CN108553086B - Method and system for removing shallow noise of functional near-infrared brain imaging - Google Patents
Method and system for removing shallow noise of functional near-infrared brain imaging Download PDFInfo
- Publication number
- CN108553086B CN108553086B CN201810316542.8A CN201810316542A CN108553086B CN 108553086 B CN108553086 B CN 108553086B CN 201810316542 A CN201810316542 A CN 201810316542A CN 108553086 B CN108553086 B CN 108553086B
- Authority
- CN
- China
- Prior art keywords
- time
- frequency
- observation
- shallow
- noise
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000002610 neuroimaging Methods 0.000 title claims abstract description 47
- 230000009466 transformation Effects 0.000 claims abstract description 24
- 238000001914 filtration Methods 0.000 claims description 39
- 239000011159 matrix material Substances 0.000 claims description 29
- 230000015654 memory Effects 0.000 claims description 23
- 238000004590 computer program Methods 0.000 claims description 16
- 230000006870 function Effects 0.000 claims description 13
- 230000000873 masking effect Effects 0.000 claims description 5
- 210000004556 brain Anatomy 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 230000008859 change Effects 0.000 description 5
- 230000001537 neural effect Effects 0.000 description 5
- 108010064719 Oxyhemoglobins Proteins 0.000 description 4
- 230000000004 hemodynamic effect Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 210000001519 tissue Anatomy 0.000 description 4
- INGWEZCOABYORO-UHFFFAOYSA-N 2-(furan-2-yl)-7-methyl-1h-1,8-naphthyridin-4-one Chemical compound N=1C2=NC(C)=CC=C2C(O)=CC=1C1=CC=CO1 INGWEZCOABYORO-UHFFFAOYSA-N 0.000 description 3
- 210000003710 cerebral cortex Anatomy 0.000 description 3
- 108010002255 deoxyhemoglobin Proteins 0.000 description 3
- 108010054147 Hemoglobins Proteins 0.000 description 2
- 102000001554 Hemoglobins Human genes 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 2
- 239000008280 blood Substances 0.000 description 2
- 210000004369 blood Anatomy 0.000 description 2
- 230000036770 blood supply Effects 0.000 description 2
- 210000005013 brain tissue Anatomy 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012880 independent component analysis Methods 0.000 description 2
- 238000001320 near-infrared absorption spectroscopy Methods 0.000 description 2
- 230000003702 neurovascular coupling effect Effects 0.000 description 2
- 229910052760 oxygen Inorganic materials 0.000 description 2
- 239000001301 oxygen Substances 0.000 description 2
- 238000000513 principal component analysis Methods 0.000 description 2
- 230000029058 respiratory gaseous exchange Effects 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 206010005746 Blood pressure fluctuation Diseases 0.000 description 1
- 238000002835 absorbance Methods 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000004872 arterial blood pressure Effects 0.000 description 1
- 230000036772 blood pressure Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 210000003128 head Anatomy 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 210000002418 meninge Anatomy 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 230000001766 physiological effect Effects 0.000 description 1
- 230000035790 physiological processes and functions Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 210000004761 scalp Anatomy 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 210000003625 skull Anatomy 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
- A61B5/0059—Measuring for diagnostic purposes; Identification of persons using light, e.g. diagnosis by transillumination, diascopy, fluorescence
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/0033—Features or image-related aspects of imaging apparatus classified in A61B5/00, e.g. for MRI, optical tomography or impedance tomography apparatus; arrangements of imaging apparatus in a room
- A61B5/004—Features or image-related aspects of imaging apparatus classified in A61B5/00, e.g. for MRI, optical tomography or impedance tomography apparatus; arrangements of imaging apparatus in a room adapted for image acquisition of a particular organ or body part
- A61B5/0042—Features or image-related aspects of imaging apparatus classified in A61B5/00, e.g. for MRI, optical tomography or impedance tomography apparatus; arrangements of imaging apparatus in a room adapted for image acquisition of a particular organ or body part for the brain
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7253—Details of waveform analysis characterised by using transforms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2576/00—Medical imaging apparatus involving image processing or analysis
- A61B2576/02—Medical imaging apparatus involving image processing or analysis specially adapted for a particular organ or body part
- A61B2576/026—Medical imaging apparatus involving image processing or analysis specially adapted for a particular organ or body part for the brain
-
- 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
- G06T2207/10048—Infrared image
-
- 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
- G06T2207/20048—Transform domain processing
- G06T2207/20064—Wavelet transform [DWT]
-
- 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
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Biophysics (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Pathology (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Radiology & Medical Imaging (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Artificial Intelligence (AREA)
- Physiology (AREA)
- Psychiatry (AREA)
- Neurology (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method and a system for removing shallow noise of functional near-infrared brain imaging, wherein the method for removing the shallow noise comprises the following steps: decomposing the functional near-infrared brain imaging observation signal of the ith observation channel into a time-frequency space through wavelet transformation to obtain time-frequency expression of the observation signal; identifying shallow noise time frequency points in the time frequency expression of the observation signals through wavelet coherence, and setting the weight of the shallow noise time frequency points to zero to obtain the time frequency expression of new observation signals; performing wavelet inverse transformation on the time-frequency expression of the new observation signal to obtain an observation signal filtered by the ith channel, wherein the observation signal is the observation signal subjected to shallow layer noise removal; and repeating the steps until all the observation channels are processed. The method fully considers the global distribution characteristic of the shallow noise, and can better describe and remove the characteristic of the shallow noise.
