CN100525701C - Imaging device - Google Patents
Imaging device Download PDFInfo
- Publication number
- CN100525701C CN100525701C CNB2003801057967A CN200380105796A CN100525701C CN 100525701 C CN100525701 C CN 100525701C CN B2003801057967 A CNB2003801057967 A CN B2003801057967A CN 200380105796 A CN200380105796 A CN 200380105796A CN 100525701 C CN100525701 C CN 100525701C
- Authority
- CN
- China
- Prior art keywords
- mentioned
- image
- scattering
- imaging device
- spectrum
- 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
- 238000003384 imaging method Methods 0.000 title claims description 52
- 238000001228 spectrum Methods 0.000 claims abstract description 77
- 238000004364 calculation method Methods 0.000 claims abstract description 22
- 230000003287 optical effect Effects 0.000 claims description 27
- 238000009826 distribution Methods 0.000 claims description 26
- 230000003595 spectral effect Effects 0.000 claims description 26
- 239000011159 matrix material Substances 0.000 claims description 24
- 238000013507 mapping Methods 0.000 claims description 21
- 210000004204 blood vessel Anatomy 0.000 claims description 16
- 230000008859 change Effects 0.000 claims description 13
- 238000005286 illumination Methods 0.000 claims description 12
- 238000010521 absorption reaction Methods 0.000 claims description 9
- 238000004458 analytical method Methods 0.000 claims description 7
- 239000006185 dispersion Substances 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 7
- 238000012545 processing Methods 0.000 abstract description 26
- 239000002245 particle Substances 0.000 description 34
- 230000000694 effects Effects 0.000 description 27
- 238000000034 method Methods 0.000 description 22
- 230000000875 corresponding effect Effects 0.000 description 19
- 206010028980 Neoplasm Diseases 0.000 description 13
- 210000004369 blood Anatomy 0.000 description 13
- 239000008280 blood Substances 0.000 description 13
- 210000000981 epithelium Anatomy 0.000 description 13
- 230000002496 gastric effect Effects 0.000 description 13
- 210000004027 cell Anatomy 0.000 description 9
- 230000006870 function Effects 0.000 description 8
- 210000001519 tissue Anatomy 0.000 description 8
- 210000005081 epithelial layer Anatomy 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 230000007170 pathology Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 239000000284 extract Substances 0.000 description 5
- 238000000605 extraction Methods 0.000 description 5
- 210000004877 mucosa Anatomy 0.000 description 5
- 206010030155 Oesophageal carcinoma Diseases 0.000 description 4
- 201000011510 cancer Diseases 0.000 description 4
- 210000003238 esophagus Anatomy 0.000 description 4
- 230000007246 mechanism Effects 0.000 description 4
- 210000004400 mucous membrane Anatomy 0.000 description 4
- 230000010287 polarization Effects 0.000 description 4
- 238000003860 storage Methods 0.000 description 4
- 208000017897 Carcinoma of esophagus Diseases 0.000 description 3
- 210000000805 cytoplasm Anatomy 0.000 description 3
- 238000005315 distribution function Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 201000005619 esophageal carcinoma Diseases 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000012821 model calculation Methods 0.000 description 3
- 230000002093 peripheral effect Effects 0.000 description 3
- 230000001360 synchronised effect Effects 0.000 description 3
- 210000001835 viscera Anatomy 0.000 description 3
- 229910052724 xenon Inorganic materials 0.000 description 3
- FHNFHKCVQCLJFQ-UHFFFAOYSA-N xenon atom Chemical compound [Xe] FHNFHKCVQCLJFQ-UHFFFAOYSA-N 0.000 description 3
- 102000001554 Hemoglobins Human genes 0.000 description 2
- 108010054147 Hemoglobins Proteins 0.000 description 2
- 241001062009 Indigofera Species 0.000 description 2
- 208000000453 Skin Neoplasms Diseases 0.000 description 2
- 208000027418 Wounds and injury Diseases 0.000 description 2
- 230000009471 action Effects 0.000 description 2
- 238000000149 argon plasma sintering Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 230000006378 damage Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 239000004744 fabric Substances 0.000 description 2
- ZZUFCTLCJUWOSV-UHFFFAOYSA-N furosemide Chemical compound C1=C(Cl)C(S(=O)(=O)N)=CC(C(O)=O)=C1NCC1=CC=CO1 ZZUFCTLCJUWOSV-UHFFFAOYSA-N 0.000 description 2
- 208000014674 injury Diseases 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 230000003902 lesion Effects 0.000 description 2
- 230000014759 maintenance of location Effects 0.000 description 2
- 210000004379 membrane Anatomy 0.000 description 2
- 239000012528 membrane Substances 0.000 description 2
- 230000001575 pathological effect Effects 0.000 description 2
- 238000013139 quantization Methods 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 201000000849 skin cancer Diseases 0.000 description 2
- 238000004611 spectroscopical analysis Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000012327 Endoscopic diagnosis Methods 0.000 description 1
- 208000000461 Esophageal Neoplasms Diseases 0.000 description 1
- 206010020880 Hypertrophy Diseases 0.000 description 1
- 208000037062 Polyps Diseases 0.000 description 1
- 230000002745 absorbent Effects 0.000 description 1
- 239000002250 absorbent Substances 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 210000000601 blood cell Anatomy 0.000 description 1
- 210000003850 cellular structure Anatomy 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000004040 coloring Methods 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000003748 differential diagnosis Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000009792 diffusion process Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 201000004101 esophageal cancer Diseases 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 210000001156 gastric mucosa Anatomy 0.000 description 1
- 210000001035 gastrointestinal tract Anatomy 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000011065 in-situ storage Methods 0.000 description 1
- 230000009545 invasion Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 210000002751 lymph Anatomy 0.000 description 1
- 230000036210 malignancy Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000015654 memory Effects 0.000 description 1
- 210000003470 mitochondria Anatomy 0.000 description 1
- 238000002715 modification method Methods 0.000 description 1
- 210000002445 nipple Anatomy 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000001874 polarisation spectroscopy Methods 0.000 description 1
- 230000037452 priming Effects 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 238000007920 subcutaneous administration Methods 0.000 description 1
- 210000004876 tela submucosa Anatomy 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 238000002834 transmittance Methods 0.000 description 1
- 230000024883 vasodilation Effects 0.000 description 1
Images
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
- Endoscopes (AREA)
Abstract
一种图像处理电路，具备：光谱推断部，输入各图像数据，并且，从推断用数据供给部得到推断光谱所需的数据，来推断各像素的光谱；散射特征计算部，根据来自光谱推断部的各像素的光谱、和来自特征计算用数据供给部的特征计算用所需的数据，计算多个散射特征；彩色图像生成部，根据来自散射特征计算部的散射特征图像进行显示颜色计算，决定作为彩色图像而应显示的各像素的RGB值，并输出RGB图像。
An image processing circuit comprising: a spectrum estimating unit that inputs each image data, and obtains data required for estimating the spectrum from an estimating data supply unit to estimate the spectrum of each pixel; The spectrum of each pixel and the data required for feature calculation from the feature calculation data supply part calculate a plurality of scattering features; the color image generation part performs display color calculation based on the scattering feature image from the scattering feature calculation part, and determines The RGB value of each pixel to be displayed as a color image, and an RGB image is output.
Description
Technical field
The present invention relates to carry out the imaging processing device of the scattering imaging processing of bio-tissue.
Background technology
As everyone knows, digestive tract tumor's diseases such as esophageal carcinoma, it is that the top layer from gastrointestinal mucosal is that basal layer in the epithelial layer begins to take place mostly.From the mutant propagation that basal layer begins to take place, its grade malignancy develops to the direction of the whole epithelial layer of displacement.Intraepithelial this tumprigenicity variation is accompanied by pathological structural change, so-called structure variation with the monomeric variation of cell, presents the irregular tissue different with normal pathological image and arranges.
The purpose of endoscopic diagnosis is to find as far as possible in early days this tumor.If can find in early days, the probability that can effect a radical cure by the operation of low invasion and attack such as endoscope treatment uprises.
But some tumor such as esophageal carcinoma is the clear and definite shape (polyp or concave shape) of stage shortage in early days, finds that it is not necessarily easy.
In order to find in early days, differentiate that these lack the tumor of change of shape, many schemes have been proposed up to now.
Think therein more effectively, consider that above-mentioned nucleus variation, structure variation cause that optical scattering changes, catching this scattering by optical means changes, carry out the technology of indiscoverable early lesion in common observation image, scattering spectroscopy, be scattering into picture (for example 2002-No. 95635 communiques of TOHKEMY).
In the past, for measure, the intraepithelial scattering properties of imaging, the scheme of a fairly large number of use polarization optical system has been proposed.It is following technology, rear single scattering light from epithelial surface is keeping polarized component, and become no polarisation from the repeatedly scattered light of propria lamina (mucous layer, tela submucosa), based on this understanding, carry out imaging according to the difference observation of level, vertical polarized component.
By the observation optical illumination bio-tissue of polarisation, thus, the backscattering light of arranging from the cell of epithelial surface is observed as the polarized component of equidirectional (for example horizontal direction) in certain direction (for example horizontal direction) in use.On the other hand,, by constituting the repeatedly scattering process of thing, thus, become no polarized condition, be observed as diffusing from tissue surface at cell and various tissue at the last light of propagating in subcutaneous.
Observe this light with direction with the polariscope of observation light different (for example vertical direction), can estimate repeatedly the size of scattered light thus.Use this value, the influence of the repeatedly scattering that is comprised in the observation light (horizontal polarisation) of roughly keeping polarisation is revised, extract single scattering light from the cell on epithelium top layer by difference operation.
The single scattering phenomenon of cell can be modeled as and float over the Mie scattering that the various spheroidal particles in the protoplasm cause.As the characteristic of Mie scattered light, the scattering spectrum shape depend on scattering particles particle diameter, with the refractive index ratio of peripheral medium (this situation mainly is a protoplasm) and observation wavelength.Particularly importantly, particle diameter and spectral shape is related.
As mentioned above, use the Mie scattering model,,, thus, can infer the particle diameter of mucous epithelium Yi Bian the spectral shape of the single scattering light that extracts with the polarisation measurement is carried out match by nonlinear least square method etc. Yi Bian change particle diameter.
At epithelial layer, nucleus is considered to mainly one of contributor be done in scattering, can think that particle diameter and the nuclear size inferred by said method have high correlation.
Above-mentioned nucleus variation is to be accompanied by nuclear enlargement (along with tumor changes, it is bigger than normal condition that nucleus becomes), therefore, infers nuclear size by this method, infers the state that epithelial tumor changes exactly.
Thereby, used the spectroscopy and the imaging of polarisation, have can the enlargement of imaging nucleus probability.
As described above, polarization spectroscopy and imaging have the probability of inferring the nucleus enlargement quantitatively, but are considering under the applicable cases of endoscope, have problem described as follows.
Adorning the special optical viewer of polarization optical system in needing.
Owing to use polarisation, need highly sensitive imaging apparatus (be used to generate, accept the optical element of polarized light: if use polariscope, then the energy of light drops to extremely low).
Because the modelling of carrying out the Mie scattering, need be used for knowing definitely or the work (at the Mie scattering model, spectral shape depends on the angle of the backscattering light that observes more greatly) of the angle (backscattering angle) of priming illumination light and observation light.
By solving above problem, might realize high performance scattering imaging device, still, also have above-mentioned problem from the angle of installation cost.
Summary of the invention
The present invention makes for addressing the above problem, and its purpose is to provide a kind of imaging device, is prerequisite to utilize existing endoscope optical system, only by improvement light supply apparatus and processor inside, can be scattering into picture easily.
Imaging device of the present invention comprises: light supply apparatus, shine more than one at least frequency band light; Camera head will be used for from the irradiates light of above-mentioned light supply apparatus observing, and organism be observed picture be transformed to picture signal; And processor, generate the organism image according to above-mentioned picture signal; It is characterized in that described processor has: spectrum deduction portion, according to using data from each view data of synchronization memorizer with from inferring that the spectrum of using the data supply unit is inferred, infer the spectrum of each pixel; The scattering signatures calculating part, according to from the spectrum of each pixel of above-mentioned spectrum deduction portion and from feature calculation with the feature calculation of data supply unit with required data, the scattering signatures of calculating bio-tissue; And coloured image generating unit, based on the above-mentioned scattering signatures that calculates by above-mentioned scattering signatures calculating part, carry out color and show calculating, decision should show the rgb value of each pixel of scattering signatures as coloured image, generation offers the display image of display part, described processor generates the organism image as described display image, and this organism image has the described scattering signatures as image information at least.
Other features of the present invention and interests will be fully aware of by the following description.
Description of drawings
Relevant the 1st embodiment of the present invention of Fig. 1 to Figure 10.
Fig. 1 is the structure chart of the structure of expression scattering imaging device.
Fig. 2 is the figure of structure of the convolutional filter of presentation graphs 1.
Fig. 3 is the figure of dichroism of the convolutional filter of presentation graphs 2.
Fig. 4 is the block diagram of structure of the image processing circuit of presentation graphs 1.
Fig. 5 is the figure that the scattering imaging device of key diagram 1 is scattering into the gastrointestinal mucosal tissue of picture.
Fig. 6 is the figure of wave filter scattering properties, that have dichroism of the expectation of the explanation gastrointestinal mucosal tissue that produces Fig. 5.
Fig. 7 is the figure of effect of the spectrum deduction portion of key diagram 4.
Fig. 8 is the flow chart of handling process of the spectrum deduction portion of presentation graphs 4.
Fig. 9 is the figure of effect of the scattering signatures calculating part of key diagram 4.
Figure 10 is the figure of structure of variation of the convolutional filter of presentation graphs 1.
Figure 11 is the block diagram of the structure of the relevant image processing circuit of expression the 2nd embodiment of the present invention.
Figure 12 and Figure 13 are about the 3rd embodiment of the present invention.
Figure 12 is the structure chart of the structure of expression scattering imaging device.
Figure 13 is the figure of structure of the convolutional filter of expression Figure 12.
Figure 14 is the figure of expression about the surface camera head of the 4th embodiment of the present invention.
Figure 15 to Figure 18 is about the 5th embodiment of the present invention.
Figure 15 is the structure chart of the structure of presentation video treatment circuit.
Figure 16 is the 1st figure of effect of the image processing circuit of explanation Figure 15.
Figure 17 is the 2nd figure of effect of the image processing circuit of explanation Figure 15.
Figure 18 is the flow chart of effect of the image processing circuit of explanation Figure 15.
Figure 19 to Figure 23 is about the 6th embodiment of the present invention.
Figure 19 is the structure chart of the structure of presentation video treatment circuit.
Figure 20 is the 1st figure of effect of the image processing circuit of explanation Figure 19.
Figure 21 is the 2nd figure of effect of the image processing circuit of explanation Figure 19.
Figure 22 is the 3rd figure of effect of the image processing circuit of explanation Figure 19.
Figure 23 is the flow chart of effect of the image processing circuit of explanation Figure 19.
The specific embodiment
In order to illustrate in greater detail the present invention, be illustrated with reference to the accompanying drawings.
The 1st embodiment:
As shown in Figure 1, the endoscope apparatus 1 that constitutes the scattering imaging device of present embodiment comprises: fujinon electronic video endoscope 3, light supply apparatus 4 and video processor 7, described fujinon electronic video endoscope 3 has CCD2, as being inserted into the image mechanism of taking body cavity inner tissue in the body cavity, described light supply apparatus 4 provides illumination light to fujinon electronic video endoscope 3,7 pairs of image pickup signals from the CCD2 of fujinon electronic video endoscope 3 of described video processor carry out signal processing, and show endoscopic images at observation display 5.
As shown in Figure 2, that convolutional filter 14 constitutes is discoid, be the structure of rotating shaft with the center, dispose C1 wave filter 14c1, C2 wave filter 14c1, C3 wave filter 14c1, be configured for exporting the bank of filters of face continuous light as shown in Figure 3 with dichroism.And as shown in Figure 1, convolutional filter 14 is to make convolutional filter motor 18 be driven control and rotation by control circuit 17.
In addition, xenon lamp 11, iris apparatus 13 and convolutional filter motor 18 are by power supply unit 10 supply capabilities.
Get back to Fig. 1, video processor 7 possesses: CCD drive circuit 20, driven CCD 2; Amplifier 22 amplifies via the image pickup signal of having taken body cavity inner tissue to object optical system 21, with CCD2; Treatment circuit 23 is to the image pickup signal that has passed through amplifier 22 be correlated with double sampling and noise removal process etc.; A/D changer 24, the image pickup signal that will pass through treatment circuit 23 is transformed to the view data of digital signal; White balance circuit 25 is handled implementing white balance from the view data of A/D changer 24; Selector 26 and synchronization memorizer 27,28,29 are used for the face continuous light of synchronous convolutional filter 14; Image processing circuit 30 reads out in each view data of the face continuous light of storage in the synchronous memories 27,28,29, and carries out γ correcting process, emphasizing contour processing, color processing etc.; D/ A circuit 31,32,33 will be transformed to analogue signal from the view data of image processing circuit 30; Time mark generator 35, input from the control circuit 17 of light supply apparatus 4, with the synchronous synchronizing signal of rotation of convolutional filter 14, export various timing signals to above-mentioned each circuit; Light adjusting circuit 43, the image pickup signal of treatment circuit 23 has been passed through in input, the iris apparatus 13 of control light supply apparatus 4, and carry out appropriate brightness control.
As shown in Figure 4, image processing circuit 30 possesses: spectrum deduction portion 52, import each view data from synchronization memorizer 27,28,29, and, from inferring that obtaining spectrum with data supply unit 51 infers required data, the spectrum of inferring each pixel; Scattering signatures calculating part 54, according to from the spectrum of each pixel of spectrum deduction portion 52 and from feature calculation with the feature calculation of data supply unit 53 with required data, calculate a plurality of scattering signatures; Coloured image generating unit 55 shows that according to the scattering signatures image from scattering signatures calculating part 54 color calculates, rgb value scattering signatures, each pixel that decision should show as coloured image, and export to D/ A circuit 31,32,33 as the RGB image.
In addition, infer with data supply unit 51 and feature calculation and be set in the interior or external module of video processor 7 with data supply unit 53.
Gastrointestinal mucosal tissues such as esophagus have roughly the structure as Fig. 5.Tumors such as esophageal carcinoma begin to take place from the basal layer that separates epithelial layer and mucous layer.The tumor that begins to take place from basal layer is replaced whole epithelial layer along with nucleus variation and structure variation with mutant, forms state through so-called variation, develops to cancer.
Epithelial layer is made of squamous epithelium, because of its fine and close cellularity shows stronger scattering properties.In addition, its scattering properties has wavelength dependency, can think and have from short wavelength to long wavelength the characteristic that reduces (thereby, can think that short-wavelength light nearly all is scattered, reflects at last Intradermal, seldom can be deep under the mucous layer below it).
Hence one can see that, and short-wavelength light is more suitable in the variation of catching intraepithelial scattering properties than long wavelength light.This is the example explanation with the illumination of narrow-band multiband for imaging too below.
As shown in Figure 6, in the convolutional filter of the face of generation continuous light,, use combination usually as the broadband filter of C-B, C-G, C-R in order to obtain the color rendering of nature.By the convolutional filter high speed rotating, 3 band of light are taken with monochromatic CCD successively by time sequence illumination subject, and are synthetic by video processor, the band image corresponding with each illumination light serves as the indigo plant of observation display, green, red channel, and is shown as a width of cloth coloured image.
But, if can make wave filter, for example make the centre wavelength of C-B, C-G such as C2, C4, C6, C-R not do bigger variation with desired characteristics, double spectrum amplitude implemented the arrowbandization, can also improve the contrast of blood-vessel image.
But,, select wave band from short-wavelength light as described above here because convolutional filter 14 is imaged as purpose with intraepithelial scattering.C1, the C2, the C3 that are equivalent to Fig. 6.A plurality of band images in these short wavelength zones can show intraepithelial structure preferably than the wave band (C4, C5, C6, C7) of long wavelength side as can be known.
Wherein, though also consider the band image corresponding with C1, C2, C3 illumination light served as indigo plant, green, red channel, reproduce on observation display as a width of cloth coloured image, but, what kind of (for example changes on colouring information and its change color and the pathology corresponding with what kind of scattering properties, the degree of the degree of structure variation or nucleus enlargement etc.) corresponding indeterminate, and be not easy to understand image.And then, even the doctor of endoscope observes this image in checking process, also be difficult to help the early discovery of tumor.
Therefore, consider as follows in the present embodiment: infer spectral reflectance according to these a plurality of band images, according to the spectral reflectance of inferring and based on the optical model of bio-tissue, be transformed to the higher feature of dependency on the pathology, the variation of its characteristic quantity in the image is shown as color information.
The purpose of present embodiment is, according to the spectral reflectance that the arrowband image is inferred each pixel, infers that the variation in image generates color information according to the characteristic quantity of being inferred based on the higher characteristic quantity of the pathology dependency of optical model.
(effect)
(be envisioned for 3 wave bands in the present embodiment, as shown in Figure 6, be S-band from the arrowband band image of synchronization memorizer 27,28,29 outputs; Corresponding to C1, C2, C3) be input in image processing circuit 30 the spectrum deduction portion 52 that is provided with.Spectrum deduction portion 52 is in image treatment part or be located at deduction the external module and obtain spectrum with data supply unit 51 and infer required data, the spectrum of inferring each pixel.
Infer spectrum, be the input value that spectrum picture becomes scattering signatures calculating part 54.Scattering signatures calculating part 54 in image treatment part or the feature calculation that is located at the external module obtain feature calculation with required data with data supply unit 53, calculate a plurality of scattering signatures.Carve at this moment, each pixel is distributed a plurality of scattering signatures.
Scattering signatures calculating part 54 is to coloured image generating unit 55 output scattering signatures images.Coloured image generating unit 55 is carried out Show Color calculating according to the scattering signatures image, determines should show as coloured image the rgb value of each pixel of scattering signatures, exports to D/ A circuit 31,32,33 as the RGB image.
The effect of each module (spectrum deduction portion 52, scattering signatures calculating part 54 and coloured image generating unit 55) then, is described.Detailed is open in document " V.Backman; R.Gurjar; K.Badizadegan, I.Itzkan, R.R.Dasari; L.T.Perelman; and M.S.Feld, ' Polarized Light Scattering Spectroscopy for Quantitative Measurementof Epithelial Cellular Structures In Situ, ' IEEE J.Sel.Top.QuntumElectron; 5,1019-1026 (1999) ".In addition, in following mathematical formulae, vector (lower case), matrix (capitalization) that " ^ " expression has a plurality of key elements.
(spectrum deduction portion 52)
The subject spectral reflectance is represented with the imaging side formula of formula (1) with the relation of the pixel value of the multi-band image of being observed.
g＾＝H＾f＾+n＾
Wherein, g^ has the dimension (N identical with the wave band number; Be 3 in the present embodiment) the pixel value column vector.F^ is a subject spectral reflectance column vector, is L on wavelength direction, and numerical value is by discretization.N^ is the noise column vector.The sytem matrix of L * N that H^ promptly is made of N capable vector for the beam split sensitivity characteristic of each wave band.
Problem is, known H^, the spectral reflectance of inferring subject according to observation g^.H^ is known as the dichroism of the imaging systems such as spectral sensitivity characteristic of the spectrophotometric transmittance characteristics of observing light spectrum, narrow band filter and imaging apparatus.
Usually, because " wave band is counted N＜wavelength hits L ", the deduction problem is ineligible (ill-condition).That is, relative with g^, there be numerous (in other words, because equational number is lacked than unknown number, have much as can be known and separate, if not additional some condition can not determine that being one separates) in the f^ that satisfies formula (1).
Therefore, need to find out appropriate deduction and separate by preparing preview information restriction solution space (be the L gt this moment) (Here it is is used for solution is decided to be 1 condition).That is, be summed up as following problem: use preview information, in the L dimension space, obtain only separating in the segment space that spectrum distributed that can become candidate solution.
As the method for using predicted condition, generally use Wiener to infer.The deduction matrix A that expression Wiener infers in formula (2).By multiply by this deduction matrix A from the right of observation vector g, infer vector.Thereby vector deduction mechanism is as having used the matrix operations device of predefined deduction matrix A and having moved.
A＾＝R
f＾H
γ＾(H＾R
f＾H
γ＾+R
n＾)
-1 …(2)
Wherein, R
f^ is autocorrelation matrix (L * L), the R of the wavelength direction of the subject spectrum that should infer
n^ is the autocorrelation matrix of the additive noise that occurs as n^ in formula (1).R
n^ can infer according to the noise characteristic of the imaging system of measuring in advance (being meant the system that has made up light source and optical viewer here), be known.Here so-called preview information is R
f^, R
n^, particularly, R
f^ is the most important parameter of the appropriate property of the decision spectral reflectance of being inferred.
In the past, this autocorrelation matrix R
fThe following situation of ^ is more: the spectrum that supposition is inferred wavelength direction be level and smooth (promptly, there are not rapid wavelength change such as bright line spectrum, on wavelength direction, be milder characteristic), use differential operator inverse of a matrix (promptly in the spatial frequency zone, being the low frequency emphasis filter); Or use is with the Markov transition matrix of statistical models such as Markov model energy performance point luminous reflectance.In the present embodiment, as autocorrelation matrix R
f^ utilizes autocorrelation matrix, and this autocorrelation matrix is obtained according to the spectrum of being inferred by the discrete particle structural model (hereinafter referred to as optical model) of bio-tissue described later.
Optical model then is described.Bio-tissue comprises that various key elements such as fibrous tissue, cell, lymph corpuscle, blood capillary, nucleus, the interior organella of cell constitute.
Because scattering is to produce in the bigger place of variations in refractive index, so can think that the main scattering object (scattering main body) of bio-tissue is an organella in the cells such as nucleus and mitochondrion.For size and the identical or slightly little particle of observation wavelength, can predict its phase function and scattering coefficient by the Mie scattering model.Phase function is represented to incide the probability that the light the scattering main body is scattered in s ' direction from direction s.In addition, scattering coefficient is the parameter that photon is subjected to scattering several times in the expression per unit distance.
This Mie scattering model has 2 π ma/ λ (λ is a wavelength, and m is a refractive index ratio, and a is the diameter of scattering main body), as the parameter of model.Can think that nucleus and plasmic refractive index ratio do not have bigger change, therefore, we can say that the Mie scattering mainly is with the model of scattering main body as the parameter prediction scattering spectrum.
On the other hand, if know the information (particle size distribution function) that the particle (organella in nucleus, the nucleus) of which kind of degree size in bio-tissue with what kind of density exists, can infer phase function and scattering function with this Mie scattering model thus.The notion of particle size distribution as shown in Figure 7.Actual particle diameter can think from about 0.4 μ m of organella in the cell to about the nuclear 4 μ m, along with structure variation begins development from normal structure, can think that particle size distribution changes (to f2 (d), d is a particle diameter from f1 (d)) like that shown in the arrow among Fig. 6.Use the Mie scattering model, according to the refractive index ratio of particle size distribution function, particle and peripheral medium (about 1.03, peripheral medium is assumed to be protoplasm), calculating phase function and scattering coefficient.Particle size distribution function can be used normal distribution or logarithm normal distribution.For and the change of particle size distribution parameter (meansigma methods or standard deviation) imagination corresponding with target, according to Mie scattering model calculating optical coefficient, based on the optical coefficient of calculating, simulating repeatedly by light propagation model, scattering process calculates spectrum.
Light propagation model can separately use according to following various situations: but utilize the method that is subjected to the diffusion equation of bigger restriction as analytic method at the degree of freedom of favourable mould shapes aspect computation time; Though with and calculate need the time, but for the bigger monte-Carlo model of the degree of freedom of mould shapes etc.
So far, if the arrangement Model Calculation as shown in Figure 8, obtains particle size distribution parameter (meansigma methods, standard deviation) at step S1.At step S1 the particle size distribution parameter is input to the Mie scattering model.At step S3, according to Mie scattering model output scattering coefficient and phase function.In fact, be that each each particle size is used the Mie scattering model, calculate scattering coefficient and phase function, particle size distribution as weighting function, is had the summation averaging of weighting, considered that the Mie scattering of particle size distribution is calculated.
By above Model Calculation, the nucleus variation, the structure variation that are accompanied by intraepithelial tumprigenicity variation are considered as the variation of particle size distribution parameter (meansigma methods, standard deviation), carry out spectrum and calculate, as knowledge consideration in advance, solution space (spectral space) is restricted with this.That is,, calculated the autocorrelation matrix R in the Wiener deduction in advance according to the spectral distribution of inferring by this Model Calculation
f^.
Particularly, at step S4,, calculate and particle size distribution parameter variation (change of meansigma methods, standard deviation) the corresponding spectrum change that obtains according to pathology knowledge by Mie scattering model+light propagation model.Result calculated in step S5, forms in spectral space with the particle size distribution parameter and changes corresponding spectral distribution.Regard this as superclass, infer the autocorrelation matrix of spectrum in wavelength direction.
As mentioned above, use the autocorrelation matrix of inferring in advance, infer intraepithelial scattering spectrum by optical model (particle size distribution model+Mie scattering model+light propagation model).Thereby, use based on H^, R
f^, R
nThe matrix A that the formula of ^ (2) calculates is stored in deduction data supply unit.
(scattering signatures calculating part 54)
Spectrum according to being inferred by spectrum deduction portion 52 can calculate various features.In the present embodiment, be conceived to the particle size distribution parameter, the method for the characteristic quantity of the spectrum of giving chapter and verse, deduction and this parameter correlation.Concept map as shown in Figure 9.
When inferring autocorrelation matrix, know the corresponding spectrum mobility scale of variation in advance with particle size distribution parameter (meansigma methods, standard deviation).
Thereby, with meansigma methods that is counted as scattering signatures and standard deviation characteristic of correspondence axle, be known as F1, F2.That is, spectrum is in the spatial distribution by F1, F2 decision.Thus, projection value (f1, f2) from the spectrum that calculates to F1, F2 and brightness (for example spectrographic area etc.) are as the 3rd value, as scattering signatures.Thereby, storing the spectrum of each feature axis in the data supply unit in feature calculation.In addition, the computing of calculating part becomes the inner product calculating of feature axis spectrum and scattering spectrum.
(coloured image generating unit 55)
For example, serve as from the scattering signatures of scattering signatures calculating part 54 output and spectrographic brightness and be blue, green, red channel, the generation coloured image.At this moment, make the corresponding also quantization suitably of the D/A performance of back levels such as pixel value and 8 bits (bit).For in picture, finding early lesion, as long as, therefore, in frame, calculate the dynamic range of scattering signatures capturing relative scattering variation under the situation mostly, being that nucleus makes a variation, the degree of structure variation is just enough, carry out quantization with 8 levels more set, export as rgb signal than top grade.
From the above mentioned, if sum up the feature of present embodiment, the deduction of spectral reflectance that will be corresponding with each pixel is more shallow as deep degree, be considered to reflect strongly the multi-band image in the short wavelength zone of intraepithelial feature, and with epithelial tissue as the discrete particle structure and modelling; Use the autocorrelation matrix of inferring according to the spectral reflectance on the model basis of calculating by Mie scattering model and light propagation model, infer scattering spectrum thus, will with the projection value of the particle size distribution parameter characteristic of correspondence axle of obtaining in advance of spectral space as scattering signatures, and these characteristic quantities of each pixel are distributed to the color channel, realize being scattering into picture with colour information.
(effect)
According to present embodiment, do not use special optical viewers such as polarization optical system, by the computing in narrow band filter and the processor, can carry out and scattering properties changes relevant imaging, can the variation of visuognosis epithelium inner structure etc. be difficult to observed feature in the past.
In addition, because can common observation image, the convolutional filter 14 of narrow band filter 14C1～14C6 of a plurality of arrowband C1～C6 (with reference to Fig. 6) that also can use as shown in figure 10 installation, at this moment, possesses the memorizer that quantity equates with the wave filter number, with the image of corresponding each narrow band filter.In addition, though it is not shown in this case, but image processing circuit 30 possesses the description of spectrum deduction 52+ of portion scattering signatures calculating part 54 in common observation image production part and the 1st embodiment, also possesses to calculate according to the output of scattering signatures calculating part 54 to emphasize that the contrast of coefficient emphasizes coefficient calculations portion.In frame, according to for example use the meansigma methods of particle size distribution or the relevant values such as composite character of standard deviation and meansigma methods, calculate a quantized value.Also can be worth, to the coefficient of emphasizing of the briliancy channel decision spatial frequency of the image that generates at common observation image production part according to this.Thus, can carry out emphasizing to common observation image based on the contrast of scattering signatures.
The 2nd embodiment:
Because the 2nd embodiment and the 1st embodiment are much at one, difference only is described, and gives identical label to identical structure, omit its explanation.
(structure and effect)
In order to infer intraepithelial scattering spectrum, in short wavelength's frequency band, use a plurality of narrow band filters.The scattering properties of this frequency band is strong, also has the absorption maximum (415nm) of hemoglobin on the other hand.For example, under the situation of considering the flat epithelium of esophagus, in normal mucosa, there is blood capillary hardly, but along with nuclear enlargement, vasodilation or in last Intradermal new vessels hypertrophy in the nipple.This blood capillary image is owing to having the special spectrum that hemoglobin is brought, so can think the main cause of error in scattering spectrum is inferred.Thereby, before inferring scattering spectrum, separate this absorption image and this blood capillary image.
Separate, can regard that these blood capillary images are altofrequency as in the spatial frequency fluctuation, dispersion image itself forms low-frequency image by scattering repeatedly.Particularly, as shown in figure 11, in the image processing circuit 30 of present embodiment, in the prime of scattering spectrum deduction portion 52 filtering portion 61,62,63 is set, its each arrowband band image with spatial filter is corresponding.The action of filtering portion 61,62,63 can realize by the convolution algorithm device of FIR wave filter, infers that by high freguency bandpass filter and scattering spectrum that the blood capillary separation of images is used the low pass filter of usefulness constitutes.
Output from the filtering portion 61,62,63 corresponding with each narrow-band band image, be separated into high frequency imaging C1H, the C2H corresponding with each band image, C3H (footnote H) and low-frequency image C1L, C2L, C3L (footnote L), low-frequency image is exported to blood capillary image forming machine structure 64 to scattering spectrum deduction portion 52, altofrequency image.
In scattering spectrum deduction portion 52, as in the explanation of the 1st embodiment,, calculate the autocorrelation matrix in the Wiener deduction according to the last Intradermal backscattering spectral distribution of inferring by the discrete particle structural model, infer scattering spectrum.
On the other hand, at blood capillary image production part 64, to the high frequency imaging of making from each wave band, by suitable noise remove or according to circumstances blood vessel structure is carried out modeled matched filter, generate the blood capillary image more brightly, and export to picture signal generating unit 65 as luminance information.
In picture signal generating unit 65, according to output, make scattering properties with colour chart from scattering spectrum deduction portion 52, on the other hand, as luminance information, scattering+blood capillary is absorbed image export to observation display by synthetic blood capillary image.
(effect)
According to present embodiment, can not use special optical viewers such as polarization optical system, and by the computing in narrow band filter and the processor, can carry out and scattering properties changes relevant imaging, can the variation of visuognosis epithelium inner structure etc. be difficult to observed feature in the past.In addition, by blood capillary image, can prevent the decline of spectral reflectance deduction precision, and can synthesize and show for Differential Diagnosis important blood capillary pattern and dispersion image by the prior separate absorbent image of space filtering mechanism.
The 3rd embodiment:
Because the 3rd embodiment and the 1st embodiment are much at one, thus difference only is described, and give identical label to identical structure, omit its explanation.
(structure and effect)
As shown in figure 13, the convolutional filter 14 of present embodiment constitutes discoid and has with the center is the double-decker of rotating shaft, radial component in the outside disposes C1 wave filter 14C1, C2 wave filter 14C2, the C3 wave filter 14C3 that constitutes the 1st bank of filters, and described the 1st bank of filters is used to export the face continuous light of arrowband of the dichroism of C1～C3 as shown in Figure 6; Radial component in the inboard disposes C4 wave filter 14C4, C5 wave filter 14C5, the C6 wave filter 14C6 that constitutes the 2nd bank of filters, and described the 2nd bank of filters is used to export the face continuous light with dichroism of C4～C6 shown in Figure 6.
And, as shown in figure 12, convolutional filter 14 makes convolutional filter motor 18 be driven control and rotation by control circuit 17, in addition, diametric move (be convolutional filter 14, move with light path is vertical, the 1st bank of filters of optionally mobile convolutional filter 14 or the 2nd bank of filters on light path), be according to from the control signal of the pattern commutation circuit 42 in the video processor 7, undertaken by pattern switching motor 19.
In addition, by power supply unit 10 to xenon lamp 11, iris apparatus 13, convolutional filter motor 18 and pattern switching motor 19 supply capabilities.
On fujinon electronic video endoscope 2, be provided with mode selector switch 41, pattern commutation circuit 42 outputs of the output of this mode selector switch 41 in video processor 7.The pattern commutation circuit 42 of video processor 7 is to the pattern switching motor 19 output control signals of light adjusting circuit 43, brightness adjustment control parameter commutation circuit 44 and light supply apparatus 4.
Brightness adjustment control parameter commutation circuit 44 is to 1st bank of filters or the 2nd bank of filters corresponding brightness adjustment control parameter of light adjusting circuit output with convolutional filter 14, light adjusting circuit 43 is according to the control signal of coming self mode commutation circuit 42 with from the brightness adjustment control parameter of brightness adjustment control parameter commutation circuit 44, the iris apparatus 13 of control light supply apparatus 4 carries out appropriate brightness control.
(effect)
As mentioned above, in the present embodiment,,, can be undertaken observing in the body cavity by common observation light by using C4 wave filter 14C4, C5 wave filter 14C5, C6 wave filter 14C6 except the effect of the 1st embodiment.
The 4th embodiment:
Because the 4th embodiment and the 1st embodiment are much at one, thus difference only is described, and give identical label to identical structure, omit its explanation.
(structure and effect)
In the respective embodiments described above, the embodiment that is scattering into picture that camera head is set in endoscope, carries out body cavity inner tissue has been described, but in the present embodiment, the irradiation narrow band light is described on the surface and scattering imaging device that can the detection of skin cancer etc.
As shown in figure 14, in the present embodiment, replace fujinon electronic video endoscope 2 to be provided with surface camera head 84, this surface camera head 84 possesses the shade 81 that contacts with skin at front end, is configured to cyclic light guide 82, by the image pickup part 83 that object optical system and CCD are constituted.
And, make shade 81 and contact skin, by the face continuous light of light guide 82 irradiations,, send image pickup signal to video processor 7 by image pickup part 83 shootings from the arrowband of dichroism shown in the C1～C3 of light supply apparatus 4.
(effect)
In this wise, in the present embodiment, also can access the action effect identical with the 1st embodiment for the surface, but detection of skin cancer etc.
The 5th embodiment:
Because the 5th embodiment and the 1st embodiment are much at one, thus difference only is described, and give identical label to identical structure, omit its explanation.
Gastrointestinal mucosals such as esophagus have hierarchy.The early cancer is mainly taken place in the top layer, development.Thereby, in order to find more early stage cancer, the pathology that take place in the mucosa top layer need be changed as image capturing.
But, because the mucosa top layer is extremely thin, generally from the reflected light contrast top layer (the 1st layer) of organism more by the reacting condition below down layer (the 2nd layer).This variation is meant scattering, absorption, specifically, is pathology structure and vessel density.
Therefore, need as far as possible in the 2nd layer changes in optical properties unaffected and emphasize a kind of algorithm of scattering signatures amount of the 1st layer optical characteristics (scattering variation) as being scattering into picture.
In the 5th embodiment,, obtain mapping to above-mentioned scattering signatures amount according to the observation of beam split image value or multi-band image value.
(structure)
As shown in figure 15, in image processing circuit 30, feature calculation storing in data supply unit 53 each internal organs described later of obtaining by multiple discriminant analysis gastrointestinal mucosal mapping (enum) data 100, be gastric mucosa mapping (enum) data, esophageal mucosa membrane injury mapping (enum) data etc., scattering signatures calculating part 54 is selected signal according to the internal organs from not shown input mechanism, with reading corresponding gastrointestinal mucosal mapping (enum) data 100 the data supply unit 53, calculate scattering signatures from feature calculation.
Then, the gastrointestinal mucosal mapping (enum) data 100 of using storage in the data supply unit 53 in feature calculation is described.The structure of gastrointestinal mucosal tissues such as esophagus has been described in Fig. 5, here with epithelial layer as the 1st layer, with the lower floor below the basal layer all as the 2nd layer, then in the spectrum that the spectrum deduction portion 52 by image processing circuit 30 infers, reflect the 2nd layer influence strongly, covered the spectrum change of the 1st layer nucleus enlargement of imagination.
Therefore, in the present embodiment, try to achieve less to the 2nd layer influence, as to emphasize the 1st layer the scattering signatures mapping of segment space in the observation spectrum space, the mapping of trying to achieve is stored in feature calculation with data supply unit 53 as gastrointestinal mucosal mapping (enum) data 100.
This mapping can be passed through known multiple discriminant analysis, as relying under the 2nd layer the condition of minimum deviation of characteristic variations, with the 1st layer characteristic variations, be to depend on the maximized Linear Mapping of spectrum change of scattering signatures and obtain at this moment.
For example at spectral space, spectrum such as Figure 16 of bio-tissue distribute.Here, class is represented the identical data acquisition system of scattering signatures of the destination layers such as epithelium in the early esophageal cancer.Described 2 classes in this Figure 16, therefore, expression has the data acquisition system of 2 kinds of different epitheliums of scattering signatures.
In each class, occur with epithelium beyond layer scattering and the distribution of the corresponding data of the variation of absorption characteristic.So-called to the spatial mapping of multiple differentiation, be meant under the minimized condition of distribution in described class Figure 16, with the maximized conversion of between class distance (be mapped as under the linear situation, as the linear identification of Fei Xier (Off イ Star シ ャ-) and known) at 2 classes and its.
By this mapping, as shown in figure 17, deviation (class in disperse) the minimized space of the spectrum of the bio-tissue of Figure 16 between class distance (disperseing between class) maximization and class is mapped.Disperse between dispersion and class in the class for example can pass through simulation of Light Scattering.
That is, mean that epithelium etc. becomes the scattering properties of bio-tissue of the layer of target, Min. ground suppressed this layer in addition absorption and the state of the influence of scattering properties under emphasized.
(effect)
In the present embodiment, as shown in figure 18, if image processing circuit 30 is imported from synchronization memorizer 27 at step S51,28, each view data of 29, at step S52, spectrum deduction portion 52 is from inferring the spectrum of inferring each pixel with data supply unit 51 acquisition organism spectrum autocorrelations certificates, at step S53, scattering signatures calculating part 54 is selected signal according to internal organs, from feature calculation with data supply unit 53 read in disperse in the class and class between the mapped gastrointestinal mucosal mapping (enum) data 100 in space of decentralized optimization, and calculating scattering signatures, at step S54, coloured image generating unit 55 shows that according to the scattering signatures image from scattering signatures calculating part 54 color calculates, decision should show the rgb value of each pixel of scattering signatures as coloured image, and as the RGB image to D/ A circuit 31,32,33 outputs.
In coloured image generating unit 55, if will be made as 3 dimensions by the scattering signatures space that multiple discriminant analysis obtains, to the rgb color channel allocation constitute scattering signatures spatial each.Pre-specified each, maximum, minimum, in its scope, distribute the scope of each color channels.
As another kind of color allocation method, make in the image contrast the biggest ground assignment information.With frame (perhaps) be unit to the arithmetic unit input picture, in picture, data map is arrived the scattering signatures space.Corresponding with the number of pixel value in the picture, DATA DISTRIBUTION is in the scattering signatures space.This maximum dispersion direction should become can reflect between class disperse, be the direction that scattering signatures changes.Thereby, use general methods such as KL expansion, try to achieve mapping value to the maximum dispersion axle.To a bit be decided to be datum mark on the maximum dispersion axle, according to its distance apart, the image branch is mixed colours.Color is distributed on hue direction etc., carries out with the performance that visuognosis is best.
In addition, also can not from multi-band image, to infer subject spectrum, but directly use the value of gain balance that this situation does not need spectrum deduction portion 52 to the correction of multi-band image value.The modification method of gain balance is the known subjects of spectral reflectance such as shooting blank, and the correction that gains, so that the volume efficiency between the multi-band image value that is observed becomes the long-pending ratio that calculates of beam split according to the subject spectral reflectance of each band characteristic and known spectral reflectance.
(effect)
In this wise, in the present embodiment, except that the effect of the 1st embodiment, the scattering properties of the bio-tissue of destination layers such as epithelium Min. ground suppressed this layer in addition absorption and the state of the influence of scattering properties under emphasized, therefore, further improve visuognosis.
The 6th embodiment:
Because the 6th embodiment and the 1st embodiment are much at one, thus difference only is described, and give identical label to identical structure, omit its explanation.
(structure)
As shown in figure 19, image processing circuit 30 possesses: blood vessel structure extraction unit 111, according to blood vessel structure information from the 2nd layer of each image data extraction of synchronization memorizer 27,28,29 (the whole lower floor that the basal layer of Fig. 5 is following); With map updating portion 112, the blood vessel structure information that extracts according to blood vessel structure extraction unit 111, deviation in the compute classes (disperseing in the class), and, be updated in the gastrointestinal mucosal mapping (enum) data 100 of feature calculation with 53 storages of data supply unit according to disperseing in the class that calculates.
(effect)
Esophageal mucosa membrane injury etc. have layer structure bio-tissue situation as shown in figure 20, it is more that the top layer has blood capillary, mid-deep strata has thicker blood vessel situation such, that have the characteristic blood vessel structure.If observe the bio-tissue of this structure with wave band with different centre wavelengths, as Figure 21, shown in Figure 22, the blood vessel on top layer reproduces at short wavelength side, and the blood vessel of mid-deep strata reproduces (for example with reference to TOHKEMY 2002-95635 communique, TOHKEMY 2002-34893 communique, TOHKEMY 2002-34908 communique etc.) at long wavelength side.
Thereby in the position that blood vessel is arranged with there is not the position of blood vessel, the characteristic that becomes epithelium is identical, the different data acquisition system of characteristic of the layer beyond it, can infer in the class according to these data acquisition systems and disperse.
Therefore, as shown in figure 23, at step S71, the image (for example B image) that blood vessel structure extraction unit 111 is used long wavelength side extracts vessel position from the RGB image.The extraction of vessel position can be used general approach such as threshold process, spatial frequency filtering processing.And, at step S72, map updating portion 112 collects the pixel of a plurality of vessel positions and does not contain the pixel of blood vessel, at step S73, calculate in the class and disperse, according to disperseing to be updated in the gastrointestinal mucosal mapping (enum) data 100 of feature calculation in the class of calculating with 53 storages of data supply unit.
In addition, the image (for example B image) that blood vessel structure extraction unit 111 is used long wavelength side extracts vessel position from the RGB image, still, in order to obtain being used to extract the band image of vessel position, also can use special band filter illumination observation portion.
(effect)
In this wise, in the present embodiment, also can access the effect identical with the 5th embodiment.
In addition, in the respective embodiments described above,, become in light source side the illumination light frequency band is separated the structure that the back is shone, still, be not limited to this, also can constitute, obtain multi-band image at shooting side service band separation filter in order to obtain multi-band image.
In this invention, different embodiments does not break away from the purport and the scope of invention in the scope of broad, clearly constitute according to the present invention.The present invention is not subjected to the restriction of its specific embodiment except the qualification that is subjected to appending claims.
The industrial possibility of utilizing
As previously discussed, the imaging device that the present invention relates to can be used as the scattering with interior tissue Characteristic imaging is the device of image information.
Claims (23)
1, a kind of imaging device comprises: light supply apparatus, shine more than one at least frequency band light; Camera head will be used for from the irradiates light of above-mentioned light supply apparatus observing, and organism be observed picture be transformed to picture signal; And processor, generate the organism image according to above-mentioned picture signal; It is characterized in that,
Described processor has:
Spectrum deduction portion according to using data from each view data of synchronization memorizer with from inferring that the spectrum of using the data supply unit is inferred, infers the spectrum of each pixel;
The scattering signatures calculating part, according to from the spectrum of each pixel of above-mentioned spectrum deduction portion and from feature calculation with the feature calculation of data supply unit with required data, the scattering signatures of calculating bio-tissue; And
The coloured image generating unit based on the above-mentioned scattering signatures that is calculated by above-mentioned scattering signatures calculating part, is carried out color and is shown calculating, determines should show as coloured image the rgb value of each pixel of scattering signatures, generates the display image that offers display part,
Described processor generates the organism image as described display image, and this organism image has the described scattering signatures as image information at least.
2, imaging device as claimed in claim 1 is characterized in that,
Above-mentioned camera head is an endoscope.
3, imaging device as claimed in claim 1 is characterized in that,
In the above-mentioned frequency band light at least one exists in the frequency band of the blue light that is used as visible wavelength region.
4, imaging device as claimed in claim 1 is characterized in that,
Above-mentioned spectrum deduction portion has at least more than one matrix operations device.
5, imaging device as claimed in claim 4 is characterized in that,
Above-mentioned at least more than one the coefficient of matrix operations device, the spectrographic scope that exists that regulation will be inferred in by the spectral space of the more than one wavelength decision of discretization.
6, imaging device as claimed in claim 5 is characterized in that,
Above-mentioned at least more than one the coefficient of matrix operations device, the light propagation model that is to use the light in the performance scattering medium to propagate designs.
7, imaging device as claimed in claim 1 is characterized in that,
Above-mentioned scattering signatures calculating part has: the unit that predefined at least more than one vector in spectral space is carried out orthographic projection.
8, imaging device as claimed in claim 1 is characterized in that,
Above-mentioned coloured image generating unit generates the absorption feature of bio-tissue as above-mentioned display image.
9, imaging device as claimed in claim 1 is characterized in that,
Above-mentioned coloured image generating unit, the image beyond above-mentioned scattering signatures and the above-mentioned scattering signatures is synthetic and generate as above-mentioned display image.
10, imaging device as claimed in claim 1 is characterized in that,
Above-mentioned coloured image generating unit, the mode with while of the image beyond above-mentioned scattering signatures and the above-mentioned scattering signatures and/or switching demonstration generates image, and controls.
11, imaging device as claimed in claim 9 is characterized in that,
Image beyond the above-mentioned scattering signatures is the image information with absorption feature of bio-tissue.
12, imaging device as claimed in claim 9 is characterized in that,
Image beyond the above-mentioned scattering signatures is equivalent to the image that obtains under the white light illumination.
13, imaging device as claimed in claim 1 is characterized in that,
Above-mentioned processor has at least more than one the unit that carries out spatial frequency filtering.
14, imaging device as claimed in claim 13 is characterized in that,
The above-mentioned unit that carries out spatial frequency filtering is arranged in according at least more than one organism image and infers and the corresponding spectrographic unit in each position of image and/or zone.
15, imaging device as claimed in claim 1 is characterized in that,
With organize the top layer and above-mentioned organize beyond the top layer interior layer this two-layer with the modeled situation of above-mentioned bio-tissue under, above-mentioned scattering signatures calculating part by use with 1 to a plurality of image values or based on the observation of these image values to the spatial mapping of scattering signatures, calculate above-mentioned scattering signatures, wherein, in described scattering signatures space, so that the minimized condition of the influence that variation brought of the observation that is produced by the change of optical property of above-mentioned interior layer makes the above-mentioned scattering signatures on top layer of organizing change maximization.
16, imaging device as claimed in claim 15 is characterized in that,
Above-mentioned mapping is a Linear Mapping.
17, imaging device as claimed in claim 16 is characterized in that,
Above-mentioned Linear Mapping is calculated by the multiple discriminant analysis in the observation space; Disperse in, the class corresponding dispersion corresponding between the class of above-mentioned multiple discriminant analysis with the difference of the optical characteristics of above-mentioned interior layer with the difference of the above-mentioned scattering signatures of organizing the top layer.
18, imaging device as claimed in claim 17 is characterized in that,
The difference of the optical characteristics of above-mentioned interior layer comprises bed thickness, absorption characteristic and the scattering properties of above-mentioned interior layer.
19, imaging device as claimed in claim 17 is characterized in that,
The above-mentioned difference of organizing the scattering properties on top layer, be utilize to constitute the above-mentioned bio-tissue of organizing the top layer the nucleus diameter probability density distribution variation and by modeled.
20, imaging device as claimed in claim 19 is characterized in that,
The above-mentioned difference of organizing the scattering properties on top layer, be utilize to constitute the above-mentioned bio-tissue of organizing the top layer the nucleus diameter probability density distribution variation and by the model modelling that constitutes by Mie scattering model and light propagation model.
21, imaging device as claimed in claim 20 is characterized in that,
Above-mentioned light propagation model is a monte-Carlo model.
22, imaging device as claimed in claim 17 is characterized in that,
Disperse in the above-mentioned class in the above-mentioned multiple discriminant analysis, infer according to the bio-tissue image.
23, imaging device as claimed in claim 22 is characterized in that,
Dispersive deduction in the above-mentioned class is to carry out according to the blood-vessel image that navigates within the bio-tissue.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP361325/2002 | 2002-12-12 | ||
| JP2002361325 | 2002-12-12 | ||
| JP314206/2003 | 2003-09-05 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1725974A CN1725974A (en) | 2006-01-25 |
| CN100525701C true CN100525701C (en) | 2009-08-12 |
Family
ID=35925123
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2003801057967A Expired - Fee Related CN100525701C (en) | 2002-12-12 | 2003-11-18 | Imaging device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100525701C (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4855868B2 (en) * | 2006-08-24 | 2012-01-18 | オリンパスメディカルシステムズ株式会社 | Medical image processing device |
| JP5879745B2 (en) * | 2011-05-20 | 2016-03-08 | セイコーエプソン株式会社 | Spectral estimation parameter generation apparatus and method, and computer program |
| WO2013042395A1 (en) * | 2011-09-20 | 2013-03-28 | オリンパスメディカルシステムズ株式会社 | Image processing equipment and endoscopic system |
| CN106256313B (en) * | 2015-06-17 | 2021-02-05 | 松下知识产权经营株式会社 | Image pickup apparatus |
| WO2017106342A1 (en) * | 2015-12-18 | 2017-06-22 | Abbott Laboratories | Spectral differentiation of histological stains |
| CN105635701B (en) * | 2016-01-21 | 2017-09-12 | 东莞全球鹰医疗科技有限责任公司 | A kind of 3CCD prism cementings positioning imaging system and its imaging method |
| CN106725275A (en) * | 2017-01-13 | 2017-05-31 | 上海市第五人民医院 | A kind of device for checking rectum anal tube mucous membrane lesion tissue |
-
2003
- 2003-11-18 CN CNB2003801057967A patent/CN100525701C/en not_active Expired - Fee Related
Non-Patent Citations (1)
| Title |
|---|
| Polarized Light Sacttering Spectroscopy forQuantitative Measurement of Epithelial Cellular Structures InSitu. Vadim Backman,Rajan Gurjar,Kamran Badizadegan, etc.IEEE JOURNAL OF TOPICS IN QUANTUM ELECTRONICS,Vol.5 No.4. 1999 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1725974A (en) | 2006-01-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP1576920B1 (en) | Imaging device | |
| CN105705075B (en) | Image processing apparatus and its method of work | |
| CN103561632B (en) | Endoscope device | |
| US5078150A (en) | Spectral diagnosing apparatus with endoscope | |
| CN101584572B (en) | Fluorescent image obtainment method and apparatus, fluorescence endoscope, and excitation-light unit | |
| CN103796566B (en) | Endoscopic system and method for displaying image | |
| EP2754381B1 (en) | Endoscope system and processor device | |
| CN100493438C (en) | Endoscope image processing device | |
| CN105266750B (en) | Medical image-processing apparatus and its working method and endoscopic system | |
| CN105212885B (en) | Medical image-processing apparatus and its method of work and endoscopic system | |
| US20130245410A1 (en) | Endoscope system, processor device thereof, and exposure control method | |
| CN105310633B (en) | Medical image-processing apparatus and its method of work and endoscopic system | |
| FR2785170A1 (en) | Device to generate data and an image representative of the degree of tissue lesion associated with human skin burns that is rapid, non-invasive and objective | |
| JP2002500907A (en) | Endoscope for fluorescence imaging | |
| WO1999001749A1 (en) | Fluorescence imaging system | |
| CN102201114A (en) | Image processing device and image processing method | |
| JP2007202589A (en) | Electronic endoscope device | |
| Eichenholz et al. | Real-time megapixel multispectral bioimaging | |
| CN105101862A (en) | Image processing device and method for operating endoscope system | |
| US8105232B2 (en) | Fluorescent endoscopic device and method of creating fluorescent endoscopic image | |
| CN102525375A (en) | Endoscope device | |
| CN100525701C (en) | Imaging device | |
| KR20120097828A (en) | Endoscope apparatus capable of providing narrow band imaging and image processing method of the endoscope | |
| RU2351276C2 (en) | Instrument to detect and analyse biological tissues | |
| Nagai et al. | Detecting colorectal lesions with image-enhanced endoscopy: an updated review from clinical trials |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090812 Termination date: 20181118 |