmits the communication picture to the display on the instrument end.

[0030] Furthermore, there are multiple remote control ends. One of the multiple remote control ends is the main control end for controlling the actions of the proximal execution end. The proximal execution end transmits the picture-in-picture to multiple remote control ends for display, and the video conferences between the multiple remote control ends are interconnected.

[0031] The present application has the following beneficial effects:

[0032] (1) This application converts the instructions controlled by the doctor's end handle into audio signals, which are parsed at the instrument end, thus realizing the plug-and-play remote medical slit lamp examination, eliminating the inconvenience of installing driver programs and making the system more compatible.

[0033] (2) Mix the instrument control instruction signal and the voice communication signal, which can send instructions and does not prevent the communication between the two parties. Description of the Drawings

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those skilled in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0035] Figure 1 It is a schematic diagram of the system provided in the first aspect of the embodiment of the present invention;

[0036] Figure 2 It is a schematic diagram of a remote slit lamp medical examination system provided in the embodiment of the present invention;

[0037] Figure 3 It is a schematic diagram of the composition of a remote control slit lamp function unit provided in the embodiment of the present invention;

[0038] Figure 4 It is a physical diagram of a remotely controlled slit lamp provided in the embodiment of the present invention, where (a) is the front view of the remotely controlled slit lamp; (b) is the side view of the remotely controlled slit lamp;

[0039] Figure 5 It is a schematic diagram of the composition unit of a control handle provided in the embodiment of the present invention;

[0040] Figure 6 It is a physical diagram of a control handle provided in the embodiment of the present invention;

[0041] Figure 7 It is a schematic diagram of the system test results provided in the embodiment of the present invention: where (a) represents the remote examination room at the hospital end; (b) represents the slit lamp examination room at the instrument end; (c) represents the front view of the system test, and the picture-in-picture technology shows the examination image and the status of the slit lamp; (d) represents the side view of the system test, and the picture-in-picture technology shows the examination image and the side status of the slit lamp;

[0042] Figure 8 It is a schematic diagram of the remote control instruction audio encoding and proximal instrument end decoding provided in the embodiment of the present invention;

[0043] Figure 9A schematic diagram of a near-end instrument-side shooting data transmission provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0045] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The sequence numbers of the operations, such as S101, S102, etc., are only used to distinguish between different operations, and the sequence numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0047] Figure 1 An ophthalmic disease diagnosis system based on a remote slit lamp is provided in an embodiment of the present invention. The system comprises a remote control end and a proximal execution end, the remote control end and the proximal execution end are connected via a network, the remote control end comprises: an acquisition unit, a remote signal conversion unit, and a remote interface unit; the proximal execution end comprises: a proximal interface unit, a proximal signal conversion unit, an execution unit, and a camera unit; the system executes the following method:

[0048] S1: The remote control end acquisition unit acquires the control command of the electrical signal and inputs it into the remote signal conversion unit, which encodes the control command into a corresponding audio signal and then sends it to the near-end execution end through the audio channel of the remote interface unit;

[0049] S2: The proximal execution end receives the audio signal through the audio channel of the proximal interface unit, and the proximal signal conversion unit decodes the audio signal into a control instruction of an electrical signal. Based on the control instruction execution unit, the slit lamp is driven to perform an action corresponding to the control instruction, and then the camera unit captures the ophthalmic examination image.

[0050] In order to realize the remote virtual diagnosis and treatment platform based on a dedicated network to control a slit lamp microscope for ophthalmology consultation, a remote slit lamp microscope examination system based on audio and video signal transmission instructions is designed.

[0051] First, the conventional digital slit lamp microscope is electrically controlled and transformed. Then, the instrument control instructions are transmitted using an audio signal, and the command signal and voice communication signal are mixed and transmitted through the audio channel of the dedicated network diagnosis and treatment platform for special post populations. Using video mixing technology, the instrument monitoring screen and the instrument examination screen are transmitted through the video channel of the dedicated network diagnosis and treatment platform.

[0052] The handle end and the instrument end of the remote slit lamp examination system are connected to the computer through a headset and a camera interface. The experimental results show that the examination system can complete the remote control of the common functions of the slit lamp. So far, 50 cases of remote examinations have been completed. The first slit lamp microscope that can be remotely controlled on the dedicated network remote virtual diagnosis and treatment platform is realized.

[0053] The above research on remotely controlled slit lamps is difficult to use existing video consultation platforms and cannot be directly applied to the current dedicated network virtual diagnosis and treatment platform. Considering the confidentiality and information security requirements of the dedicated network, the system should avoid installing software and device drivers as much as possible. Aiming at the characteristics of imbalance, dispersion, confidentiality, and security of medical resources on the dedicated network virtual diagnosis and treatment platform and for special post populations, this paper designs a remote slit lamp diagnosis system based on the dedicated network remote virtual diagnosis and treatment platform.

[0054] First, the conventional digital slit lamp microscope is electrically controlled and transformed. Then, the instrument control instructions are transmitted using an audio signal, and the command signal and voice communication signal are mixed and transmitted through the audio channel of the dedicated network diagnosis and treatment platform. Using video mixing technology, the instrument monitoring screen and the instrument examination screen are transmitted through the video channel of the dedicated network diagnosis and treatment platform. The handle end and the instrument end of the remote slit lamp medical examination system are connected to the computer through a headset and a camera interface, and no software and device drivers need to be installed. The first slit lamp microscope that can be remotely controlled on the dedicated network remote virtual diagnosis and treatment platform is realized.

[0055] 1. System Overview

[0056] 1.1 System Architecture

[0057] The overall block diagram of the remote slit lamp medical examination system is as Figure 2As shown in the figure. The system consists of three parts: the doctor operation terminal, the instrument execution terminal, and the video conferencing platform. At the doctor operation terminal, the doctor's operating habits are retained, and an HID (Human Interface Device) handle controller is designed. At the instrument execution terminal, the existing slit lamp microscope is modified to achieve remote control of 11 common functions of the slit lamp. In terms of the video conferencing platform, in response to the confidentiality requirements of the dedicated network and the characteristics of information security, the existing video consultation system in the hospital is utilized, and instructions and data are transmitted through the audio and video channels of the virtual diagnosis and treatment platform without the need to install proprietary software and driver programs.

[0058] Figure 2 It is a schematic diagram of a remote slit lamp medical examination system provided by an embodiment of the present invention;

[0059] 1.2 Remote control principle based on a dedicated network

[0060] At the doctor's end, the control handle modulates the control instructions into an audio signal that can be transmitted, accesses through the audio interface of the computer connected to the dedicated network, and utilizes the audio channel of the existing diagnosis and treatment platform for transmission through the established video consultation. The system mixes the instrument control instruction signal and the voice communication signal through an audio mixing and signal modulation circuit, sending instructions and enabling communication between the two parties;

[0061] At the instrument end, the remotely controlled slit lamp is connected to the audio interface of the "Dongwang" computer to receive instructions. The instrument uses a synchronization detection circuit, a signal demodulation circuit, and a motor drive circuit to decode the synchronization signal, extract the control instructions, and drive the slit lamp for examination. The doctor operation end and the instrument execution end of the remote slit lamp examination system are connected to the computer through the synchronization video interface, and there is no need to install device driver programs and video consultation software. The remote virtual diagnosis and treatment platform can only transmit one video. In this paper, video mixing technology is used to transmit the instrument's consulting room monitoring video and the slit lamp examination video to the doctor's end through the video channel of the diagnosis and treatment platform.

[0062] 1.3 Composition of the remotely controlled slit lamp execution end

[0063] The remotely controlled slit lamp execution is realized by modifying a conventional digital slit lamp (S390L, Shanghai Meiwo) ( Figure 9 ). The remotely controlled slit lamp receives the instruction signal and drives the slit lamp body to complete the medical examination.

[0064] As follows Figure 3 It is the functional unit composition of the remotely controlled slit lamp. The audio signal at the doctor's end is connected to the audio input port of the execution end interface unit;

[0065] The interface unit separates the input audio signal into two paths through a filter: one path is sent to the MCU core unit, and the MCU core unit converts the audio signal into an operation instruction to control the motor drive;

[0066] Another path is transmitted to the external speaker through the audio output port for playing the voice guidance of the doctor to the subject.

[0067] The execution - end sensor unit is used to detect the status of the slit lamp body.

[0068] The MCU core unit decodes the audio signal from the interface unit, controls the video unit to switch the picture according to the instruction, and controls the motor drive unit to drive the slit lamp body to complete the instruction action.

[0069] The operator remotely controls the brightness of the slit light band, the irradiation angle of the slit illumination light source, the imaging magnification, etc., to examine the patient's eyes; the operator observes the consulting room where the patient is located and the operating status of the instrument to complete the remote interaction with the patient. Design the transmission system to adjust its physical mechanism to complete the transformation of the mechanical part, and all motors are driven by the core control unit. Set corresponding sensors such as Hall sensors, angle sensors, etc. to monitor the operating status of the instrument and realize the closed - loop control of the stepper motor. The physical object of the remotely controlled slit lamp is as shown, and the function description of the remote slit lamp is listed in Table 1:

[0070] Table 1 Function description of the remote - controlled slit lamp

[0071]

[0072]

[0073] The sampling rate of the sound card of a general computer is 44KHz, that is, the selected frequency theoretically cannot exceed 22KHz at most.

[0074] Among them, the imaging magnification, the width of the slit light spot, the color of the slit light spot, the angle of the slit light spot, and the diffuse light are discrete audio frequencies;

[0075] Others are based on continuous audio - frequency control.