Description
Technical Field
The invention relates to a method for removing functional near-infrared brain imaging shallow noise, in particular to a method for removing functional near-infrared brain imaging shallow noise based on wavelet coherence and wavelet transformation; and also relates to a shallow noise removing system for implementing the method.
Background
Functional near-infrared brain imaging (fNIRS) is an important non-invasive brain imaging technique. The principle of fNIRS is to use near infrared light to detect changes in hemoglobin concentration in the cerebral cortex and to infer neural activity in the cerebral cortex. Specifically, brain neural activity causes localized hemodynamic changes in the brain, which changes (including oxyhemoglobin concentration changes and deoxyhemoglobin concentration changes) cause localized brain tissue to change absorbance of near infrared light within the wavelength range of about 600-. Therefore, the fNIRS can calculate the oxyhemoglobin concentration change and the deoxyhemoglobin concentration change in the local brain tissue in turn by detecting the absorption rate change of the near-infrared light, and further infer the neural activity of the brain through the neuro-vascular coupling (neuro-vascular coupling) rule. For example, when a subject is assigned to a right-handed finger exercise task, the left motor area of the cerebral cortex discharges, consuming oxygen and energy. At this time, the overcompensation mechanism of the brain blood supply system greatly inputs blood containing abundant oxyhemoglobin into the local area, so that the concentration of oxyhemoglobin in the local area increases, and the concentration of deoxyhemoglobin decreases. In the fNIRS experiment, an experimenter makes a subject perform a task according to a certain experimental paradigm, and simultaneously observes the concentration change of hemoglobin at different positions of the brain by using the fNIRS.
One important issue faced by fNIRS is the shallow noise interference problem. In the fNIRS observation process, near infrared light enters the brain through superficial tissues such as scalp, skull, meninges and the like, and a blood supply system also exists in the superficial tissues, so that hemodynamic changes in the superficial tissues can be detected by the fNIRS, and the hemodynamic changes can be mixed into the hemodynamic changes related to neural activity as noise components, so that the specificity and sensitivity of fNIRS in brain neural activity observation are influenced. The shallow noise has a global distribution characteristic and large energy, so that the interference on the fNIRS signal is large. The physiological sources of superficial noise are diverse, including heartbeat, respiration, arterial blood pressure fluctuations, heart rate fluctuations, very low frequency physiological activity, etc., with very complex time and frequency characteristics, and thus its removal is an important but very difficult problem in fNIRS.
The existing method for removing the fNIRS shallow noise mainly comprises the following steps: one is a band-pass filtering method, which uses a band-pass filter (band-pass filter) technique to retain signal components within a certain passband frequency and remove other components. The main problem with this approach is that it is difficult to predetermine the frequency of the shallow noise component and once the frequency of the shallow noise component overlaps with the frequency of the component of interest, the useful component in the signal will be removed. The other is a short-interval channel recording method, which uses a channel with a short emitter-receiver spacing to record signals in shallow tissues specially, so that the signals are regressed from a common observation channel to achieve the denoising effect. The problem with this approach is that additional channels need to be used, wasting valuable fNIRS observation channel numbers and reducing the fNIRS observable area. The other methods of additional auxiliary equipment, such as ultrasonic doppler instrument, etc., record shallow signals and remove them, and have the problems of physical property difference between the signals recorded by the additional equipment and the blood oxygen signals, difficulty in establishing corresponding relationship, complex operation, and greatly increased preparation time and discomfort of the subject, thereby increasing difficulty in use. And fourthly, a data-driven method is adopted, such methods comprise Independent component analysis (Independent component analysis) and principal component analysis (principal component analysis), and the like, and the methods can decompose different components in the signal, identify noise components in the signal and remove the noise components. The problem is that the noise identification process needs to depend on subjective selection of an operator, needs abundant experience, lacks objectivity, and is difficult to master and use by general primary users.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a method for removing shallow noise of functional near-infrared brain imaging.
Another technical problem to be solved by the present invention is to provide a system for removing shallow noise in functional near-infrared brain imaging.
In order to achieve the purpose, the invention adopts the following technical scheme:
according to a first aspect of the embodiments of the present invention, there is provided a method for removing shallow noise in functional near-infrared brain imaging, including the following steps:
decomposing the functional near-infrared brain imaging observation signal of the ith observation channel into a time-frequency space through wavelet transformation to obtain time-frequency expression of the observation signal; wherein i is a positive integer;
identifying shallow noise time frequency points in the time frequency expression of the observation signals through wavelet coherence, and setting the weight of the shallow noise time frequency points to zero to obtain the time frequency expression of new observation signals;
performing wavelet inverse transformation on the time-frequency expression of the new observation signal to obtain an observation signal filtered by the ith channel, wherein the observation signal is the observation signal subjected to shallow layer noise removal;
and repeating the steps until all the K observation channels are processed.
Wherein preferably, the time-frequency expression of the observation signal at the time point n and the scale s
The following formula is adopted:
wherein s represents the scale of the wavelet transform; n is a time point; n is the length of the time sequence; δ t is a unit time step;
is a wavelet mother function; n' is a time point used to traverse all time points of the time series.
Preferably, the method for identifying the shallow noise time frequency point in the time-frequency expression of the observation signal through wavelet coherence and setting the weight of the shallow noise time frequency point to zero includes the following steps:
a shallow noise time-frequency filtering template of the near-infrared brain imaging observation signal is identified through wavelet coherence;
and masking the time-frequency expression of the observation signal by adopting a shallow noise time-frequency filtering template, and setting the weight of the shallow noise time-frequency point to zero to obtain the new time-frequency expression of the observation signal.
Preferably, the shallow noise time-frequency filtering template for the near-infrared brain imaging observation signal with the wavelet coherence identification function comprises the following steps:
s211, sequentially calculating the wavelet coherence values of the fNIRS observation signals of the ith observation channel and the fNIRS observation signals of all the observation channels through wavelet coherence;
s212, superposing the K wavelet coherence values obtained in the step S211 after significance binarization to obtain a co-variational time-frequency matrix of the ith observation channel;
and S213, carrying out binarization on the co-variational time-frequency matrix to obtain a shallow noise time-frequency filtering template of the ith channel.
Wherein preferably, all K wavelet coherence values wcoh (x) obtained in step S211 are usedi,xj) Overlapping the significant binaryzation to obtain a co-variational time-frequency matrix M of the ith channeliThe following formula is adopted:
the significant function returns 1 when the statistic represented by the argument is significant, and returns 0 when the statistic represented by the argument is not significant.
Preferably, the covariation time-frequency matrix is binarized to obtain the shallow noise time-frequency filtering template of the ith channel, each element in the covariation time-frequency matrix is compared with a shallow noise global threshold, when the element is larger than the shallow noise global threshold, the value of the element is replaced by 1, otherwise, the value of the element is replaced by 0, and the shallow noise time-frequency filtering template of the ith channel is obtained.
According to a second aspect of the embodiments of the present invention, there is provided a system for removing shallow noise in functional near-infrared brain imaging, including a processor and a memory; the memory having stored thereon a computer program operable on the processor, the computer program when executed by the processor implementing the steps of:
decomposing the functional near-infrared brain imaging observation signal of the ith observation channel into a time-frequency space through wavelet transformation to obtain time-frequency expression of the observation signal; wherein i is a positive integer;
identifying shallow noise time frequency points in the time frequency expression of the observation signals through wavelet coherence, and setting the weight of the shallow noise time frequency points to zero to obtain the time frequency expression of new observation signals;
performing wavelet inverse transformation on the time-frequency expression of the new observation signal to obtain an observation signal filtered by the ith channel, wherein the observation signal is the observation signal subjected to shallow layer noise removal;
and repeating the steps until all the observation channels are processed.
Preferably, when the shallow noise time frequency point in the time-frequency expression of the observation signal is identified through wavelet coherence and the weight of the shallow noise time frequency point is reset to zero, the computer program is executed by the processor, and the following steps are also realized;
a shallow noise time-frequency filtering template of the near-infrared brain imaging observation signal is identified through wavelet coherence;
and masking the time-frequency expression of the observation signal by adopting a shallow noise time-frequency filtering template, and setting the weight of the shallow noise time-frequency point to zero to obtain the new time-frequency expression of the observation signal.
Preferably, when the shallow noise time frequency point in the time-frequency expression of the observation signal is identified through wavelet coherence and the weight of the shallow noise time frequency point is reset to zero, the computer program is executed by the processor, and the following steps are also realized;
sequentially calculating the wavelet coherence values of the fNIRS observation signals of the ith observation channel and the fNIRS observation signals of all the K observation channels through wavelet coherence;
superposing the K wavelet coherence values after significance binarization to obtain a co-variational time-frequency matrix of the ith observation channel;
and carrying out binarization on the co-variational time-frequency matrix to obtain a shallow noise time-frequency filtering template of the ith channel.
Wherein preferably, when said computer program is executed by said processor, the following steps are also implemented;
and carrying out binarization on the co-variation time-frequency matrix to obtain a shallow noise time-frequency filtering template of the ith channel, comparing each element in the co-variation time-frequency matrix with a shallow noise global threshold, and replacing the value of the element by 1 when the element is greater than the shallow noise global threshold, or replacing the value of the element by 0 to obtain the shallow noise time-frequency filtering template of the ith channel.
The method for removing the shallow noise of the functional near-infrared brain imaging fully considers the global distribution characteristic of the shallow noise and utilizes wavelet coherence (wavelet transform coherence) technology to identify the time frequency point of the shallow noise. Wavelet coherence analysis is carried out on signals of any two observation channels, and the covariant characteristics of the two channels at each time-frequency point can be obtained. If a certain channel and most other channels have common variation on a certain time frequency point, the time frequency point of the channel can be considered to belong to a shallow noise time frequency point. And carrying out the analysis and the denoising channel by channel to realize the removal of the shallow noise.
