CN116309336A - Extraction method of key image marker of vascular cognitive impairment - Google Patents
Extraction method of key image marker of vascular cognitive impairment Download PDFInfo
- Publication number
- CN116309336A CN116309336A CN202310098020.6A CN202310098020A CN116309336A CN 116309336 A CN116309336 A CN 116309336A CN 202310098020 A CN202310098020 A CN 202310098020A CN 116309336 A CN116309336 A CN 116309336A
- Authority
- CN
- China
- Prior art keywords
- image
- magnetic resonance
- index
- indexes
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 208000037820 vascular cognitive impairment Diseases 0.000 title claims abstract description 39
- 239000003550 marker Substances 0.000 title claims abstract description 30
- 238000000605 extraction Methods 0.000 title abstract description 18
- 238000000034 method Methods 0.000 claims abstract description 28
- 238000002598 diffusion tensor imaging Methods 0.000 claims abstract description 21
- 230000001537 neural effect Effects 0.000 claims abstract description 15
- 238000007781 pre-processing Methods 0.000 claims abstract description 15
- 238000003064 k means clustering Methods 0.000 claims abstract description 11
- 238000011160 research Methods 0.000 claims abstract description 11
- 238000002599 functional magnetic resonance imaging Methods 0.000 claims abstract description 9
- 230000008309 brain mechanism Effects 0.000 claims abstract description 5
- 238000000611 regression analysis Methods 0.000 claims abstract description 4
- 210000004556 brain Anatomy 0.000 claims description 22
- 238000002595 magnetic resonance imaging Methods 0.000 claims description 16
- 238000012360 testing method Methods 0.000 claims description 16
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims description 15
- 238000009499 grossing Methods 0.000 claims description 13
- 230000000284 resting effect Effects 0.000 claims description 13
- 230000006870 function Effects 0.000 claims description 12
- 238000012549 training Methods 0.000 claims description 12
- 208000010877 cognitive disease Diseases 0.000 claims description 11
- 238000001914 filtration Methods 0.000 claims description 10
- 238000012216 screening Methods 0.000 claims description 10
- 238000009792 diffusion process Methods 0.000 claims description 9
- 230000002792 vascular Effects 0.000 claims description 9
- 238000012937 correction Methods 0.000 claims description 7
- 238000009826 distribution Methods 0.000 claims description 7
- 238000003384 imaging method Methods 0.000 claims description 7
- 208000025174 PANDAS Diseases 0.000 claims description 6
- 208000021155 Paediatric autoimmune neuropsychiatric disorders associated with streptococcal infection Diseases 0.000 claims description 6
- 240000004718 Panda Species 0.000 claims description 6
- 235000016496 Panda oleosa Nutrition 0.000 claims description 6
- 208000035475 disorder Diseases 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 230000004886 head movement Effects 0.000 claims description 5
- 238000012800 visualization Methods 0.000 claims description 5
- 238000002474 experimental method Methods 0.000 claims description 4
- 210000004885 white matter Anatomy 0.000 claims description 4
- 241001286408 Gretna Species 0.000 claims description 3
- 238000001790 Welch's t-test Methods 0.000 claims description 3
- 210000001175 cerebrospinal fluid Anatomy 0.000 claims description 3
- 230000007717 exclusion Effects 0.000 claims description 3
- 210000004884 grey matter Anatomy 0.000 claims description 3
- 238000012417 linear regression Methods 0.000 claims description 3
- 230000003862 health status Effects 0.000 claims description 2
- 230000008569 process Effects 0.000 claims description 2
- 239000000945 filler Substances 0.000 claims 1
- 238000003745 diagnosis Methods 0.000 abstract description 5
- 201000010099 disease Diseases 0.000 description 9
- 230000003925 brain function Effects 0.000 description 8
- 238000010801 machine learning Methods 0.000 description 5
- 230000002159 abnormal effect Effects 0.000 description 4
- 238000002610 neuroimaging Methods 0.000 description 4
- 206010012289 Dementia Diseases 0.000 description 3
- 201000004810 Vascular dementia Diseases 0.000 description 3
- 230000035945 sensitivity Effects 0.000 description 3
- 230000032683 aging Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000013399 early diagnosis Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000005311 nuclear magnetism Effects 0.000 description 2
- 230000003950 pathogenic mechanism Effects 0.000 description 2
- 208000020016 psychiatric disease Diseases 0.000 description 2
- 208000028698 Cognitive impairment Diseases 0.000 description 1
- 208000026106 cerebrovascular disease Diseases 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004138 cluster model Methods 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 230000003340 mental effect Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 230000003557 neuropsychological effect Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 208000011580 syndromic disease Diseases 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/05—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves
- A61B5/055—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
- G06V10/763—Non-hierarchical techniques, e.g. based on statistics of modelling distributions
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10088—Magnetic resonance imaging [MRI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30016—Brain
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Theoretical Computer Science (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pathology (AREA)
- Software Systems (AREA)
- Radiology & Medical Imaging (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Primary Health Care (AREA)
- Biophysics (AREA)
- High Energy & Nuclear Physics (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Epidemiology (AREA)
- Veterinary Medicine (AREA)
- Probability & Statistics with Applications (AREA)
- Heart & Thoracic Surgery (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Multimedia (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The application provides a method for extracting key image markers of vascular cognitive impairment, which comprises the steps of 1) acquiring resting-state functional magnetic resonance imaging data and magnetic resonance diffusion tensor imaging data of normal people and vascular cognitive impairment patients, and analyzing and extracting the resting-state functional magnetic resonance imaging data and the magnetic resonance diffusion tensor imaging data to obtain image indexes of multi-mode magnetic resonance neural image data; 2) Preprocessing image indexes of multi-mode magnetic resonance neural image data; 3) Selecting an image index and constructing a model; 4) Extracting an image marker; 5) Regression analysis of image markers with neurocognitive scale. The method develops a multi-modal neural image marker extraction method by using an unsupervised K-means clustering mode, and finds a key image marker in a plurality of indexes of multi-modal neural image data; provides assistance and basis for the early accurate diagnosis and treatment of VCI and the research of clinical VCI brain mechanism.
Description
The invention field:
the invention relates to the field of neuroimaging markers, and particularly provides a method for extracting a key imaging marker of vascular cognitive impairment.
The background technology is as follows:
vascular cognitive dysfunction (vascular cognitive impairment, VCI) refers to a class of cognitive impairment syndromes caused by cerebrovascular risk factors and cerebrovascular disease. With aging population, the prevalence of VCI in China is increasing, which can seriously affect the daily life quality of patients, so that the family members of the patients bear heavy mental and economic burden.
VCI encompasses all disease stages from mild cognitive dysfunction to vascular dementia originating from cerebrovascular lesions, and is divided into three subtypes according to clinical manifestations: vascular cognitive impairment of the non-demented type (vascular cognitive impairment no dementia, VCIND), vascular dementia (vascular dementia, vaD) and Mixed Dementia (MD), with VCIND being the most common subtype of VCI. According to the Chinese cognition and aging study of the team lead, VCIND accounts for 42% of the total number of patients with Chinese vascular cognitive impairment, and is the most common subtype. The Canadian senior study center found, through 5 years of follow-up, that 46% of VCIND patients would progress to VaD. Thus, it is demonstrated that early detection, early diagnosis and early intervention are performed on VCI patients,
at present, the diagnosis and typing of VCI are still mainly based on clinical manifestations and neuropsychological scales, and have large subjectivity, and are not beneficial to clinical early diagnosis and prevention. A large number of researches report that the brain structure and brain function of the patients with vascular cognitive impairment are obviously different from those of normal test, which is beneficial to the diagnosis of VCI, and the extraction of simple and objective imaging markers has a certain significance. Different mode data of magnetic resonance imaging can be used for objectively quantifying the changes of brain functions and structures, but the characteristics of brain functions, brain structures and the like are complicated and redundant. The machine learning method can well integrate and analyze the multi-modal redundancy characteristics, set constraint models of machine learning objective functions and extract indexes with contribution. The common method is usually a supervised method, and the robustness of the model cannot be sufficiently verified. The conventional machine learning method has the defects that the reasoning process is difficult to establish human-comprehensible meanings according to indexes such as the neural image characteristics of VCI and the like, and the interpretability is often not strong. In addition, the research shows that no effective method for extracting and diagnosing the VCI image marker exists at present. Therefore, there is a need to develop an unsupervised, easy to interpret, simple and objective method for extracting multimodal neuroimaging markers.
Disclosure of Invention
The image marker extraction method only uses the nuclear magnetism of the patient structure, diffusion tensor imaging and resting state functional nuclear magnetism image data. The application of the patent is applicable to extraction of different crowds regardless of the age, sex, region and race of the subject, because the patent is only related to abnormal changes of brain structure, brain network and brain function in the image data of the subject, such as the age, sex, occupation, region and race of the subject. Meanwhile, the method provided by the invention is based on the analysis of the structure, network and functions of the conventional brain image, and combines a machine learning algorithm to analyze the interaction among high-dimensional features and find out the features specific to VCI. It is noted that this image marker extraction method is not only suitable for analysis of VCI, a disease. For other mental diseases, the pathogenic mechanism is usually abnormal of brain functions, structures and networks, and the image marker extraction method can be used for excavating image markers of other diseases, has wider application and can be used for supporting the research of other diseases.
In order to solve the defects of the prior art, the invention innovatively develops a multi-modal neural image marker extraction method by using an unsupervised K-means clustering mode, and finds a key image marker in a plurality of indexes of multi-modal neural image data. And the regression of the extracted markers and the clinical scale proves that whether the selected image markers can characterize the change of brain structure and brain function or not and the correlation with clinical practice. The invention aims to provide an extraction method of a vascular cognitive impairment image marker, and the whole extraction method and system of the image marker provide assistance and basis for early and accurate diagnosis and treatment of VCI and research of clinical VCI brain mechanisms.
The basic idea of the method is as follows: the data of the multi-mode neural image is applied, and firstly, indexes which can represent brain structures, networks and brain functions in the neural image data are extracted and preprocessed. Integrating all indexes, inputting the indexes into a minimum absolute value convergence and selection operator (Least Absolute Shrinkage and Selection Operator, LASSO) to give weights to different image indexes, and selecting the threshold value of the clustering quantity and the weights according to the loss of a clustering algorithm. The clustering quantity and the weight threshold value are input into a clustering algorithm, effective image markers are extracted through constraint of classification tasks, final weights of the markers are given, and extraction of the image markers is achieved. The extracted image markers were input into a correlation vector regression (Relevant Vector Regression, RVR) model to predict the scores of the neuroscale, assessing their ability to be applied clinically.
In one aspect, the present application provides a method for extracting a key image marker of a vascular cognitive disorder, the method comprising:
1) Acquiring resting-state functional magnetic resonance imaging data and magnetic resonance diffusion tensor imaging data of a vascular cognitive disorder patient and a normal experiment group, and analyzing and extracting the resting-state functional magnetic resonance imaging data and the magnetic resonance diffusion tensor imaging data to obtain image indexes of multi-mode magnetic resonance neural image data;
2) Preprocessing image indexes of multi-mode magnetic resonance neural image data;
3) Selecting an image index and constructing a model;
4) Extracting an image marker;
5) Regression analysis of image markers with neurocognitive scale.
Further, step 1) includes:
1-1) acquiring magnetic resonance imaging data of a patient with vascular cognitive impairment and a normal experimental group
Patients with vascular cognitive impairment who met the conditions and normal experimental groups were enrolled according to inclusion exclusion criteria. Filling around the head during the acquisition prevents head movements and informs all subjects to empty the brain but not to sleep. And acquiring images of the head by using Siemens 3T magnetic resonance imaging equipment and adopting a 32-channel head coil to acquire resting state magnetic resonance imaging data and magnetic resonance diffusion tensor imaging data.
1-2) analyzing and extracting resting state functional image magnetic resonance imaging data:
the first step, preprocessing the resting state functional image magnetic resonance imaging data by using a standard pipeline in conn toolkit in Matlab, and comprises the following steps: rearranging and expanding functional images, correcting a time layer, identifying outliers, indirectly segmenting and standardizing, jointly registering functions and structures and smoothing based on Gaussian kernels with the full width and half height of 6 mm;
the second step, inputting the data before smoothing and after smoothing into a RESTPlus tool package for further processing, comprising: detrending, friston 24, gray matter and cerebrospinal fluid covariates regression and filtering;
thirdly, based on the result of the second step, combining with the AAL map to obtain the index of the functional image: low frequency vibration amplitude, fractional amplitude of low frequency fluctuation, percent amplitude of fluctuation and Kendell region consistency coefficient;
fourth, based on the result of the second step, inputting the processed data into Gretna toolkit to construct the functional connection of the ROIs-ROIs; extracting graph theory image indexes of brain ROIs based on functional connection and AAL (analog to digital) atlas: homography, mesocenter, isocenter, network efficiency, node clustering coefficients, node efficiency, node local efficiency, node shortest path length, rich club and small world indexes;
1-3) analyzing and extracting magnetic resonance diffusion tensor imaging data:
the first step, preprocessing magnetic resonance diffusion tensor imaging data based on a standard pipeline of a PANDA kit in Matlab in a Linux system, wherein the preprocessing comprises the following steps: estimating brain mask, image cropping, turbine correction and head movement correction;
secondly, calculating indexes of diffusion tensor parameters, and registering the diffusion tensor indexes from an individual space to an MNI standard space;
thirdly, combining the manually segmented white matter atlas provided in the PANDA kit to obtain a final diffuse image index: fractional anisotropy, average diffusivity, axial diffusivity, radial diffusivity, and local diffusion uniformity;
further, step 2) includes:
when a certain imaging index or a certain patient data is lacking by more than 20%, the index or the patient is excluded; after the data are removed, the overall mode of the index is used for making up for other missing values; combining all indexes together, and calculating according to the following formula to obtain standardized image indexes, so as to ensure that variables are in a unified dimension; carrying out disorder treatment on the integral index;
wherein x is i As the i-th value in the index, u x Std is the mean value of the whole index x Is the standard deviation of the index.
Further, step 3) includes:
firstly, screening the characteristics of all image indexes, verifying whether each index accords with variance homogeneity, if the variances are equal, filtering out the characteristics with obvious differences by using T test, otherwise, using Welch T test;
secondly, on the premise of preliminarily screening out image indexes, further screening out important features by using a LASSO regression algorithm, and avoiding the phenomenon of over-fitting of the model; the weight of each image index, i.e. the degree of contribution of the index, is obtained using the following formula:
wherein m is the number of features, y i For the actual label, x i Corresponding to y i W is the weight of the image index,
as a regular term of the term, I W i || 1 Is 1 norm.
Thirdly, setting threshold values of the number and the weight of clusters, and organizing parameters by using a grid search mode; the clustering number is 1-11, and the weight threshold value
Wherein min and max are minimum and maximum functions respectively, and coef is an index weight value output by LASSO; combining the two parameters into a grid, defining parameters by combining an elbow rule and a loss function of a K-means clustering algorithm, and taking an abscissa value of a loss curve as an index threshold value or clustering number when the loss curve has a first inflection point:
where m is the number of samples, x i For the (i) th sample,
representing the center point corresponding to the cluster; after defining the number of clusters and the characteristic threshold value, establishing a final K-means clustering model.
Further, step 4) includes:
after the K-means clustering model is established, a clustering model visualization tool package is applied to visualize the image indexes, and finally the contributed image markers and the corresponding weight values are visualized according to different weight distributions.
Further, step 5) includes:
firstly, extracting contributed image markers from multi-mode image indexes, and carrying out disorder and distribution on features and labels according to the proportion of 7:3, wherein 70% is used as a training set, and 30% is used as a testing set; in order to prevent possible data leakage, respectively carrying out standardized processing on the training set and the testing set;
secondly, inputting standardized data into an RVR model, training by using a training set, and testing by using a testing set; the scores of the predicted neurocognitive scale are compared to the actual scores and the correlation between the two is verified by linear regression and Pearson.
In another aspect, the application provides the application of the method in research of brain mechanisms of vascular cognitive impairment.
In another aspect, the present application provides the use of the above method in macroscopic statistics or research of the health status of a population, said use not comprising diagnostic purposes.
Examples of such applications include, but are not limited to, studies of the likelihood of onset of vascular cognitive dysfunction in a particular population, studies of the association of vascular cognitive dysfunction with other diseases, and the like.
The method for combining fMRI and DTI has highest sensitivity and specificity, and the extracted image marker has higher correlation with a nerve scale in clinical screening, so that the sensitivity and the sensitivity of the extracted image marker are disclosed. The extracted image marker and the extraction method can provide assistance and basis for early and accurate diagnosis and treatment service and for the research of clinical VCI brain mechanism, and have certain popularization significance and value.
Drawings
Image markers with influence in the cluster model of fig. 1 and importance: A. clustering is considered as the weight condition corresponding to the normal experiment group; B. clustering is considered as the weight case corresponding to VCI. The method comprises the steps of carrying out a first treatment on the surface of the
FIG. 2; in the resting state functional image and diffusion tensor imaging mode, the brain region of the image marker is visualized: A. visualization of brain regions where the image markers are located under the resting state functional images; B. and visualizing the brain region where the image marker is located under diffusion tensor imaging.
FIG. 3 predicts the results of neurocognitive scale based on imaging markers (A-G are different predicted results);
fig. 4 is a technical roadmap of the method of the present application.
Detailed Description
The image marker extraction method is irrelevant to the conditions of age, sex, occupation, region, whether the basic disease is combined or not, and the like, and is only relevant to abnormal changes of brain structure, network and functions in image data. Meanwhile, the method provided by the design is based on the analysis of the structure, the network and the functions of the conventional brain image, and combines a machine learning algorithm to analyze the interaction among high-dimensional features and find out the features specific to VCI. It is noted that this image marker extraction method is not only suitable for analysis of VCI, a disease. For other mental diseases, the pathogenic mechanism is usually abnormal of brain functions, structures and networks, and the image marker extraction method can be used for excavating image markers of other diseases, has wider application and can be used for supporting the research of other diseases.
Example 1 specific procedure for the extraction of neuroimaging markers for VCI:
the basic flow of the neural image marker extraction method of the VCI is shown in FIG. 4:
1) Acquiring resting-state functional magnetic resonance imaging data and magnetic resonance diffusion tensor imaging data of a vascular cognitive disorder patient and a normal experiment group, and analyzing and extracting the resting-state functional magnetic resonance imaging data and the magnetic resonance diffusion tensor imaging data to obtain image indexes of multi-mode magnetic resonance neural image data;
1-1) acquiring magnetic resonance imaging data of a patient with vascular cognitive impairment and a normal experimental group
Patients with vascular cognitive impairment who met the conditions and normal experimental groups were enrolled according to inclusion exclusion criteria. Filling around the head during the acquisition prevents head movements and informs all subjects to empty the brain but not to sleep. And acquiring images of the head by using Siemens 3T magnetic resonance imaging equipment and adopting a 32-channel head coil to acquire resting state magnetic resonance imaging data and magnetic resonance diffusion tensor imaging data.
1-2) analyzing and extracting resting state functional image magnetic resonance imaging data:
the first step, preprocessing the resting state functional image magnetic resonance imaging data by using a standard pipeline in conn toolkit in Matlab, and comprises the following steps: functional image rearrangement and expansion, temporal layer correction, outlier identification, indirect segmentation and normalization, functional and structural joint registration, smoothing based on 6mm full width half height gaussian kernels.
And secondly, preprocessing different resting state magnetic resonance image indexes has certain difference, and data before and after smoothing is selected according to indexes for further processing. The smoothed pre/post data is input into the rest tool package for further processing, mainly comprising: detrence, covariate regression (Friston 24, gray matter and cerebrospinal fluid) and filtering.
Thirdly, based on the treatment in the second step, combining with the AAL map to obtain the index of the functional image: low frequency vibration amplitude (pre-trending smoothing but not filtering), fractional amplitude of low frequency fluctuations (pre-trending smoothing but not filtering), percent amplitude of fluctuations (pre-trending smoothing and filtering), and kendel region consistency coefficients (final filtering and smoothing).
And fourthly, performing trend removal, covariate regression and filtering operation on the data after smoothing in the pretreatment, and inputting the processed data into a Gretna tool kit to construct the functional connection of the ROIs-ROIs. Graph theory image indexes of brain ROIs can be extracted based on functional connection and AAL (analytical instrument) maps: homography, mesocenter, isocenter, network efficiency, node cluster coefficients, node efficiency, node local efficiency, node shortest path length, rich Club (Rich Club), and small world indicators.
1-3) analyzing and extracting magnetic resonance diffusion tensor imaging data:
the first step, preprocessing magnetic resonance diffusion tensor imaging data based on a standard pipeline of a PANDA kit in Matlab in a Linux system, wherein the preprocessing comprises the following steps: brain mask estimation, image cropping, turbine correction, and head motion correction.
And secondly, calculating the index of the diffusion tensor parameter, and registering the diffusion tensor index from the individual space to the MNI standard space.
Thirdly, combining the manually segmented white matter atlas provided in the PANDA kit to obtain a final diffuse image index: fractional anisotropy, average diffusivity, axial diffusivity, radial diffusivity, and local diffusion uniformity.
2) Preprocessing image indexes of multi-mode magnetic resonance neuroimaging data:
it is difficult to achieve good results by directly inputting the multi-modal imaging index into the model, and preprocessing is a necessary step before modeling. When a certain imaging index or a certain patient data is missing more than 20%, the index or patient will be excluded. After deleting the data, the overall mode of the index is used to compensate for other missing values. And then, combining all indexes together, and calculating according to the following formula to obtain a standardized image index, so as to ensure that the variables are in a unified dimension.
Wherein x is i As the i-th value in the index, u x Std is the mean value of the whole index x Is the standard deviation of the index. And finally, carrying out disorder treatment on the whole index.
3) Selecting image indexes and constructing a model:
firstly, screening the characteristics of all the image indexes, firstly verifying whether each index accords with variance homogeneity, if the variances are equal, using T test to filter out the characteristics with obvious differences, otherwise using Welch T test.
And secondly, on the premise of preliminarily screening out the image indexes, further screening out important features by using a LASSO regression algorithm, and avoiding the phenomenon of over-fitting of the model. And optimizing the following formula to obtain the weight of each image index, namely the contribution degree of the index.
Wherein m is a characteristic individualNumber, y i For the actual label, x i Corresponding to y i W is the weight of the image index,
as a regular term of the term, I W i || 1 Is 1 norm.
And thirdly, adjusting and establishing model parameters. The parameters to be set are mainly the number of clusters and the threshold of weights, and the parameters are organized by using a grid search mode. Specific parameter selection range: cluster number 1-11, weight threshold
Wherein min and max are minimum and maximum functions, respectively, and coef is an index weight value output by LASSO. Combining the two parameters into a grid, defining the parameters by combining an elbow rule and a loss function of a K-means clustering algorithm, and taking an abscissa value of a point as an index threshold value or clustering number when a first larger inflection point appears on a loss curve.
Where m is the number of samples, x i For the (i) th sample,
representing the center point corresponding to the cluster. After defining the number of clusters and the characteristic threshold value, establishing a final K-means clustering model.
4) Extracting an image marker:
after the K-means clustering model is established, a clustering model visualization tool package (https:// github. Com/you fGh/kmmeans-feature-import) is applied to visualize the image indexes, and finally the contributed image markers and the corresponding weight values are visualized according to different weight distributions. The distribution of the weight values is shown in fig. 1. The brain region where the image marker is located is further visualized, and the visualization result is shown in fig. 2.
5) Regression analysis of image markers with neurocognitive scale:
firstly, extracting a contributed image marker from a multi-mode image index, and carrying out disorder and distribution on the characteristics and the labels according to the proportion of 7:3, wherein 70% is used as a training set, and 30% is used as a testing set. In order to prevent possible data leakage, the training set and the test set are respectively standardized.
And secondly, inputting standardized data into the RVR model, training by using a training set, and testing by using a testing set. The scores of the predicted neurocognitive scale were compared to the actual scores and the correlation between the two was verified by linear regression and Pearson, the results are shown in fig. 3.
Claims (8)
1. A method of extracting a key image marker of a vascular cognitive disorder, the method comprising:
1) Acquiring resting-state functional magnetic resonance imaging data and magnetic resonance diffusion tensor imaging data of a vascular cognitive disorder patient and a normal experiment group, and analyzing and extracting the resting-state functional magnetic resonance imaging data and the magnetic resonance diffusion tensor imaging data to obtain image indexes of multi-mode magnetic resonance neural image data;
2) Preprocessing image indexes of multi-mode magnetic resonance neural image data;
3) Selecting an image index and constructing a model;
4) Extracting an image marker;
5) Regression analysis of image markers with neurocognitive scale.
2. The method of claim 1, wherein step 1) comprises:
1-1) acquiring magnetic resonance imaging data of a patient with vascular cognitive impairment and a normal experimental group
Filling fillers around the head to prevent head movement in the acquisition process according to inclusion exclusion standard inclusion-conforming vascular cognitive disorder patients and normal experimental groups, informing all subjects to empty the brain but not sleep, acquiring images of the head by adopting 32-channel head coils by using Siemens 3T magnetic resonance imaging equipment, and acquiring resting state magnetic resonance imaging data and magnetic resonance diffusion tensor imaging data;
1-2) analyzing and extracting resting state functional image magnetic resonance imaging data:
the first step, preprocessing the resting state functional image magnetic resonance imaging data by using a standard pipeline in a CONN toolkit in Matlab, and comprises the following steps: rearranging and expanding functional images, correcting a time layer, identifying outliers, indirectly segmenting and standardizing, jointly registering functions and structures and smoothing based on Gaussian kernels with the full width and half height of 6 mm;
the second step, inputting the data before smoothing and after smoothing into a RESTPlus tool package for further processing, comprising: detrending, friston 24, gray matter, white matter, and cerebrospinal fluid covariates regression and filtering;
thirdly, based on the result of the second step, combining with the AAL map to obtain the index of the functional image: low frequency vibration amplitude, fractional amplitude of low frequency fluctuation, percent amplitude of fluctuation and Kendell region consistency coefficient;
fourth, based on the result of the second step, inputting the processed data into Gretna toolkit to construct the functional connection of the ROIs-ROIs; extracting graph theory image indexes of brain ROIs based on functional connection and AAL (analog to digital) atlas: homography, mesocenter, isocenter, network efficiency, node clustering coefficients, node efficiency, node local efficiency, node shortest path length, rich club and small world indexes;
1-3) analyzing and extracting magnetic resonance diffusion tensor imaging data:
the first step, preprocessing magnetic resonance diffusion tensor imaging data based on a standard pipeline of a PANDA kit in Matlab in a Linux system, wherein the preprocessing comprises the following steps: estimating brain mask, image cropping, turbine correction and head movement correction;
secondly, calculating indexes of diffusion tensor parameters, and registering the diffusion tensor indexes from an individual space to an MNI standard space;
thirdly, combining the manually segmented white matter atlas provided in the PANDA kit to obtain a final diffuse image index: fractional anisotropy, average diffusivity, axial diffusivity, radial diffusivity, and local diffusion uniformity.
3. The method of claim 1, wherein step 2) comprises:
when a certain imaging index or a certain patient data is lacking by more than 20%, the index or the patient is excluded; after the data are removed, the overall mode of the index is used for making up for other missing values; combining all indexes together, calculating and obtaining standardized image indexes according to the following formula, and ensuring that variables are in a unified dimension; carrying out disorder treatment on the integral index;
wherein x is i As the i-th value in the index, u x Std is the mean value of the whole index x Is the standard deviation of the index.
4. The method of claim 1, wherein step 3) comprises:
firstly, screening the characteristics of all image indexes, verifying whether each index accords with variance homogeneity, if the variances are equal, filtering out the characteristics with obvious differences by using T test, otherwise, using Welch T test;
secondly, on the premise of preliminarily screening out image indexes, further screening out important features by using a LASSO regression algorithm, and avoiding the phenomenon of over-fitting of the model; using the formula in (2), obtaining the weight of each image index, namely the contribution degree of the index:
wherein m is the number of features, y i For the actual label, x i Corresponding to y i W is the weight of the image index,
as a regular term of the term, I W i || 1 Is 1 norm;
thirdly, setting threshold values of the number and the weight of clusters, and organizing parameters by using a grid search mode; the clustering number is 1-11, and the weight threshold value
Wherein min and max are minimum and maximum functions respectively, and coef is an index weight value output by LASSO; combining the two parameters into a grid, defining the parameters by combining an elbow rule and a loss function of a K-means clustering algorithm, and taking an abscissa value of a loss curve as an index threshold value or clustering number when the loss curve has a first inflection point:
where m is the number of samples, x i For the (i) th sample,
representing the center point corresponding to the cluster; after defining the number of clusters and the characteristic threshold value, establishing a final K-means clustering model.
5. The method of claim 1, wherein step 4) comprises:
after the K-means clustering model is established, a clustering model visualization tool package is applied to visualize the image indexes, and finally the contributed image markers and the corresponding weight values are visualized according to different weight distributions.
6. The method of claim 1, wherein step 5) comprises:
firstly, extracting contributed image markers from multi-mode image indexes, and carrying out disorder and distribution on the image markers and the labels according to the proportion of 7:3, wherein 70% is used as a training set, and 30% is used as a test set; in order to prevent possible data leakage, respectively carrying out standardized processing on the training set and the testing set;
secondly, inputting standardized data into a related vector regression model, training by using a training set, and testing by using a testing set; the scores of the predicted neurocognitive scale are compared to the actual scores and the correlation between the two is verified by linear regression and Pearson.
7. Use of the method according to any one of claims 1-6 in research of brain mechanisms of vascular cognitive impairment.
8. Use of the method according to any one of claims 1-6 in macroscopic statistics or research of the health status of a population, said use not comprising diagnostic purposes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310098020.6A CN116309336B (en) | 2023-01-20 | 2023-01-20 | Extraction method of key image marker of vascular cognitive impairment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310098020.6A CN116309336B (en) | 2023-01-20 | 2023-01-20 | Extraction method of key image marker of vascular cognitive impairment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116309336A true CN116309336A (en) | 2023-06-23 |
| CN116309336B CN116309336B (en) | 2024-06-11 |
Family
ID=86780608
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310098020.6A Active CN116309336B (en) | 2023-01-20 | 2023-01-20 | Extraction method of key image marker of vascular cognitive impairment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116309336B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150018664A1 (en) * | 2013-07-12 | 2015-01-15 | Francisco Pereira | Assessment of Traumatic Brain Injury |
| CN112651955A (en) * | 2020-12-30 | 2021-04-13 | 深圳大学 | Intestinal tract image identification method and terminal device |
| CN112990266A (en) * | 2021-02-07 | 2021-06-18 | 西安电子科技大学 | Method, device, equipment and storage medium for processing multi-modal brain image data |
| CN115205011A (en) * | 2022-06-15 | 2022-10-18 | 海南大学 | Bank user portrait model generation method based on LSF-FC algorithm |
| CN115359045A (en) * | 2022-10-19 | 2022-11-18 | 之江实验室 | Image convolution neural network disease prediction system based on multi-mode magnetic resonance imaging |
-
2023
- 2023-01-20 CN CN202310098020.6A patent/CN116309336B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150018664A1 (en) * | 2013-07-12 | 2015-01-15 | Francisco Pereira | Assessment of Traumatic Brain Injury |
| CN112651955A (en) * | 2020-12-30 | 2021-04-13 | 深圳大学 | Intestinal tract image identification method and terminal device |
| CN112990266A (en) * | 2021-02-07 | 2021-06-18 | 西安电子科技大学 | Method, device, equipment and storage medium for processing multi-modal brain image data |
| CN115205011A (en) * | 2022-06-15 | 2022-10-18 | 海南大学 | Bank user portrait model generation method based on LSF-FC algorithm |
| CN115359045A (en) * | 2022-10-19 | 2022-11-18 | 之江实验室 | Image convolution neural network disease prediction system based on multi-mode magnetic resonance imaging |
Non-Patent Citations (1)
| Title |
|---|
| 秦琪 等: "利用弥散张量成像技术观察小血管病所致非痴呆型血管性认知障碍患者白质结构损害与认知下降的关系", 首都医科大学学报, vol. 42, no. 3, pages 373 - 384 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116309336B (en) | 2024-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhu et al. | Dual attention multi-instance deep learning for Alzheimer’s disease diagnosis with structural MRI | |
| Liu et al. | Anatomical landmark based deep feature representation for MR images in brain disease diagnosis | |
| Liu et al. | Classification of Alzheimer's disease using whole brain hierarchical network | |
| Buvaneswari et al. | Deep learning-based segmentation in classification of Alzheimer’s disease | |
| WO2022256063A9 (en) | Predictive prognosis based on multimodal analysis | |
| Minhas et al. | A nonparametric approach for mild cognitive impairment to ad conversion prediction: Results on longitudinal data | |
| CN112348785B (en) | Epileptic focus positioning method and system | |
| CN112927187A (en) | Method for automatically identifying and positioning focal cortical dysplasia epileptic focus | |
| Zhu et al. | Discovering dense and consistent landmarks in the brain | |
| CN116312772A (en) | Early treatment effect prediction method for antidepressant for depression based on rs-fMRI and P11 gene DNA methylation | |
| Xi et al. | Brain Functional Networks with Dynamic Hypergraph Manifold Regularization for Classification of End-Stage Renal Disease Associated with Mild Cognitive Impairment. | |
| Xi et al. | Hypergraph representation of multimodal brain networks for patients with end-stage renal disease associated with mild cognitive impairment | |
| CN114373095A (en) | Alzheimer disease classification system and method based on image information | |
| Feng et al. | Improved detection of focal cortical dysplasia in normal‐appearing FLAIR images using a Bayesian classifier | |
| Jeong et al. | Neonatal encephalopathy prediction of poor outcome with diffusion-weighted imaging connectome and fixel-based analysis | |
| CN116309336B (en) | Extraction method of key image marker of vascular cognitive impairment | |
| CN112200810A (en) | Multi-modal automated ventricular segmentation system and method of use thereof | |
| Supriyanti et al. | Coronal slice segmentation using a watershed method for early identification of people with Alzheimer's | |
| CN111652843A (en) | Decision support method and system based on nuclear magnetic resonance structure image weighting | |
| CN109242816B (en) | Brain glioma pathological grade imaging auxiliary judgment method based on resting state brain function contralateral positioning | |
| CN114847922A (en) | Brain age prediction method based on automatic fiber bundle identification | |
| TW202111617A (en) | Method, non-transitory computer-readable media and apparatus for age prediction of regional brain area | |
| CN113571148B (en) | One-key mental image individualized brain function report generation system, equipment and storage medium | |
| Supriyanti et al. | Coronal slices segmentation of mri images using active contour method on initial identification of alzheimer severity level based on clinical dementia rating (CDR) | |
| CN112837807B (en) | Early intelligent high-precision auxiliary diagnosis method for T2DM brain aging cognitive impairment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information |
Inventor after: Qin Qi Inventor after: Tang Yi Inventor after: Li Chunlin Inventor after: Xing Yi Inventor after: Qu Junda Inventor after: Yin Junsi Inventor before: Tang Yi Inventor before: Qin Qi Inventor before: Li Chunlin Inventor before: Xing Yi Inventor before: Qu Junda Inventor before: Yin Junsi |
|
| CB03 | Change of inventor or designer information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |