CN-122025002-A - Visual rehabilitation evaluation system and method based on non-invasive mode

CN122025002ACN 122025002 ACN122025002 ACN 122025002ACN-122025002-A

Abstract

The invention discloses a non-invasive visual rehabilitation evaluation system and method, which can objectively and accurately reflect the activities and responses of brains under visual stimulation and avoid evaluation deviation caused by subjective factors of patients by adopting a non-invasive bimodal nerve signal acquisition module and synchronously acquiring brain electricity and functional near infrared spectrum signals, and further, the visual cortex characteristic rules are deeply mined by carrying out signal reinforcement preprocessing and fusion decoding on the signals through a bimodal signal fusion processing module, so that the visual functions of the patients are comprehensively known, meanwhile, the technology realizes real-time acquisition and rapid processing of the signals, and timely obtains visual perception state evaluation results, and in addition, a closed loop feedback mechanism is formed by dynamically adjusting visual stimulation parameters through a closed loop parameter optimization module based on the comparison of the evaluation results and preset thresholds, so that the requirements of users are met in training, the pertinence and the effectiveness of training are enhanced, and the visual rehabilitation training effect is further improved.

Inventors

ZHOU JIE
HU DONGYU
Wu meichen
LI RONG
Hong Zichao
CHEN QIANYIN
ZHANG JINGLIN

Assignees

暨南大学附属第一医院（广州华侨医院）

Dates

Publication Date: 20260512
Application Date: 20260123

Claims (10)

1. A non-invasive based visual rehabilitation assessment system, comprising: the non-invasive bimodal nerve signal acquisition module is used for synchronously acquiring brain electrical signals and functional near infrared spectrum signals based on visual stimulus signals; The bimodal signal fusion processing module is used for carrying out strengthening pretreatment and fusion decoding processing on the acquired electroencephalogram signals and the functional near infrared spectrum signals to obtain a visual perception state evaluation result; And the closed-loop parameter optimization module is used for dynamically calculating the parameter adjustment amount based on the visual perception state evaluation result and generating a new visual stimulus signal based on the parameter adjustment amount.
2. The non-invasive based visual rehabilitation assessment system according to claim 1, wherein the non-invasive bimodal neural signal acquisition module comprises: the electroencephalogram acquisition unit is used for acquiring electroencephalogram signals and carrying out band-pass filtering and notch filtering on the electroencephalogram signals; the functional near infrared spectrum signal acquisition unit is used for acquiring functional near infrared spectrum signals; The triaxial acceleration sensing unit is used for removing motion artifacts of the electroencephalogram signals and the functional near infrared spectrum signals by combining the triaxial acceleration sensor, the independent component analysis algorithm and the motion compensation model.
3. The non-invasive based visual rehabilitation assessment system according to claim 2, wherein the tri-axial acceleration sensing unit comprises: the three-axis acceleration sensor is used for monitoring the head movement state of a user in real time and acquiring acceleration data in a three-dimensional space; The time alignment subunit is used for performing time alignment on the acceleration data and the synchronously acquired electroencephalogram signals and the functional near infrared spectrum signals; The independent component identification subunit is used for carrying out blind source separation on the electroencephalogram signals and the functional near infrared spectrum signals by adopting an independent component analysis algorithm, and identifying and separating out independent components related to movement; And the motion component separation subunit is used for combining the motion compensation model to correct or remove the separated motion related components so as to obtain an electroencephalogram signal and a functional near infrared spectrum signal with motion artifacts removed.
4. The non-invasive based visual rehabilitation assessment system according to claim 3, wherein the bimodal signal fusion processing module comprises: The strengthening pretreatment unit is used for strengthening pretreatment of the collected electroencephalogram signals and the functional near infrared spectrum signals to obtain the electroencephalogram signals and the functional near infrared spectrum signals after strengthening pretreatment; The characteristic extraction unit is used for extracting steady-state visual evoked potential characteristics of the brain electrical signals after the strengthening pretreatment and extracting visual cortex first visual area brain blood oxygen saturation characteristics of the functional near infrared spectrum signals after the strengthening pretreatment; And the fusion decoding processing unit is used for inputting a CNN-LSTM mixed model based on an attention mechanism into a steady-state visual evoked potential characteristic and a visual cortex first visual area cerebral blood oxygen saturation characteristic and outputting a visual perception state evaluation result.
5. The non-invasive based visual rehabilitation assessment system according to claim 4, wherein the fusion decoding processing unit comprises: The input layer is used for receiving steady-state visual evoked potential characteristics and brain blood oxygen saturation characteristics of the visual cortex first visual area; the weight distribution subunit is used for dynamically distributing the weights of the steady-state visual evoked potential characteristics and the cerebral blood oxygen saturation characteristics of the first visual area through an attention mechanism, and carrying out characteristic fusion based on the weights to obtain a fusion vector sequence; the CNN-LSTM hybrid model is used for extracting spatial hierarchy features in the fusion vector sequence by adopting a CNN module and capturing time hierarchy features in the fusion vector sequence by combining with the LSTM module; and the output layer is used for generating a multidimensional visual perception state evaluation result comprising the concentration degree, the stimulus recognition accuracy, the steady-state visual evoked potential signal-to-noise ratio and the first visual area cerebral blood oxygen activation intensity based on the processing result of the CNN-LSTM hybrid model.
6. The non-invasive based visual rehabilitation assessment system according to claim 5, wherein the closed-loop parameter optimization module comprises: the parameter adjustment amount calculation unit is used for carrying out comparison calculation based on the visual perception state evaluation result and a preset threshold value to obtain a parameter adjustment amount; And the visual stimulus signal generating unit is used for generating a new visual stimulus signal based on the parameter adjustment quantity by adopting a fuzzy PID closed-loop regulation algorithm.
7. The non-invasive based visual rehabilitation assessment system according to claim 6, wherein the system further comprises: the training management module is used for recording training data and generating a training report; and the feedback interaction module is used for generating a visual rehabilitation evaluation result based on the training report, and visually presenting the training state and the evaluation result in real time.
8. A non-invasive visual rehabilitation assessment method, which is characterized by comprising the following steps: synchronously acquiring brain electrical signals and functional near infrared spectrum signals based on visual stimulus signals by using a non-invasive bimodal nerve signal acquisition module; the acquired electroencephalogram signals and the functional near infrared spectrum signals are subjected to strengthening pretreatment and fusion decoding treatment through a bimodal signal fusion processing module, so that a visual perception state evaluation result is obtained; and a closed-loop parameter optimization module is adopted, the parameter adjustment amount is dynamically calculated based on the visual perception state evaluation result, and a new visual stimulation signal is generated according to the parameter adjustment amount.
9. A computing device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the non-invasive visual rehabilitation assessment method according to claim 8 when the computer program is executed.
10. A computer readable storage medium, characterized in that the storage medium has stored thereon a computer program which, when executed by a processor, implements the steps of the non-invasive based visual rehabilitation assessment method according to claim 8.

Description

Visual rehabilitation evaluation system and method based on non-invasive mode Technical Field The invention relates to the technical field of visual rehabilitation evaluation, in particular to a non-invasive visual rehabilitation evaluation system and method. Background Visual rehabilitation assessment is of vital importance for treatment and rehabilitation of patients with visual impairment, and can help medical staff accurately solve the visual function condition of the patients, so that a personalized rehabilitation treatment scheme is formulated. The traditional visual rehabilitation evaluation method takes fixed graphic stimulus training, visual tracking training, contrast sensitivity training instrument and the like as typical representatives, and has the core realization logic that a preset specific visual stimulus (such as a black-and-white chessboard, a moving light spot and the like) is repeatedly presented through training equipment to guide a user to carry out visual concentration, tracking or resolution training, in the training process, the training effect is evaluated only by relying on behavioural feedback information (such as user fixation stability, subjective feeling report, trained eyesight, reading speed, contrast sensitivity test data and the like), and parameters (such as stimulus graphic complexity, training time length, training frequency and the like) of a subsequent training scheme are manually adjusted based on the evaluation result, but the traditional visual rehabilitation evaluation method has the defects that (1) the traditional visual training completely depends on behavioural feedback (such as eyesight and user feeling) to evaluate the training effect, the acquisition and analysis of a user visual cortex neural response signal are not involved, and the real-time sensing effect cannot be optimized due to the fact that the visual function state is directly related to brain neural activity, the objective data of the neural layer leads to the fact that the visual performance is on the surface, the subjective feeling is subject to the subjective, the user is easily disturbed, the user is easily influenced by the user's cognitive state, the real-time is difficult to realize the real-time, and the real-time performance cannot be well as the real-time dynamic response is difficult to realize when the training is difficult to realize in real-time, and the real-time response is difficult to realize in the training is difficult to realize in real-time, the real-time response is difficult to realize in the training is due to the real-time, visual fatigue or insensitivity to current stimulus and the like, the stimulus parameters cannot be adjusted in time, so that training efficiency is affected, and user interference emotion can be possibly caused by continuous ineffective stimulus. Disclosure of Invention In view of the above, the invention provides a non-invasive visual rehabilitation evaluation system and method, which can effectively solve the defects that the prior art relies on behavioral feedback to evaluate training effects, does not relate to acquisition and analysis of user visual cortex neural response signals, cannot acquire user visual perception states and physiological response data in real time, and cannot realize dynamic optimization in the training process. The technical scheme of the invention is realized as follows: a non-invasive based visual rehabilitation assessment system, comprising: the non-invasive bimodal nerve signal acquisition module is used for synchronously acquiring brain electrical signals and functional near infrared spectrum signals based on visual stimulus signals; The bimodal signal fusion processing module is used for carrying out strengthening pretreatment and fusion decoding processing on the acquired electroencephalogram signals and the functional near infrared spectrum signals to obtain a visual perception state evaluation result; And the closed-loop parameter optimization module is used for dynamically calculating the parameter adjustment amount based on the visual perception state evaluation result and generating a new visual stimulus signal based on the parameter adjustment amount. As a further alternative to the non-invasive based visual rehabilitation assessment system, the non-invasive bimodal neural signal acquisition module comprises: the electroencephalogram acquisition unit is used for acquiring electroencephalogram signals and carrying out band-pass filtering and notch filtering on the electroencephalogram signals; the functional near infrared spectrum signal acquisition unit is used for acquiring functional near infrared spectrum signals; The triaxial acceleration sensing unit is used for removing motion artifacts of the electroencephalogram signals and the functional near infrared spectrum signals by combining the triaxial acceleration sensor, the independent component analysis algorithm and the motion compensation model. As a fu