CN103006220B - Method for detecting brain function communicated area based on signal sparse approximation - Google Patents
Method for detecting brain function communicated area based on signal sparse approximation Download PDFInfo
- Publication number
- CN103006220B CN103006220B CN201210582691.1A CN201210582691A CN103006220B CN 103006220 B CN103006220 B CN 103006220B CN 201210582691 A CN201210582691 A CN 201210582691A CN 103006220 B CN103006220 B CN 103006220B
- Authority
- CN
- China
- Prior art keywords
- signal
- brain function
- functional
- data
- wavelet
- 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.)
- Expired - Fee Related
Links
Landscapes
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention discloses a method for detecting a brain function communicated area based on signal sparse approximation. The method comprises the following steps: 1, performing wavelet packet decomposition on each time point data in a functional magnetic resonance signal to obtain a wavelet tree on original time point data; 2, carrying out sparsity measurement on nodes in the wavelet tree, and selecting the node with strongest sparsity in the wavelet tree to form an effective sparse approximation data set on original functional magnetic resonance data; and 3, carrying out hybrid matrix optimization on the formed sparse approximation data set by adopting independent content analysis, and reconstructing a source signal in a functional area in combination with an original functional magnetic resonance mixed signal to finish accurate positioning and detection of the functional area. According to the method, sparse approximation of original mixed data is obtained by utilizing the more general supposed sparsity of the functional magnetic resonance signal, and the source signal is separated through the independent content analysis and the signal reconstruction, so that the purpose of accurately positioning the brain function communicated area is achieved.
Description
Technical field
The present invention relates to a kind of human brain function connected region detection method based on Functional magnetic resonance imaging, be specifically related to a kind of based on the sparse approximate blind source signal separation method of functional MRI mixed signal to the accurate location of human brain function connected region and detection method.
Background technology
Functional magnetic resonance imaging is a kind of Novel magnetic resonance imaging technique starting the nineties in 20th century to rise.This combine with technique function, dissection and image tripartite surface information, study from single morphosis the systematic study combined with function for traditional mr techniques and strong technical support is provided, simultaneously for the diagnosis etc. of the research to human brain function connected region, neuro-cognitive, brain section disease and mental illness provides favourable technical guarantee.Utilizing Functional magnetic resonance imaging to carry out in the research of human brain function connected region, effectively, the research of method to human brain function connected region that processes higher-dimension MR data accurately serves vital effect.But current Data Management Analysis method, although can complete the detection of functional area to a certain extent, all deposits more deficiency and defect, and the accuracy of its Orientation of functions also treats further raising.Such as, method of fuzzy cluster analysis is limited by the restriction that number is estimated in iteration speed, Fuzzy Exponential and functional areas; Independent component analysis is limited by the separate hypothesis of stronger functional areas source signal completely, limits the detection in Functional connectivity region.
Therefore, existing human brain function connected region detection technique is also waited to further develop and raising.
Summary of the invention
The invention provides a kind of based on the sparse approximate brain function connected region detection method of signal, sparse approximate by carrying out functional MRI mixed signal, obtain sparse approximate data collection, adopt Independent Component Analysis to carry out the blind segmentation of brain function signal again, realize more accurately detection and localization brain function connected region.
For achieving the above object, the invention provides a kind of based on the sparse approximate brain function connected region detection method of signal, be characterized in, the method includes the steps of:
Step 1, in functional magnetic resonance signal each time point data perform WAVELET PACKET DECOMPOSITION, obtain the wavelet tree about former time point data;
Step 2, sparsity metric is carried out to the node in wavelet tree, and select the strongest openness wavelet tree node, so that form the effective openness approximate data collection about former functional MRI data;
Step 3, hybrid matrix optimizing is carried out to the openness approximate data centralized procurement independent component analysis formed, and carry out the reconstruct of functional areas source signal in conjunction with former functional MRI mixed signal, to complete accurate location and the detection of functional area.
In above-mentioned steps 1, three layers of one dimension WAVELET PACKET DECOMPOSITION are carried out respectively to time point data.