Drawings
FIG. 1 is a flow chart of a method for removing shallow noise in functional near-infrared brain imaging according to the present invention;
fig. 2 is a schematic structural diagram of a system for removing shallow noise in functional near-infrared brain imaging provided by the present invention.
Detailed Description
The technical contents of the invention are described in detail below with reference to the accompanying drawings and specific embodiments.
In view of the problems of the existing method, the invention provides a functional near-infrared brain imaging shallow noise removal method based on wavelet coherence and wavelet transformation. The basic principle of the method is that a wavelet transform (wavelet transform) is utilized to decompose an fNIRS signal into a time-frequency space, then time-frequency points belonging to shallow noise are identified, the weight of the time-frequency points is set to zero, and then the time-frequency points are inversely transformed back to an original signal space, so that the aim of removing the shallow noise is fulfilled. The most key core innovation is the identification algorithm of the frequency points when shallow noise exists. The method fully considers the global distribution characteristic of the shallow noise, and utilizes wavelet coherence (wavelet transform coherence) technology to identify the frequency points of the shallow noise. Wavelet coherence analysis is carried out on signals of any two observation channels, and the covariant characteristics of the two channels at each time-frequency point can be obtained. If a certain channel and most other channels have common variation on a certain time frequency point, the time frequency point of the channel can be considered to belong to a shallow noise time frequency point. And carrying out the analysis and the denoising channel by channel to realize the removal of the shallow noise.
As shown in fig. 1, the method for removing the superficial noise in functional near-infrared brain imaging provided by the present invention comprises the following steps: firstly, decomposing the fNIRS observation signal of the ith observation channel into a time-frequency space through wavelet transformation to obtain time-frequency expression of the observation signal; secondly, identifying shallow noise time frequency points in the time frequency expression of the observation signal through wavelet coherence, and setting the weight of the shallow noise time frequency points to zero to obtain the time frequency expression of a new observation signal; then, performing wavelet inverse transformation on the time-frequency expression of the new observation signal to obtain an observation signal filtered by the ith channel, wherein the observation signal at the moment is the observation signal with the shallow layer noise removed; and repeating the steps until all the observation channels are processed and the shallow noise is removed. This process is described in detail below.
And S1, decomposing the fNIRS observation signal of the ith observation channel into a time-frequency space through wavelet transformation to obtain the time-frequency expression of the observation signal.
In the embodiment provided by the present invention, let the fNIRS observation signals of all K observation channels be xkWhere K is 1,2, …, and K is each observation channel. For a certain observation channel i belongs to {1,2, …, K }, the fNIRS observation signal x of the ith observation channel is subjected to wavelet transformationiDecomposing the time-frequency space to obtain the time-frequency expression of the observation signal at the time point n and the scale s
The following formula is adopted:
where s represents the scale of the wavelet transform, corresponding to frequency; n is a time point; n is the length of the time sequence; δ t is a unit time step;
is a wavelet mother function; n' is a time point used to traverse all time points of the time series. x is the number ofn’fNIRS observed signals at time point n'.
S2, identifying the shallow noise time frequency point in the time frequency expression of the observation signal through wavelet coherence, and setting the weight of the shallow noise time frequency point to zero to obtain the time frequency expression of a new observation signal.
The method comprises the following steps of identifying shallow noise time frequency points in time-frequency expression of an observation signal through wavelet coherence, and setting the weight of the shallow noise time frequency points to zero, wherein the method specifically comprises the following steps:
and S21, identifying the shallow noise time-frequency filtering template of the fNIRS observation signal through wavelet coherence.
The shallow noise time-frequency filtering template for identifying the fNIRS observation signals through wavelet coherence specifically comprises the following steps:
s211, sequentially calculating the fNIRS observation signals x of the ith observation channel through wavelet coherenceiAnd f NIRS observation signal wavelet coherence values with all observation channels.
For a certain observation channel i belongs to {1,2, …, N }, traversing all observation channels j to 1,2, …, K, and sequentially calculating the fNIRS observation signal x of the ith observation channel through wavelet coherenceiAnd f NIRS observation signal wavelet coherence values of all observation channels. Wherein, fNIRS observation signal x of the ith observation channel is calculated by wavelet coherenceiAnd calculating the wavelet coherence value of the fNIRS observation signal of the jth observation channel by adopting the following formula:
wherein, wcoh (x)i,xj) The wavelet coherence value of the fNIRS observation signals of the i and j observation channels is represented, W represents wavelet transformation, S represents a smoothing operator, n is a time sampling point, and S represents a wavelet transformation scale.
By performing wavelet coherence analysis on signals of any two observation channels, the covariant characteristics of the two channels at each time-frequency point can be obtained. If a certain channel and most other channels have common variation on a certain time frequency point, the time frequency point of the channel can be considered to belong to a shallow noise time frequency point.
And S212, superposing the K wavelet coherence values obtained in the step S211 after significance binarization to obtain a co-variational time-frequency matrix of the ith observation channel.
All K wcoh (x) obtained in step S211i,xj) Overlapping the significant binaryzation to obtain a co-variational time-frequency matrix M of the ith channeliI.e. by
The significant function returns 1 when the statistic represented by the argument is significant, and returns 0 when the statistic represented by the argument is not significant. In the embodiment provided by the invention, the fNIRS observation signal x of the ith observation channel is calculated by wavelet coherenceiWhen the wavelet coherence value of the fNIRS observation signal of the jth observation channel is compared with the wavelet coherence value of the fNIRS observation signal of the jth observation channel, the wavelet coherence value of the fNIRS observation signal of the jth observation channel is automatically aligned with wcoh (x)i,xj) Whether the representative statistic is significant or not is judged, and signal (wcoh (x) is returnedi,xj) ) to generate a matrix of 0's and 1's.
S213, for the co-variational time-frequency matrix MiAnd carrying out binarization to obtain a shallow noise time-frequency filtering template of the ith channel.
For common degree-of-variation time frequency matrix MiBinarizing to obtain shallow noise time-frequency filtering template B of ith channeliI.e. by
Bi＝boolean(Mi＞threshold)；
Wherein, the bolean (status) function is to judge whether the expression status is true or false, if true, the function value is 1, otherwise, the function value is 0. the threshold is a global threshold of shallow noise, and may be obtained experimentally or set as needed. MiGreater than threshold is the common degree time frequency matrix MiComparing each element with a threshold, when the element is greater than the threshold, replacing the value of the element by 1, otherwise, replacing the value of the element by 0 to obtain a shallow noise time-frequency filtering template B of the ith channeli。
S22, adopting shallow noise time-frequency filtering template BiTime-frequency representation of observation signals
Masking to zero the weight of the frequency point in shallow noise time to obtain the time-frequency expression of new observation signal
To pair
Applying shallow noise time-frequency filtering template BiMasking (mask) is carried out, and the energy of the frequency point when the mask is marked as shallow noise is set to be 0, namely
And S3, performing inverse wavelet transform on the time-frequency expression of the new observation signal to obtain the observation signal filtered by the ith channel, wherein the observation signal at the moment is the observation signal with the shallow layer noise removed.
Performing wavelet inverse transformation on the time-frequency expression of the new observation signal to obtain a signal x 'after the ith channel is filtered'iI.e. by
Wherein iwt (-) denotes the inverse wavelet transform corresponding to the wavelet transform in step S1. Wherein, the wavelet transformation and the wavelet inverse transformation adopt the conventional processing method in the field, and are not described in detail herein.
And S4, repeating the steps S1-S3 until all the observation channels are processed and the shallow noise is removed.
Traversing i, namely, repeating the steps S1 to S3 by respectively setting i to 1,2, … and K for all K channels to obtain filtered signals x 'of all K channels from which the shallow layer noise is removed'nN is 1,2, …, N. And finishing the shallow noise removing process.
In summary, the method for removing the shallow noise in the functional near-infrared brain imaging provided by the invention decomposes the fNIRS observation signal of the ith observation channel into the time-frequency space through wavelet transform to obtain the time-frequency expression of the observation signal; shallow noise time frequency points in the time frequency expression of the observation signal are obtained through wavelet coherence identification, and the weight of the shallow noise time frequency points is set to be zero to obtain the time frequency expression of a new observation signal; then, performing wavelet inverse transformation on the time-frequency expression of the new observation signal to obtain an observation signal filtered by the ith channel, wherein the observation signal at the moment is the observation signal with the shallow layer noise removed; and repeating the steps until all the observation channels are processed and the shallow noise is removed. The method fully considers the global distribution characteristic of the shallow noise and can effectively remove the shallow noise.
The method for removing the shallow noise of the functional near-infrared brain imaging provided by the invention has the following advantages:
firstly, in the aspect of operation:
(1) and the data is driven, no additional auxiliary equipment or short-interval channel is needed, and the operation is simple. The method can be suitable for various models of fNIRS machines produced by various manufacturers.
(2) Filtering parameters or noise components do not need to be manually set or selected, experience does not need to be enriched, and the result is more objective.
Secondly, in the aspect of denoising effect:
(1) compared with methods such as band-pass filtering, short-channel denoising and additional auxiliary equipment, the method makes full use of the global characteristics of the shallow noise, and can better depict and remove the characteristics of the shallow noise.
(2) Due to the time-frequency characteristic of wavelet transformation, the time information of shallow noise interference can be accurately positioned, and the time-varying problem of the shallow noise (namely, the physiological state of a human body such as respiration rate, heart rate, blood pressure and the like can be constantly changed along with time, so that the frequency characteristic of the shallow noise can be changed along with time) can be solved, and the shallow noise can be accurately and specifically removed in the time dimension.
(3) Because the wavelet coherence analysis can describe the covariant characteristic of the band delay between two signals, the method can solve the time-lag problem when the shallow signal is transmitted at different positions of the head, thereby better removing the shallow noise.
The invention also provides a system for removing the shallow noise of the functional near-infrared brain imaging. As shown in fig. 2, the system includes a processor 22 and a memory 21 storing instructions executable by the processor 22;
The memory 21 is used for storing the program codes and transmitting the program codes to the CPU. Memory 21 may include volatile memory, such as Random Access Memory (RAM); the memory 21 may also include non-volatile memory, such as read-only memory, flash memory, a hard disk, or a solid state disk; the memory 21 may also comprise a combination of memories of the kind described above.
Specifically, the calibration-free positioning system provided by the embodiment of the present invention includes a processor 22 and a memory 21; the memory 21 has stored thereon a computer program operable on the processor 22, which when executed by the processor 22, performs the steps of:
decomposing the fNIRS observation signal of the ith observation channel into a time-frequency space through wavelet transformation to obtain time-frequency expression of the observation signal;
identifying shallow noise time frequency points in the time frequency expression of the observation signals through wavelet coherence, and setting the weight of the shallow noise time frequency points to zero to obtain the time frequency expression of new observation signals;
performing wavelet inverse transformation on the time-frequency expression of the new observation signal to obtain an observation signal filtered by the ith channel, wherein the observation signal is the observation signal subjected to shallow layer noise removal;
and repeating the steps until all the observation channels are processed and the shallow noise is removed.
Wherein the computer program is executed by the processor 22 to carry out the following steps;
time-frequency representation of observation signals
The following formula is adopted:
where s represents the scale of the wavelet transform, corresponding to frequency; n is a time point; n is the length of the time sequence; δ t is a unit time step;
is a wavelet mother function; n' is a time point used to traverse all time points of the time series.
Wherein, when the shallow noise time frequency point in the time-frequency expression of the observation signal is identified through wavelet coherence, and the weight of the shallow noise time frequency point is reset to zero, the computer program is executed by the processor 22 to realize the following steps;
a shallow noise time-frequency filtering template of the near-infrared brain imaging observation signal is identified through wavelet coherence;
and masking the time-frequency expression of the observation signal by adopting a shallow noise time-frequency filtering template, and setting the weight of the shallow noise time-frequency point to zero to obtain the new time-frequency expression of the observation signal.
Wherein, when the shallow noise time-frequency filtering template of the near-infrared brain imaging observation signal is identified by the wavelet coherence, the computer program is executed by the processor 22 to realize the following steps;
sequentially calculating the wavelet coherence values of the fNIRS observation signals of the ith observation channel and the fNIRS observation signals of all the observation channels through wavelet coherence;
superposing the K wavelet coherence values after significance binarization to obtain a common-variation time-frequency matrix of the ith observation channel;
and carrying out binarization on the co-variational time-frequency matrix to obtain a shallow noise time-frequency filtering template of the ith channel.
Wherein the computer program is executed by the processor 22 to carry out the following steps;
all the obtained N wavelet coherence values wcoh (x)i,xj) Overlapping the significant binaryzation to obtain a co-variational time-frequency matrix M of the ith channeliThe following formula is adopted:
the significant function returns 1 when the statistic represented by the argument is significant, and returns 0 when the statistic represented by the argument is not significant.
Wherein the computer program is executed by the processor 22 to carry out the following steps;
and carrying out binarization on the co-variation time-frequency matrix to obtain a shallow noise time-frequency filtering template of the ith channel, comparing each element in the co-variation time-frequency matrix with a shallow noise global threshold, and replacing the value of the element by 1 when the element is greater than the shallow noise global threshold, or replacing the value of the element by 0 to obtain the shallow noise time-frequency filtering template of the ith channel.
The embodiment of the invention also provides a computer readable storage medium. The computer-readable storage medium herein stores one or more programs. Among other things, computer-readable storage media may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as read-only memory, flash memory, a hard disk, or a solid state disk; the memory may also comprise a combination of memories of the kind described above. When the one or more programs in the computer readable storage medium are executed by one or more processors, part of or all of the steps of the method for removing the shallow noise in the functional near-infrared brain imaging in the above embodiments of the method are implemented.
The method and system for removing the shallow noise in functional near-infrared brain imaging provided by the invention are explained in detail above. Any obvious modifications to the invention, which would occur to those skilled in the art, without departing from the true spirit of the invention, would constitute a violation of the patent rights of the invention and would carry a corresponding legal responsibility.
Claims (8)
1. A method for removing shallow noise of functional near-infrared brain imaging is characterized by comprising the following steps:
decomposing the functional near-infrared brain imaging observation signal of the ith observation channel into a time-frequency space through wavelet transformation to obtain time-frequency expression of the observation signal; wherein i is a positive integer and is belonged to {1,2, …, K }, and K is each observation channel;
a shallow noise time-frequency filtering template of the near-infrared brain imaging observation signal is identified through wavelet coherence; masking the time-frequency expression of the observation signal by adopting a shallow noise time-frequency filtering template, and setting the weight of a shallow noise time-frequency point to zero to obtain the time-frequency expression of a new observation signal; the shallow noise time-frequency filtering template of the near-infrared brain imaging observation signal with the wavelet coherence identification function comprises the following substeps: s211, calculating wavelet coherence values of the functional near-infrared brain imaging observation signals of the ith observation channel and the functional near-infrared brain imaging observation signals of all the observation channels in sequence through wavelet coherence; s212, superposing the K wavelet coherence values obtained in the step S211 after significance binarization to obtain a co-variational time-frequency matrix of the ith observation channel; s213, binarizing the common-variable-degree time-frequency matrix to obtain a shallow noise time-frequency filtering template of the ith channel;
performing wavelet inverse transformation on the time-frequency expression of the new observation signal to obtain an observation signal filtered by the ith channel, wherein the observation signal is the observation signal subjected to shallow layer noise removal;
and repeating the steps until all the observation channels are processed.