[0076] In some embodiments, 8 frequencies ( Figure 8 as shown) or more, f1 - f8, are selected. Each frequency represents one bit in binary, and the presence or absence of the frequency corresponds to 1 and 0. The encoding of f1 - f4 represents functions, such as the intensity of the slit (slit light spot) being 0001 and the width of the slit being 0010. The encoding of f5 - f8 represents the amount of action, such as the width of the slit (slit light spot) being 5mm, then the encoding is 00100101, and the three frequencies f3, f6, and f8 exist in the generated signal. The specific values of f1 - f8 can be set according to the actual situation as long as they can be transmitted through the audio channel of the video conferencing system.

[0077] In some embodiments, the frequencies used are 697, 770, 852, 941, 1209, 1336, 1477, and 1633 Hz;

[0078] In some embodiments, if the overall frequency used increases, such as to 8000 Hz, then the frequencies are 8697, 8770, 8852, 8941, 9209, 9336, 9477, and 9633 Hz.

[0079] 1.4 Composition of the Doctor's Remote Control Terminal

[0080] The function of the doctor's end control handle is to convert the doctor's operations into control instructions and modulate them into audio signals that can be transmitted. As follows Figure 3 The following figure shows a schematic diagram of the composition units of the control handle. The control handle consists of a sensor unit, an MCU core unit, and an interface unit. The sensor unit contains a rotary potentiometer, a gear switch, a lever, and a push-button switch, which convert the doctor's operations into electrical signals; the MCU core unit collects the electrical signals output by the sensors and generates instruction signals (audio) for controlling the slit lamp according to the coding rules: the interface unit completes the driving of the microphone at the doctor's end and the reverberation of the microphone signal and the control instruction audio signal. Ergonomic design has been carried out for the control handle at the doctor's end, maximizing the continuation of the doctor's operating habits, making the new system easy to use. The physical object of the doctor's end handle is as Figure 6 shown.

[0081] The remote execution end of the slit lamp is deployed in the ophthalmology clinic of the hospital, and the doctor's control end is deployed in the remote clinic of the Department of Ophthalmology, Medical College of the hospital in Beijing. The two parties are connected through the hospital's dedicated network remote virtual diagnosis and treatment platform. The doctor uses the handle to remotely control the slit lamp to complete 11 common functions listed in Table 1.

[0082] The system runs on the hospital's dedicated network remote virtual diagnosis and treatment platform, can obtain clear images for diagnosis, and realizes good real-time interaction for remote consultation. The system test results are as Figure 7 shown. So far, the system has carried out remote diagnosis and treatment for 50 patients.

[0083] Remote medical consultation in the form of video conferencing has become a way to improve the accessibility of medical resources, and remotely controlling medical examination instruments has become a way to improve the quality of remote consultation. The deployment of the existing remote slit lamp microscope examination system in the literature is relatively complex, cannot be directly applied to the existing video consultation platform, requires the installation of device driver programs, and has low convenience and information security. This paper proposes a slit lamp examination system based on a dedicated network remote virtual diagnosis and treatment platform. Using the existing dedicated network remote virtual diagnosis and treatment platform and transmitting data through audio and video, it does not require the installation of any software and device driver programs, improving the convenience of system deployment.

[0084] In actual tests, the remotely controlled slit lamp can achieve remote control of 11 common basic functions, obtain images that meet the diagnostic requirements, and both the video clarity and smoothness can meet the diagnostic requirements; the delay of the command transmission of the slit lamp microscope is less than 1 s. The slit lamp examination system based on the remote virtual diagnosis and treatment platform of the dedicated network proposed in this paper improves the convenience and network security of the remote diagnosis system and realizes the plug-and-play of the system.

[0085] 2. Intelligent system optimization

[0086] For the slit lamp examination system, it can be further optimized. On the basis of closed-loop control, a remote control mathematical model is established by using a control algorithm to improve the system control performance and dynamic characteristics; artificial intelligence technology is integrated into the system to initially diagnose the images obtained by this system and improve the examination efficiency and accuracy.

[0087] Some proximal execution ends also include an adaptive control unit. The adaptive control unit automatically executes a set of control instructions based on the received control instructions and the adaptive control strategy to adjust the slit lamp to the next examination state, and the adaptive control strategy is obtained based on reinforcement learning.

[0088] Reinforcement learning consists of two agents and four parts. The two agents are:

[0089] Agent: The agent, that is, the individual who takes action;

[0090] Environment: The environment, which can generate feedback on the action;

[0091] The four parts include: A, S, R, P; A represents the action space, that is, all the action spaces for the agent to take actions, S represents the state space, the feedback state after the agent takes actions, R represents the reward, which is a real number, a positive value represents a reward, a negative value represents a punishment, and P represents the strategy, that is, what action a∈A the agent will take in the state s∈S (sometimes A is used to represent the strategy).

