The present invention relates to a technique to separate and remove influence of surface layer components such as skin blood flow components mixed in signal components using a biophotonic measurement apparatus using visible light or near infrared light.
It is reported that optical detection signals and biological signals (hereinafter referred to as NIRS signals) obtained from non-invasive optical brain function imaging using NIRS including optical topography may be affected by variations in skin blood flow in the scalp since irradiation with light is performed on/over the scalp. In consideration of such influence of skin blood flow, methods to extract and remove components thereof are studied. Most of such methods acquire signal components from portions with different depths by method using a plurality of irradiator-detector (source-detector) distances (hereinafter referred to as SD distance) and intend to remove, using the signal components, skin blood flow-derived signals which may affect measurement signals of a shallow layer part. A method of measuring with a plurality of SD distances is hereinafter referred to as a multi-SD method.
For example, there is a method where absorption coefficients in the scalp and brain (grey matter) are determined by system of equations using optical path lengths in the scalp and brain (grey matter) with each of a short SD distance and long SD distance (e.g. refer to NPL 2). In this method, a configuration of the head is assumed as a two-layered model and a partial mean optical path length in each of the layers is further required to be assumed; however, assuming an optical path length of a subject is difficult.
Also, a subtraction method using an adaptive filter has been proposed. In this method, by subtracting, from a measurement signal with a long SD distance (hereinafter referred to as long SD signal), a value obtained by multiplying a measurement signal with a short SD distance (hereinafter referred to as short SD signal) by an appropriate coefficient, a skin blood flow-derived signal is removed (e.g. refer to NPL 3). Moreover, as the subtraction method using linear regression, a method is proposed where brain activity signals are obtained by subtracting, from a long SD signal, a fitting signal where short SD signals are fitted to linear regression of long SD signals (e.g. refer to NPL 4).
As techniques related to the above, the methods below are disclosed.
PTL 2 discloses a method to dispose a plurality of pairs of irradiator-detector such that midpoints thereof are equivalent with each other, to perform measurement thereby, and to remove unnecessary information by arithmetic processing with an object of providing an optical measurement apparatus capable of removing unnecessary information derived from skin blood flow or the like using an optical emitting/receiving unit having a plurality of light emitting probes and a plurality of light receiving probes. Also, PTL 3 discloses a method to obtain a result characterizing a state mainly in the brain tissue itself free from influence derived from overlapping adjacent tissues or the like by employing a device configuration where two detectors are used for one light source and appropriately discriminating information obtained from the two detectors. Furthermore, PTLs 4, 5, and 6 disclose methods to calculate variations in absorbance and to perform calculation such as subtraction with a long SD signal and short SD signal. These methods, however, have problems as described below.
Firstly, there is a problem that it is difficult to determine various coefficients in calculations such as subtraction among measurement signals of respective SD distances. In such calculation, various coefficients affect the result and thus it is required to set appropriate values. Moreover, when short SD signals are acquired, an SD distance of 10 mm is often provided and thereby signal components not affected by brain blood flow but by variations in absorbance only in the skin are acquired. Therefore, an amplitude ratio of brain blood flow-derived components and skin blood flow-derived components is unknown and thus determining an appropriate coefficient for the calculation is difficult. In order to appropriately correct long SD measurement signals including contributions from the skin and brain, it is required to know contribution ratios of each of the skin and brain and their optical path length ratio.
Furthermore, fitting short SD signals to long SD signals may disadvantageously remove brain blood flow-derived signals from the long SD signals when skin blood flow-derived signals and brain blood flow-derived signals are not independent, that is, the skin blood flow-derived signals and brain blood flow-derived signals are correlated.
As a means to solve the above problems, PTL 7 discloses a method to separate, for each independent component calculated using independent component analysis, signals measured with a plurality of SD distances into brain blood flow-derived signal components and skin blood flow-derived signal components using SD distance dependency and to restructure the brain blood flow-derived signals and skin blood flow-derived signals for each of the SD distances.
However, since this method uses the independent component analysis, measurement signals are required to be time series signals of a certain period of time. Thus, processing to separate skin blood flow-derived signals and brain blood flow-derived signals is performed after measurement and no result can be obtained during measurement.
PTL 1: JP 09-019408 A
PTL 2: JP 2008-64675 A
PTL 3: Japanese National Publication of International Patent Application No. 2002-527134
PTL 4: U.S. Pat. No. 7,072,701 B2
PTL 5: U.S. Pat. No. 5,349,961
PTL 6: U.S. Pat. No. 5,902,235
PTL 7: WO 2012005303 A1
PTL 8: JP 2001-178708 A
An object of the present invention is to separate and extract brain blood flow-derived signals and skin blood flow-derived signals included in NIRS signals during measurement.
In the present invention, in particular, measured values of variations in hemoglobin concentration-length measured at a plurality of SD distances at timings that can be deemed as simultaneous are separated into skin blood flow-derived signals and brain blood flow-derived signals in real-time using SD distance dependency.
The present invention allows for performing measurement while brain blood flow-derived signals and skin blood flow-derived signals are separated and extracted in real time during measurement. This allows for interrupting and redoing the measurement when the brain blood flow-derived signals cannot be measured with a sufficient accuracy, thereby enabling efficient and sound data acquisition.
Embodiments of the present invention will be described below with reference to the drawings.
An exemplary configuration of an apparatus of the present invention is illustrated in
Here, the one or more light sources 101 are only required to be semiconductor lasers (LDs), light-emitting diodes (LEDs), or the like and the one or more optical detectors are only required to be avalanche photodiodes (APDs), photodiodes (PDs), photomultiplier tubes (PMTS), or the like. Also, the waveguide 40 is only required to be an optical fiber, glass, light guide, or the like.
The light source 101 is driven by a light source driving device 103. A gain of the one or more optical detectors 102 is controlled by a control/analysis unit 106. The control/analysis unit 106 also controls the light source driving device 103 and receives input of conditions or the like from an input unit 107.
Electrical signals photoelectrically converted by the optical detector 102 are amplified by an amplifier 104, subjected to analog-digital conversion by an analog-digital converter 105, transmitted to the control/analysis unit 106, and processed thereat.
The control/analysis unit 106 has a control unit to control the light source driving device and/or optical detector and an analysis unit to analyze signals obtained from the optical detector and executes analysis based on signals detected by the optical detector 102. Specifically, with digital signals obtained by conversion by the analog-digital converter 105, variations in oxy-hemoglobin (oxy-Hb) concentration-length and/or deoxy-hemoglobin (deoxy-Hb) concentration-length are calculated from variations in detected amount of light or variations in absorbance based on, for example, a method described in NPL 1. Here, the variation in concentration-length refers to an amount of variation in a product of concentration and optical path length.
Descriptions herein assume that the control/analysis unit 106 performs all of driving the light source 101, gain control of the optical detector 102, and signal processing from the analog-digital converter 105; however, the same functions can be implemented by providing separate control units and further providing a means to integrate these control units.
Furthermore, detection signals of received light amount and signals of variations in oxy-hemoglobin or deoxy-hemoglobin concentration-length calculated using the detection signals are stored in a storage unit 108. Measurement results can be displayed on a display unit 109 based on an analysis result and/or stored data.
Although a source 50 and detector 60 are not illustrated in
Next, a method to separate and extract the brain blood flow-derived signals and skin blood flow-derived signals will be described. Hereinafter, of variations in hemoglobin concentration-length obtained from NIRS measurement, brain blood flow-derived will be described only for variations in oxy-hemoglobin concentration-length with a large amplitude; however, variations in deoxy-hemoglobin (deoxy-Hb) concentration-length or variations in total hemoglobin (oxy-Hb+deoxy-Hb) concentration-length may be used. When it is simply noted as variations in hemoglobin concentration-length, the term is used in a generic sense and may include any of the above.
An exemplary measurement cross-sectional view of a multi-SD method is illustrated in
Hereinafter, a method to separate and extract brain blood flow-derived signals and skin blood flow-derived signals will be described.
[Formula 1]
y=c (1)
[Formula 2]
y=a(x−xs0) (2)
A measurement signal of a variation in the hemoglobin concentration-length can be expressed as formula 3 as the sum of the two.
[Formula 3]
y=a(x−xs0)+c (3)
Incidentally, here the letter y represents variations in hemoglobin concentration-length, x represents the SD distance, xs0 represents an x-intercept, a represents the inclination, and c represents a variation in the hemoglobin concentration-length derived from the skin blood flow. Since a plurality of variation values of the hemoglobin concentration-length with different SD distances is observed at timings deemed as simultaneous, the value xs0 is given. Performing linear regression according to formula 3 using these signals gives values at time t, namely a(t) and c(t). Assigning the values to formula 1 and formula 2 gives a value of skin blood flow-derived signal amplitude at that time and a value of brain blood flow-derived signal amplitude with an arbitrary SD. Therefore, repeating the above each time an observation value is obtained results in skin blood flow-derived signals in time series and brain blood flow-derived signals in time series. Here, the value xs0 corresponds to the shortest SD distance where light can reach the brain and can be deemed as constant irrespective of time. In this manner, in the present invention, the skin blood flow-derived signal and/or brain blood flow-derived signal associated with the measurement time is calculated from a plurality of measurement signals measured in association as signals at certain measurement time using SD distance dependency of the measurement signals. The value of xs0 may be obtained by the Monte Carlo simulation using a head structure of the subject. Alternatively, a value obtained from the Monte Carlo simulation using a standard head structure or a value empirically obtained may be preset. Further alternatively, an arbitrary value may be set by inputting the value externally. Still alternatively, the value xs0 may be actually measured and thereby set by blocking skin blood flow by applying pressure to the skin and measuring brain activities with a plurality of SD distances.
For example, a case of performing two types of measurement with SD distances of 15 mm and 30 mm will be described. When, at time t, a variation in the oxy-hemoglobin concentration-length measured with an SD distance of 30 mm is y30 and a variation in the oxy-hemoglobin concentration-length measured with an SD distance of 15 mm is y15, an inclination at time t, a(t), is given by formula 4.
a(t)=(y30(t)−y15(t))/(30−15) [Formula 4]
In the variation in the oxy-hemoglobin concentration-length measured with the SD distance of 30 mm, a value of brain blood flow-derived signal amplitude at time (t) is a(t)*(30−xs0) and a value of skin blood flow-derived signal amplitude at time (t) is c(t). Assigning these values to the formula 5 and formula 6 allows for obtaining a brain contribution ratio and skin contribution ratio at time t. Also with variations in the deoxy-hemoglobin concentration-length, a value of brain blood flow-derived signal amplitude, a value of skin blood flow-derived signal amplitude, brain contribution ratio, and skin contribution ratio at time t can be obtained in a similar manner.
A flowchart of measurement is illustrated in
Next, a case where the above method is applied to measurement with actual arrangement of probes will be described. Exemplary arrangement of probes on the human head is illustrated in
Arrangement of probes and arrangement of measurement points of doubled density are illustrated in
Here, to extract skin blood flow-derived signals, measurement signals at measurement points with a plurality of SD distances are used. When measurement signals with the same SD distance are used for mapping by interpolation, for example an SD distance is approximately 15 to 20 mm, obtained is a map with large contribution of signals derived from shallow parts including the skin.
Here, depending on an SD distance, there are cases where resolution becomes low due to a small number of measurement points when imaging is performed only with signals with the same SD distance. In the example of
A test configuration diagram using an optical brain function measurement apparatus 90 of a whole head measurement type is illustrated in
Provided are a plurality of light sources 402a to 402d of different wavelengths (if two types of wavelengths are used, for example the light sources 402a and 402c are 695 nm and the light sources 402b and 402d are 830 nm), modulators or oscillators 401a and 401b (401c and 401d) to modulate intensity of light from the plurality of light sources 402a and 402b (402c and 402d) via driving signal lines 416a and 416b (416c and 416d) at frequencies different from one another, a plurality of light irradiation means to irradiate the scalp of the subject 10, via a light emitting optical fiber 405a (405b) with light from a coupler 404a (404b) coupling rays of the intensity-modulated light via the respective optical fibers 403a and 403b (403c and 403d), and a plurality of light receiving means formed by detectors 408a and 408b provided to each of light receiving optical fibers 407a and 407b such that tips thereof are positioned near light irradiation points of the plurality of light irradiation means at preset distances from the light irradiation points (e.g. 15 mm and 30 mm). The light passed through the living body is collected by the light receiving optical fibers 407a and 407b, then subjected to photoelectric conversion by the detectors 408a and 408b, and amplified. Here, tips of the light emitting optical fibers 405a and 405b and light receiving optical fibers 407a and 407b have light emitting probes 501a and 501b and light receiving probes 502a and 502b, respectively, for retaining the optical fibers and appropriately mounting on the subject 10. Moreover, a probe holder 503 is fixed to the subject 10 in order to retain the plurality of probes.
The light receiving means detects light reflected and/or transmitted by the inside of the subject 10 and converts the light into electrical signals. As the detector 408, a photoelectric conversion element represented by a photomultiplier tube or photodiode is used. A case of using two types of wavelengths is described in
The electric signals representing intensity of light passed through the living body and subjected to photoelectric conversion by the detectors 408a and 408b are respectively input to lock-in amplifiers 409a to 409d. The lock-in amplifiers 409a to 409d are also input with reference signals 417a to 417d from the oscillators [modulators] 401a and 401b (401c and 401d). For example, the lock-in amplifiers 409a and 409b output separated light of 695 nm from the light sources 402a and 402c and extracted by lock-in processing and the lock-in amplifiers 409c and 409d output separated light of 830 nm from the light sources 402b and 402d. Here, in
Intensity signals, of the passed light of separated respective wavelengths, which are output from the lock-in amplifiers 409a to 409d are subjected to analog-digital conversion by an analog-digital converter 410 and then transmitted to a computer for measurement control 411. The computer for measurement control 411 calculates variations in the oxy-hemoglobin concentration, deoxy-hemoglobin concentration-length, and total hemoglobin concentration-length from detection signals at the respective detection points using the intensity signals of the passed light by a well-known procedure described in NPL 1 or the like and stores in a storage device as time series information at the plurality of measurement points. Incidentally, an example of performing the lock-in processing and then performing the analog-digital conversion has been described here; however, it is also possible to perform the analog-digital conversion of the signals from the detectors and then digitally performing the lock-in processing. Furthermore, the example of separating a plurality of rays of light by a modulation method has been described; however, without limiting thereto, it is also possible to use, for example a time division method where a plurality of rays of light is discriminated by shifting, in terms of time, timings of irradiating with the plurality of rays of light. In this case, if the shifted time of timings is set sufficiently short to allow approximation thereof as long as a value of hemoglobin concentration-length does not vary, measurement values can be deemed as simultaneous measurement values.
Even when timings of irradiation or detection are shifted in order to avoid saturation of the detectors, variation values of the hemoglobin concentration-length that can be deemed as simultaneous values can be measured in a similar manner.
The computer 412 includes an input unit, analysis unit, storage unit, and an extraction unit and analyzes, in the analysis unit, a result calculated by the computer for measurement control 411. The input unit is input externally with settings such as analysis conditions. Note that when the computer 412 has a display function, the display unit 413 may not be included therein. The analysis result from the analysis unit is stored in the storage unit. The extraction unit extracts information of local brain hemodynamics of the subject 10 from signals analyzed by the analysis unit. The information of local brain hemodynamics of the subject 10 extracted by the extraction unit is displayed on a display unit 413. The computer for measurement control 411 and computer 412 are separately illustrated in
Exemplary display is illustrated in
Exemplary display is illustrated in
The skin contribution ratio and brain contribution ratio will be described.
[Formula 5]
Brain contribution ratio=100×value of brain blood flow-derived component signal amplitude/(value of brain blood flow-derived signal amplitude+value of skin blood flow-derived signal amplitude)[%] (13)
[Formula 6]
Skin contribution ratio=100×value of skin blood flow-derived component signal amplitude/(value of brain blood flow-derived signal amplitude+value of skin blood flow-derived signal amplitude)[%] (14)
The separated signals are displayed in
The present invention allows for separating and extracting, from measurement signals, brain blood flow-derived components and skin blood flow-derived components in real-time depending on an object with a human head optical measurement apparatus using visible light or near infrared light, thereby enhancing accuracy and reproducibility of human brain function measurement.
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/JP2013/083817 | 12/18/2013 | WO | 00 |