Method and system for measuring color vision based on pupil reflex

Abstract

The invention discloses a method and system for measuring color vision based on pupillary reflex, which stimulates pupils based on multi-wavelength colors, covers various wavelengths of visible light, and can trigger various pupillary reflex modes due to different absorption and reflection characteristics of light with different wavelengths in retina and different activation conditions of visual cells. The processing information of the visual system on the color light signal can be completely obtained by comprehensively collecting the dynamic parameters, the overall state of the color vision is accurately reflected, and the problem of insufficient wavelength coverage in the prior art is solved; secondly, in analysis and evaluation, feature extraction is performed on standardized dynamic parameters to obtain feature parameters, the integrated model takes the feature parameters as input, and color vision disorder qualitative results and quantitative results are output at the same time according to internal logic in a qualitative and quantitative organic combination mode, so that color vision disorder evaluation is more comprehensive, accurate and deep; the requirement for accurately evaluating the color vision state is met, and the defect that qualitative analysis and quantitative analysis cannot be combined in the prior art is overcome.

Description

Technical Field

[0001] This invention relates to the field of color vision measurement technology, and in particular to a method and system for measuring color vision based on pupil reflex. Background Technology

[0002] Color vision is an important visual function for humans to perceive color information from the outside world. Color vision disorders can affect an individual's ability to distinguish colors, causing many inconveniences in daily life, study and work. Therefore, accurate measurement and assessment of color vision status is of great significance for clinical diagnosis, occupational adaptation and visual science research.

[0003] In recent years, some studies have begun to focus on the potential link between pupillary reflexes and color vision. Theoretically, light stimuli of different wavelengths will trigger specific pupillary reflex patterns, and individuals with normal color vision and those with color vision disorders may differ in these reflex patterns. However, research on measuring color vision based on pupillary reflexes is still in its developmental stage, lacking systematic and standardized methods. Most existing related studies only focus on pupillary reflexes under single-wavelength or a few-wavelength stimuli, failing to comprehensively cover multiple wavelengths in the visible light range, making it difficult to accurately reflect the overall state of color vision. In addition, there is a lack of suitable analytical models to organically combine pupillary reflex characteristics with qualitative and quantitative analysis of color vision disorders. Summary of the Invention

[0004] In view of this, the present invention proposes a method and system for measuring color vision based on pupillary reflex, which can effectively solve the shortcomings of existing technologies that fail to fully cover multiple wavelengths in the visible light range and cannot organically combine qualitative and quantitative analysis.

[0005] The technical solution of this invention is implemented as follows:

[0006] A method for measuring color vision based on pupillary reflex, specifically including:

[0007] The pupil is stimulated by multiple wavelengths of color, and dynamic parameters of pupil reflection are collected.

[0008] The dynamic parameters of pupillary reflex are preprocessed to obtain standardized dynamic parameters of pupillary reflex.

[0009] Feature extraction is performed on the standardized dynamic parameters of pupillary reflex to obtain pupillary reflex feature parameters;

[0010] Based on the pupillary reflex characteristic parameters and integrated model, qualitative and quantitative analysis results of color vision deficiency were obtained.

[0011] As a further alternative to the method for measuring color vision based on pupillary reflex, the method of stimulating the pupil with multi-wavelength color specifically includes:

[0012] Dark adaptation procedures are performed on subjects in a standardized darkroom;

[0013] After dark adaptation, the pupils are stimulated using a unified stimulation scheme based on the DKL color space.

[0014] As a further optional embodiment of the method for measuring color vision based on pupillary reflex, the acquisition of dynamic parameters of pupillary reflex specifically includes:

[0015] An infrared camera and a coaxial infrared LED light source are used to synchronously collect dynamic parameters of pupillary reflex, including changes in pupil diameter, changes in pupil area, contraction latency, peak contraction rate, contraction speed, and time to peak, and the eye movement position is recorded as quality control data.

[0016] As a further optional embodiment of the method for measuring color vision based on pupillary reflex, the preprocessing of the dynamic parameters of pupillary reflex to obtain standardized dynamic parameters of pupillary reflex specifically includes:

[0017] The isolated forest algorithm was used to remove artifacts from the dynamic parameters of pupil reflexes, resulting in dynamic parameters of pupil reflexes after removing outlier data.

[0018] Baseline calibration was performed on the dynamic parameters of pupillary reflexes after removing outlier data to obtain standardized dynamic parameters of pupillary reflexes.

[0019] As a further optional embodiment of the method for measuring color vision based on pupil reflex, the step of extracting features from standardized dynamic parameters of pupil reflex to obtain pupil reflex feature parameters specifically includes:

[0020] Basic reflex features are extracted from standardized pupillary reflex dynamic parameters to obtain constriction intensity features, temporal features, and stability features;

[0021] Based on the characteristics of contraction intensity, temporal sequence, and stability, a multi-dimensional color vision sensitivity feature vector is constructed;

[0022] The multidimensional color vision sensitivity feature vector is standardized to obtain pupil reflex feature parameters.

[0023] As a further optional embodiment of the method for measuring color vision based on pupillary reflex, the analysis based on pupillary reflex characteristic parameters and an integrated model to obtain qualitative and quantitative analysis results of color vision deficiency specifically includes:

[0024] Based on the CNN+random forest ensemble model, the type of color vision deficiency is determined by inputting pupil reflex feature parameters.

[0025] Based on the LSTM+multivariate linear regression model, pupillary reflex feature parameters are mapped to color vision sensitivity scores to obtain color vision sensitivity scores.

[0026] As a further alternative to the method for measuring color vision based on pupillary reflex, the method further includes:

[0027] Based on the type of color vision deficiency and color vision sensitivity score, a structured report is generated by associating fundus risk.

[0028] A system for measuring color vision based on pupillary reflex, comprising:

[0029] The stimulation and acquisition module is used to stimulate the pupil based on multi-wavelength colors and acquire dynamic parameters of pupillary reflection.

[0030] The preprocessing module is used to preprocess the dynamic parameters of pupillary reflexes to obtain standardized dynamic parameters of pupillary reflexes.

[0031] The feature extraction module is used to extract features based on standardized pupil reflex dynamic parameters to obtain pupil reflex feature parameters;

[0032] The intelligent analysis module is used to perform analysis based on pupillary reflex characteristic parameters and integrated models to obtain qualitative and quantitative analysis results of color vision deficiency.

[0033] A computing device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of any of the above-described methods for measuring color vision based on pupillary reflex.

[0034] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of any of the above-described methods for measuring color vision based on pupillary reflex.

[0035] The beneficial effects of this invention are as follows: By stimulating the pupil with multiple wavelengths of color, covering a variety of wavelengths within the visible light range, different wavelengths of light have different absorption and reflection characteristics on the retina, which can activate different cone cells and rod cells, triggering complex and diverse pupillary reflex patterns. By comprehensively collecting the dynamic parameters of pupillary reflexes under these multi-wavelength stimuli, the processing information of the visual system for different color light signals can be obtained more completely, thereby accurately reflecting the overall state of color vision and effectively making up for the shortcomings of existing technologies in wavelength coverage. Secondly, feature extraction is performed on the standardized dynamic parameters of pupillary reflexes to obtain pupillary reflex feature parameters. The integrated model receives the pupillary reflex feature parameters as input and, based on internal calculation logic and classification regression mechanism, simultaneously outputs qualitative results (clarifying the type of color vision impairment) and quantitative results (providing an indicator of the degree of color vision impairment). This organic combination of qualitative and quantitative analysis makes the assessment of color vision impairment more comprehensive, accurate, and in-depth, meeting the needs for precise assessment of color vision status and effectively solving the deficiency of existing technologies in being unable to organically combine qualitative and quantitative analysis. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 This is a flowchart illustrating a method for measuring color vision based on pupillary reflex according to the present invention.

[0038] Figure 2 This is a schematic diagram of the composition of a system for measuring color vision based on pupillary reflex according to the present invention;

[0039] Figure 3 This is a schematic diagram of the composition of a computing device according to the present invention. Detailed Implementation

[0040] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] refer to Figures 1 to 3 A method for measuring color vision based on pupillary reflex, specifically including:

[0042] The pupil is stimulated by multiple wavelengths of color, and dynamic parameters of pupillary reflection are collected. In some embodiments, the stimulation of the pupil by multiple wavelengths of color specifically includes:

[0043] Dark adaptation procedures were performed on subjects in a standardized darkroom: Subjects entered a standardized darkroom with an illuminance of <1 lux, sat 58cm away from the stimulus presentation device (17-inch OLED display, resolution 1280×1024, refresh rate 60Hz, color deviation ΔE <0.5), with their heads lightly resting on a support pad to reduce movement interference, and the non-test eye was shielded with a light-blocking sheet to avoid cross-interference and accommodation reflection interference. A 1-minute dark adaptation procedure was performed, and after confirming that the eye condition was stable (pupil diameter change <0.2mm / min), the stimulus presentation stage began.

[0044] After dark adaptation, the pupils were stimulated using a unified stimulation scheme based on the DKL color space. The color space was designed based on the DKL color space, covering the red-green and blue-yellow opposing axes. The wavelength combinations included monochromatic stimuli of red (630nm), blue (470nm), green (550nm), and yellow (580nm), and mixed-color stimuli of red-green and blue-yellow gratings. The light intensity and timing included a fixed stimulation light intensity of 200kcd / m², with each stimulus lasting 13 seconds. After stimulation, the screen lights were turned off, and dark adaptation was performed after a 60-second interval. The stimulation was presented in a cyclical manner through an OLED display, displaying monochromatic and mixed-color stimuli in a preset sequence.

[0045] Specifically, the test was conducted in a standardized darkroom with an illuminance of <1 lux, minimizing interference from external light and providing a stable and controllable environment for pupillary reflex measurement. The subject's head rested lightly against a support pad, effectively reducing interference caused by body movement. The non-test eye was shielded with a light-blocking sheet to avoid cross-interference and accommodation reflection interference, making the measurement results more accurately reflect the true response of the test eye to color stimuli. Secondly, the 1-minute dark adaptation process and the standard of confirming pupil diameter change <0.2 mm / min ensured that the subject's eyes reached a stable state before entering the stimulus presentation stage. This helps improve the accuracy and repeatability of subsequent pupillary reflex measurements and reduces measurement errors caused by unstable eye conditions.

[0046] Stimulus schemes designed based on the DKL color space, covering the red-green and blue-yellow opposing axes, can more comprehensively simulate the human visual system's color perception characteristics. This is because the DKL color space is more compatible with the physiological mechanisms of the visual system, and can better activate different cone cell channels, thereby obtaining richer and more physiologically meaningful pupillary reflex information. Secondly, wavelength combinations include monochromatic stimuli (red 630nm, blue 470nm, green 550nm, yellow 580nm) and mixed-color stimuli (red-green, blue-yellow gratings). Multiple wavelength combinations allow for the examination of pupillary reflex information from different perspectives. The aperture's response to color can detect not only the perception of a single color but also the response to color mixing, providing a more comprehensive assessment of color vision function. Furthermore, the light intensity is fixed at 200 kcd / m², each stimulus lasts 13 seconds, and the screen light is turned off after stimulation, followed by a 60-second dark adaptation period. The stimulus presentation method uses an OLED display to cyclically present monochrome and mixed color stimuli according to a preset time sequence. The standardized stimulus parameters and presentation method ensure consistency in each stimulus, making the results comparable between different subjects and between different measurements of the same subject, thus improving the reliability and accuracy of the measurement.

[0047] In some embodiments, the acquisition of dynamic parameters of pupillary reflex specifically includes:

[0048] An infrared camera and a coaxial infrared LED light source are used to synchronously collect dynamic parameters of pupillary reflex, including changes in pupil diameter, changes in pupil area, contraction latency, peak contraction rate, contraction speed, and time to peak, and the eye movement position is recorded as quality control data.

[0049] Specifically, by simultaneously acquiring multi-dimensional dynamic parameters of pupillary reflex, including changes in pupil diameter, pupil area, contraction latency, peak contraction rate, contraction speed, and time to peak, the dynamic response process of the pupil after stimulation can be captured comprehensively and meticulously. These parameter changes reflect the response characteristics of the visual system to light stimulation from different perspectives. For example, changes in diameter and area directly reflect changes in the physical shape of the pupil; contraction latency reflects the timeliness of nerve conduction and response; peak contraction rate and contraction speed reflect the intensity and speed of pupil contraction, etc. By acquiring such comprehensive parameters, rich data support is provided for subsequent in-depth analysis of color vision status, which helps to more completely assess color vision function and avoid missing important information due to single parameters.

[0050] Using an infrared camera with a coaxial infrared LED light source for data acquisition has significant advantages. Infrared light is relatively safe for the human eye and is less likely to cause significant visual interference. It can be used to take pictures without affecting the normal visual response of the subject. At the same time, the infrared camera can clearly capture images of the pupil under different lighting conditions (especially in low-light environments), ensuring the stability and clarity of the images. This allows for the accurate measurement of various parameters of the pupil, reducing measurement errors caused by image quality issues and improving the accuracy of data acquisition.

[0051] Recording eye movement position as quality control data is a key design feature. During the measurement process, the subject's eye movements may affect the measurement results of pupil parameters. By recording the eye movement position, the data can be quality controlled, and data that is inaccurate due to excessive eye movement can be filtered out. Alternatively, eye movement factors can be corrected in subsequent analysis. This helps ensure that the collected pupillary reflex dynamic parameters truly reflect the pupil's response to color stimuli, rather than being interfered with by irrelevant factors such as eye movement, thereby greatly improving the reliability and validity of the measurement results.

[0052] The dynamic parameters of pupillary reflex are preprocessed to obtain standardized dynamic parameters of pupillary reflex, specifically including:

[0053] The isolated forest algorithm was used to remove artifacts from the dynamic parameters of pupil reflexes, resulting in dynamic parameters of pupil reflexes after removing outlier data.

[0054] Baseline calibration was performed on the dynamic parameters of pupillary reflex after removing outlier data. The average pupil area from 100 ms before stimulation to 100 ms after stimulation was used as the benchmark to calculate the standardized parameters of pupillary reflex.

[0055] Specifically, the isolated forest algorithm is used to remove artifacts from the dynamic parameters of pupil reflex. This effectively identifies and removes outliers in the data. During pupil reflex measurement, due to various interference factors (such as blinking, brief changes in external light, equipment noise, etc.), the collected data may contain some abnormal data points that deviate from the normal range. These abnormal data are called artifacts. The isolated forest algorithm quickly and accurately detects the abnormal points in the data by constructing isolated trees. After removing them, the obtained dynamic parameters of pupil reflex are purer and can more realistically reflect the normal physiological response of the pupil to color stimuli.

[0056] Baseline calibration was performed on the dynamic parameters of pupillary reflexes after removing outlier data. Standardized parameters were calculated based on the mean pupil area from 100 ms before stimulation to 100 ms after stimulation. This operation is of great significance because the baseline pupillary state may vary among different subjects. Even under the same experimental conditions, the initial pupillary size and response amplitude may differ. Through baseline calibration, the data of each subject are adjusted to a relatively uniform standard, eliminating the influence of individual differences in baseline state on the data. In this way, the dynamic parameters of pupillary reflexes of different subjects have better comparability. Whether conducting comparative analysis between groups or tracking data of the same subject at different time points, it is possible to more accurately capture changes in pupillary reflexes caused by color stimulation, thereby improving the accuracy and reliability of data analysis.

[0057] Feature extraction is performed on the standardized dynamic parameters of pupillary reflex to obtain pupillary reflex feature parameters, specifically including:

[0058] Basic reflex features were extracted from standardized pupillary reflex dynamic parameters to obtain constriction intensity, temporal, and stability features. The constriction intensity feature used the pupil area change rate (% / ms) from 100 ms after stimulus presentation to peak constriction as the core indicator, where the area change rate = (baseline area - instantaneous area) / baseline area × 100%. The temporal features extracted the constriction latency (time from stimulus presentation to the start of pupil constriction, ms) and peak time (time from the start of pupil constriction to reaching minimum area, ms). The stability feature calculated the standard deviation of pupil diameter (mm) during the stimulus duration to quantify the fluctuation of pupillary response.

[0059] Based on the contraction intensity characteristics, temporal characteristics, and stability characteristics, a multi-dimensional color vision sensitivity feature vector is constructed. This includes weighting and summing the contraction intensity characteristics of each monochromatic stimulus according to weight allocation (red: 30%, blue: 40%, green: 20%, yellow: 10%) to generate a comprehensive color vision sensitivity score (range 0-100 points). Simultaneously, the contraction intensity difference under mixed color grating stimulation (difference in peak contraction rate between red and green stimuli, and difference in peak contraction rate between blue and yellow stimuli) is extracted as an evaluation index for color vision fusion function. The above features are integrated into a 12-dimensional feature vector, including the contraction intensity and temporal characteristics (latency + peak time) of 4 monochromatic stimuli and the contraction intensity difference of 2 mixed color stimuli.

[0060] The multidimensional color vision sensitivity feature vector is Z-score standardized to obtain pupillary reflex feature parameters, which include monochromatic stimulus parameters (contraction intensity, latency, peak time, and diameter standard deviation of red / blue / green / yellow stimuli), mixed color stimulus parameters (contraction intensity difference between red-green / blue-yellow stimuli), and comprehensive evaluation parameters (comprehensive score of color vision sensitivity and score of color vision fusion function).

[0061] Specifically, extracting contraction intensity, temporal, and stability features from standardized pupillary reflex dynamic parameters allows for precise quantification of the pupil's response to color stimuli. The contraction intensity feature uses the pupillary area change rate from 100ms after stimulus presentation to the peak contraction time as the core indicator. Calculated as (baseline area - instantaneous area) / baseline area × 100%, it accurately reflects the intensity change of pupillary contraction. The temporal features extract the contraction latency and peak time, clarifying the time progression from stimulus reception to the onset of contraction and reaching minimum pupillary area, helping to understand the timeliness of visual information transmission and pupillary response. The stability feature quantifies the fluctuation of the pupillary response by calculating the standard deviation of the pupillary diameter during the stimulus duration, providing an objective basis for assessing the stability of the pupillary response.

[0062] A multi-dimensional color vision sensitivity feature vector was constructed based on contraction intensity, temporal, and stability characteristics. This vector was further used to generate a comprehensive color vision sensitivity score and extract an evaluation index for color vision fusion function, achieving a comprehensive assessment of color vision sensitivity. The comprehensive score was generated by weighting and summing the contraction intensity characteristics of each monochromatic stimulus according to specific weights, taking into account the pupil's response to different color stimuli and thus reflecting an individual's overall sensitivity to multiple colors. Simultaneously, the contraction intensity difference under mixed-color grating stimulation was extracted as an evaluation index for color vision fusion function, assessing an individual's ability to perceive color mixing. This multi-dimensional, multi-index evaluation method, compared to single-feature evaluation, can more comprehensively and accurately reflect various aspects of color vision sensitivity.

[0063] Standardizing the multidimensional color vision sensitivity feature vector yields pupillary reflex feature parameters, which effectively enhances the universality and comparability of the data. Different subjects may have significant differences in pupillary responses. Standardization can adjust these feature parameters to a uniform dimension and range, eliminating the influence of different data dimensions and orders of magnitude.

[0064] Based on pupillary reflex characteristic parameters and an integrated model, qualitative and quantitative analysis results of color vision deficiency were obtained, including:

[0065] Based on the CNN+Random Forest ensemble model, the input pupil reflex feature parameters are used to determine the type of color vision deficiency. The input of CNN is the time series of pupil diameter, and the output is the preliminary probability of the type of color vision deficiency (red-green color blindness / blue-yellow color blindness / normal). The input of Random Forest is statistical features (contraction intensity, latency, difference), and the output is the type-corrected probability. The final classification result is determined by weighted voting (CNN weight 60%, Random Forest weight 40%).

[0066] Based on the LSTM+multivariate linear regression model, pupil reflex feature parameters are mapped to color vision sensitivity scores to obtain color vision sensitivity scores. The LSTM input is the pupil diameter time series, and the output is the time series feature encoding vector (dimension 16). The multivariate linear regression input is the time series encoding vector + statistical features, and the output is the color vision sensitivity score (range 0-100). The final score is generated by weighted summation of the LSTM output (70% weight) and the regression output (30% weight).

[0067] Specifically, an ensemble model combining CNN and random forest is used to determine the type of color vision deficiency. This approach fully leverages the strengths of both models. CNN excels at processing time-series data, automatically learning complex patterns and features in pupil diameter time sequences and capturing subtle differences in pupil response over time. Random forest, on the other hand, has excellent processing capabilities for statistical features (contraction intensity, latency, and difference), effectively analyzing the distribution and relationships of these features. By integrating the results of the two models through weighted voting (CNN weight 60%, random forest weight 40%), the advantages of different models can be comprehensively utilized, improving the accuracy of color vision deficiency classification and reducing the risk of misclassification that may occur with a single model. Secondly, this ensemble method increases the reliability of the classification results. Different models may analyze the data from different perspectives. When the results of the two models are consistent, the confidence of the classification can be enhanced. When the results differ, the weighted voting mechanism can comprehensively consider the performance and importance of the models, providing a more reasonable final classification and avoiding misclassification caused by the limitations of a single model.

[0068] The color vision sensitivity score is calculated based on an LSTM + multiple linear regression model. This method combines the ability of LSTM to process pupil diameter time series with the ability of multiple linear regression to analyze time series encoding vectors and statistical features. LSTM can deeply explore long-term dependencies in time series and capture the dynamic changes in pupil response; multiple linear regression can comprehensively consider time series feature encoding vectors and statistical features, effectively integrating these different types of information to more comprehensively reflect various aspects of color vision sensitivity. Secondly, the final score is generated by weighted summation of LSTM output (70% weight) and regression output (30% weight). This weighting method fully considers the importance of time series features in color vision sensitivity assessment while also taking into account the role of statistical features, making the generated color vision sensitivity score more accurate and able to more accurately quantify an individual's sensitivity to color.

[0069] In some embodiments, the method further includes:

[0070] Based on the type of color vision deficiency and color vision sensitivity score, and by associating it with fundus risk, a structured report is generated. The specific steps include:

[0071] The logistic regression model was used, with input variables including color vision deficiency type (coded value: red-green color blindness = 1, blue-yellow color blindness = 2, normal = 0) and color vision sensitivity score (continuous value, range 0-100 points), and output variable being the probability of fundus lesion risk (range 0%-100%).

[0072] Risk levels are categorized based on probability: low risk (probability < 30%), medium risk (30% ≤ probability ≤ 70%), and high risk (probability > 70%).

[0073] Generate a structured report, including color vision analysis results and fundus risk results. The color vision analysis results include type (e.g., "blue-yellow blindness"), sensitivity score (e.g., "38 points"), and bias rate (e.g., "-62%)". The fundus risk results include risk level (e.g., "high"), associated diseases (e.g., "age-related macular degeneration"), and risk probability (e.g., "82%)".

[0074] Specifically, a logistic regression model is used, taking color vision deficiency type (represented by coded values, such as red-green color blindness = 1, blue-yellow color blindness = 2, normal = 0) and color vision sensitivity score (continuous value, range 0-100) as input variables, and outputting the probability of fundus disease risk (range 0%-100%). The logistic regression model can scientifically analyze the intrinsic relationship between color vision-related factors and fundus disease risk through data training, find the correlation weights between them, and thus achieve a quantitative assessment of the probability of fundus disease risk. This quantitative method makes the assessment of fundus risk more objective and accurate, avoiding errors from subjective judgment. Secondly, according to the risk probability, three levels are divided: low risk (probability < 30%), medium risk (30% ≤ probability ≤ 70%), and high risk (probability > 70%), making the assessment results of fundus risk more intuitive and easy to understand. Different levels help to quickly identify the patient's risk level. For low-risk patients, the frequency of examinations can be appropriately reduced, reducing medical costs and the burden on patients; for medium- and high-risk patients, monitoring and intervention measures can be strengthened to detect problems in a timely manner and take treatment measures to improve the prevention and control of the disease.

[0075] A system for measuring color vision based on pupillary reflex, comprising:

[0076] The stimulation and acquisition module is used to stimulate the pupil based on multi-wavelength colors and acquire dynamic parameters of pupillary reflection.

[0077] The preprocessing module is used to preprocess the dynamic parameters of pupillary reflexes to obtain standardized dynamic parameters of pupillary reflexes.

[0078] The feature extraction module is used to extract features based on standardized pupil reflex dynamic parameters to obtain pupil reflex feature parameters;

[0079] The intelligent analysis module is used to perform analysis based on pupillary reflex characteristic parameters and integrated models to obtain qualitative and quantitative analysis results of color vision deficiency.

[0080] A computing device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of any of the above-described methods for measuring color vision based on pupillary reflex.

[0081] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of any of the above-described methods for measuring color vision based on pupillary reflex.

[0082] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for measuring color vision based on pupillary reflex, characterized in that, Specifically, it includes: The pupil is stimulated by multiple wavelengths of color, and dynamic parameters of pupil reflection are collected. The dynamic parameters of pupillary reflex are preprocessed to obtain standardized dynamic parameters of pupillary reflex. Feature extraction is performed on the standardized dynamic parameters of pupillary reflex to obtain pupillary reflex feature parameters; Based on the pupillary reflex characteristic parameters and integrated model, qualitative and quantitative analysis results of color vision deficiency were obtained.

2. The method for measuring color vision based on pupillary reflex according to claim 1, characterized in that, The method of stimulating the pupil with multi-wavelength color specifically includes: Dark adaptation procedures are performed on subjects in a standardized darkroom; After dark adaptation, the pupils are stimulated using a unified stimulation scheme based on the DKL color space.

3. The method for measuring color vision based on pupillary reflex according to claim 2, characterized in that, The acquisition of dynamic parameters of pupillary reflex specifically includes: An infrared camera and a coaxial infrared LED light source are used to synchronously collect dynamic parameters of pupillary reflex, including changes in pupil diameter, changes in pupil area, contraction latency, peak contraction rate, contraction speed, and time to peak, and the eye movement position is recorded as quality control data.

4. The method for measuring color vision based on pupillary reflex according to claim 3, characterized in that, The preprocessing of the pupillary reflex dynamic parameters to obtain standardized pupillary reflex dynamic parameters specifically includes: The isolated forest algorithm was used to remove artifacts from the dynamic parameters of pupil reflexes, resulting in dynamic parameters of pupil reflexes after removing outlier data. Baseline calibration was performed on the dynamic parameters of pupillary reflexes after removing outlier data to obtain standardized dynamic parameters of pupillary reflexes.

5. The method for measuring color vision based on pupillary reflex according to claim 4, characterized in that, The process of extracting features from standardized pupillary reflex dynamic parameters to obtain pupillary reflex feature parameters specifically includes: Basic reflex features are extracted from standardized pupillary reflex dynamic parameters to obtain constriction intensity features, temporal features, and stability features; Based on the characteristics of contraction intensity, temporal sequence, and stability, a multi-dimensional color vision sensitivity feature vector is constructed; The multidimensional color vision sensitivity feature vector is standardized to obtain pupil reflex feature parameters.

6. The method for measuring color vision based on pupillary reflex according to claim 5, characterized in that, The analysis based on pupillary reflex characteristic parameters and an integrated model yields qualitative and quantitative analysis results of color vision deficiency, specifically including: Based on the CNN+random forest ensemble model, the type of color vision deficiency is determined by inputting pupil reflex feature parameters. Based on the LSTM+multivariate linear regression model, pupillary reflex feature parameters are mapped to color vision sensitivity scores to obtain color vision sensitivity scores.

7. The method for measuring color vision based on pupillary reflex according to claim 6, characterized in that, The method further includes: Based on the type of color vision deficiency and color vision sensitivity score, a structured report is generated by associating fundus risk.

8. A system for measuring color vision based on pupillary reflex, characterized in that, include: The stimulation and acquisition module is used to stimulate the pupil based on multi-wavelength colors and acquire dynamic parameters of pupillary reflection. The preprocessing module is used to preprocess the dynamic parameters of pupillary reflexes to obtain standardized dynamic parameters of pupillary reflexes. The feature extraction module is used to extract features based on standardized pupil reflex dynamic parameters to obtain pupil reflex feature parameters; The intelligent analysis module is used to perform analysis based on pupillary reflex characteristic parameters and integrated models to obtain qualitative and quantitative analysis results of color vision deficiency.

9. A computing device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of the method for measuring color vision based on pupillary reflex as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the method for measuring color vision based on pupillary reflex as described in any one of claims 1-7.

Images