[0092] The algorithm interacts through the agent and the environment, uses the rewards and states feedback by the environment to guide the agent to update actions, aims to maximize the long-term reward, and then obtains an optimal strategy. The overview of Q-learning in typical reinforcement learning is as follows. First, the agent collects data through sampling, {(s i , a i , s′ i , r i )}, where s i executes the action a in the current state i and the state changes to s ′ i , and at this time, the reward r is obtainedi ,

[0093] Then, the Q function (action-value function) is as follows: Among them, r(s,a) in the Q function represents the current instantaneous reward, represents the action that can obtain the maximum reward by taking different actions at the next state s'; γ represents the discount factor;

[0094] And our target is:

[0095] Then the objective function is: Q-learning updates the Q-value table by continuously sampling to obtain different rewards. Thus, it is finally iterated to the point where the best actions in different states can be determined according to the Q-table;

[0096] When the state is simple, the Q-table can be updated by exhaustive search. However, when the state space and action space become increasingly complex, Q-learning gradually becomes incompetent. For this reason, by introducing a neural network in deep learning to maintain a neural network representing the Q-table, the DQN algorithm is generated, and thus reinforcement learning evolves to deep reinforcement learning (Deep Reinforcement Learning, DRL).

[0097] Deep reinforcement learning is a process in which, under the environment E, the Agent takes actions A according to the state S and continuously trains to generate a policy P in order to obtain the maximum reward R.

[0098] Symbolic representations in deep reinforcement learning:

[0099] π: The function symbol represents the policy function. The parameter is often the state returned by the environment, and the output is a specific action. It can be known that a = π(s) ∈ A.

[0100] r: The immediate reward obtained after taking action a.

[0101] G: Cumulative reward, G t represents the cumulative reward from time t to the end of the action;

[0102] Q π (s,a): State-action value function, sometimes simply referred to as the action-value function, that is, at time t, in state s, taking action a, the expected value of the cumulative reward expected to be obtained using the policy π.

[0103] V π(s): State value function. That is, at time t, in state s, when taking action a, the expected value of the cumulative reward obtained using policy π.

[0104] For policy P, the methods for DRL to generate policy P are roughly divided into two types, namely policy-based and value-based. These two policies are not completely mutually exclusive and can be used together.

[0105] Among them, for the value-based policy, it calculates the policy π that maximizes the reward according to the action value function Q π (s, a) and the state value function V π (s), that is, by traversing all states and actions, finding the policy that maximizes the value functions Q π (s, a) and V π (s). The main methods are as follows: Dynamic Programming: Dynamic programming, Monte-Carlo Methods: Monte Carlo algorithm, Temporal-Difference Learning: Temporal difference learning. Among them, the aforementioned Q-learning and DQN are both temporal difference learning, that is, they adopt value-based policies.

[0106] All methods of value-based ultimately learn through value functions, and these value functions may be related to actions or states. While policy-based methods directly learn the policy using the gradient method.

[0107] In value-based methods, the policy is π(s), which is only related to the state. We modify the policy through the feedback of the value function and then calculate the value function according to the policy.

[0108] In policy-based methods, the policy is π(s|a, θ), with a set of parameters θ added. Our goal is no longer to indirectly optimize π(s) based on Q and V functions, but to directly optimize the policy π θ by optimizing the parameter θ.

[0109] To optimize the learning objective, a loss function is required. Next, we introduce two loss functions in different spaces:

[0110] The loss function in discrete states is:

[0111]

[0112] Among them, S1 represents the initial state, and V1 is all the cumulative rewards that can be obtained from the initial state to the end. In other words, since the Loss function fixes θ in each game, just by looking at the initial state S1, it can know how much reward the policy under this set of parameters can obtain. However, since the actions in some games are not deterministic but add some randomness (for example, for the same state and action, the game itself generates some noise to prevent the same result from occurring), the loss function here is the expected value.

[0113] What we need to do is to maximize this J(θ) by learning this set of parameters θ, which is what the policy-based method does.

[0114] For the Loss function in the continuous space, we have:

[0115]

[0116] Among them, is a stationary distribution of π(θ) in the Markov chain. It can be simply understood that J(θ) is the expected value of the V value in various states.

[0117] The value-based method scores the actions in each state, like training a referee (critic). According to the scores of the referee for the state actions, the highest score is selected to achieve the optimization result.

[0118] The policy-based method doesn't care about the scores of state actions and focuses on optimizing itself. The policy-based method is like an actor. As long as it learns its own parameters well, it naturally knows what action to choose in what state.

[0119] In actual training, the policy-based method also uses V(s) or Q(s,a), but when actually choosing actions, it doesn't depend on these two functions. This is the difference between the policy-based method and the value-based method.

[0120] The Actor-Critic method combines the two, adding the value-based method on the basis of the policy-based method, estimating the cumulative reward with the QV function, reducing the number of samplings, and improving the training effect and efficiency.

[0121] The Soft Actor-Critic (SAC) algorithm introduces maximum entropy. When obtaining the maximum expected cumulative reward, it requires the entropy of each output action to be maximized, using maximum entropy to increase exploration. At the same time, the "actor-critic" (AC) framework used in the algorithm has an on-policy strategy when updating the Q-function return value, that is, the behavior policy is the same as the target policy, which has stability and solves the instability of the off-policy in DQN.

[0122] The maximization of entropy can ensure policy randomization, that is, the probability of each output action is as dispersed as possible, rather than concentrated on a specific action. The algorithm also supports automatic adjustment of the entropy temperature coefficient. When the initial temperature coefficient is large, it encourages the agent to explore. As the agent converges slowly, the temperature coefficient can decay adaptively.

[0123] The SAC algorithm can be divided into two parts: policy evaluation and policy improvement:

[0124] In the policy evaluation part, the state value function Vsoft and the action value function Qsoft are defined respectively as:

[0125] V soft (s t ) = E[Q(s t , a t )] - lgπ(a t |s t )

[0126] Q soft (s t , a t ) = r(s t , a t ) + γE[V_soft(s t+1 )

[0127] In the policy improvement part, relative entropy is used to optimize the policy. Sampling policy and reward function: The sampling policy refers to the selection of DNN samples during the deep reinforcement learning process, usually the current environmental state s t . To improve the convergence and fitting ability of deep reinforcement learning,

[0128] SAC (Soft Actor-Critic) is a model-based deep reinforcement learning algorithm that combines the advantages of the Actor-Critic algorithm and the model-based policy optimization algorithm. The SAC algorithm can efficiently solve the reinforcement learning problem in the continuous action space, especially suitable for problems with high-dimensional state spaces and action spaces.

[0129] The action space in this study includes: adjusting the slit lamp intensity, slit width, slit color, slit lamp diffusion, lamp irradiation angle, imaging system angle, left - right movement of the microscope, front - back movement of the microscope, up - down movement of the microscope, and imaging magnification. Among them, for continuous variables such as slit lamp intensity, slit width, and lamp irradiation angle, discrete representation with fixed intervals is performed.

[0130] Since the adjustment of the slit lamp may involve multiple actions, the action sequence is defined as the sequential combination of actions in multiple time steps: The model generates a series of action combinations (i.e., action sequences) through multiple decisions (multiple time steps), gradually approaching the target state. For example: Step 1: Adjust the slit width and intensity; Step 2: Correct the imaging angle and mechanical position; Step 3: Optimize the diffusion and color. Implementation method: The Policy Network dynamically selects the action for each step based on the current state and historical information, and finally forms a sequence. Without explicit design of combinations, the reinforcement learning algorithm will automatically learn the dependency relationships and sequential logic between parameters.

[0131] The state is represented as: the current inspection settings of the slit lamp after performing an action on the slit lamp;

[0132] The feedback is represented as: the diagnostic image obtained based on the current inspection settings of the slit lamp;

[0133] Reward function design: Since during remote slit lamp examination, doctors need to obtain eye images under various detection conditions, the diagnostic image must be clear, and the diagnostic image can meet the inspection condition requirements for ophthalmic examination. At the same time, it is also necessary to infer the image under the next inspection condition based on the eye images that have been examined. Therefore, the design of the reward function incorporates the requirements of image clarity, compliance with detection conditions, and the adjustment actions of the slit lamp meeting expectations;

[0134] R t = α1·ImageClarity + α2·ImageRequire + α3·NextRequire

[0135] Among them, ImageClarity represents the clarity score of the image in the current state, ImageRequire represents the degree score of the diagnostic image in the current state meeting the diagnostic requirements, and NextRequire represents the degree to which the movement direction of the current inspection image conforms to expectations compared to the previous time step;

[0136] In some embodiments, the clarity score of an image is obtained by training a clarity scoring network with images of which the clarity has been labeled using a contrastive learning network; ImageRequire is a scoring network for meeting the diagnostic requirements, which is trained with a set of images labeled with the degree of meeting the diagnostic requirements using deep learning, and NextRequire is the score of the action taken compared to the state at the previous time step, and the action score is derived from the history:

[0137] During the interaction with the environment, the agent will obtain many observations. At each observation, it will take an action and also receive a reward. Therefore, the history is a sequence of observations, actions, and rewards:

[0138] H t = O1, R1, A1, …, A t-1 , O t , R t

[0139] When taking the current action, the agent will rely on the history it obtained previously. Therefore, the entire current state can be regarded as a function of this history:

[0140] S t = f(H t )

[0141] After actions at multiple time steps, when the diagnostic image obtained by the slit lamp is the next state that meets the diagnostic requirements, the reward NextRequire for the actions in this history is relatively high, and its weight is higher than the weight of clarity and the weight of meeting the diagnostic requirements;

[0142] In some embodiments, α1 = 0.15, α2 = 0.3, α3 = 0.55.

[0143] In some embodiments, NextRequire is obtained based on the consistency level between the current action and the action path of the next state that meets the diagnostic requirements.

[0144] In some embodiments, we use a DQN network to complete the training of the adaptive strategy:

[0145] First, continuous changes in the action space, such as angles, are discretized into angular degrees of 1 degree to adapt to the discrete action space of the DQN;

[0146] Now, which action we choose is determined by the DQN. Therefore, we need to train the DQN to enable it to achieve satisfactory performance.

[0147] First, the output value of our DQN is the Q value that can be obtained after performing each action under a given state. However, in fact, we don't know what the optimal Q value is in many cases. We obtain the optimal Q value through the Reward:

[0148] That is, for state s, when performing action a, the obtained reward is certain. Therefore, starting from the reward, the problem of comparing the predicted Q value and the true Q value is converted into the problem of essentially making the model fit the reward.

[0149] The loss function of the model is expressed as:

[0150]

[0151] DQN includes two deep neural networks. They have exactly the same structure but different parameters, that is, the weights, biases, etc. of each layer in the neural network. The parameters of one are while those of the other are θ i . In each iteration, what is updated is θ i and not and it is stipulated that after running C steps, let and the network where it is located is called the target network, and the network where θ i is located is the Q network.

[0152] Now the problem is converted to the need for a set of training sets that can provide a batch of quadruples (s, a, r, s'), where s' is the next state after s performs a.

[0153] Replay Buffer: Since DQN uses a replay buffer to train the network and samples a batch of quadruples (s, a, r, s') for training in each batch, we need to cache a batch of such quadruples in the experience pool for training. Since each time an action is performed, it can transfer to the next state and obtain a reward, we can obtain a such quadruple every time an action is performed, and this quadruple can also be directly put into the experience pool.

[0154] Step 1: Obtain the initial manipulator action and the initial state corresponding to the initial action, where the initial state is the current inspection setting of the slit lamp;

[0155] Step 2: Reinforcement learning takes the next action in the action space, and the action space includes adjusting the slit lamp intensity, slit width, slit color, slit lamp diffusion, lamp illumination angle, imaging system angle, microscope left - right movement, microscope front - back movement, microscope up - down movement, imaging magnification;

[0156] Step 3: After the action is executed, the slit lamp switches to the next examination setting, and an updated examination image obtained based on the current examination setting.

[0157] Step 4: Obtain the reward function of the action based on the updated examination image, and obtain the adaptive control strategy of the robotic arm through the training iteration of the reward function.

[0158] Among them, the examination setting is a state vector composed of the current slit lamp intensity, slit width, slit color, slit lamp diffusivity, lamp irradiation angle, imaging system angle, microscope coordinates (x, y, z coordinates), and imaging magnification.

[0159] In some embodiments, DQN uses a buffer pool to obtain a batch of quadruples, and the adaptive control strategy is obtained after the iteration is completed.

[0160] When a remote doctor operates the slit lamp, the control unit at the proximal instrument end acts according to the received doctor's instructions, adjusts to the next examination position based on the adaptive strategy. After the adaptive adjustment, the doctor can also make fine-tuning according to the current state.

[0161] In some embodiments, record the position of the adaptive strategy adjustment and the doctor's fine-tuning operation, and the model autonomously learns and optimizes.

[0162] In some embodiments, for the remote slit lamp examination system: the control instrument at the doctor end encodes the operation of the instrument into an audio signal and then transmits it to the interaction platform. The interaction platform transmits it to the instrument end. After the audio signal is decoded by the driving device at the instrument end, the slit lamp is driven to perform an examination. After the camera captures an image, it is transmitted to the interaction platform, and the interaction platform sends it to the monitor at the doctor end to present the examination image.

[0163] In some embodiments, for the remote slit lamp examination system: the doctor at the doctor end controls the movement of the slit lamp at the instrument end through the controller unit. The control electrical signal is converted into an audio signal and then transmitted to the interface unit. After being transmitted through a dedicated network, it reaches the instrument end. The interface unit at the instrument end transmits the received audio signal to the core control unit. The core control unit also includes decoding the received doctor control signal into a control instruction to drive the motor to move, so that the slit lamp examination instrument bound by the mechanism makes corresponding actions to perform an examination on the patient. After that, the video image captured by the electronic eyepiece on the microscope is transmitted to the interface unit at the instrument end and input to the interaction platform, and then displayed on the interface unit at the doctor end.

[0164] In some embodiments, the proximal execution end includes an image encoder, and the examination image is sent to the remote control end through the video channel of the proximal interface unit after being encoded; the remote control end includes an image decoder, and the encoded image is decoded after being received through the video channel of the distal interface unit to obtain the examination image; the image encoder and decoder are trained for the auto-encoding task based on the slit lamp diagnostic image.

[0165] In some embodiments, the encoder of the image is trained by a convolutional neural network. The doctor end reconstructs the image through the decoder for the doctor to view and make a diagnosis. At the same time, the features of the encoder are input into the auxiliary diagnosis system to output the auxiliary diagnosis results, such as conjunctivitis, dry eye, glaucoma, keratitis, cataract, eye trauma, and eye tumors. Among them, the incidence of diseases such as iritis, eye trauma, and eye tumors is relatively low; keratitis, iritis, and dry eye are relatively common eye diseases. Since the conventional cross-entropy loss is difficult to effectively distinguish eye diseases; the present invention adopts a weighted focal loss that fuses image features. By combining the score of the slit lamp microscopy examination criteria and the weighted focal loss, the loss function not only takes into account the class imbalance problem of various ophthalmic diseases but also enhances the consistency between the model and the slit lamp prior knowledge through the KL divergence, which helps to more accurately predict ophthalmic diseases and provide higher classification accuracy. The calculation method of the loss function is expressed as:

[0166]

[0167] In the formula, L is the loss function of the convolutional neural network, which is used to guide the training of the convolutional neural network;

[0168] λ i = 1 + Score i / max(Score) is the weight of the i-th sample, and Score i is the slit lamp microscopy examination criteria score of the i-th sample, and max(Score) represents taking the maximum value of the slit lamp microscopy examination criteria score;

[0169] p i is the highest probability value of the ophthalmic disease predicted based on the examination image features,

[0170] γ is the focal factor, preferably set to 2, which is used to reduce the loss contribution of easy-to-classify samples;

[0171] q i is the probability distribution of the ophthalmic disease predicted by the convolutional neural network for the i-th sample;

[0172] r i is the prior distribution of the ophthalmic disease based on the image features of the i-th sample;

[0173] α is the weight coefficient of the KL divergence term;

[0174] D KL (·||·) is the KL divergence term, which is used to quantify the difference between the predicted distribution and the prior distribution, forcing the network to align the predicted probabilities with the distribution of the radiomics scores and enhancing the consistency between the model and the domain knowledge.

[0175] In some embodiments, the slit lamp microscopy score is obtained by a doctor's grading and scoring based on the grading criteria of slit lamp examination indicators.

[0176] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system for performing the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0177] Generally speaking, the various example embodiments of the present disclosure may be implemented in hardware or dedicated circuits, software, firmware, logic, or any combination thereof. Some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software that can be executed by a controller, a microprocessor, or other computing devices. When aspects of the embodiments of the present disclosure are illustrated or described as block diagrams, flowcharts, or using some other graphical representation, it will be understood that the blocks, devices, systems, techniques, or methods described herein may be implemented as non-limiting examples in hardware, software, firmware, dedicated circuits or logic, general hardware or a controller or other computing devices, or some combination thereof.

[0178] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above may refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0179] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0180] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0181] In addition, in each embodiment of the present invention, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0182] The exemplary embodiments of the present disclosure described in detail above are merely illustrative and not restrictive. Those skilled in the art should understand that various modifications and combinations can be made to these embodiments or their features without departing from the principles and spirit of the present disclosure, and such modifications should fall within the scope of the present disclosure.

Claims

1. An ophthalmic disease diagnosis and examination system based on a remote slit lamp, characterized in that, The system comprises a remote control end and a near-end execution end, wherein the remote control end and the near-end execution end are connected via a network, wherein the remote control end comprises: an acquisition unit, a far-end signal conversion unit, and a far-end interface unit; wherein the near-end execution end comprises: a near-end interface unit, a near-end signal conversion unit, an execution unit, and a camera unit; and wherein the system executes the following method: S1: The remote control end acquisition unit acquires the control command of the electrical signal and inputs it into the remote signal conversion unit, which encodes the control command into a corresponding audio signal and then sends it to the near-end execution end through the audio channel of the remote interface unit; S2: The proximal execution end receives the audio signal through the audio channel of the proximal interface unit, and the proximal signal conversion unit decodes the audio signal into a control instruction of an electrical signal. Based on the control instruction execution unit, the slit lamp is driven to perform an action corresponding to the control instruction, and then the camera unit captures the ophthalmic examination image.

2. The ophthalmic disease diagnosis and examination system based on a remote slit lamp according to claim 1, wherein The method performed by the system also includes: S3: sending the ophthalmic examination image to the remote control end through the video channel of the proximal interface unit, S4: The remote control end receives the diagnosis image through the video channel of the remote interface unit, and obtains an updated control instruction from the acquisition unit according to the diagnosis image; Repeat S1-S4 until the remote control terminal receives the required examination image for ophthalmic disease examination, and diagnoses the ophthalmic disease based on the examination image; Optionally, the proximal execution end includes an image encoder, and the diagnosis image is encoded and sent to the video channel of the proximal interface unit of the remote control end; the remote control end includes an image decoder, and the video channel of the remote interface unit receives the encoded image and decodes it to obtain the diagnosis image; Optionally, the image encoder and decoder are trained based on a self-encoding task of slit lamp diagnostic images.

3. The ophthalmic disease diagnosis and examination system based on a remote slit lamp according to claim 1, wherein The method executed by the system further includes: the proximal execution end further includes an adaptive control unit, the adaptive control unit automatically executes a set of control instructions based on an adaptive control strategy according to the received control instructions to adjust the slit lamp to the next inspection state, and the method for obtaining the adaptive control strategy includes: Step 1: Obtain an initial robot arm motion and an initial state corresponding to the initial motion, wherein the initial state is a current inspection setting of the slit lamp; Step 2: Reinforcement learning takes the next action in the action space, which includes adjusting the slit lamp intensity, slit width, slit color, slit lamp diffusion, lamp illumination angle, imaging system angle, microscope left and right movement, microscope front and back movement, microscope lifting movement, and imaging magnification; Step 3: After the action is performed, the slit lamp switches to the next examination setting and an updated examination image is obtained based on the current examination setting. Step 4: Get the reward function of the action based on the updated inspection image, and get the adaptive control strategy of the robot arm through training iterations of the reward function.

4. The ophthalmic disease diagnosis and examination system based on a remote slit lamp according to claim 1, wherein, The acquisition unit uses the controller handle to acquire control instructions for the electrical signal of the handle operation; Optionally, the remote signal conversion unit is built into the controller handle; Optionally, the distal signal conversion unit is connected to the controller handle and accesses the computer through the audio interface of the computer at the remote control end.

5. The ophthalmic disease diagnosis and examination system based on a remote slit lamp according to claim 1, characterized in that, The control instruction or the action corresponding to the control instruction includes adjusting any one or more of the following: slit intensity, slit width, slit color, slit angle, slit light diffusion, slit lamp irradiation angle, imaging system angle, left-right movement of the microscope, front-back movement of the microscope, up-down movement of the microscope, magnification of the imaging system.

6. The ophthalmic disease diagnosis and examination system based on a remote slit lamp according to claim 5, wherein, Encoding the control instruction as an audio signal means using different audio frequencies to represent different control instructions, and different control instructions correspond to corresponding actions at the instrument end; Optionally, the distal interface unit and the proximal interface unit are two computers connected through a network.

7. The ophthalmic disease diagnosis and examination system based on a remote slit lamp according to claim 1, characterized in that, The camera unit of the proximal execution end includes a slit lamp microscope examination camera and a monitoring camera. The slit lamp microscope examination camera is used to capture the examination image of the microscope, and the monitoring camera is used to capture the instrument operation state image during the examination by the slit lamp microscope; the proximal execution end transmits the examination image and the instrument operation state image to the remote control end for display through the proximal interface unit; Optionally, the slit lamp microscope examination camera is an electronic eyepiece; Optionally, the proximal execution end further includes a video unit. The video unit combines the examination image of the microscope and the instrument operation state image as the main picture and the secondary picture of the picture-in-picture respectively into one image and then transmits it to the instrument end interface unit, and transmits the diagnosis image to the remote control end through the video interface; the remote control end acquisition unit also acquires the switching instruction for the main picture and the secondary picture of the picture-in-picture and then transmits it to the distal interface unit. The distal interface unit transmits the switching instruction and the audio signal to the proximal interface unit and then separates the audio signal and the switching instruction. The video unit sets the original main picture of the picture-in-picture as the secondary picture of the new picture-in-picture according to the received switching instruction, sets the original primary and secondary pictures of the picture-in-picture as the main picture of the new picture-in-picture, and then transmits the new picture-in-picture to the proximal interface unit and sends it to the remote control end.

8. The ophthalmic disease diagnosis and examination system based on a remote slit lamp according to claim 1, wherein The acquisition unit further includes a microphone module for acquiring the voice communication signal of the doctor to the patient, transmitting the voice communication signal to the distal interface unit. The control end interface unit transmits the reverberation signal obtained by reverberating the audio signal and the voice communication signal to the proximal interface unit. The proximal interface unit separates the reverberation signal to obtain the audio signal and the voice communication signal, and the voice communication signal is played by the voice playback device.

9. The ophthalmic disease diagnosis and examination system based on a remote slit lamp according to claim 7, characterized in that, The remote control end further includes a camera unit for acquiring the communication picture of the doctor to the patient, transmitting the communication picture to the control end interface unit, and transmitting the communication picture to the instrument end interface unit through the data interface. The instrument end interface unit transmits the communication picture to the display on the instrument end for display.

10. The ophthalmic disease diagnosis and examination system based on a remote slit lamp according to claim 8, wherein, There are multiple remote control ends; one of the multiple remote control ends is the main control end for controlling the actions of the proximal execution end. The proximal execution end transmits the picture-in-picture to multiple remote control ends for display, and the multiple remote control ends are interconnected through a video conference.

Images