2. The method for removing superficial noise in functional near-infrared brain imaging according to claim 1, wherein:
time-frequency representation of the observation signal
The following formula is adopted:
where s represents the scale of the wavelet transform, corresponding to frequency; n is a time point; n is the length of the time sequence; δ t is a unit time step;
is a wavelet mother function; n' is a time point used to traverse all time points of the time series.
3. The method for removing superficial noise in functional near-infrared brain imaging according to claim 1, wherein:
all K wavelet coherence values wcoh (x) obtained in step S211i，xj) Overlapping the significant binaryzation to obtain a co-variational time-frequency matrix M of the ith channeliThe following formula is adopted:
wherein, wcoh (x)i，xj) And the wavelet coherence value of the functional near-infrared brain imaging observation signals of the two observation channels i and j is represented, and the signaliciance function returns 1 when the statistic represented by the independent variable is significant and returns 0 when the statistic represented by the independent variable is not significant.
4. The method for removing superficial noise in functional near-infrared brain imaging according to claim 1, wherein:
and carrying out binarization on the co-variation time-frequency matrix to obtain a shallow noise time-frequency filtering template of the ith channel, comparing each element in the co-variation time-frequency matrix with a shallow noise global threshold, and replacing the value of the element by 1 when the element is greater than the shallow noise global threshold, or replacing the value of the element by 0 to obtain the shallow noise time-frequency filtering template of the ith channel.
5. A system for removing shallow noise in functional near-infrared brain imaging is used for realizing the method for removing the shallow noise in any one of claims 1-4, and is characterized by comprising a processor and a memory; the memory having stored thereon a computer program operable on the processor, the computer program when executed by the processor implementing the steps of:
decomposing the functional near-infrared brain imaging observation signal of the ith observation channel into a time-frequency space through wavelet transformation to obtain time-frequency expression of the observation signal; wherein i is a positive integer;
identifying shallow noise time frequency points in the time frequency expression of the observation signals through wavelet coherence, and setting the weight of the shallow noise time frequency points to zero to obtain the time frequency expression of new observation signals;
performing wavelet inverse transformation on the time-frequency expression of the new observation signal to obtain an observation signal filtered by the ith channel, wherein the observation signal is the observation signal subjected to shallow layer noise removal;
and repeating the steps until all the observation channels are processed.
6. The system according to claim 5, wherein when the shallow noise time bin in the time-frequency expression of the observation signal is identified by wavelet coherence and the weight of the shallow noise time bin is reset to zero, the computer program is executed by the processor, and further implementing the following steps:
a shallow noise time-frequency filtering template of the near-infrared brain imaging observation signal is identified through wavelet coherence;
and masking the time-frequency expression of the observation signal by adopting a shallow noise time-frequency filtering template, and setting the weight of the shallow noise time-frequency point to zero to obtain the new time-frequency expression of the observation signal.
7. The system for shallow noise removal for functional near-infrared brain imaging according to claim 6, wherein when said shallow noise time-frequency filtering template for functional near-infrared brain imaging observation signals is identified by wavelet coherence, said computer program is executed by said processor, and further implements the following steps:
s211, calculating wavelet coherence values of the functional near-infrared brain imaging observation signals of the ith observation channel and the functional near-infrared brain imaging observation signals of all the observation channels in sequence through wavelet coherence;
s212, superposing the K wavelet coherence values obtained in the step S211 after significance binarization to obtain a co-variational time-frequency matrix of the ith observation channel;
and S213, carrying out binarization on the co-variational time-frequency matrix to obtain a shallow noise time-frequency filtering template of the ith channel.
8. The system for shallow noise removal for functional near-infrared brain imaging of claim 7, wherein said computer program, when executed by said processor, further performs the steps of:
and carrying out binarization on the co-variation time-frequency matrix to obtain a shallow noise time-frequency filtering template of the ith channel, comparing each element in the co-variation time-frequency matrix with a shallow noise global threshold, and replacing the value of the element by 1 when the element is greater than the shallow noise global threshold, or replacing the value of the element by 0 to obtain the shallow noise time-frequency filtering template of the ith channel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810316542.8A CN108553086B (en) | 2018-04-10 | 2018-04-10 | Method and system for removing shallow noise of functional near-infrared brain imaging |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810316542.8A CN108553086B (en) | 2018-04-10 | 2018-04-10 | Method and system for removing shallow noise of functional near-infrared brain imaging |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108553086A CN108553086A (en) | 2018-09-21 |
| CN108553086B true CN108553086B (en) | 2021-05-04 |
Family
ID=63534496
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810316542.8A Active CN108553086B (en) | 2018-04-10 | 2018-04-10 | Method and system for removing shallow noise of functional near-infrared brain imaging |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108553086B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111012319B (en) * | 2019-12-05 | 2023-03-14 | 广东省医疗器械研究所 | Method, system and storage medium for monitoring and imaging skin blood flow and blood vessel |
| CN111195393A (en) * | 2020-01-08 | 2020-05-26 | 国家康复辅具研究中心 | Near-infrared brain region activation mode-oriented peripheral electrical stimulation system and use method thereof |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104020136A (en) * | 2014-06-25 | 2014-09-03 | 中国科学院重庆绿色智能技术研究院 | Method and device for denoising near infrared spectrum by wavelet mid-value |
| CN104055524A (en) * | 2014-06-30 | 2014-09-24 | 山东大学 | Brain function connection detection method and system based on near infrared spectrum |
| CN105069285A (en) * | 2015-07-22 | 2015-11-18 | 中国地质大学（武汉） | Wavelet coherence based multi-neural signal correlation analysis method |
| CN106419850A (en) * | 2016-11-03 | 2017-02-22 | 国家康复辅具研究中心 | Dynamic brain function detection method and system based on near infrared spectrum and blood pressure information |
| CN107280685A (en) * | 2017-07-21 | 2017-10-24 | 国家康复辅具研究中心 | Top layer physiological noise minimizing technology and system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9598011B2 (en) * | 2014-01-09 | 2017-03-21 | Northrop Grumman Systems Corporation | Artificial vision system |
-
2018
- 2018-04-10 CN CN201810316542.8A patent/CN108553086B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104020136A (en) * | 2014-06-25 | 2014-09-03 | 中国科学院重庆绿色智能技术研究院 | Method and device for denoising near infrared spectrum by wavelet mid-value |
| CN104055524A (en) * | 2014-06-30 | 2014-09-24 | 山东大学 | Brain function connection detection method and system based on near infrared spectrum |
| CN105069285A (en) * | 2015-07-22 | 2015-11-18 | 中国地质大学（武汉） | Wavelet coherence based multi-neural signal correlation analysis method |
| CN106419850A (en) * | 2016-11-03 | 2017-02-22 | 国家康复辅具研究中心 | Dynamic brain function detection method and system based on near infrared spectrum and blood pressure information |
| CN107280685A (en) * | 2017-07-21 | 2017-10-24 | 国家康复辅具研究中心 | Top layer physiological noise minimizing technology and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108553086A (en) | 2018-09-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| von Lühmann et al. | Using the general linear model to improve performance in fNIRS single trial analysis and classification: a perspective | |
| Gabard-Durnam et al. | The Harvard Automated Processing Pipeline for Electroencephalography (HAPPE): standardized processing software for developmental and high-artifact data | |
| Sangaiah et al. | An intelligent learning approach for improving ECG signal classification and arrhythmia analysis | |
| Hamaneh et al. | Automated removal of EKG artifact from EEG data using independent component analysis and continuous wavelet transformation | |
| Xu et al. | Cluster-based analysis for personalized stress evaluation using physiological signals | |
| Todd et al. | The identification of peaks in physiological signals | |
| Azami et al. | Fuzzy entropy metrics for the analysis of biomedical signals: Assessment and comparison | |
| Maddirala et al. | Eye-blink artifact removal from single channel EEG with k-means and SSA | |
| US20160367198A1 (en) | Apparatus and method for detecting and removing artifacts in optically acquired biological signals | |
| Asgari et al. | Consistent changes in intracranial pressure waveform morphology induced by acute hypercapnic cerebral vasodilatation | |
| Saalasti | Neural networks for heart rate time series analysis | |
| Leeuwis et al. | Functional connectivity analysis in motor-imagery brain computer interfaces | |
| Zhang et al. | Correcting physiological noise in whole-head functional near-infrared spectroscopy | |
| Zheng et al. | Human emotional stress assessment through Heart Rate Detection in a customized protocol experiment | |
| Einalou et al. | Effective channels in classification and functional connectivity pattern of prefrontal cortex by functional near infrared spectroscopy signals | |
| Ruiz-Gómez et al. | Measuring alterations of spontaneous EEG neural coupling in Alzheimer's disease and mild cognitive impairment by means of cross-entropy metrics | |
| CN108553086B (en) | Method and system for removing shallow noise of functional near-infrared brain imaging | |
| Hossain et al. | Motion artifacts correction from EEG and fNIRS signals using novel multiresolution analysis | |
| Li et al. | Cortical statistical correlation tomography of EEG resting state networks | |
| Pavlova et al. | Effects of missing data on characterization of complex dynamics from time series | |
| Ang et al. | Extracting and selecting discriminative features from high density NIRS-based BCI for numerical cognition | |
| Toulni et al. | Electrocardiogram signals classification using discrete wavelet transform and support vector machine classifier | |
| Ghobadi Azbari et al. | A novel approach to the extraction of fetal electrocardiogram based on empirical mode decomposition and correlation analysis | |
| Hosni et al. | A graph-based nonlinear dynamic characterization of motor imagery toward an enhanced hybrid bci | |
| Truong et al. | Deep convolutional neural network applied to electroencephalography: Raw Data vs spectral features |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |