Artificial intelligence diagnosis assisted fundoscopy device

Abstract

The invention relates to a fundoscopy device that enables early detection of various retinal and optic nerve diseases and follow-up of the course of the disease with the help of an artificial intelligence algorithm.

Description

[0001] DESCRIPTION

[0002] ARTIFICIAL INTELLIGENCE DIAGNOSIS ASSISTED FUNDOSCOPY

[0003] DEVICE

[0004] Technical Field

[0005] The invention relates to a fundoscopy device that enables early detection of various retinal and optic nerve diseases and follow-up of the course of the disease with the help of an artificial intelligence algorithm.

[0006] Prior Art

[0007] Fundoscopy, also called ophthalmoscopy, is a method of examining the back of the eye. It is used to evaluate the eye layer (retina), optic nerve and vascular structure. This procedure is especially important for monitoring and early diagnosis of conditions that may affect eye health such as diabetes, hypertension, eye pressure (glaucoma). Fundoscopy is performed using an instrument called an ophthalmoscope that focuses on the back of the eye. This instrument contains a light source and magnifying lenses. The ophthalmologist or specialist can examine the condition of the retina, optic nerve and vasculature by looking into the patient's eye with fundoscopy. During this examination, factors such as the condition of the retina, whether the blood vessels are normal, and the appearance of the optic nerve are observed. This data can be used to provide information about the general state of eye health and, if necessary, to make treatment or follow-up plans. These fundoscopic diagnoses are performed based on the experience and expertise of ophthalmologists.

[0008] Currently, in the Department of Neurology, the retina of the patient's eye is examined directly with an ophthalmoscope. In this case, the doctor looks at the patient's eye by placing the ophthalmoscope against the eye. This often presents a non-ergonomic working environment. In addition, instantaneous movement or play may result in the inability to focus on the image that can be taken into consideration. However, since the image archive of the patient cannot be taken due to instant imaging, a data bank cannot be created. In addition, in the diagnostic field application with students trained in this field, students cannot see the same image as the educators.

[0009] TR2021 / 011214 discloses an electronic and software device that provides decision support for the detection of eye diseases as an alternative to manual blood group detection. In this system, the eye images taken are analysed by the developed device and artificial intelligence based detection of red eye diseases is performed. The system has the capacity to send decision suggestions to various software platforms or defined servers in data format if desired. The decision support device for the detection of red eye diseases, which has been developed with these features and can be used as mobile, can support the relevant expert by providing decision support for red eye disease with high accuracy without requiring additional hardware.

[0010] EP3948773A1 discloses image-based detection of eye and systemic diseases. The method comprises the steps of acquiring ophthalmic image data, applying a machine learning classifier trained to classify the acquired ophthalmic image data into at least one of a plurality of classifications using a field dataset of ophthalmic images labelled with one or more of the plurality of classifications. The system comprises a hardware processor and one or more software modules configured to perform the above methods when executed by at least one hardware processor.

[0011] WO2022153320A1 discloses a system and method for imaging blood vessels of the eye. The method comprises the steps of acquiring a stream of image data of an anterior portion of an eye of a subject at a rate of at least 30 frames per second, applying a spatio-temporal analysis to the stream to detect the flow of individual blood cells in the limbal or conjunctival blood vessels of the eye, and determining the condition of the subject based on the detected flow.

[0012] When the studies in the prior art are examined, it was necessary to develop a fundoscopy device that enables early detection of various retinal and optic nerve diseases and follow-up of the course of the disease with the help of an artificial intelligence algorithm.

[0013] Objectives of the Invention The object of the present invention is to provide a fundoscopy device that enables early detection of various retinal and optic nerve diseases and follow-up of the course of the disease with the help of an artificial intelligence algorithm.

[0014] Another object of the present invention is to provide a fundoscopy device for magnifying and focusing images by means of buttons on a hand terminal.

[0015] Another object of the present invention to provide a fundoscopy device in which the parts included are of a modular construction.

[0016] Another object of the present invention is to provide a fundoscopy device which enables the image and data to be carried by means of a adjustable screen.

[0017] Another object of the present invention is to provide a fundoscopy device in which the images can be recorded to a removable card in the form of pictures or videos or to a phone, tablet, PC etc. via protocols such as bluetooth, Wi-Fi etc.

[0018] Deatiled Description of the Invention

[0019] The fundoscopy device provided to achieve the objects of the present invention are shown in the attached figures.

[0020] Figures;

[0021] Figure 1: A side view of the fundoscopy device according to the invention.

[0022] Figure 2: A perspective view of the fundoscopy device according to the invention.

[0023] Figure 3: A rear view (from the screen side) of the fundoscopy device according to the invention.

[0024] Figure 4: An exploded view of the fundoscopy device of the invention.

[0025] Figure 5a: A schematic view of the front side of the circuit board in the fundoscopy device of the invention.

[0026] Figure 5b: A schematic view of the back side of the circuit board in the fundoscopy device of the invention. Figure 6: A side view of an alternative configuration of the inventive fundoscopy device.

[0027] Figure 7: A perspective view of an alternative configuration of the inventive fundoscopy device.

[0028] Figure 8: An exploded view of an alternative configuration of the inventive fundoscopy device.

[0029] Figure 9: A rear (screen side) view of an alternative configuration of the inventive fundoscopy device.

[0030] The parts in the figures are numbered individually, and the corresponding descriptions are given below.