The present invention is a kind of to be compared with the Data Management Analysis method of prior art based on the sparse approximate brain function connected region detection method of signal, its advantage is, first the present invention by utilizing the more general hypothesis of functional magnetic resonance signal openness, obtain the sparse approximate of former blended data, again by independent component analysis with signal reconstruction to realize being separated of source signal, thus reach the object of accurately location brain function connected region, be conducive to the research to aspects such as brain science, neuroscience and brain section diseases.
Accompanying drawing explanation
Fig. 1 is a kind of method flow diagram based on the sparse approximate brain function connected region detection method of signal of the present invention;
Fig. 2 is a kind of flow chart based on the sparse approximate brain function connected region detection method of signal of the present invention.
Detailed description of the invention
Below in conjunction with accompanying drawing, further illustrate specific embodiments of the invention.
As Fig. 1 and shown in composition graphs 2, the present invention discloses a kind of based on the sparse approximate brain function connected region detection method of signal, and the method includes the steps of:
Step 1, WAVELET PACKET DECOMPOSITION: to each time point data of functional MRI data, carry out 3 layers of one dimension WAVELET PACKET DECOMPOSITION respectively, the wherein selection of wavelet basis, have chosen the Daubechies(simultaneously possessing anti symmetry, orthogonality and biorthogonality in wavelet systems and is slightly written as in the present invention: db2 wavelet basis db) in family; After WAVELET PACKET DECOMPOSITION, can obtain the corresponding wavelet tree of each time point data (as shown in Figure 2), in wavelet tree, each node has different sparse attribute., and the one defined about original signal based on the wavelet tree node of gained is effective sparse approximate
Step 2, form sparse approximate data collection operation: this operation mainly includes the work of two aspects.
Step 2.1, one are the sparsity metric device of structure one based on the wavelet tree node of the distance metric norm (lp norm) of normed linear space, each the openness of wavelet tree node is measured, obtain an openness quality vector Q about each node, it meets, and wherein it represents the wavelet packet tree the about each time point of former functional MRI data
lthe wavelet coefficient vector that individual wavelet tree node forms jointly, and n represents the length of vector.
Step 2.2, its two according to openness size, select openness maximum wavelet tree node to form the effective sparse approximate set of one about former mixed signal, thus improve reconstruct and the stationkeeping ability of brain function signal.
Step 3, independent component analysis and reconstruct: this flow process mainly carries out independent component analysis to the sparse approximate set about former mixed signal of the formation in step 2, so that estimate about each source signal mixed coefficint A in former functional MRI data and separate mixed sparse W, combine former mixed function MR data signal again, carry out the reconstruct of source signal (or being called functional network), finally obtain required Functional connectivity network, i.e. independent element image and corresponding time course.
Although content of the present invention has done detailed introduction by above preferred embodiment, will be appreciated that above-mentioned description should not be considered to limitation of the present invention.After those skilled in the art have read foregoing, for multiple amendment of the present invention and substitute will be all apparent.Therefore, protection scope of the present invention should be limited to the appended claims.
Claims (2)
1., based on the sparse approximate brain function connected region detection method of signal, it is characterized in that, the method includes the steps of:
Step 1, in functional MRI mixed signal each time point data perform WAVELET PACKET DECOMPOSITION, obtain the wavelet tree about these time point data;
Step 2, sparsity metric is carried out to the node in wavelet tree, and select the strongest openness wavelet tree node, so that form the effective openness approximate data collection about functional MRI mixed signal;
Step 3, carry out hybrid matrix optimizing to the openness approximate data centralized procurement independent component analysis formed, and combined function magnetic resonance mixed signal carries out the reconstruct of functional areas source signal, to complete accurate location and the detection of functional area.
