Image processing method and image processing device

Abstract

The present application provides an image processing method and an image processing device, wherein the method includes: acquiring multimodal neuroimaging data detected by a test subject from a brain detection instrument in real time, the multimodal neuroimaging data including multiple brain structure indicators and multiple brain function indicators for characterizing dynamic changes in the brain of the test subject; generating multiple brain health deviation images based on the basic information of the test subject, the device model of the brain detection instrument, multiple brain function indicators, and multiple brain structure indicators to characterize the brain health status of the test subject, wherein each brain health deviation image includes multiple brain chart curves and individualized data points, the multiple brain chart curves are generated based on standard brain chart data under corresponding brain indicators, and the individualized data points are used to characterize the ranking of the standard brain chart data detected for the test subject under corresponding brain indicators. Through this application, the accuracy of brain health status assessment is improved.

Description

Technical Field

[0001] The present application relates to the field of image processing technology, and in particular to an image processing method and an image processing device. Background Art

[0002] At present, with the continuous accumulation of neuroimaging data and the continuous advancement of standardized modeling methods, our demand for judging dynamic changes in the brain is increasing. Commonly used standardized modeling methods mainly include parametric methods, non-parametric methods, semi-parametric methods and Bayesian statistical methods. However, there are still some challenges and shortcomings in constructing personalized brain charts.

[0003] First, most of the existing brain charts are based on data from European and American populations, while the sample size of the Chinese population is relatively small and lacks representativeness. This results in the existing brain charts being unable to fully reflect the uniqueness of the Chinese population in brain development and aging, and there is also a lack of representative brain charts specifically for the Chinese population.

[0004] In addition, existing brain diagrams also have some limitations in modeling. Most brain diagrams mainly focus on the modeling of brain structure, while the modeling of important functional indicators of the brain, such as blood perfusion in brain regions, blood-brain barrier permeability, neural activity patterns and other dynamic changes, is still imperfect. Summary of the Invention

[0005] In view of this, an object of the present application is to provide an image processing method and an image processing device to overcome at least one of the above-mentioned drawbacks.

[0006] In a first aspect, an embodiment of the present application provides an image processing method, the method comprising: acquiring, in real time, multimodal neuroimaging data detected for a subject to be tested from a brain detection instrument, the multimodal neuroimaging data comprising multiple brain structural indicators and multiple brain function indicators for characterizing dynamic changes in the brain of the subject to be tested; and generating, based on basic information of the subject to be tested, a device model of the brain detection instrument, the multiple brain function indicators, and the multiple brain structural indicators, multiple brain health deviation reports for characterizing the brain health status of the subject to be tested, wherein the multiple brain health deviation reports comprise multiple first brain health deviation images and multiple second brain health deviation images, each first brain health deviation image corresponding to a brain function indicator, each second brain health deviation image corresponding to a brain structure indicator, each brain health deviation image comprising multiple brain chart curves and individualized data points, the multiple brain chart curves being generated based on standard brain chart data for corresponding brain indicators, and the individualized data points for characterizing the ranking of the standard brain chart data detected for the subject to be tested for the corresponding brain indicators.

[0007] In an optional embodiment of the present application, the basic information of the subject to be tested includes age and gender, the horizontal coordinate of each brain health deviation image is age, the vertical coordinate of each brain health deviation image is the unit of the corresponding brain indicator, and the multiple brain chart curves of each brain health deviation image include a first brain indicator development curve and at least one second brain indicator development curve. The first brain indicator development curve is used to represent the change curve of the average value of all detection values ​​of the corresponding brain indicators of all age groups under the gender of the subject to be tested and the device model of the brain detection instrument used by the subject to be tested for brain detection. There is a preset gap between each data point of the second brain indicator development curve and each data point in the first brain indicator development curve.

[0008] In an optional embodiment of the present application, multiple brain health deviation images of the subject to be tested are determined in the following manner: the age, gender, device model of the brain detection instrument, the multiple brain function indicators, and the multiple brain structure indicators of the subject to be tested are input into a brain health assessment model to obtain multiple brain health deviation images, wherein the brain health assessment model is used to characterize the relationship between target parameters and brain health status, and the target parameters are the age, gender, multiple brain indicators obtained by brain testing of the subject to be tested, and device model of the brain detection instrument.

[0009] In an optional embodiment of the present application, the brain health deviation image also includes percentiles, wherein, for each brain health deviation image, the associated area of ​​the individualized data point in the brain health deviation image displays the corresponding percentile, and the percentile is used to characterize the proportion of the individualized data point associated with the percentile compared with the standard brain chart data at the corresponding age and corresponding brain index.

[0010] In an optional embodiment of the present application, the brain health assessment model is determined in the following manner: determining a response variable and a covariate, the response variable including a brain health deviation image characterizing the brain health status of the subject to be tested, and the covariate including age, gender, multiple brain function indicators, and multiple brain structure indicators; fitting the response variable and the covariate according to a generalized additive model to obtain the brain health assessment model.

[0011] In an optional embodiment of the present application, a brain health assessment model is trained in the following manner: a training sample set is obtained, the training sample set including multiple training samples, each training sample including a sample age, a sample gender, a device model of a sample brain detection instrument, multiple sample brain function indicators, multiple sample brain structure indicators, and multiple sample brain health deviation images; the sample age, the sample gender, the device model of the sample brain detection instrument, the multiple sample brain function indicators, and the multiple sample brain structure indicators are used as inputs of an initial brain health assessment model, and the multiple sample brain health deviation images are used as outputs of the initial brain health assessment model to train the initial brain health assessment model.

[0012] In an optional embodiment of the present application, the multiple brain structure indicators include structural magnetic resonance imaging data and diffusion tensor weighted imaging data, and the multiple brain function indicators include functional magnetic resonance imaging data, arterial spin labeling data, and vascular-water exchange imaging data.

[0013] In a second aspect, an embodiment of the present application further provides an image processing device, comprising: an acquisition module for acquiring, in real time, multimodal neuroimaging data detected for a subject to be tested from a brain detection instrument, the multimodal neuroimaging data comprising multiple brain structure indicators and multiple brain function indicators for characterizing dynamic changes in the brain of the subject to be tested; and a generation module for generating, based on basic information of the subject to be tested, a device model of the brain detection instrument, the multiple brain function indicators, and the multiple brain structure indicators, multiple brain health deviation reports for characterizing the brain health status of the subject to be tested, wherein the multiple brain health deviation reports comprise multiple first brain health deviation images and multiple second brain health deviation images, each first brain health deviation image corresponding to a brain function indicator, each second brain health deviation image corresponding to a brain structure indicator, each brain health deviation image comprising multiple brain chart curves and individualized data points, the multiple brain chart curves being generated based on standard brain chart data under corresponding brain indicators, and the individualized data points being used to characterize the ranking of the standard brain chart data detected for the subject to be tested under corresponding brain indicators.

[0014] In a third aspect, an embodiment of the present application further provides an electronic device comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, the steps of the method described above are performed.

[0015] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method described above are executed.

[0016] The image processing method and image processing device provided in the embodiments of the present application obtain multimodal neuroimaging data detected by the test subject from a brain detection instrument in real time. The multimodal neuroimaging data includes multiple brain structure indicators and multiple brain function indicators for characterizing the dynamic changes in the brain of the test subject; based on the basic information of the test subject, the device model of the brain detection instrument, multiple brain function indicators and multiple brain structure indicators, multiple brain health deviation images are generated to characterize the brain health status of the test subject, wherein each brain health deviation image includes multiple brain chart curves and individualized data points. The multiple brain chart curves are generated based on standard brain chart data under corresponding brain indicators, and the individualized data points are used to characterize the ranking of the standard brain chart data detected for the test subject under corresponding brain indicators. Through this application, the accuracy, efficiency and personalization of brain health assessment are improved, providing strong support for clinical diagnosis and treatment.

[0017] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 A flowchart of the image processing method provided in an embodiment of the present application;

[0020] Figure 2 A flowchart of determining a brain health assessment model provided in an embodiment of the present application;

[0021] Figure 3 A schematic diagram of the structure of an image processing device provided in an embodiment of the present application;

[0022] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for which protection is claimed, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, each other embodiment obtained by those skilled in the art without making creative work falls within the scope of protection of the present application.

[0024] First, the application scenarios to which this application is applicable are introduced. This application can be applied in the field of image processing technology.

[0025] Research has found that with the accumulation of neuroimaging data and the advancement of standardized modeling methods and technologies, people's demand for fitting the dynamic changes of the brain is increasing.

[0026] Currently, commonly used standardized modeling methods mainly include parametric methods, nonparametric methods, semiparametric methods, and Bayesian statistical methods. However, existing brain charts still have some shortcomings. Most brain charts are based on populations in Europe and the United States, while the sample size of the Chinese population is small and lacks representativeness (sample size is less than 3%). Therefore, existing brain charts cannot fully reflect the uniqueness of the Chinese population, and there is a lack of representative brain charts for the Chinese population. In addition, most existing brain charts focus on modeling brain anatomical structures, and the modeling of important functional indicators of the brain, such as blood perfusion of brain regions, blood-brain barrier permeability, and neural activity patterns, is still incomplete.

[0027] Based on this, an embodiment of the present application provides an image processing method and an image processing device, which obtain multimodal neuroimaging data for a test subject from a brain detection instrument in real time, wherein the multimodal neuroimaging data includes multiple brain structure indicators and multiple brain function indicators for characterizing the dynamic changes in the brain of the test subject; based on the basic information of the test subject, the device model of the brain detection instrument, multiple brain function indicators and multiple brain structure indicators, multiple brain health deviation images are generated to characterize the brain health status of the test subject, wherein each brain health deviation image includes multiple brain chart curves and individualized data points, wherein the multiple brain chart curves are generated based on standard brain chart data under corresponding brain indicators, and the individualized data points are used to characterize the sorting status of the standard brain chart data detected for the test subject under corresponding brain indicators. By simultaneously considering multiple brain function indicators and multiple brain structure indicators, the accuracy of brain health assessment is improved.

[0028] See also Figure 1 , Figure 1 This is a flow chart of the image processing method provided in the embodiment of the present application. Figure 1 As shown in , the image processing method provided by the embodiment of the present application includes:

[0029] S101. Acquire multimodal neuroimaging data for a subject to be tested from a brain detection instrument in real time.

[0030] Here, all multimodal neuroimaging data are in the Digital Imaging and Communications in Medicine (DICOM) data format, which is converted to the NIFTI format using the dcm2niix tool for easy reading and writing.

[0031] The multimodal neuroimaging data includes multiple brain structure indicators and multiple brain function indicators used to characterize the dynamic changes in the brain of the test subject.

[0032] Preferably, the multiple brain structure indicators include structural magnetic resonance imaging data and diffusion tensor weighted imaging data, and the multiple brain function indicators include functional magnetic resonance imaging data, arterial spin labeling data, and vascular-water exchange imaging data.

[0033] Among them, brain structure indicators were extracted from multimodal neuroimaging data, and the following data were obtained:

[0034] Structural MRI data (3D T1WI): Primarily used to obtain structural information about the brain, such as brain tissue distribution, brain region delineation, and white and gray matter boundaries. The FreeSurfer segmentation process begins with data preprocessing, including image registration, bias field correction, and denoising. Brain regions are separated through brain tissue identification (brain extraction). Cortical reconstruction then identifies white and gray matter boundaries and generates inner and outer cortical surfaces. Subcortical structures, such as the basal ganglia, thalamus, and ventricles, are further segmented, and standardized brain region segmentation is performed, often using templates such as the Desikan-Killiany Atlas. The segmentation process not only covers cortical and subcortical structures but also includes the demarcation of gyri and sulci. The resulting segmentation output includes the cortical surface, brain region labels, and anatomical feature data.

[0035] Diffusion tensor-weighted imaging (DKI) data provides information on the orientation, density, and integrity of white matter fiber tracts, helping to understand the brain's microstructure. Through image denoising, correction, and spatial registration, the data are fitted using a diffusion kurtosis model to calculate parameters such as the diffusion tensor and kurtosis tensor, thereby obtaining diffusion characteristics in different directions. Fractional anisotropy (FA), a key indicator, reflects the anisotropy of water molecules in a specific direction. FA values ​​are then used to analyze tissue microstructure, particularly in areas such as white matter fiber tracts.

[0036] Brain function indicators were extracted from multimodal neuroimaging data, and the following data were obtained:

[0037] Functional magnetic resonance imaging (fMRI) data reflects the brain's functional activity during specific tasks or in a resting state, revealing activation patterns and functional connectivity across different brain regions. Data preprocessing begins with artifact and noise removal, motion correction, and alignment of data across all time points. Each subject's brain image is spatially normalized and mapped to a standard brain template to facilitate cross-individual comparisons. Smoothing is then performed to enhance the signal-to-noise ratio, and time series analysis is performed to extract activation signals from different brain regions and identify brain activity associated with specific tasks or stimuli.

[0038] Arterial spin labeling (ASL) data is used to quantify brain blood perfusion, namely cerebral blood flow (CBF), which reflects changes in blood supply to different brain regions. First, data preprocessing is performed, including removing artifacts and noise, and then head motion correction is used to ensure image alignment at all time points. Next, spatial registration is performed to align images at different time points or from different subjects to a standard brain template. Then, the baseline signal is removed and the difference between the arterial spin labeling signal and the control signal is calculated to obtain a quantitative value of cerebral blood flow (CBF). Through these steps, regional cerebral blood flow information can be extracted from ASL data, revealing changes in blood supply to different brain regions.

[0039] Vascular water exchange imaging (VEXI) data provide information on blood-brain barrier permeability, helping to understand the exchange of substances between cerebral vasculature and brain tissue. Diffusion imaging data with b = 0 s / mm² and VEXI data with bf = 0 s / mm² and bd = 0 s / mm² were linearly registered to T1WI using the Advanced Normalization Tools (ANTs) normalization tool. T1WI were then registered from native space to a study-specific template using a symmetric image normalization algorithm (ANTs (Advanced Normalization Tools) algorithm). Parametric maps (such as the apparent water exchange across the blood-brain barrier, AXRBBB) were then iteratively aligned to the template by applying a linear registration transformation matrix and a previously generated deformation field. A Gaussian filter with a standard deviation of 1.5 voxels was applied to the preprocessed VEXI data. Finally, the MNI152 standard brain T1-weighted images (MNI152 T1WI) were nonlinearly registered to the corresponding templates to transform some MNI152 template-based maps into the study-specific template space, thereby deriving the water-vascular imaging permeability parameters (AXRBBB) of specific regions.

[0040] S102. Generate multiple brain health deviation reports based on the basic information of the subject to be tested, the device model of the brain detection instrument, multiple brain function indicators, and multiple brain structure indicators to characterize the brain health status of the subject to be tested.

[0041] Among them, multiple brain health deviation reports include multiple first brain health deviation images and multiple second brain health deviation images, each first brain health deviation image corresponds to a brain function indicator, each second brain health deviation image corresponds to a brain structure indicator, each brain health deviation image includes multiple brain chart curves and individualized data points, the multiple brain chart curves are generated based on standard brain chart data under corresponding brain indicators, and the individualized data points are used to represent the ranking status of the standard brain chart data under corresponding brain indicators detected for the test subject.

[0042] This application acquires multimodal neuroimaging data from brain detection instruments in real time and generates multiple brain health deviation images, which can significantly improve the accuracy, efficiency and personalization of brain health assessment, and provide strong support for clinical diagnosis and treatment. At the same time, these data and images also provide important data resources for brain science research.

[0043] Here, the basic information of the subject to be tested includes age and gender, the horizontal axis of each brain health deviation image is age, the vertical axis of each brain health deviation image is the unit of the corresponding brain indicator, and the multiple brain chart curves of each brain health deviation image include a first brain indicator development curve and at least one second brain indicator development curve. The first brain indicator development curve is used to represent the change curve of the average value of all detection values ​​of the corresponding brain indicators of all age groups under the gender of the subject to be tested and the device model of the brain detection instrument used by the subject to be tested for brain detection. There is a preset gap between each data point of the second brain indicator development curve and each data point in the first brain indicator development curve.

[0044] In an optional embodiment, based on standard brain chart data for the corresponding brain indicator (which can be the average or range derived from the test results of a large number of normal people), a first brain indicator development curve (solid black line) represents the average value of all test values ​​for the corresponding brain indicator across all age groups, depending on the test subject's gender and testing instrument. This can be considered the average human level for that indicator. A second brain indicator development curve (dashed black line) has a preset difference from the first brain indicator development curve, indicating the threshold between exceeding or falling below the average level. There is a preset interval. For example, the lower limit of the preset interval may be 60% of the average human brain health level, and the upper limit of the preset interval may be 100% of the average human brain health level. If a personalized data point is above the second brain indicator development curve representing 100% of the average human brain health level, it indicates that the test subject's brain health exceeds 100% of the average human level and is in a healthy state. Similarly, if a personalized data point is below the second brain indicator development curve representing 60% of the average human brain health level, it indicates that the test subject's brain health is below 60% of the average human level and is in a sub-healthy state, requiring further diagnosis. These thresholds are used to indicate whether the test subject's brain indicator deviates from the normal range.

[0045] Here, multiple brain structure and brain function indicators of a large number of healthy people are used to create a reference standard range for normal people across the life cycle that changes with age, which is called normal population brain chart data (standard brain chart data).

[0046] The personalized data points represent the test results of the test subject under the corresponding brain indicators. The ranking of the personalized data points in the standard brain chart data and the position of the personalized data points relative to the position of the brain chart curve can intuitively show whether the brain indicators of the test subject are higher or lower than the average level, as well as the degree of deviation.

[0047] The horizontal axis of the brain health deviation image is age, indicating the conditions of different age groups. The vertical axis of the brain health deviation image is the unit of the corresponding brain indicator, such as fraction, percentage, volume, etc. By observing the image, you can intuitively understand the health status of the tested subjects in different brain indicators, as well as the changing trends with age.

[0048] Specifically, the brain health deviation image also includes percentiles, wherein, for each brain health deviation image, the corresponding percentile is displayed in the associated area of ​​each data point in the brain health deviation image, and the percentile is used to characterize the proportion of the individualized data points associated with the percentile at the corresponding age and corresponding brain indicators compared with the standard brain chart data.

[0049] Here, a percentile is a statistic that indicates what proportion of data points in a set of data are less than or equal to a specific value. For example, the 50th percentile (median) means that half of the data points are less than or equal to that value, and the other half are greater than that value.

[0050] For each brain health deviation image, the corresponding percentile is calculated and displayed for each personalized data point. This percentile reflects the relative position between the data point of the tested subject and the standard brain chart data (usually the average value or range based on the test results of a large number of normal people) under the same age and brain indicators.

[0051] In the brain health deviation image, the associated area of ​​each personalized data point (which could be a small area around the data point or the data point itself) will show the corresponding percentile. This area can highlight the percentile information through color coding, numerical labels or other visual elements.

[0052] If the percentile of a data point is high (for example, close to 100%), it indicates that the data point is at a higher level relative to the standard brain chart data at the same age and brain indicators. On the contrary, if the percentile is low (for example, close to 0%), it indicates that the data point is at a lower level. By comparing the percentiles of different data points, we can intuitively understand whether the health status of the test subject in different brain indicators deviates from the normal range and the degree of deviation.

[0053] Doctors and researchers can use these percentile-based brain health deviation images to more accurately assess the brain health of test subjects. These images can serve as an important basis for developing personalized treatment plans, monitoring disease progression, and evaluating treatment effectiveness.

[0054] Determine a plurality of brain health deviation images of the subject to be tested by:

[0055] The age, gender, device model of the brain detection instrument of the test subject, multiple brain function indicators, and multiple brain structure indicators are input into the brain health assessment model to obtain multiple brain health deviation images, wherein the brain health assessment model is used to characterize the relationship between target parameters and brain health status, wherein the target parameters are the age, gender, multiple brain indicators obtained from the brain detection of the test subject, and the device model of the brain detection instrument.

[0056] First, collect the basic information and brain index data of the test subjects: Age: record the actual age of the test subjects; Gender: record the gender of the test subjects (male / female); Device model of brain testing instrument: record the device number of the specific instrument used for brain testing to ensure data traceability and accuracy; Multiple brain function indicators: These indicators may include quantitative assessment results of cognitive function, memory, attention, etc.; Multiple brain structure indicators: These indicators may involve quantitative data on structural characteristics such as brain volume, gray matter / white matter ratio, and sulcal morphology;

[0057] Then, a brain health assessment model is applied, which can characterize the relationship between target parameters (age, gender, brain indicators, device model) and brain health status. Multiple brain health deviation images are obtained from the model. The model will generate a series of brain health deviation images based on the input data. These images may be presented in two-dimensional or three-dimensional form and show the degree of deviation from the brain health status. These deviation images are analyzed through visual inspection or automated analysis tools to determine the specific deviation area and degree of brain health status. This helps to identify potential brain function or structural abnormalities. Finally, experts in the field of neuroscience or medical imaging interpret the deviation images to determine specific brain health problems and possible causes, and formulate a treatment plan: Based on the interpretation results, a personalized treatment or intervention plan is formulated for the test subject to improve brain health status. Regular brain testing is recommended to track changes in brain health status and adjust the treatment plan.

[0058] See also Figure 2 , Figure 2 This is a flow chart of determining a brain health assessment model provided in an embodiment of the present application. Figure 2 As shown in , the image processing method provided by the embodiment of the present application includes:

[0059] S201. Determine the response variable and covariates.

[0060] Here, the response variable includes a brain health deviation image representing the brain health status of the subject to be tested, and the covariates include age, gender, multiple brain function indicators, and multiple brain structure indicators.

[0061] The response variable here isn't a single numeric value in the traditional sense, but rather a series of images, each representing the degree of deviation from a specific brain health indicator for the test subject. However, in conventional applications of the GAMLSS model, the response variable is typically numeric. To incorporate images of deviation from brain health into the GAMLSS model, we need to extract numeric features from these images. These features can be the average deviation, maximum deviation, or sum of deviations for a specific region of the image, or other quantitative metrics derived through image analysis algorithms.

[0062] The covariates here include the following parameters:

[0063] Age: The actual age of the subject to be tested, which is a numerical variable;

[0064] Gender: The gender of the test subject, usually treated as a categorical variable, can be coded as 0 (female) and 1 (male), or other appropriate coding methods;

[0065] Multiple brain function indicators: These indicators may include cognitive function scores, memory test scores, etc., all of which are numerical variables;

[0066] Multiple brain structure indicators: may involve brain volume, gray matter / white matter ratio, etc., which are also numerical variables.

[0067] S202. Fitting the response variable and the covariate according to a generalized additive model to obtain the brain health assessment model.

[0068] First, data preprocessing was performed to extract numerical features from the brain health deviation images as the response variables of the GAMLSS model; numerical covariates (such as age, brain function indicators, and brain structure indicators) were standardized to eliminate the impact of dimensional differences on the model; and categorical variables such as gender were appropriately encoded for use in the model.

[0069] Next, we select an appropriate probability distribution based on the characteristics of the response variable. Since the response variable is a numerical feature extracted from the image, it may conform to a normal distribution, a lognormal distribution, a gamma distribution, or other distributions. For each parameter (θ1, θ2, θ3, θ4), we define a regression function gj(xi). These functions can be linear or more complex nonlinear, such as smoothing functions in generalized additive models (GAMs). For example, the four parameters (θ1, θ2, θ3, θ4) typically represent: a location parameter (such as the mean or median), a scale parameter (such as the standard deviation or variance), a shape parameter (such as skewness), and a hypershape parameter (such as kurtosis).

[0070] Use statistical software (such as the gamlss package in R) to fit the model. During the fitting process, specify the distribution type of the response variable and the regression function for each parameter. Use residual analysis and other methods to check the model's fit and ensure the rationality of the model's assumptions. Based on the diagnostic results, adjust the model structure, such as changing the distribution type or adjusting the regression function, to improve the model's predictive performance.

[0071] Modeling brain structure and function indices with age and deriving corresponding percentiles. GAMLSS allows modeling distribution parameters such as location, scale, and shape as functions of covariates. Therefore, it is possible to predict percentiles for brain volume at different ages based on the relationship between brain structure and function indices and age.

[0072] In an alternative embodiment, the basic form of the GAMLSS model is as follows:

[0073] Yi ~ f (yi∣θ1(xi), θ2(xi), θ3(xi), θ4(xi))

[0074] Here, yi is the response variable, and f (yi|θ1(xi),θ2(xi),θ3(xi),θ4(xi)) is the probability distribution given a set of location, scale, and shape parameters. Specifically, f represents a probability density function (or probability mass function), where the parameters θ1, θ2, θ3, and θ4 are modeled by the covariates. These four parameters (θ1, θ2, θ3, and θ4) typically represent: the location parameter (such as the mean or median), the scale parameter (such as the standard deviation or variance), the shape parameter (such as skewness), and the hypershape parameter (such as kurtosis), respectively.

[0075] Each parameter can be expressed as a function of a covariate through the regression model:

[0076] θj(xi)=gj(xi), where j=1, 2, 3, 4, gj(xi) is the regression function, usually a linear regression model or a more complex nonlinear function.

[0077] To model brain volume percentiles at different ages based on brain structure and function indicators (as covariates) and age (also as a covariate), one first needs to select an appropriate probability distribution to describe the brain volume data. For example, one can choose a normal, lognormal, or gamma distribution. Next, a GAMLSS model is used to model the location parameter (mean or median), scale parameter (standard deviation or variance), shape parameter (skewness), and hypershape parameter (kurtosis). These parameters can be expressed as functions of brain structure and function indicators and age. The GAMLSS model is then fitted using statistical software (such as the R package gamlss). Finally, once the model is fitted, it can be used to predict brain volume percentiles at different ages. This typically involves calculating the quantile function for a given probability p.

[0078] For normally distributed data, percentiles can be calculated using the following formula:

[0079] P(p)=μ+σ×Φ−1(p)

[0080] Where: μ is the estimated mean (location parameter), σ is the estimated standard deviation (scale parameter), and Φ−1(p) is the quantile function of the standard normal distribution, that is, the Z value corresponding to a given probability p.

[0081] The GAMLSS model provides a flexible method to model and predict the distribution parameters of complex data. By selecting appropriate distribution and regression functions, it can accurately describe and predict the relationship between brain structure and function indicators and age, and derive the percentiles of brain volume in different age groups.

[0082] Specifically, the brain health assessment model can be trained in the following ways:

[0083] Get the training sample set.

[0084] The training sample set includes multiple training samples, each training sample includes a sample age, a sample gender, a device model of a sample brain detection instrument, multiple sample brain function indicators, multiple sample brain structure indicators, and multiple sample brain health deviation images;

[0085] First, a large amount of brain health data needs to be obtained from reliable medical databases or research institutions. This data should include individuals of different ages, genders, and brain health conditions. After collecting the data, it needs to be cleaned to remove incomplete, abnormal, or duplicate data to ensure data accuracy and consistency. For each training sample, its brain health status needs to be labeled, which is usually based on a doctor's diagnosis or professional brain health assessment results.

[0086] Next, features useful for predicting brain health are extracted from the raw data, including sample age, sample gender, model of the brain monitoring instrument used, multiple sample brain function indicators (such as memory and attention), multiple sample brain structure indicators (such as brain volume and sulcus depth), and possible genetic information. Multiple samples' brain health deviation images are used as the model's output. These images can be brain scans such as MRI and CT scans, and features related to health status are extracted through image processing techniques.

[0087] The sample age, sample gender, device model of the sample brain detection instrument, multiple sample brain function indicators, and multiple sample brain structure indicators are used as inputs of the initial brain health assessment model, and multiple sample brain health deviation images are used as outputs of the initial brain health assessment model to train the initial brain health assessment model.

[0088] Choose the appropriate machine learning or deep learning model based on the complexity of the problem and the characteristics of the data. For example, convolutional neural networks (CNNs) excel at processing image data, while recurrent neural networks (RNNs) are suitable for processing time series data.

[0089] The sample age, sample gender, device model of the sample brain detection instrument, multiple sample brain function indicators, and multiple sample brain structure indicators are used as input features of the model, and multiple sample brain health deviation images are used as output targets of the model. The initial model is trained using the training sample set, and the prediction error is minimized by adjusting the model parameters. During the training process, cross-validation and other techniques can be used to evaluate the performance of the model to prevent overfitting.

[0090] Use the test dataset to evaluate the performance of the model, including indicators such as accuracy, recall rate, and F1 score. Based on the evaluation results, fine-tune the model parameters to improve the model performance. Through feature importance analysis, understand which features have the greatest impact on the model's prediction results, which helps to further optimize feature selection and model design.

[0091] The image processing method and image processing device provided in the embodiments of the present application acquire multimodal neuroimaging data detected by a brain detection instrument in real time for a subject to be tested. The multimodal neuroimaging data includes multiple brain structure indicators and multiple brain function indicators for characterizing the dynamic changes in the brain of the subject to be tested; based on the basic information of the subject to be tested, the device model of the brain detection instrument, multiple brain function indicators, and multiple brain structure indicators, multiple brain health deviation images are generated to characterize the brain health status of the subject to be tested, wherein each brain health deviation image includes multiple brain chart curves and personalized data points. The multiple brain chart curves are generated based on standard brain chart data under corresponding brain indicators, and the personalized data points are used to characterize the sorting of the standard brain chart data detected for the subject to be tested under corresponding brain indicators. Through this application, the accuracy of brain health assessment is improved, providing strong support for clinical diagnosis and treatment.

[0092] This application can flexibly respond to the challenges of uneven data distribution and multimodal fitting. First, the present invention is based on a large sample of Chinese population neuroimaging data modeling. Secondly, the standardized modeling method adopted in this application can effectively improve the accuracy of fitting when dealing with sparse data distribution, and can reasonably fuse data of different modalities, making up for the shortcomings of traditional methods in multimodal data fitting. In this way, this application can accurately depict the changing trajectory of brain structure and function, covering the change process of an individual throughout his life cycle, and forming an accurate brain chart.

[0093] Compared with the existing technology, the present application has significant advantages. First, for the first time, a brain structure and function brain diagram unique to the normal population in China has been established, which is innovative and representative. Secondly, through the optimization of the standardized modeling method, the problem of uneven data distribution can be better solved, thereby improving the accuracy and stability of the brain structure and function change model. In addition, with regard to the fusion of multimodal data, the present invention can effectively unify the modeling of different neuroimaging modalities (such as structural magnetic resonance imaging, functional magnetic resonance imaging, etc.), eliminating the difficulties of traditional methods in modal conversion. This makes the construction of brain diagrams more comprehensive and accurate. Finally, the present invention has high clinical application value and can provide reliable support for the diagnosis and treatment of neurological diseases.

[0094] This application establishes a multimodal fitting normal reference value of brain structure and function that is unique to the Chinese population (i.e., a brain chart). In addition, the brain structure and function chart has good adaptability and good transfer and generalization capabilities and can be applied to new data.

[0095] Since multimodal neuroimaging data covers multiple brain structure indicators and brain function indicators, this application can provide a more comprehensive and detailed brain health assessment, which helps to identify a variety of potential problems in brain health, including but not limited to structural abnormalities, functional disorders, etc.; real-time acquisition and processing of data means that feedback on the brain health status of the test subject can be obtained immediately. This is particularly important for situations where medical decisions need to be made quickly, such as emergency treatment during acute brain injury or disease onset; by combining the basic information of the test subject and the device model of the brain detection instrument, the generated brain health deviation image can reflect the differences between individuals, which helps to achieve more personalized brain health assessment and treatment plan formulation; the brain health deviation image intuitively shows the proportion of the test subject in the standard brain chart data under different brain indicators through multiple brain chart curves and data points. This visualization method makes it easier for doctors and other medical professionals to understand and interpret the data, thereby making more accurate diagnoses.

[0096] The images of brain health deviations provided in this application can serve as an important reference for clinical decision-making. Doctors can use the information provided by these images, combined with other clinical data, to develop more precise and effective treatment plans. This approach not only aids clinical practice but also provides new tools and perspectives for neuroscience research. By deeply analyzing these multimodal neuroimaging data, researchers can further understand the structure and function of the brain and their relationship to various neurological diseases.

[0097] Based on the same inventive concept, an image processing device corresponding to the image processing method is also provided in the embodiment of the present application. Since the principle of solving the problem by the device in the embodiment of the present application is similar to the above-mentioned image processing method in the embodiment of the present application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0098] See also Figure 3 , Figure 3 This is a schematic diagram of the structure of the image processing device provided in the embodiment of the present application. Figure 3 As shown in , the image processing device 300 includes:

[0099] An acquisition module 301 is configured to acquire multimodal neuroimaging data of a subject to be tested from a brain detection instrument in real time, wherein the multimodal neuroimaging data includes a plurality of brain structure indicators and a plurality of brain function indicators used to characterize dynamic changes in the brain of the subject to be tested;

[0100] A generation module 302 is configured to generate multiple brain health deviation reports based on the basic information of the subject to be tested, the device model of the brain detection instrument, the multiple brain function indicators, and the multiple brain structure indicators, to characterize the brain health status of the subject to be tested. The multiple brain health deviation reports include multiple first brain health deviation images and multiple second brain health deviation images, each first brain health deviation image corresponding to a brain function indicator, and each second brain health deviation image corresponding to a brain structure indicator. Each brain health deviation image includes multiple brain chart curves and individualized data points. The multiple brain chart curves are generated based on standard brain chart data under corresponding brain indicators. The individualized data points are used to characterize the ranking of the standard brain chart data under corresponding brain indicators detected for the subject to be tested.

[0101] See also Figure 4 , Figure 4 This is a schematic diagram of the structure of the electronic device provided in the embodiment of the present application. Figure 4 As shown in FIG, the electronic device 500 includes a processor 510, a memory 520 and a bus 530.

[0102] The memory 520 stores machine-readable instructions executable by the processor 510. When the electronic device 500 is running, the processor 510 communicates with the memory 520 via the bus 530. When the machine-readable instructions are executed by the processor 510, the above-mentioned Figure 1 The specific implementation of the steps of the image processing method in the method embodiment shown can be found in the method embodiment and will not be repeated here.

[0103] The embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the computer program can execute the above-mentioned Figure 1 The specific implementation of the steps of the image processing method in the method embodiment shown can be found in the method embodiment and will not be repeated here.

[0104] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0105] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. There may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some communication interface, indirect coupling or communication connection of devices or units, which may be electrical, mechanical or other forms.

[0106] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0107] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0108] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0109] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The scope of protection of the present application is not limited thereto. Although the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily conceive of changes to the technical solutions described in the above-mentioned embodiments within the technical scope disclosed in the present application, or perform equivalent replacements for some of the technical features thereof. These modifications, changes, or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. An image processing method, characterized in that: include: Acquiring multimodal neuroimaging data of a subject to be tested from a brain detection instrument in real time, wherein the multimodal neuroimaging data includes a plurality of brain structure indicators and a plurality of brain function indicators for characterizing dynamic changes in the brain of the subject to be tested; generating, based on the basic information of the subject to be tested, the device model of the brain detection instrument, the multiple brain function indicators, and the multiple brain structure indicators, multiple brain health deviation reports for characterizing the brain health status of the subject to be tested, wherein the multiple brain health deviation reports include multiple first brain health deviation images and multiple second brain health deviation images, each first brain health deviation image corresponding to a brain function indicator, each second brain health deviation image corresponding to a brain structure indicator, each brain health deviation image including multiple brain chart curves and individualized data points, the multiple brain chart curves being generated based on standard brain chart data under corresponding brain indicators, and the individualized data points being used to characterize the ranking of the standard brain chart data under corresponding brain indicators detected for the subject to be tested; The basic information of the subject to be tested includes age and gender, the abscissa of each brain health deviation image is age, the ordinate of each brain health deviation image is the unit of the corresponding brain indicator, and the multiple brain chart curves of each brain health deviation image include a first brain indicator development curve and at least one second brain indicator development curve, the first brain indicator development curve is used to represent a change curve of the average value of all detection values ​​of the corresponding brain indicator for all age groups under the gender of the subject to be tested and the device model of the brain detection instrument used by the subject to be tested for brain testing, and there is a preset gap between each data point of the second brain indicator development curve and each data point of the first brain indicator development curve; Determine a plurality of brain health deviation images of the subject to be tested by: The age, gender, device model of the brain detection instrument, the multiple brain function indicators, and the multiple brain structure indicators of the subject to be tested are input into a brain health assessment model to obtain multiple brain health deviation images, wherein the brain health assessment model is used to characterize the relationship between target parameters and brain health status, wherein the target parameters are the age, gender, multiple brain indicators obtained from brain testing of the subject to be tested, and the device model of the brain detection instrument.

2. The method according to claim 1, characterized in that The brain health deviation image also includes percentiles, wherein, for each brain health deviation image, the corresponding percentile is displayed in the associated area of ​​the individualized data point in the brain health deviation image, and the percentile is used to represent the proportion of the individualized data point associated with the percentile compared with the standard brain chart data under the corresponding age and corresponding brain index.

3. The method according to claim 2, characterized in that The brain health assessment model is determined by: Determining a response variable and covariates, wherein the response variable includes a brain health deviation image representing a brain health status of the subject to be tested, and the covariates include age, gender, multiple brain function indicators, and multiple brain structure indicators; The response variable and the covariate are fitted according to a generalized additive model to obtain the brain health assessment model.

4. The method according to claim 1, wherein The brain health assessment model is trained by: Obtaining a training sample set, wherein the training sample set includes a plurality of training samples, each training sample including a sample age, a sample gender, a device model of a sample brain detection instrument, a plurality of sample brain function indicators, a plurality of sample brain structure indicators, and a plurality of sample brain health deviation images; The sample age, the sample gender, the device model of the sample brain detection instrument, the multiple sample brain function indicators, and the multiple sample brain structure indicators are used as inputs of the initial brain health assessment model, and the multiple sample brain health deviation images are used as outputs of the initial brain health assessment model to train the initial brain health assessment model.

5. The method according to claim 1, wherein The multiple brain structure indicators include structural magnetic resonance imaging data and diffusion tensor weighted imaging data, and the multiple brain function indicators include functional magnetic resonance imaging data, arterial spin labeling data, and vascular-water exchange imaging data.

6. An image processing device, characterized in that: include: an acquisition module, configured to acquire multimodal neuroimaging data of a subject to be tested from a brain detection instrument in real time, wherein the multimodal neuroimaging data includes a plurality of brain structure indicators and a plurality of brain function indicators for characterizing dynamic changes in the brain of the subject to be tested; a generation module, configured to generate a plurality of brain health deviation reports based on the basic information of the subject to be tested, the device model of the brain detection instrument, the plurality of brain function indicators, and the plurality of brain structure indicators, to characterize the brain health status of the subject to be tested, wherein the plurality of brain health deviation reports include a plurality of first brain health deviation images and a plurality of second brain health deviation images, each first brain health deviation image corresponding to a brain function indicator, each second brain health deviation image corresponding to a brain structure indicator, each brain health deviation image including a plurality of brain chart curves and individualized data points, the plurality of brain chart curves being generated based on standard brain chart data under corresponding brain indicators, and the individualized data points being used to characterize the ranking of the standard brain chart data under corresponding brain indicators detected for the subject to be tested; The basic information of the subject to be tested includes age and gender, the abscissa of each brain health deviation image is age, the ordinate of each brain health deviation image is the unit of the corresponding brain indicator, and the multiple brain chart curves of each brain health deviation image include a first brain indicator development curve and at least one second brain indicator development curve, the first brain indicator development curve is used to represent a change curve of the average value of all detection values ​​of the corresponding brain indicator for all age groups under the gender of the subject to be tested and the device model of the brain detection instrument used by the subject to be tested for brain testing, and there is a preset gap between each data point of the second brain indicator development curve and each data point of the first brain indicator development curve; The generating module determines a plurality of brain health deviation images of the subject to be tested in the following manner: The age, gender, device model of the brain detection instrument, the multiple brain function indicators, and the multiple brain structure indicators of the subject to be tested are input into a brain health assessment model to obtain multiple brain health deviation images, wherein the brain health assessment model is used to characterize the relationship between target parameters and brain health status, wherein the target parameters are the age, gender, multiple brain indicators obtained from brain testing of the subject to be tested, and the device model of the brain detection instrument.

7. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate via the bus, and the processor executes the machine-readable instructions to perform the steps of any one of the methods described in claims 1 to 5.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are executed.

Images