[0031] 1. Hand terminal

[0032] 2. LCD screen

[0033] 3. Connector

[0034] 4. Imaging device housing

[0035] 5. Imaging device

[0036] 6. Lenticular adapter

[0037] 7. Record button

[0038] 8. Zoom button

[0039] 9. Circuir board

[0040] 10. Memory card

[0041] 11. Artifical intelligence based processor

[0042] A fundoscopy device that enables the early detection of various retinal and optic nerve diseases and the follow-up of the course of the disease with the help of an artificial intelligence algorithm, comprises,

[0043] - Hand terminal (1) for holding the device,

[0044] - LCD screen (2) for displaying captured images and videos to the user, Connector (3) for modular mounting and removal of the LCD screen (2) on the hand terminal (1), Imaging device housing (4) fixed to the slide structure on the hand terminal (1),

[0045] The imaging device (5), which allows the user's eye to be visualised by inserting it into the imaging device housing (4),

[0046] - Lenticular adapter (6) connected to the imaging device (5) and enabling the sharpness of the image to be adjusted and the user's eye to be focused,

[0047] - Recording button (7) on the hand-held terminal (1) for recording video images,

[0048] - Zoom button (8) on the hand terminal (1), which can be turned left or right to enlarge or reduce the image,

[0049] Circuit board (9) connected to the LCD screen (2),

[0050] The memory card (10) located on the circuit board (9), which allows the captured images or videos to be saved and exported on demand,

[0051] Artificial intelligence-based processor (11) located on the circuit board (9) and enabling data processing.

[0052] In the inventive device, during use, the doctor holds the device to the patient's eye with the help of the hand terminal (1) and looks at the retina of the patient. When the device is placed on the patient's eye, an image is formed on the LCD screen (2) with the help of the imaging device (5). Imaging device (5) refers to a device such as a camera, digital microscope, etc. The connection of the LCD screen (2) with the hand terminal (1) is provided by a connector (3). The connector (3) may have a structure in the form of a pin or magnetic holder. In addition, with the lenticular adapter (6) in the inventive device, it is possible to adjust the clarity of the image during use and to focus on the patient's eye. In the same way, in order to record the image in the form of video, it can take video recording by pressing and holding the record button (7). Although the record button (7) is located on the hand terminal (1), in alternative applications it can also be located on the LCD screen (2) as a touch screen. Video and image recordings are saved to the memory card (10) on the circuit board (9) adapted to the LCD screen (2). If it is desired to enlarge the image taken through the eye, the image can be enlarged and reduced by turning the zoom button (8) on the hand terminal (1) to the right and left. As with the record button (7), the zoom button (8) can be positioned on the LCD screen (2) in alternative applications. At the same time, the light source on the imaging device (5) helps to sharpen the image. The recorded images enable early detection of various retinal and optic nerve diseases and follow-up of the course of the disease with deep learning and convolutional neural network algorithms using an artificial intelligence-based processor (11). The images and videos recorded in the inventive device can be exported via wireless connections (Wi-Fi, Bluetooth, etc.) or micro USB support.

[0053] In the inventive device, there are two alternative options for artificial intelligence diagnosis. Firstly, the data received via the memory card (10) can be processed and evaluated via a computer. The second is the evaluation of the snapshots taken through the artificial intelligence-based processor (11) adapted into the LCD screen (2) through the device.

[0054] The method that the device will use at the point of diagnosis is artificial intelligence techniques. Artificial intelligence (Al) diagnosis is a process based on the analysis of eye images provided by medical devices such as digital ophthalmoscopes. Artificial intelligence techniques such as artificial neural networks and deep learning algorithms are used in this diagnostic process. The learning ability of artificial intelligence models are improved by training them on image data labelled as eye diseases as well as focusing on large amounts of retina, eye vessels, optic disc, etc. Thus, artificial intelligence can identify and classify various retinal and optic nerve diseases. It also plays an important role in predicting future eye and optic nerve disease risks. Artificial neural networks, one of the most basic components of deep learning, is an artificial intelligence algorithm that mimics the biological nervous system. These deep neural networks can be designed to perform complex tasks by processing eye data, can be improved thanks to their learning capabilities, and can provide more precise results as they are fed with more data.

[0055] All components of the device are mounted on the hand terminal (1) by means of mechanical slides or screws. There are also screw assemblies for modularity.

[0056] In an alternative configuration of the inventive device, the positions and structures of the record button (7) and the zoom button (8) may change. In addition, the dimensions and structure of the LCD screen (2) used may change. An exemplary alternative design is shown in the attached figures (Figure 6-9).

Claims

CLAIMS1. A fundoscopy device that enables the early detection of various retinal and optic nerve diseases and the follow-up of the course of the disease with the help of an artificial intelligence algorithm, characterize by, it comprises,- Hand terminal (1) for holding the device,- LCD screen (2) for displaying captured images and videos to the user,Connector (3) for modular mounting and removal of the LCD screen (2) on the hand terminal (1),Imaging device housing (4) fixed to the slide structure on the hand terminal (1),The imaging device (5), which allows the user's eye to be visualised by inserting it into the imaging device housing (4),- Lenticular adapter (6) connected to the imaging device (5) and enabling the sharpness of the image to be adjusted and the user's eye to be focused,- Recording button (7) on the hand-held terminal (1) for recording video images,- Zoom button (8) on the hand terminal (1), which can be turned left or right to enlarge or reduce the image,Circuit board (9) connected to the LCD screen (2), Artificial intelligence-based processor (11) located on the circuit board (9) and enabling data processing.

2. A fundoscopy device according to the claim 1, characterized by, it comprises, the memory card (10) located on the circuit board (9), which allows the captured images or videos to be saved and exported on demand.

Images