Parkinson's disease diagnostic systems and devices based on virtual reality and eye-tracking technology

Abstract

This invention relates to the field of medical diagnostic technology, specifically disclosing a Parkinson's disease diagnostic system and device based on virtual reality and eye-tracking technology. The diagnostic system includes an eye movement trajectory acquisition module, an eye movement behavior feature extraction module, an eye movement function analysis module, an eye movement function evaluation module, an AI-based Parkinson's disease diagnosis module, and a diagnostic result fusion analysis module. The eye movement trajectory acquisition module acquires the current subject's eye movement trajectory; the eye movement behavior feature extraction module extracts eye movement behavior features; the eye movement function analysis module obtains the eye movement behavior feature analysis results; the eye movement function evaluation module outputs Parkinson's-specific eye movement disorder diagnostic results; the AI-based Parkinson's disease diagnosis module outputs preliminary Parkinson's disease diagnostic results; and the diagnostic result fusion analysis module obtains the final Parkinson's disease diagnostic result. This invention can non-invasively, rapidly, and accurately detect Parkinson's disease patients, and has high application value and practicality.

Description

Technical Field

[0001] This invention relates to the field of medical diagnostic technology, and in particular to a Parkinson's disease diagnostic system based on virtual reality and eye-tracking technology, and a Parkinson's disease diagnostic device based on virtual reality and eye-tracking technology including the virtual reality and eye-tracking technology-based Parkinson's disease diagnostic system. Background Technology

[0002] Parkinson's disease (PD) is the second most common neurodegenerative disease, affecting 400 to 1900 people worldwide. Data shows that the number of PD patients in my country is projected to rise to approximately 5 million by 2030. PD has an insidious onset and slow progression, affecting multiple systems. Its main clinical manifestations include resting tremor, bradykinesia, rigidity, and gait abnormalities. The disease originates from a reduction in dopaminergic neurons in the basal ganglia, leading to dysfunction of the substantia nigra-striatum. As an intermediate circuit for eye movement, damage to the basal ganglia can also cause eye movement disorders in PD patients. Eye movement abnormalities are sensitive and specific to PD patients and high-risk groups; therefore, detailed assessment of eye movements can provide valuable information for early detection of the disease.

[0003] Clinically, there are no reliable diagnostic indicators for Parkinson's disease (PD), nor are there effective measures to prevent or halt its development. Furthermore, there are no specific drugs for PD, and it cannot be completely cured. Treatment focuses on improving symptoms and alleviating pain through methods such as dopamine or anticholinergic supplementation. As the disease progresses, PD symptoms gradually worsen, impacting patients' daily activities and placing a heavy burden on their families and society. Therefore, early detection and management can significantly improve the prognosis for PD patients.

[0004] Traditional methods for diagnosing Parkinson's disease include clinical manifestations, imaging examinations, and dopaminergic drug trials, but these methods have certain limitations, such as requiring expensive equipment, having high technical barriers, and being time-consuming. Summary of the Invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art, and to provide a Parkinson's disease diagnostic system and device based on virtual reality and eye-tracking technology, which can not only provide a more convenient diagnostic method, but also provide more accurate diagnostic results by monitoring the patient's eye movement behavior.

[0006] As a first aspect of the present invention, a Parkinson's disease diagnostic system based on virtual reality and eye-tracking technology is provided. The Parkinson's disease diagnostic system based on virtual reality and eye-tracking technology includes: an eye movement trajectory acquisition module, an eye movement behavior feature extraction module, an eye movement function analysis module, an eye movement function evaluation module, an AI Parkinson's disease diagnostic module, and a diagnostic result fusion analysis module.

[0007] The eye movement trajectory acquisition module is used to acquire the current subject's eye movement trajectory;

[0008] The eye movement behavior feature extraction module is used to extract eye movement behavior features from the current subject's eye movement trajectory;

[0009] The eye movement function analysis module is used to analyze the eye movement behavior characteristics of the current subject in order to obtain the eye movement behavior characteristic analysis results of the current subject.

[0010] The eye movement function evaluation module is used to evaluate the current subject's eye movement function status based on the analysis results of the current subject's eye movement behavior characteristics, so as to output the current subject's Parkinson's-specific eye movement disorder diagnosis result;

[0011] The AI ​​Parkinson's disease diagnosis module is used to make predictions based on the current subject's Parkinson's-specific oculomotor disorder diagnosis results and the current subject's medical history and basic information, so as to output the current subject's preliminary Parkinson's disease diagnosis results.

[0012] The diagnostic result fusion analysis module is used to fuse and analyze the current subject's Parkinson's-specific oculomotor disorder diagnosis result and the current subject's preliminary Parkinson's disease diagnosis result to obtain the current subject's final Parkinson's disease diagnosis result, and update the current subject's detailed test report in real time based on the current subject's final Parkinson's disease diagnosis result.

[0013] Preferably, the eye movement trajectory acquisition module is specifically used to employ virtual reality technology to provide a standardized visual stimulation environment and record the current subject's gaze trajectory under different test tasks, so as to acquire the current subject's eye movement trajectory under different test tasks; wherein, the different test tasks include gaze stabilization tasks, dynamic tracking tasks, and visual search tasks.

[0014] Preferably, the eye movement function analysis module is specifically used to extract eye movement behavior features from the current subject's eye movement trajectory using an eye tracker, wherein the eye movement behavior features include fixation features, saccade features, tracking features, gaze path features, fixation deviation features, fixation transition features, and saccade transition features;

[0015] Fixation characteristics include the duration, frequency, location, and drift of fixation; saccade characteristics include the latency, speed, amplitude, and accuracy of saccades; tracking characteristics include the smoothness, accuracy, and reaction speed of tracking; fixation deviation characteristics include the speed and path of fixation deviation; fixation transition characteristics include the speed and accuracy of fixation transition; saccade transition characteristics include the speed and accuracy of saccade transition.

[0016] Preferably, the eye movement function evaluation module is specifically used to compare the eye movement behavior characteristic analysis results of the current subject with the data of normal people and the data of Parkinson's patients, respectively, evaluate the eye movement function status of the current subject based on the comparison results, and output the diagnosis result of Parkinson's-specific eye movement disorder of the current subject.

[0017] Preferably, the AI ​​Parkinson's disease diagnosis module is specifically used to input the current subject's Parkinson's-specific oculomotor disorder diagnosis result, along with the current subject's medical history and basic information, into the AI ​​algorithm diagnosis model for prediction, so as to output the current subject's preliminary Parkinson's disease diagnosis result;

[0018] The AI ​​algorithm diagnostic model is trained using support vector machines, convolutional neural networks, random forests, logistic regression, deep reinforcement learning, or Naive Bayes algorithms on a large amount of eye movement behavior feature data, medical history, and basic information of Parkinson's disease patients, and is formed after rigorous verification and testing.

[0019] As a second aspect of the present invention, a Parkinson's disease diagnostic device based on virtual reality and eye-tracking technology is provided. The device includes a display, an eye-tracking camera, a processor, a network interface, and a storage system. The processor includes a computer program and the aforementioned Parkinson's disease diagnostic system based on virtual reality and eye-tracking technology. The display is used to display different test tasks designed by the computer program. The eye-tracking camera is used to acquire video images of the subject's eyeballs under different test tasks. The processor is used to identify the eyeball position in the video images of the subject's eyeballs, calculate the coordinate position of the eyeballs at each time point, generate the subject's eye movement trajectory, extract eye movement behavior features from the subject's eye movement trajectory, and analyze the subject's eye movement behavior features to obtain the current subject's... The system analyzes the eye movement behavior characteristics of the current subject; evaluates the current subject's eye movement function status based on the analysis results, and outputs a diagnosis of Parkinson's-specific oculomotor disorder; predicts a preliminary diagnosis of Parkinson's disease based on the current subject's Parkinson's-specific oculomotor disorder diagnosis, medical history, and basic information; fuses the current subject's Parkinson's-specific oculomotor disorder diagnosis and preliminary diagnosis of Parkinson's disease to obtain a final diagnosis of Parkinson's disease, and updates the subject's detailed test report in real time based on the final diagnosis of Parkinson's disease; the storage system stores all data in the Parkinson's disease diagnosis process; the network interface enables communication between the system or device and the network, allowing the processor to process data using AI algorithms.

[0020] The Parkinson's disease diagnostic system based on virtual reality and eye-tracking technology provided by this invention can detect Parkinson's disease patients rapidly and accurately in a non-invasive manner, reducing the rates of misdiagnosis and missed diagnosis. It also greatly simplifies the diagnostic process and reduces the workload of doctors. It effectively addresses the shortcomings of existing diagnostic systems and equipment, and has high application value and practicality. Attached Figure Description

[0021] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the following detailed description to explain the invention, but do not constitute a limitation thereof. In the drawings:

[0022] Figure 1 The structural block diagram of the Parkinson's disease diagnostic system based on virtual reality and eye-tracking technology provided by the present invention.

[0023] Figure 2The structural block diagram of the Parkinson's disease diagnostic device based on virtual reality and eye-tracking technology provided by the present invention. Detailed Implementation

[0024] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the Parkinson's disease diagnostic system and device based on virtual reality and eye-tracking technology proposed in this invention. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the protection scope of this invention.

[0025] As a first aspect of the present invention, a Parkinson's disease diagnostic system based on virtual reality and eye-tracking technology is provided, wherein, as Figure 1 As shown, the Parkinson's disease diagnostic system based on virtual reality and eye-tracking technology includes: an eye movement trajectory acquisition module, an eye movement behavior feature extraction module, an eye movement function analysis module, an eye movement function evaluation module, an AI Parkinson's disease diagnosis module, and a diagnosis result fusion analysis module.

[0026] The eye movement trajectory acquisition module is used to acquire the current subject's eye movement trajectory;

[0027] The eye movement behavior feature extraction module is used to extract eye movement behavior features from the current subject's eye movement trajectory;

[0028] The eye movement function analysis module is used to analyze the eye movement behavior characteristics of the current subject in order to obtain the eye movement behavior characteristic analysis results of the current subject.

[0029] The eye movement function evaluation module is used to evaluate the current subject's eye movement function status based on the analysis results of the current subject's eye movement behavior characteristics, so as to output the current subject's Parkinson's-specific eye movement disorder diagnosis result;

[0030] The AI ​​Parkinson's disease diagnosis module is used to make predictions based on the current subject's Parkinson's-specific oculomotor disorder diagnosis results and the current subject's medical history and basic information, so as to output the current subject's preliminary Parkinson's disease diagnosis results.

[0031] The diagnostic result fusion analysis module is used to fuse and analyze the current subject's Parkinson's-specific oculomotor disorder diagnosis result and the current subject's preliminary Parkinson's disease diagnosis result to obtain the current subject's final Parkinson's disease diagnosis result, and update the current subject's detailed test report in real time based on the current subject's final Parkinson's disease diagnosis result, so that medical staff and subjects can understand the disease status in real time.

[0032] Preferably, the eye movement trajectory acquisition module is specifically used to employ virtual reality technology to provide a standardized visual stimulation environment and record the current subject's gaze trajectory under different test tasks, so as to acquire the current subject's eye movement trajectory under different test tasks; wherein, the different test tasks include gaze stabilization tasks, dynamic tracking tasks, and visual search tasks.

[0033] Fixation stability task: Participants need to fixate on a fixed target point or object in a VR environment to assess relevant indicators of eye stability, such as fixation duration and fixation drift.

[0034] Visual search task: Multiple target objects are set up in a VR environment, and participants are required to quickly find and gaze at these objects. This task can assess participants' saccade amplitude, visual search speed, and accuracy.

[0035] Dynamic tracking task: A moving target object is presented in a VR environment, and the subject is required to track and watch the movement of the object to assess the subject's dynamic visual tracking ability and reaction speed.

[0036] Preferably, the eye movement function analysis module is specifically used to extract eye movement behavior features from the current subject's eye movement trajectory using an eye tracker, and to preprocess them to remove noise and artifacts. The eye movement behavior features include fixation features, saccade features, tracking features, gaze path features, fixation deviation features, fixation transition features, and saccade transition features. Each eye movement behavior feature has its specific function and physiological basis, reflecting different cognitive and neural control processes.

[0037] Fixation characteristics include the duration, frequency, location, and drift of fixation; saccade characteristics include the latency, speed, amplitude, and accuracy of saccades; tracking characteristics include the smoothness, accuracy, and reaction speed of tracking; fixation deviation characteristics include the speed and path of fixation deviation; fixation transition characteristics include the speed and accuracy of fixation transition; saccade transition characteristics include the speed and accuracy of saccade transition.

[0038] Preferably, the eye movement function evaluation module is specifically used to compare the eye movement behavior characteristic analysis results of the current subject with the data of normal people and the data of Parkinson's patients, respectively, evaluate the eye movement function status of the current subject based on the comparison results, and output the diagnosis result of Parkinson's-specific eye movement disorder of the current subject.

[0039] It should be noted that normal individuals do not exhibit sustained nystagmus when performing paradigmatic tasks, while all PD patients demonstrate sustained nystagmus.

[0040] Preferably, the AI ​​Parkinson's disease diagnosis module is specifically used to input the current subject's Parkinson's-specific oculomotor disorder diagnosis result, along with the current subject's medical history and basic information, into the AI ​​algorithm diagnosis model for prediction, so as to output the current subject's preliminary Parkinson's disease diagnosis result;

[0041] The AI ​​algorithm diagnostic model is trained using algorithms such as support vector machine, convolutional neural network, random forest, logistic regression, deep reinforcement learning, and Naive Bayes on a large amount of eye movement behavior data, medical history (history of head trauma, cerebrovascular disease, medication use, etc.) and basic information (age, gender, genetic factors, environmental factors, etc.) of Parkinson's disease patients. After rigorous verification and testing, a more complex and accurate AI algorithm diagnostic model is constructed to improve the accuracy and reliability of Parkinson's disease diagnosis.

[0042] As a second aspect of the present invention, a Parkinson's disease diagnostic device based on virtual reality and eye-tracking technology is provided, wherein, as Figure 2As shown, the Parkinson's disease diagnostic device based on virtual reality and eye-tracking technology includes: a display, an eye-tracking camera, a processor, a network interface, and a storage system. The processor includes a computer program, an information acquisition system, and an analysis system. The information acquisition system includes an eye movement trajectory acquisition module and an eye movement behavior feature extraction module. The analysis system includes an eye movement function analysis module, an eye movement function evaluation module, an AI Parkinson's disease diagnosis module, and a diagnostic result fusion analysis module. The display is used to show different test tasks designed by the computer program. The eye-tracking camera is used to acquire video images of the current subject's eyes under different test tasks. The processor is used to identify the eye position in the current subject's eye video images, calculate the coordinate position of the eye at each time point, generate the current subject's eye movement trajectory, extract eye movement behavior features from the current subject's eye movement trajectory, and analyze the current subject's eye movement behavior features to obtain the current subject's... The system analyzes the subject's eye movement behavior characteristics; evaluates the subject's eye movement function status based on the analysis results, and outputs a diagnosis of Parkinson's-specific eye movement disorder; predicts a preliminary diagnosis of Parkinson's disease based on the diagnosis of Parkinson's-specific eye movement disorder and the subject's medical history and basic information; fuses the diagnosis of Parkinson's-specific eye movement disorder and the preliminary diagnosis of Parkinson's disease to obtain a final diagnosis of Parkinson's disease, and updates the subject's detailed test report in real time based on the final diagnosis of Parkinson's disease; the storage system stores all data in the Parkinson's disease diagnosis process (recording data during collection, analysis, and diagnosis, and the final results); the network interface enables communication between the system or device and the network, allowing the processor to process data using AI algorithms.

[0043] The working process of the Parkinson's disease diagnostic device based on virtual reality and eye-tracking technology provided by this invention is as follows:

[0044] After powering on and connecting to the network, the subject wears a VR acquisition device, and the eye-tracking camera calibrates the pupils. Subsequently, the eye movement trajectory acquisition module within the processor begins running a standardized virtual reality test task. Simultaneously, the display shows the computer-programmed virtual reality task, and the eye-tracking camera tracks the pupils, acquiring video images. Using conventional image processing techniques, the pupil position in the video images is identified, and the coordinates of the pupils at each time point are calculated, generating the current subject's eye movement trajectory. The eye movement behavior feature extraction module within the processor extracts time-domain and frequency-domain eye movement behavior features from the current subject's eye movement trajectory. These features are then preprocessed through data calibration, filtering, data interpolation, outlier removal, and drift removal to ensure data reliability. The eye movement function analysis module within the processor classifies and analyzes the eye movement behavior features extracted under different task paradigms to obtain the current subject's eye movement behavior feature analysis results. The eye-tracking function evaluation module within the processor compares the eye-tracking behavior feature analysis results output by the eye-tracking function analysis module with pre-recorded data from normal individuals and Parkinson's patients to evaluate the subject's eye-tracking function status and output a Parkinson's-specific oculomotor disorder diagnosis. Then, the AI ​​Parkinson's disease diagnosis module within the processor inputs the Parkinson's-specific oculomotor disorder diagnosis result from the eye-tracking function evaluation module, along with the subject's medical history and basic information, into the AI ​​algorithm diagnostic model for prediction, thus obtaining a preliminary diagnosis of Parkinson's disease for the current subject. Finally, the diagnostic result fusion analysis module within the processor fuses the Parkinson's-specific oculomotor disorder diagnosis result from the eye-tracking function evaluation module and the preliminary diagnosis result from the AI ​​Parkinson's disease diagnosis module to provide a final diagnosis of Parkinson's disease for the current subject. Based on this final diagnosis, the module updates the subject's detailed test report in real time, allowing medical staff and the subject to monitor their condition.

[0045] Specifically, the display, located inside the eye tracker, is a key component used to display virtual reality scenes. It transmits computer-generated images to the subject's eyes, allowing the subject to experience an important role in the virtual environment. The test concludes with the display of results.

[0046] Specifically, the eye-tracking camera illuminates the pupil with an infrared light source, then captures changes in the pupil's position to infer the subject's gaze direction and eye movement trajectory, and transmits the data to a computer system.

[0047] Specifically, the computer program includes a designed virtual reality task that provides a standardized visual stimulus environment. This includes specific images, videos, or virtual scenes used to guide the subject's eye movements.

[0048] This invention provides a Parkinson's disease diagnostic system and device based on virtual reality eye-tracking technology. Significant differences exist in eye-tracking behaviors exhibited by normal individuals and Parkinson's patients in different eye-tracking paradigm tasks, reflected in paradigm completion ability and specific eye-tracking abnormalities. This invention first extracts the subject's eye-tracking behavior, uses signal processing methods to extract temporal and frequency domain physical features of the eye movements, and analyzes Parkinson's-specific eye-tracking disorders. Eye-tracking features are extracted according to paradigms and analyzed in depth using an AI Parkinson's diagnostic tool. Finally, the analysis from each module is integrated to provide diagnostic results, generating a real-time updated test report for medical personnel and subjects to understand their condition in real time.

[0049] It is understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, and the present invention is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A Parkinson's disease diagnostic system based on virtual reality and eye-tracking technology, characterized in that, The Parkinson's disease diagnostic system based on virtual reality and eye-tracking technology includes: an eye movement trajectory acquisition module, an eye movement behavior feature extraction module, an eye movement function analysis module, an eye movement function evaluation module, an AI Parkinson's disease diagnosis module, and a diagnosis result fusion analysis module. The eye movement trajectory acquisition module is used to acquire the current subject's eye movement trajectory; The eye movement behavior feature extraction module is used to extract eye movement behavior features from the current subject's eye movement trajectory; The eye movement function analysis module is used to analyze the eye movement behavior characteristics of the current subject in order to obtain the eye movement behavior characteristic analysis results of the current subject. The eye movement function evaluation module is used to evaluate the current subject's eye movement function status based on the analysis results of the current subject's eye movement behavior characteristics, so as to output the current subject's Parkinson's-specific eye movement disorder diagnosis result; The AI ​​Parkinson's disease diagnosis module is used to make predictions based on the current subject's Parkinson's-specific oculomotor disorder diagnosis results and the current subject's medical history and basic information, so as to output the current subject's preliminary Parkinson's disease diagnosis results. The diagnostic result fusion analysis module is used to fuse and analyze the current subject's Parkinson's specific oculomotor disorder diagnosis result and the current subject's preliminary Parkinson's disease diagnosis result to obtain the current subject's final Parkinson's disease diagnosis result, and update the current subject's detailed test report in real time based on the current subject's final Parkinson's disease diagnosis result. Specifically, the eye-tracking function analysis module is used to extract eye-tracking behavior features from the current subject's eye movement trajectory using an eye tracker. These eye-tracking behavior features include fixation features, saccade features, tracking features, gaze path features, fixation deviation features, fixation transition features, and saccade transition features. Fixation features include the duration, frequency, location, and drift of fixation; saccade features include the latency, speed, amplitude, and accuracy of saccades; tracking features include the smoothness, accuracy, and reaction speed of tracking; fixation deviation features include the speed and path of fixation deviation; fixation transition features include the speed and accuracy of fixation transitions; and saccade transition features include the speed and accuracy of saccade transitions. Specifically, the eye movement trajectory acquisition module is used to provide a standardized visual stimulation environment using virtual reality technology, and to record the current subject's gaze trajectory under different test tasks, so as to acquire the current subject's eye movement trajectory under different test tasks; wherein, different test tasks include gaze stabilization task, dynamic tracking task and visual search task; Specifically, the eye movement function evaluation module is used to compare the eye movement behavior characteristic analysis results of the current subject with the data of normal people and Parkinson's patients, respectively, evaluate the eye movement function status of the current subject based on the comparison results, and output the Parkinson's-specific eye movement disorder diagnosis result of the current subject.

2. The Parkinson's disease diagnostic system based on virtual reality and eye-tracking technology according to claim 1, characterized in that, The AI ​​Parkinson's disease diagnosis module is specifically used to input the current subject's Parkinson's-specific oculomotor disorder diagnosis result, along with the current subject's medical history and basic information, into the AI ​​algorithm diagnosis model for prediction, so as to output the current subject's preliminary Parkinson's disease diagnosis result. The AI ​​algorithm diagnostic model is trained using support vector machines, convolutional neural networks, random forests, logistic regression, deep reinforcement learning, or Naive Bayes algorithms on a large amount of eye movement behavior feature data, medical history, and basic information of Parkinson's disease patients, and is formed after rigorous verification and testing.

3. A Parkinson's disease diagnostic device based on virtual reality and eye-tracking technology, characterized in that, The Parkinson's disease diagnostic device based on virtual reality and eye-tracking technology includes: a display, an eye-tracking camera, a processor, a network interface, and a storage system. The processor includes a computer program and the Parkinson's disease diagnostic system based on virtual reality and eye-tracking technology as described in any one of claims 1 to 2. The display is used to display different test tasks designed by the computer program. The eye-tracking camera is used to acquire video images of the subject's eyeballs under different test tasks. The processor is used to identify the eyeball position in the video images of the subject's eyeballs, calculate the coordinate position of the eyeballs at each time point, generate the subject's eye movement trajectory, extract eye movement behavior features from the subject's eye movement trajectory, analyze the subject's eye movement behavior features to obtain the subject's eye movement behavior feature analysis results, and according to... The system analyzes the eye movement behavior characteristics of the current subject to evaluate the current subject's eye movement function status and outputs a diagnosis of Parkinson's-specific oculomotor disorder. Based on the current subject's Parkinson's-specific oculomotor disorder diagnosis and medical history and basic information, a preliminary diagnosis of Parkinson's disease is predicted and output. The current subject's Parkinson's-specific oculomotor disorder diagnosis and preliminary Parkinson's disease diagnosis are then fused and analyzed to obtain a final diagnosis of Parkinson's disease. The system updates the current subject's detailed test report in real time based on the final diagnosis. The storage system stores all data from the Parkinson's disease diagnosis process. The network interface enables communication between the system or device and the network, allowing the processor to process data using AI algorithms.

Images