CN103838788A - Data analyzing method based on fMRI brain activation data warehouse - Google Patents
Data analyzing method based on fMRI brain activation data warehouse Download PDFInfo
- Publication number
- CN103838788A CN103838788A CN201210491021.9A CN201210491021A CN103838788A CN 103838788 A CN103838788 A CN 103838788A CN 201210491021 A CN201210491021 A CN 201210491021A CN 103838788 A CN103838788 A CN 103838788A
- Authority
- CN
- China
- Prior art keywords
- data
- bold
- fmri
- brain
- data warehouse
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/283—Multi-dimensional databases or data warehouses, e.g. MOLAP or ROLAP
Abstract
The invention discloses a data analyzing method based on an fMRI brain activation data warehouse. The method includes the following steps that a basic framework of the fMRI data warehouse is built, data extraction is carried out, and dimensions and a fact table are introduced; data processing and multi-dimensional analysis are carried out. Basic principles and technologies of the data warehouse are used for secondary processing of data in brain activation researches, and a multi-dimensional analysis technology is introduced into analysis and processing of result data of brain activation, so that secondary analysis, processing and integration can be carried out, new rules are found, and deep mechanisms of advanced brain activities are disclosed. Accordingly, scattered brain function imaging data and results in the prior art can be used orderly.
Description
Technical field
The present invention relates to a kind of Data warehouse and data mining technology be introduced to the data analysing method based on fMRI brain activation data warehouse, belong to field of medical image processing.
Background technology
With the change of Functional MRI (functional Magnetic Resonance Imaging, fMRI) energy real-time follow-up signal.The thinking activities that for example occur within several seconds only, or the variation of signal in cognitive experiment, temporal resolution reaches 1s.Along with continuous maturation and the development of this technology, increasing scholar uses the higher nervous activity of fMRI research brain.FMRI has greatly strengthened the ability that we carry out brain research, be widely used in recent years the brain function researches such as sensory perception, motion control and language acknowledging, and these research activitiess has also produced a large amount of experimental datas.SPM is to be cerebral function imaging data analysis specially, and he can solve the comparison problem between different images data, has provided the result with statistical significance.
The concept of " data warehouse " proposing first mid-term 1980's is that William Mr. H.Inmon defines in its " setting up data warehouse " book.He thinks: data warehouse in management and decision subject-oriented, integrated, with time correlation, not revisable data acquisition.But along with data warehouse development and perfect, its service also should break through the scope of management and decision, extends to other field.In the face of the new demand of cerebral function imaging research, many functions of data warehouse, as the excavation of data deep layer, multidimensional analysis and dynamic queries etc., can be served as strong instrument, for it provides brand-new method.
Although SPM provides a large amount of data and results for experiment, along with deepening continuously of brain science research, the excavation of data has constantly been produced to new demand, SPM cannot meet the requirement of brain science research completely.This just need to introduce new technology, finds new rule thereby data are carried out to secondary analysis, processing and integration, discloses the deeper mechanisms of brain high-grade movable, makes cerebral function imaging data at present at random and result obtain ordering and uses.Therefore, our target is exactly data and the result providing in conjunction with SPM, how data warehouse theory and scheme is transplanted in the middle of the brain science research of FMRI.
Summary of the invention
The problems referred to above that exist for solving available data analytical technology, the present invention will design a kind of data analysing method based on fMRI brain activation data warehouse, comprises the following steps:
The method for building up of A, FMRI data warehouse;
The basic framework of A1, FMRI data warehouse, holds the overall framework of data warehouse on the whole;
A2, data pick-up; The raw data that the data of data warehouse not directly obtain, but through SPM data after treatment;
A3, dimension and the fact; Mainly comprise: tested personnel's dimension, test mode dimension, brain region dimension, brain coordinate dimension, experimental period sequence dimension, fact table;
B, carry out data processing;
Weighting picture and time have formed the function of, are designated as y=bold(x, y, z, n), wherein x, y, z is respectively the three-dimensional coordinate of tissue points; N is the time, and its temporal resolution is determined according to the sweep velocity of machinery and equipment;
For a tissue points, its three-dimensional coordinate is fixed, so function can be write again as y=bold(x0, y0, z0, n);
B1, normalization: the time series S=application of formula 1 to each voxel is carried out normalization
Here bold(x0, y0, z0, n) be the fMRI signal intensity of this voxel of tissue points (x0, y0, z0) n two field picture.
Described B1 is that sequence merges:
Application of formula 2 is asked poor and is formed a new sequence two different task sequences; Thus each tissue points is analyzed along the variation of the difference of time shaft;
Δ bold (x, y, z, n)=bold
1(x, y, z, n)-bold
2(x, y, z, n) formula 2
Wherein, bold
1(x, y, z, n) and bold
2(x, y, z, n) is not the signal intensity of two different tasks;
Described B1 is that sequence is calculated: the object that sequence is calculated is the more complicated calculations of carrying out between different sequences that may occur from now in order to process, and allows using multiple sequences as variable, carries out the calculating of various complexity; Calculative sequence is chosen, and be a variable of this sequence name, then in expression formula, write the computing formula of deal with data, system provides many middle functions, as trigonometric function, exponential sum logarithm etc.
Compared with prior art, the present invention has following beneficial effect:
1, set up the basic framework of fMRI data warehouse due to the present invention, carry out data pick-up and introduce dimension and fact table, so can carry out integration, optimization again to result on original information data processing basis, improve the efficiency of processing;
2, because usage data of the present invention warehouse ultimate principle and technology are carried out secondary treating brain and are activated the data in research, multidimensional analysis technology is introduced in the result data analyzing and processing of brain activation, thereby find new rule so can carry out secondary analysis, processing and integration, disclose the deeper mechanisms of brain high-grade movable, and then can make cerebral function imaging data at present at random and result obtain ordering utilization;
If 3 use original disposal route, by SPM system, result is carried out to simple analyzing and processing, it can not obtain the analysis result specific to point, reduces the practicality of data; And the present invention proposes set up data warehouse and data are excavated can accomplish to obtain potential logical laws in to experimental subjects analysis, the experimental result that brain is activated is more scientific.
Accompanying drawing explanation
2, the total accompanying drawing of the present invention, wherein:
Fig. 1 is the overall frame structure schematic diagram of data warehouse of the present invention.
Fig. 2 multi-dimensional data cube.
Fig. 3 is the two dimension slicing of Fig. 2.
Embodiment
Below in conjunction with accompanying drawing, the present invention is described further.
For a fixed body vegetarian refreshments or fixed area, y=bold(x0, y0, z0, n) variation tendency of function curve reflected the situation of change of this point or region signal in whole fMRI experimentation.Obviously, get different points or region and will draw different function curves.Can be by the bold(x of zones of different relatively, y, z, n in research) function curve identifies functional response Nao district.In the cortical area that the point here or region generally have signal to occur in prediction, choose, the process of choosing also combines statistical analysis method.
(1) newly-built analytical database: analyzing before data, set up a new analysis data word bank.By in data input database and utilize database to analyze it.
(2) data importing: the data that analyze are imported to database from file, and put into data channel.Native system can provide at most 64 data channel.
It can import four kinds of data sources: the occurrence in region.It is under MNI coordinate system, to each moment, by brain certain a bit centered by, the mean value of the numerical value of all tissue points that R is radius; Maximal value number.It is in brain to each moment, exceed some peaked all tissue points quantity; Voxel point value and.It is in brain to each moment, all voxel point values and; Auxiliary function value.Analyze may need to use in data procedures as 01 or trigonometric function etc. carry out assistant analysis.Therefore also can be such function numerical value input database.
(3) editor calculates: for the treatment of the situation of complicated calculations more between the different pieces of information passage that may occur.
(4) graphical analysis: data are shown with the form of figure.
(5) statistical treatment: select one group of data in data channel or computing formula, and make various statistics by each tissue points along time shaft in time series file, as correlation analysis and t check etc.
A certain region of interest tissue points is that a function is y=bold(x0 in value in the same time not, y0, z0, n), if want that the variation in the blood oxygenation level Shi Yugai district of understanding those parts of cerebral cortex is closely related, just can use correlation analysis, obtain the related coefficient between each tissue points function in this district and full brain, it will change in-1～+ 1 scope, and its absolute value is larger, show to be correlated with more remarkable, vice versa.Result can obtain which part Yu Gai district work in brain is correlated with.
In SPM, check to represent with t the difference that the signal of picture element changes between functional control and task.Analyzing before SPMt file, determine in advance that a level of significance test P(is as 0.01 or 0.05) and with this, statistic of every bit is carried out to significance tests, only have t value analyzed higher than the tissue points ability of t check region of rejection.
(1) coordinate inquiry: by the name storage in all tissue points three-dimensional coordinates under MNI conventional coordinates and Talarich coordinate He Nao district corresponding to this coordinate in same table of database, and by these numberings of storage order according to three-dimensional array, each point (x like this, y, z) corresponding coordinate numberings all.Therefore,, as long as user provides centre coordinate and the search radius that will search for, system just draws all coordinate, t value and affiliated brain districts that meet the requirements a little in this region, and arranges out by order from big to small.
(2) multidimensional analysis: for watching under different situations (different testees or different task), the volume of the activation of t value in Ge Nao district, t value exceedes the tissue points number of t check region of rejection.These data can also be gathered along brain region dimension, obtain upper level brain district activation volume, or select Yi Genao district to carry out test, thereby obtain the activation volume etc. in the next stage Ge Nao district of this district's jurisdiction.
In dimension, select tested personnel's dimension, brain region peacekeeping test mode dimension observed data, experimental data can form a data cube.As Fig. 2 (left side), cubical X-axis represents test mode, and Y-axis represents tested personnel, and Z axis represents different brain region (figure Midbrain Area is jurisdiction Ge Nao district under ground floor " Left Cerebrum " brain district).Numeral in cube is the volume activating, and t value exceedes the tissue points number of t check region of rejection.All corresponding coordinates (x, y, z) of the numeral of any lattice in cube, it represents tested personnel y, the z region in brain, the volume of the activation in the time of test x task.After data cube establishes, just can cut into slices to it, merge and add up.
For example, get a scale in tested personnel dimension, along test mode and the dimension section of two, brain region, as Fig. 2 (right side), the 2-D data now obtaining section represent these personnel in different test modes, the value of each brain district activation volume.Can further analyze on this basis or make it to generate corresponding chart visual representation.These data can also be gathered along brain region dimension simultaneously, obtain " Left Cerebrum " brain district activation volume, or select Yi Genao district to carry out test, thereby obtain the activation volume in the next stage Ge Nao district of this district's jurisdiction.
Claims (3)
1. the data analysing method based on fMRI brain activation data warehouse, comprises the following steps:
The method for building up of A, FMRI data warehouse;
The basic framework of A1, FMRI data warehouse, holds the overall framework of data warehouse on the whole;
A2, data pick-up; The raw data that the data of data warehouse not directly obtain, but through SPM data after treatment;
A3, dimension and the fact; Mainly comprise: tested personnel's dimension, test mode dimension, brain region dimension, brain coordinate dimension, experimental period sequence dimension, fact table;
B, carry out data processing;
Weighting picture and time have formed the function of, are designated as y=bold(x, y, z, n), wherein x, y, z is respectively the three-dimensional coordinate of tissue points; N is the time, and its temporal resolution is determined according to the sweep velocity of machinery and equipment;
For a tissue points, its three-dimensional coordinate is fixed, so function can be write again as y=bold(x0, y0, z0, n);
B1, normalization: the time series S=application of formula 1 to each voxel is carried out normalization
formula 1
Here bold(x0, y0, z0, n) be the fMRI signal intensity of this voxel of tissue points (x0, y0, z0) n two field picture.
2. a kind of data analysing method based on fMRI brain activation data warehouse according to claim 1, is further characterized in that: described B1 is that sequence merges:
Application of formula 2 is asked poor and is formed a new sequence two different task sequences; Thus each tissue points is analyzed along the variation of the difference of time shaft;
Δ bold (x, y, z, n)=bold
1(x, y, z, n)-bold
2(x, y, z, n) formula 2
Wherein, bold
1(x, y, z, n) and bold
2(x, y, z, n) is not the signal intensity of two different tasks.
3. a kind of data analysing method based on fMRI brain activation data warehouse according to claim 1, is further characterized in that:
Described B1 is that sequence is calculated: the object that sequence is calculated is the more complicated calculations of carrying out between different sequences that may occur from now in order to process, and allows using multiple sequences as variable, carries out the calculating of various complexity; Calculative sequence is chosen, and be a variable of this sequence name, then in expression formula, write the computing formula of deal with data, system provides many middle functions, as trigonometric function, exponential sum logarithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210491021.9A CN103838788A (en) | 2012-11-27 | 2012-11-27 | Data analyzing method based on fMRI brain activation data warehouse |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210491021.9A CN103838788A (en) | 2012-11-27 | 2012-11-27 | Data analyzing method based on fMRI brain activation data warehouse |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103838788A true CN103838788A (en) | 2014-06-04 |
Family
ID=50802297
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210491021.9A Pending CN103838788A (en) | 2012-11-27 | 2012-11-27 | Data analyzing method based on fMRI brain activation data warehouse |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103838788A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182489A (en) * | 2014-08-11 | 2014-12-03 | 同济大学 | Query processing method for text big data |
| CN105653861A (en) * | 2015-12-31 | 2016-06-08 | 首都医科大学宣武医院 | Functional magnetic resonance imaging (fMRI) based category concepts similarity calculation method |
| CN107330948A (en) * | 2017-06-28 | 2017-11-07 | 电子科技大学 | A kind of fMRI data two-dimensional visualization methods based on popular learning algorithm |
| CN109522894A (en) * | 2018-11-12 | 2019-03-26 | 电子科技大学 | A method of detection fMRI brain network dynamic covariant |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6003036A (en) * | 1998-02-12 | 1999-12-14 | Martin; Michael W. | Interval-partitioning method for multidimensional data |
| CN1564160A (en) * | 2004-04-22 | 2005-01-12 | 重庆市弘越科技有限公司 | Method of seting up and inquirying multiple-demensional data cube |
| CN101042746A (en) * | 2006-03-21 | 2007-09-26 | 上海浦东国际集装箱码头有限公司 | Container pier intellectualized report forms based on store house and method |
-
2012
- 2012-11-27 CN CN201210491021.9A patent/CN103838788A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6003036A (en) * | 1998-02-12 | 1999-12-14 | Martin; Michael W. | Interval-partitioning method for multidimensional data |
| CN1564160A (en) * | 2004-04-22 | 2005-01-12 | 重庆市弘越科技有限公司 | Method of seting up and inquirying multiple-demensional data cube |
| CN101042746A (en) * | 2006-03-21 | 2007-09-26 | 上海浦东国际集装箱码头有限公司 | Container pier intellectualized report forms based on store house and method |
Non-Patent Citations (2)
| Title |
|---|
| 孙波: "脑认知状态fMRI数据的分析及分类方法研究", 《中国优秀硕士学位论文全文数据库 医药卫生科技辑》 * |
| 汲业等: "数据仓库在脑功能成像中的应用", 《中国医学影像技术》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182489A (en) * | 2014-08-11 | 2014-12-03 | 同济大学 | Query processing method for text big data |
| CN104182489B (en) * | 2014-08-11 | 2018-04-27 | 同济大学 | A kind of inquiry processing method of text big data |
| CN105653861A (en) * | 2015-12-31 | 2016-06-08 | 首都医科大学宣武医院 | Functional magnetic resonance imaging (fMRI) based category concepts similarity calculation method |
| CN105653861B (en) * | 2015-12-31 | 2018-03-23 | 首都医科大学宣武医院 | Similarity calculation method between a kind of class concepts based on fMRI |
| CN107330948A (en) * | 2017-06-28 | 2017-11-07 | 电子科技大学 | A kind of fMRI data two-dimensional visualization methods based on popular learning algorithm |
| CN107330948B (en) * | 2017-06-28 | 2020-05-12 | 电子科技大学 | fMRI data two-dimensional visualization method based on popular learning algorithm |
| CN109522894A (en) * | 2018-11-12 | 2019-03-26 | 电子科技大学 | A method of detection fMRI brain network dynamic covariant |
| CN109522894B (en) * | 2018-11-12 | 2021-08-27 | 电子科技大学 | Method for detecting dynamic covariation of fMRI brain network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Mensen et al. | Advanced EEG analysis using threshold-free cluster-enhancement and non-parametric statistics | |
| CN102663414B (en) | Cerebrum cognitive status recognition method based on cerebral function imaging | |
| CN103838788A (en) | Data analyzing method based on fMRI brain activation data warehouse | |
| CN105842642A (en) | Fractional anisotropy microstructure characteristic extraction method based on kurtosis tensor and apparatus thereof | |
| Zakharova et al. | Visual and cognitive interpretation of heterogeneous data | |
| CN105395198A (en) | Method for acquiring brand new diffusion magnetic resonance imaging contrast ratio and application of method | |
| CN105894493A (en) | FMRI data feature selection method based on stability selection | |
| CN109917759B (en) | Information presentation format and function layout method for intelligent control production line system | |
| CN103838736A (en) | Establishment method for FMRI brain activation data warehouse | |
| Kaiser et al. | Rapid contextualization of fragmented scene information in the human visual system | |
| Ltifi et al. | Using perspective wall to visualize medical data in the Intensive Care Unit | |
| EP2679147A2 (en) | Method and device for encoding eye and viewing direction data | |
| CN102525468A (en) | Cerebral function imaging-based method for analyzing intracerebral activation trend | |
| Green et al. | Topology of our cosmology with persistent homology | |
| Szymczyk et al. | A Comparison of the Entrepreneurial Performance of Asian-Oceanian Countries via the Multi-Criteria Decision-Making Techniques of Critic, Aras, Waspas, Mairca and Borda Count Methods | |
| Abd Karim et al. | Bibliometric Analysis of Data Mining on Medical Imaging | |
| CN103646160B (en) | A kind of denoising bearing calibration in fMRI task state data prediction | |
| Kolahkaj et al. | A recommender system for web mining using neural network and fuzzy algorithm | |
| Chao et al. | Revisiting Alwyn H. Gentry’s forest transect data: latitudinal beta diversity patterns are revealed using a statistical sampling-model-based approach | |
| Varanasi et al. | Energy landscape analysis based sliding window studies of brain dynamics in young and old subjects | |
| Devalingam et al. | Global research output and growth rate study in the field of plastics technology | |
| Tan et al. | On Driver's Workload in Dangerous Scenes Based on EEG Data | |
| Wang et al. | Investigating the co-relationship between housing price and inflation by assembling wavelet based time-frequency analysis | |
| Klapka et al. | Visual analysis of search results in Scopus database focused on sustainable tourism | |
| Bhat et al. | Parametric Analysis of Fixations During Object Search using AI |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140604 |
|
| WD01 | Invention patent application deemed withdrawn after publication |