2. as claimed in claim 1 based on the sparse approximate brain function connected region detection method of signal, it is characterized in that, in described step 1, three layers of one dimension WAVELET PACKET DECOMPOSITION are carried out respectively to time point data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210582691.1A CN103006220B (en) | 2012-12-28 | 2012-12-28 | Method for detecting brain function communicated area based on signal sparse approximation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210582691.1A CN103006220B (en) | 2012-12-28 | 2012-12-28 | Method for detecting brain function communicated area based on signal sparse approximation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103006220A CN103006220A (en) | 2013-04-03 |
| CN103006220B true CN103006220B (en) | 2015-02-11 |
Family
ID=47955716
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210582691.1A Expired - Fee Related CN103006220B (en) | 2012-12-28 | 2012-12-28 | Method for detecting brain function communicated area based on signal sparse approximation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103006220B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103211597B (en) * | 2013-04-27 | 2014-12-17 | 上海海事大学 | Resting brain function connected region detecting method based on affine clustering |
| CN103961103B (en) * | 2014-05-07 | 2015-12-30 | 大连理工大学 | A kind of ICA to plural fMRI data is estimated the method being divided into line phase and correcting |
| CN104850863B (en) * | 2015-05-29 | 2017-11-17 | 上海海事大学 | A kind of human brain function activity state classification method |
| CN108197661B (en) * | 2018-01-19 | 2022-03-01 | 上海海事大学 | Cognitive activity state classification system and method based on voxel horizontal brain function information |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090264785A1 (en) * | 2008-04-18 | 2009-10-22 | Brainscope Company, Inc. | Method and Apparatus For Assessing Brain Function Using Diffusion Geometric Analysis |
| WO2010045724A1 (en) * | 2008-10-20 | 2010-04-29 | Neurochip Corporation | Method and rhythm extractor for detecting and isolating rhythmic signal features from an input signal using the wavelet packet transform |
| US8364254B2 (en) * | 2009-01-28 | 2013-01-29 | Brainscope Company, Inc. | Method and device for probabilistic objective assessment of brain function |
| US8838226B2 (en) * | 2009-12-01 | 2014-09-16 | Neuro Wave Systems Inc | Multi-channel brain or cortical activity monitoring and method |
-
2012
- 2012-12-28 CN CN201210582691.1A patent/CN103006220B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN103006220A (en) | 2013-04-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Deep convolutional neural network for automatically segmenting acute ischemic stroke lesion in multi-modality MRI | |
| Goceri | Skin disease diagnosis from photographs using deep learning | |
| Johnson et al. | Improving the speed of MRI with artificial intelligence | |
| Billot et al. | Robust machine learning segmentation for large-scale analysis of heterogeneous clinical brain MRI datasets | |
| CN103006220B (en) | Method for detecting brain function communicated area based on signal sparse approximation | |
| Catani et al. | Comment on “The geometric structure of the brain fiber pathways” | |
| Bian et al. | Pyramid network with online hard example mining for accurate left atrium segmentation | |
| Yoo et al. | An example-based multi-atlas approach to automatic labeling of white matter tracts | |
| CN110619641A (en) | Automatic segmentation method of three-dimensional breast cancer nuclear magnetic resonance image tumor region based on deep learning | |
| Xu et al. | Brain imaging-based machine learning in autism spectrum disorder: methods and applications | |
| Hossain et al. | Brain Tumor Auto-Segmentation on Multimodal Imaging Modalities Using Deep Neural Network. | |
| CN108875819A (en) | A kind of object and component associated detecting method based on shot and long term memory network | |
| CN106709867A (en) | Medical image registration method based on improved SURF and improved mutual information | |
| Kushwaha et al. | 3D medical image fusion using dual tree complex wavelet transform | |
| Liang et al. | Development of a framework for large scale three-dimensional pathology and biomarker imaging and spatial analytics | |
| Shah et al. | EMED-UNet: an efficient multi-encoder-decoder based UNet for medical image segmentation | |
| Liu et al. | Adaptive squeeze-and-shrink image denoising for improving deep detection of cerebral microbleeds | |
| Pan et al. | Collaborative image synthesis and disease diagnosis for classification of neurodegenerative disorders with incomplete multi-modal neuroimages | |
| Zou et al. | SwinE-UNet3+: swin transformer encoder network for medical image segmentation | |
| Wang et al. | An end‐to‐end infant brain parcellation pipeline | |
| CN115984257A (en) | Multi-modal medical image fusion method based on multi-scale transform | |
| Ng et al. | Large-scale neuroinformatics for in situ hybridization data in the mouse brain | |
| Niu et al. | Fusion of multiple spatial networks derived from complex-valued fMRI data via CNN classification | |
| CN115578361A (en) | Brain image fusion method based on convolution sparse representation | |
| CN103211597B (en) | Resting brain function connected region detecting method based on affine clustering |
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 |
Granted publication date: 20150211 Termination date: 20181228 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |