CN104207775A - Obese patient functional image analysis method based on mutual sample entropy - Google Patents
Obese patient functional image analysis method based on mutual sample entropy Download PDFInfo
- Publication number
- CN104207775A CN104207775A CN201410387786.7A CN201410387786A CN104207775A CN 104207775 A CN104207775 A CN 104207775A CN 201410387786 A CN201410387786 A CN 201410387786A CN 104207775 A CN104207775 A CN 104207775A
- Authority
- CN
- China
- Prior art keywords
- sample entropy
- mutual sample
- area
- voxel
- mutual
- 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
Abstract
The invention discloses an obese patient functional image analysis method based on mutual sample entropy. The method is characterized by comprising the following steps of collecting resting-state functional magnetic resonance data of a brain; selecting interest areas; extracting a time sequence of voxels in each interest area, up-sampling the time sequence of each voxel, and calculating the mutual sample entropy value of any two interest areas; comparing the mutual sample entropy value of any two interested area of the obese patient with the mutual sample entropy value of corresponding two interested areas of a normal testee, and determining the reason for obesity or worsening of the obesity. The method has the beneficial effects that starting with the physiological line of the obese patient through the resting-state imaging way, the changing of physiological activity of the brain of the patient can be accurately reflected; the EEG signal brain network establishing method is applied to fMRI, and the problem that the fMRI time resolution is not high can be overcome by utilizing the up-sampling method; the mutual sample entropy overcomes the matching problem of the mutual approximate entropy.
Description
Technical field
The present invention relates to a kind of image analysis methods, be specifically related to a kind of adiposis patient functional image analytical method based on mutual Sample Entropy, belong to Medical Image Processing and analysis technical field.
Background technology
At present, the research of brain network has become a focus of brain science research field.Brain network connects can be divided into structural brain network, functional brain network and because of validity brain network.Based on MRI and DTI etc., structural brain network mainly can reflect that the imaging methods of the physiological structure of brain is studied, and functional brain network and mainly can reflect that the means of brain function imaging are studied based on EEG, MEG and fMRI etc. because of validity brain network.Wherein, functional brain network is a kind of Undirected networks, and because of validity brain network be a kind of special functional brain network, its functional connection is oriented.
Functional study based on the brain network of EEG has been carried out a lot, also achieves many achievements.As found the EEG signal research of epileptic patient: corresponding to before epilepsy-outbreak in-outbreak after these 3 stages, the functional brain network of patient once shows the characteristic trending towards random networking-trend towards regular network-trend towards random network, and also namely the topological property of functional brain network shows the dynamic behavior changed with disease state.In AD (Alzheimer Disease, Alzheimer) pathological study, the people such as Stam build networking based on EEG signal and find that the brain network shortest path that patient AD builds at wave band is significantly greater than normal control.In schizoid research, find that the node connectivity of schizophrenia patients reduces, brain network clustering coefficient also reduces, and shortest path increases, and the course of disease length of the variation of these brain network parameters and schizophrenia patients is closely related.In research process, many diverse ways of constructing function brain network have also been obtained utilization, such as Phase synchronization, cross-correlation, Mutual information entropy, mutually approximate entropy etc.
Meanwhile, the functional brain network based on fMRI also has a lot of achievement.Such as the research of AD is found that the overall situation of patient's brain network and local (two hippocampus) cluster level all significantly decline, this means that disorder of overall importance appears in the functional connection of patient's brain.Another report points out that schizophrenia patients presents the functional connection of increase relative to normal control.
At present, countries in the world all face this global problem fat, it is defined as disease by World Health Organization (WHO) (WHO), it is the third-largest factor most threatening to human health after cardiovascular diseases and cancer, the fat number in the whole world (accounts for 6% of total population) more than 400,000,000, super severe one number about 1,600,000,000 (accounting for 24% of total population), WHO predict: to 2015 by have 2,300,000,000 adult overweight and 700,000,000 overweight people, and the end of the year 2008, China's obese people has broken through 9,000 ten thousand, and overweight number is more than 200,000,000; Expectation Future Ten year China's population of being obese will considerably beyond 200,000,000, and overweight population will more than 6.5 hundred million.The disease harm that obesity is brought mainly contains: type Ⅱdiabetes mellitus, coronary heart disease, hypertension, fatty liver, apoplexy, digestive tract disease, osteoarthritis and cancer (colon and rectum carcinoma, breast carcinoma, uterus carcinoma etc.), in addition, obesity has a strong impact on the quality of life of people, social acceptance is reduced, income reduces, psychological burden increases the weight of, and increases the burden of publilc health system.
Can be found by introduction above, the method much used at EEG is not also applied in fMRI.Meanwhile, the research of functional brain network is mainly carried out for disease of brain, and being applied in fMRI of non-disease of brain as fat so is not also launched.
Summary of the invention
The object of the present invention is to provide a kind of adiposis patient functional image analytical method based on mutual Sample Entropy, be intended to the problem solving the obesity human brain network change not yet understood now, the method can be used for being interconnected the description of interactively between core group, thus provides radiological evidence for the Physiological Mechanism of obesity.
In order to realize above-mentioned target, the present invention adopts following technical scheme:
Based on an adiposis patient functional image analytical method for mutual Sample Entropy, it is characterized in that, comprise the following steps:
Step one: the functional MRI data gathering brain with magnetic resonance tool with tranquillization state scan pattern;
Step 2: the difference section obtaining adiposis patient and normal subjects brain function under tranquillization state, picks out the cerebral nucleus group fat relevant to diet, aforementioned cerebral nucleus group is defined as area-of-interest;
Step 3: the time series extracting voxel in each area-of-interest, and up-sampling is carried out to the time series of the voxel of each area-of-interest, then utilize the time series after up-sampling to calculate the mutual sample entropy of any two area-of-interests;
Step 4: the mutual sample entropy of two corresponding to normal subjects for the mutual sample entropy of any two area-of-interests of adiposis patient area-of-interests is compared, when the mutual sample entropy of the two is unequal, then determine that the interaction relationship between two loops that these two area-of-interests of adiposis patient represent there occurs change, the change of the interaction relationship between these two loops is exactly the fat reason occurring or increase the weight of.
The aforesaid adiposis patient functional image analytical method based on mutual Sample Entropy, it is characterized in that, the detailed process of step 3 is:
(1) from all area-of-interests, select arbitrarily two regions to partner, be designated as area-of-interest X, area-of-interest Y respectively, suppose that area-of-interest X is made up of i voxel, then each voxel is designated as voxel x respectively
1, voxel x
2, voxel x
3..., voxel x
i, suppose that area-of-interest Y is made up of j voxel, then each voxel is designated as voxel y respectively
1, voxel y
2, voxel y
3..., voxel y
j;
(2) voxel x is extracted respectively
1with the time series of voxel y, then carry out up-sampling to the time series extracted, the time series after up-sampling is designated as time series x respectively
1(t), time series y
1(t), time series y
2(t), time series y
3(t) ..., time series y
j(t);
(3) voxel x is calculated
1up-sampling after time series x
1mutual sample entropy Q between (t) and area-of-interest Y
1Y:
1. sequence x computation time is distinguished
1(t) and time series y
1(t), time series y
2(t), time series y
3(t) ..., time series y
jt the mutual sample entropy between (), mutual sample entropy Q made by meter respectively
11, mutual sample entropy Q
12, mutual sample entropy Q
13..., mutual sample entropy Q
1j;
2. by mutual sample entropy Q
11, mutual sample entropy Q
12, mutual sample entropy Q
13..., mutual sample entropy Q
1jsummation, then obtain meansigma methods, this meansigma methods is time series x
1t the mutual sample entropy between () and area-of-interest Y, is designated as mutual sample entropy Q
1Y;
(4) the mutual sample entropy used the same method between the time series after the up-sampling calculating other voxels in area-of-interest X and area-of-interest Y, this mutual sample entropy is designated as mutual sample entropy Q respectively
2Y, mutual sample entropy Q
3Y..., mutual sample entropy Q
iY;
(5) the mutual sample entropy between area-of-interest X and area-of-interest Y is calculated:
By mutual sample entropy Q
2Y, mutual sample entropy Q
3Y..., mutual sample entropy Q
iYsummation, then obtain meansigma methods, this meansigma methods is the mutual sample entropy between area-of-interest X and area-of-interest Y, is designated as mutual sample entropy Q
xY.
The aforesaid adiposis patient functional image analytical method based on mutual Sample Entropy, it is characterized in that, in step one, also comprise the process of aforementioned data being carried out to Preprocessing, aforementioned Preprocessing comprises:
(1) time rectification is carried out to the brain function MR data collected;
(2) data after correcting the time carry out the dynamic rectification of head;
(3) data after correct dynamic rectification use EPI template to carry out Spatial normalization;
(4) space smoothing is carried out to the data after Spatial normalization.
The aforesaid adiposis patient functional image analytical method based on mutual Sample Entropy, is characterized in that, the process that the data after correct dynamic rectification carry out Spatial normalization comprises the following steps:
1. head being moved the affine transformation of 12 parameters of the data acquisition after rectification is registrated in MNI standard form;
2. the image obtained by registration is heavily cut to the voxel of 3mm × 3mm × 3mm, and by MNI coordinate transform in Talairach coordinate system.
The aforesaid adiposis patient functional image analytical method based on mutual Sample Entropy, is characterized in that, in step 2, the process obtaining adiposis patient and normal subjects difference section of brain function under tranquillization state is:
(1) low frequency amplitude slow wave concussion analyzing and processing is carried out to the data obtained;
(2) low frequency amplitude slow wave concussion numerical value is calculated;
(3) the two sample t-test based on voxel are carried out to the low frequency amplitude slow wave concussion numerical value obtained, thus obtain the difference section of adiposis patient and normal subjects brain function under tranquillization state.
The aforesaid adiposis patient functional image analytical method based on mutual Sample Entropy, is characterized in that, the process of pretreated data being carried out to low frequency amplitude slow wave concussion analyzing and processing comprises the following steps:
1. linear drift process is gone to pretreated data;
2. to go the data after linear drift to carry out bandpass filtering that frequency range is 0.01Hz-0.08Hz.
Usefulness of the present invention is:
1, with the imaging mode of tranquillization state, start with from adiposis patient physiology baseline values, accurately can reflect the change of the inherent physiological brain activity of patient body;
2, this method has been applied to the method building EEG signal brain network in functional mri (fMRI), and utilize the method for up-sampling to overcome the fMRI problem not high enough relative to EEG signal temporal resolution, make use of the advantage that fMRI spatial resolution is high simultaneously;
3, the method design using for reference Sample Entropy and mutual approximate entropy has gone out mutual Sample Entropy, and builds network with it, thus overcomes the problem that mutual approximate entropy counts self coupling.
Accompanying drawing explanation
Fig. 1 is the main flow figure of the adiposis patient functional image analytical method that the present invention is based on mutual Sample Entropy;
Fig. 2 is the sub-process figure that in method of the present invention, ALFF analyzes and mutual Sample Entropy is analyzed.
Detailed description of the invention
In order to make object of the present invention, technical scheme and advantage clearly understand, below in conjunction with the drawings and specific embodiments, concrete introduction is done to the present invention.
With reference to Fig. 1, the adiposis patient functional image analytical method based on mutual Sample Entropy of the present invention comprises the following steps:
Step one: image data
Gather the functional MRI data of adiposis patient and normal subjects's brain with tranquillization state scan pattern with magnetic resonance tool.
In order to make result more accurate, carry out Preprocessing to the data collected, this Preprocessing comprises:
1, time rectification is carried out to the brain function MR data collected
Due to the time point difference of data acquisition, so need to carry out time rectification to the brain function MR data collected, corrected by the time and can reduce this difference.
It is exactly the difference of acquisition time between layers in rectification 1 volume that time corrects, thus effectively ensures that each layer is all obtain from the identical time.
2, the data after correcting the time carry out the dynamic rectification of head
Because nuclear magnetic resonance duration of experiment is long, the meeting that the physiologic factor such as breathing, blood flow, pulsation of subjects is unavoidable causes head movement, so need to carry out the dynamic rectification of head to the data of correcting through the time.
Each two field picture of a sequence is all carried out registration with the first two field picture of this sequence by head dynamic rectification exactly, under being registrated to the same coordinate system, move to correct head, and then the data after correct dynamic rectification carry out hand inspection, if be translating beyond 1mm, rotate over 1 °, so just get rid of the data that this was corrected through the time, will not analyze.
3, the data after correct dynamic rectification carry out Spatial normalization
Owing to there is multiple subjects in experiment, and there is certain difference in the brain shape between subjects, in order to follow-up statistical analysis, the normalization of brain shape must be carried out, the brain registration of subjects is normalized in the brain template of standard, namely correctly must move the data after correcting and carry out Spatial normalization process.
Data after correct dynamic rectification are carried out Spatial normalization process main process and are comprised the following steps:
(1), by head move the affine transformation of 12 parameters of the data acquisition after rectification to be registrated in MNI (Montreal Neurological Institute, Montreal neurological institute) standard form.
(2), the brain image after registration is heavily cut to the voxel of 3mm × 3mm × 3mm, and by MNI coordinate transform in Talairach coordinate system.
4, space smoothing is carried out to the data after Spatial normalization
In order to reach reduce noise, improve noise when eliminate the nuance between different tested brain structure object, Spatial normalization adopts the gaussian kernel function of 6 millimeters of full width at half maximum (FWHM, Full Width at Half Maximum) smoothing to the data after Spatial normalization.
Step 2: pick out area-of-interest
With reference to Fig. 2, select area-of-interest and mainly comprise the following steps:
1, carry out the concussion of low frequency amplitude slow wave to the data obtained to analyze
Based on MATLAB (Matrix Lab, matrix labotstory) REST (the Resting-State Statistic Toolkit of platform, tranquillization state analysis software package) software, low frequency amplitude slow wave concussion analyzing and processing is carried out to the data obtained, is called for short ALFF (Amplitude of low-frequency fluctuation) analyzing and processing.
The process of ALFF analyzing and processing mainly comprises the following steps:
(1), linear drift process is gone to the data obtained;
(2), to go the data after linear drift to carry out bandpass filtering that frequency range is 0.01Hz-0.08Hz, in order to remove the impact of low frequency wonder and high-frequency noise.
2, the concussion of low frequency amplitude slow wave numerical value, i.e. ALFF value is calculated
The computational process of the ALFF value of each voxel is as follows:
(1) power spectrum is obtained after filter result being carried out fast Fourier transform;
(2) by power spectrum evolution;
(3) calculate the meansigma methods of the power spectrum in 0.01-0.08Hz, this meansigma methods is ALFF value;
(4) namely ALFF value is obtained standardized ALFF value (mALFF) divided by the average A LFF value of all voxels of full brain.
In 2004, the generation of electrophysiologic studies display low-frequency oscillation may be due to spontaneous neuron activity, this spontaneous neuron activity has physiological significance, be presented as that brain district produces its own circadian sexual activity pattern by there being with it the information interaction in the brain interval of connection, so have reason to think that ALFF value can as the feature of a reaction cerebral activity.
3, area-of-interest is selected
To the low frequency amplitude slow wave concussion numerical value obtained, on SPM5 (mapping of Statistic Parameter Mapping5 statistical parameter) software platform, carry out the two sample t-test based on voxel, the relatively difference of adiposis patient and normal subjects ALFF value, thus obtain the difference section of adiposis patient and normal subjects brain function under tranquillization state.
From the region with significant difference, pick out the cerebral nucleus group fat relevant to diet, this cerebral nucleus group is defined as area-of-interest, this area-of-interest is used for the mutual Sample Entropy analysis of follow-up adiposis patient and normal subjects.
Step 3: calculate mutual Sample Entropy
Extract the time series of voxel in each area-of-interest, and up-sampling is carried out to the time series of the voxel of each area-of-interest, then utilize the time series after up-sampling to calculate the mutual sample entropy of any two area-of-interests.Detailed process is as follows:
1, from all area-of-interests, select arbitrarily two regions to partner, be designated as area-of-interest X, area-of-interest Y respectively, the voxel of composition area-of-interest X is designated as voxel x, and suppose that area-of-interest X is made up of i voxel, then each voxel is designated as voxel x respectively
1, voxel x
2, voxel x
3..., voxel x
i; Equally, the voxel of composition area-of-interest Y is designated as voxel y, and suppose that area-of-interest Y is made up of j voxel, then each voxel is designated as voxel y respectively
1, voxel y
2, voxel y
3..., voxel y
j.
2, voxel x is extracted respectively
1(voxel y is comprised with voxel y
1, voxel y
2, voxel y
3..., voxel y
j) time series, then to extract time series carry out up-sampling, the time series after up-sampling is designated as time series x respectively
1(t), time series y
1(t), time series y
2(t), time series y
3(t) ..., time series y
j(t).
3, the mutual sample entropy between two area-of-interests can reflect seasonal effect in time series similarity degree between two area-of-interests, characterize the connection compactness of two area-of-interests, so utilize first order autoregressive model, from the mutual sample entropy Q between time-domain calculation area-of-interest X and area-of-interest Y
xY, thus characterize the connection compactness of two area-of-interests, this process is specific as follows:
(1) voxel x is calculated respectively
1up-sampling after time series x
1in (t) and area-of-interest Y each voxel up-sampling after time series between mutual sample entropy:
1. the time series x in area-of-interest X is calculated
1(t) and the time series y in area-of-interest Y
1t the mutual sample entropy between (), this mutual sample entropy is designated as Q
11;
2. the time series x in area-of-interest X is calculated
1(t) and the time series y in area-of-interest Y
2t the mutual sample entropy between (), this mutual sample entropy is designated as Q
12;
3. the time series x in area-of-interest X is calculated
1(t) and the time series y in area-of-interest Y
3t the mutual sample entropy between (), this mutual sample entropy is designated as Q
13;
4. by that analogy, until calculate the time series x in area-of-interest X
1(t) and the time series y in area-of-interest Y
jt the mutual sample entropy between (), this mutual sample entropy is designated as Q
1j.
(2) voxel x is calculated
1up-sampling after time series x
1mutual sample entropy Q between (t) and area-of-interest Y
1Y:
By mutual sample entropy Q
11, mutual sample entropy Q
12, mutual sample entropy Q
13..., mutual sample entropy Q
1jsummation, then obtain meansigma methods, this meansigma methods is time series x
1t the mutual sample entropy between () and area-of-interest Y, is designated as mutual sample entropy Q
1Y.
(3) use the same method and calculate other voxels in area-of-interest X and (comprise voxel x
2, voxel x
3, voxel x
4..., voxel x
i) up-sampling after time series (comprise time series x
2(t), time series x
3(t) ..., time series x
i(t)) and area-of-interest Y between mutual sample entropy, be designated as mutual sample entropy Q respectively
2Y, mutual sample entropy Q
3Y..., mutual sample entropy Q
iY.
(4) the mutual sample entropy between area-of-interest X and area-of-interest Y is calculated:
By mutual sample entropy Q
2Y, mutual sample entropy Q
3Y..., mutual sample entropy Q
iYsummation, then obtain meansigma methods, this meansigma methods is the mutual sample entropy between area-of-interest X and area-of-interest Y, is designated as mutual sample entropy Q
xY.
According to the method described above, travel through all area-of-interests, obtain the mutual sample entropy of any two area-of-interests.
Step 4: determine the fat reason occurring or increase the weight of based on mutual Sample Entropy
The mutual sample entropy of two corresponding to normal subjects for the mutual sample entropy of any two area-of-interests of adiposis patient area-of-interests is compared, when the mutual sample entropy of the two is unequal, then determine that the interaction relationship between two loops that these two area-of-interests of adiposis patient represent there occurs change, the change of the interaction relationship between these two loops is exactly the fat reason occurring or increase the weight of.
Mutual sample entropy between two area-of-interests can reflect seasonal effect in time series similarity degree between two area-of-interests, thus the connection compactness both characterizing, therefore, the brain network change of adiposis patient can be studied with the mutual sample entropy between two area-of-interests.Specifically see two examples below.
Example one: the mutual sample entropy Q between adiposis patient OFC (orbitofrontal cortex orbitfrontal cortex) and VTA (ventral tegmental area ventral tegmental area) these two regions
fatsize be 0.3527, and the mutual sample entropy Q between these two regions of OFC and VTA of normal subjects
normallysize be 0.2948, mutual sample entropy Q
fatthan mutual sample entropy Q
normally23.18% is exceeded when P value is less than 0.05
The mutual sample entropy of adiposis patient raises, and illustrates that adiposis patient reduces compared with the connection compactness between normal person OFC and VTA two area-of-interests.
OFC is the part in brain frontal cortex region, belongs to 10,11 and 47 in Brodman subregion, and primary responsibility driving effect, belongs to a part for drive circuit.And VTA is the important component part in reward circuit, in the trophic behavior of regulation and control human body, play vital effect.
Due to long-term a large amount of feeds, adiposis patient there occurs change relative to the neurophysiological mechanism of normal person, drive circuit reduces the driving dynamics of reward circuit, make reward circuit active reduction under quiescent condition, the satisfaction of food intake is reduced, need more feed just can reach the satisfaction of normal level, therefore adiposis patient can strengthen food ration, thus aggravation of causeing fat.
Example two: the mutual sample entropy Q between adiposis patient Caudate (caudatum) and Putamen (shell core) two regions
fatsize be 0.3279, the mutual sample entropy Q between normal subjects Caudate and Putamen two regions
normallysize be 0.3358, mutual sample entropy Q
fatthan mutual sample entropy Q
normally2.35% is reduced when P value is less than 0.05
The mutual sample entropy of adiposis patient reduces, and illustrates that adiposis patient increases compared with the connection compactness between normal person Caudate and Putamen two area-of-interests.
From the angle of nervous physiology, Caudate is the striatal part of veutro, belongs to the reward circuit of brain.And Putamen is the important component part of brain learning memory circuit.
Studied explanation in the past, based under tranquillization state compared with normal controls, the reward circuit level of activation of adiposis patient reduces, this shows mainly because the reward circuit of adiposis patient occurs abnormal, under identical diet control condition, adiposis patient is lower due to the level of reward circuit activity, is not easy to reach to meet state, therefore meeting by increasing to ingest, finally causeing fat.In this test, find that the interaction between adiposis patient Caudate and Putamen is strengthened to some extent, find in the research to addiction mechanism, the ability of addiction patient to the addicted thing such as learning and memory of drugs, game, food etc. strengthens.These all show in the angle of physiological mechanism, and the increase of adiposis patient to food ability of learning and memory causes the increase of the demand to food, thus order about reward circuit to acquisition food after express corresponding gratification.
In example one, example two, mutual Sample Entropy is applied to adiposis patient cerebral nucleus group change, Granger causal analysis method relevant to Pearson came in the past and has concordance, demonstrate the research that mutual Sample Entropy method not only may be used for EEG signal, also can be successfully applied in fMRI date processing and analysis.
As can be seen here, method of the present invention is the evidence that the research of fat physiological mechanism provides iconography.
In addition, method of the present invention not only utilizes up-sampling to overcome the fMRI problem not high enough relative to EEG signal temporal resolution, and overcomes with mutual Sample Entropy structure network the problem that mutual approximate entropy counts self coupling.
It should be noted that, above-described embodiment is only in order to explain the present invention, and it does not limit the present invention in any form, the technical scheme that the mode that all employings are equal to replacement or equivalent transformation obtains, and all drops in protection scope of the present invention.
Claims (6)
1., based on an adiposis patient functional image analytical method for mutual Sample Entropy, it is characterized in that, comprise the following steps:
Step one: the functional MRI data gathering brain with magnetic resonance tool with tranquillization state scan pattern;
Step 2: the difference section obtaining adiposis patient and normal subjects brain function under tranquillization state, picks out the cerebral nucleus group fat relevant to diet, described cerebral nucleus group is defined as area-of-interest;
Step 3: the time series extracting voxel in each area-of-interest, and up-sampling is carried out to the time series of the voxel of each area-of-interest, then utilize the time series after up-sampling to calculate the mutual sample entropy of any two area-of-interests;
Step 4: the mutual sample entropy of two corresponding to normal subjects for the mutual sample entropy of any two area-of-interests of adiposis patient area-of-interests is compared, when the mutual sample entropy of the two is unequal, then determine that the interaction relationship between two loops that these two area-of-interests of adiposis patient represent there occurs change, the change of the interaction relationship between these two loops is exactly the fat reason occurring or increase the weight of.
2. the adiposis patient functional image analytical method based on mutual Sample Entropy according to claim 1, it is characterized in that, the detailed process of step 3 is:
(1) from all area-of-interests, select arbitrarily two regions to partner, be designated as area-of-interest X, area-of-interest Y respectively, suppose that area-of-interest X is made up of i voxel, then each voxel is designated as voxel x respectively
1, voxel x
2, voxel x
3..., voxel x
i, suppose that area-of-interest Y is made up of j voxel, then each voxel is designated as voxel y respectively
1, voxel y
2, voxel y
3..., voxel y
j;
(2) voxel x is extracted respectively
1with the time series of voxel y, then carry out up-sampling to the time series extracted, the time series after up-sampling is designated as time series x respectively
1(t), time series y
1(t), time series y
2(t), time series y
3(t) ..., time series y
j(t);
(3) voxel x is calculated
1up-sampling after time series x
1mutual sample entropy Q between (t) and area-of-interest Y
1Y:
1. sequence x computation time is distinguished
1(t) and time series y
1(t), time series y
2(t), time series y
3(t) ..., time series y
jt the mutual sample entropy between (), mutual sample entropy Q made by meter respectively
11, mutual sample entropy Q
12, mutual sample entropy Q
13..., mutual sample entropy Q
1j;
2. by mutual sample entropy Q
11, mutual sample entropy Q
12, mutual sample entropy Q
13..., mutual sample entropy Q
1jsummation, then obtain meansigma methods, this meansigma methods is time series x
1t the mutual sample entropy between () and area-of-interest Y, is designated as mutual sample entropy Q
1Y;
(4) the mutual sample entropy used the same method between the time series after the up-sampling calculating other voxels in area-of-interest X and area-of-interest Y, this mutual sample entropy is designated as mutual sample entropy Q respectively
2Y, mutual sample entropy Q
3Y..., mutual sample entropy Q
iY;
(5) the mutual sample entropy between area-of-interest X and area-of-interest Y is calculated:
By mutual sample entropy Q
2Y, mutual sample entropy Q
3Y..., mutual sample entropy Q
iYsummation, then obtain meansigma methods, this meansigma methods is the mutual sample entropy between area-of-interest X and area-of-interest Y, is designated as mutual sample entropy Q
xY.
3. the adiposis patient functional image analytical method based on mutual Sample Entropy according to claim 1, it is characterized in that, in step one, also comprise the process of described data being carried out to Preprocessing, described Preprocessing comprises:
(1) time rectification is carried out to the brain function MR data collected;
(2) data after correcting the time carry out the dynamic rectification of head;
(3) data after correct dynamic rectification use EPI template to carry out Spatial normalization;
(4) space smoothing is carried out to the data after Spatial normalization.
4. the adiposis patient functional image analytical method based on mutual Sample Entropy according to claim 3, is characterized in that, the process that the data after correct dynamic rectification carry out Spatial normalization comprises the following steps:
1. head being moved the affine transformation of 12 parameters of the data acquisition after rectification is registrated in MNI standard form;
2. the image obtained by registration is heavily cut to the voxel of 3mm × 3mm × 3mm, and by MNI coordinate transform in Talairach coordinate system.
5. the adiposis patient functional image analytical method based on mutual Sample Entropy according to claim 1 or 3, is characterized in that, in step 2, the process obtaining adiposis patient and normal subjects difference section of brain function under tranquillization state is:
(1) low frequency amplitude slow wave concussion analyzing and processing is carried out to the data obtained;
(2) low frequency amplitude slow wave concussion numerical value is calculated;
(3) the two sample t-test based on voxel are carried out to the low frequency amplitude slow wave concussion numerical value obtained, thus obtain the difference section of adiposis patient and normal subjects brain function under tranquillization state.
6. the adiposis patient functional image analytical method based on mutual Sample Entropy according to claim 5, is characterized in that, the process of pretreated data being carried out to low frequency amplitude slow wave concussion analyzing and processing comprises the following steps:
1. linear drift process is gone to pretreated data;
2. to go the data after linear drift to carry out bandpass filtering that frequency range is 0.01Hz-0.08Hz.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410387786.7A CN104207775B (en) | 2014-08-07 | 2014-08-07 | A kind of adiposis patient functional image analytical method based on mutual Sample Entropy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410387786.7A CN104207775B (en) | 2014-08-07 | 2014-08-07 | A kind of adiposis patient functional image analytical method based on mutual Sample Entropy |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104207775A true CN104207775A (en) | 2014-12-17 |
| CN104207775B CN104207775B (en) | 2016-04-06 |
Family
ID=52089967
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410387786.7A Expired - Fee Related CN104207775B (en) | 2014-08-07 | 2014-08-07 | A kind of adiposis patient functional image analytical method based on mutual Sample Entropy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104207775B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105395194A (en) * | 2015-12-14 | 2016-03-16 | 中国人民解放军信息工程大学 | Electroencephalogram (EEG) channel selection method assisted by functional magnetic resonance imaging |
| CN106066930A (en) * | 2015-04-16 | 2016-11-02 | 西门子公司 | Utilize the peculiar assessment of experimenter of the neurological disorder of biochemistry mechanism of transmission |
| CN106709244A (en) * | 2016-12-12 | 2017-05-24 | 西北工业大学 | Brain function network modeling method for resting state synchronization EEG-fMRI |
| CN108065934A (en) * | 2017-11-23 | 2018-05-25 | 西安电子科技大学 | Loss of weight operation based on LME models causes the iconography statistical analysis technique that brain structure changes |
| CN112116984A (en) * | 2020-09-16 | 2020-12-22 | 无锡职业技术学院 | Obesity analysis intervention method facing obese college student population |
| CN113827246A (en) * | 2021-11-25 | 2021-12-24 | 北京航空航天大学杭州创新研究院 | Magnetoencephalography data acquisition and analysis method and system |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013097118A1 (en) * | 2011-12-28 | 2013-07-04 | 中国科学院自动化研究所 | Method for processing brain function magnetic resonance data |
-
2014
- 2014-08-07 CN CN201410387786.7A patent/CN104207775B/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013097118A1 (en) * | 2011-12-28 | 2013-07-04 | 中国科学院自动化研究所 | Method for processing brain function magnetic resonance data |
Non-Patent Citations (5)
| Title |
|---|
| JOSHUA S. RICHMAN, ET AL.: "Physiological time-series analysis using approximate entropy and sample entropy", 《AM J PHYSIOL HEART CIRC PHYSIOL》, vol. 278, 31 December 2000 (2000-12-31) * |
| MOSES O.SOKUNBI: "Sample entropy reveals high discriminative power between young and elderly adults in short fMRI data sets", 《FRONTIERS IN NEUROINFORMATICS》, vol. 8, 23 July 2014 (2014-07-23) * |
| WALTER S. PRITCHARD, ET AL.: "Functional Brain Networks Formed Using Cross-Sample Entropy Are Scale Free", 《BRAIN CONNECTIVITY》, vol. 4, no. 6, 1 August 2014 (2014-08-01) * |
| 冯志燕等: "基于互样本熵的脑电功能网络构建及其应用", 《电子设计工程》, vol. 21, no. 8, 30 April 2013 (2013-04-30) * |
| 刘慧等: "生物时间序列的近似熵和样本熵方法比较", 《仪器仪表学报》, vol. 25, no. 4, 31 August 2004 (2004-08-31) * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106066930A (en) * | 2015-04-16 | 2016-11-02 | 西门子公司 | Utilize the peculiar assessment of experimenter of the neurological disorder of biochemistry mechanism of transmission |
| US10311373B2 (en) | 2015-04-16 | 2019-06-04 | Siemens Healthcare Gmbh | Subject-specific assessment of neurological disorders with biochemical propagation mechanisms |
| CN106066930B (en) * | 2015-04-16 | 2020-06-09 | 西门子保健有限责任公司 | Subject-specific assessment of neurological disorders using biochemical transmission mechanisms |
| CN105395194A (en) * | 2015-12-14 | 2016-03-16 | 中国人民解放军信息工程大学 | Electroencephalogram (EEG) channel selection method assisted by functional magnetic resonance imaging |
| CN105395194B (en) * | 2015-12-14 | 2018-03-16 | 中国人民解放军信息工程大学 | A kind of brain electric channel system of selection of functional mri auxiliary |
| CN106709244A (en) * | 2016-12-12 | 2017-05-24 | 西北工业大学 | Brain function network modeling method for resting state synchronization EEG-fMRI |
| CN106709244B (en) * | 2016-12-12 | 2019-08-13 | 西北工业大学 | A kind of tranquillization state synchronizes the brain function network modeling method of EEG-fMRI |
| CN108065934A (en) * | 2017-11-23 | 2018-05-25 | 西安电子科技大学 | Loss of weight operation based on LME models causes the iconography statistical analysis technique that brain structure changes |
| CN108065934B (en) * | 2017-11-23 | 2021-01-12 | 西安电子科技大学 | LME model-based imaging statistical analysis method for brain structure change caused by weight loss surgery |
| CN112116984A (en) * | 2020-09-16 | 2020-12-22 | 无锡职业技术学院 | Obesity analysis intervention method facing obese college student population |
| CN112116984B (en) * | 2020-09-16 | 2023-08-29 | 无锡职业技术学院 | Obesity analysis intervention method facing to obese university student group |
| CN113827246A (en) * | 2021-11-25 | 2021-12-24 | 北京航空航天大学杭州创新研究院 | Magnetoencephalography data acquisition and analysis method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104207775B (en) | 2016-04-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104207775B (en) | A kind of adiposis patient functional image analytical method based on mutual Sample Entropy | |
| CN106023194B (en) | Amygdaloid nucleus spectral clustering dividing method based on tranquillization state function connects | |
| Wee et al. | Resting-state multi-spectrum functional connectivity networks for identification of MCI patients | |
| Peng et al. | Altered intrinsic regional brain activity in male patients with severe obstructive sleep apnea: a resting-state functional magnetic resonance imaging study | |
| US9839392B2 (en) | Method and system for estimating brain concussion | |
| Diykh et al. | EEG sleep stages identification based on weighted undirected complex networks | |
| CN106447023B (en) | Brain default network directed connection analysis method based on die body structure | |
| CN106264510B (en) | Modeling method for screening surgical patients | |
| CN106127769A (en) | A kind of brain Forecasting Methodology in age connecting network based on brain | |
| CN109920550A (en) | A method of teenager's lafora's disease is studied based on dMRI | |
| CN113768519B (en) | Method for analyzing consciousness level of patient based on deep learning and resting state electroencephalogram data | |
| WO2018036368A1 (en) | Method for analyzing heart rate variability, apparatus and use thereof | |
| Moridian et al. | Automatic diagnosis of sleep apnea from biomedical signals using artificial intelligence techniques: Methods, challenges, and future works | |
| Perez-Valero et al. | A review of automated techniques for assisting the early detection of Alzheimer’s disease with a focus on EEG | |
| CN110458832B (en) | Resting state brain function symmetry analysis method | |
| CN113362944B (en) | Assessment method of functional dyspepsia and needling curative effect prediction model based on machine learning | |
| Siuly et al. | SchizoGoogLeNet: the googlenet-based deep feature extraction design for automatic detection of schizophrenia | |
| CN115937092A (en) | Cognitive memory dysfunction auxiliary diagnosis device based on memory curve | |
| Bahador et al. | Multimodal spatio-temporal-spectral fusion for deep learning applications in physiological time series processing: A case study in monitoring the depth of anesthesia | |
| CN103235869B (en) | Brain loop based on heroin addiction model causation relationship analysis method | |
| Li et al. | Dataset of whole-brain resting-state fMRI of 227 young and elderly adults acquired at 3T | |
| CN103886591A (en) | Brain nuclei Granger causal analysis method based on RYGB surgery weight losing | |
| Zhao et al. | Detection of epileptic foci based on interictal iEEG by using convolutional neural network | |
| Yuan et al. | Gk-bsc: Graph kernel-based brain states construction with dynamic brain networks and application to schizophrenia identification | |
| Dauwels et al. | Diagnosis of Alzheimer’s disease using electric signals of the brain. A grand challenge |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160406 Termination date: 20210807 |