JP2008309662A - Image processor and image processing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To certainly classify the respective image positions of specimen images, which are formed by imaging specimens dyed by a plurality of coloring matters in a short time at every principal element to which each of the respective image positions belongs. <P>SOLUTION: A coloring matter amount estimation part 152 estimates the coloring matter amounts of hemotoxylin, eosin and eosin dyed by a red corpuscle at the respective image positions of the target specimen images. A coloring matter space distribution diagram forming part 153 converts the respective image positions of the target specimen images to a coloring matter spaces having the respective coloring matter amounts as coloring matter feature axes on the basis of the respective estimated coloring matter amounts to form a coloring matter space distribution diagram. A coloring matter amount threshold value determination part 154 determines the coloring matter amount threshold values of the respective coloring matter amounts on the basis of the formed coloring matter space distribution diagram and a space dividing part 155 divides a coloring matter amount space into a plurality of partial regions by the plane determined by the determined coloring matter threshold values of the respective coloring matter amounts. A class sorting part 156 classifies the respective image positions of the target specimen images at every principal element according to the division result. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image processing apparatus and an image processing program for processing a specimen image including a specimen including at least one of a plurality of predetermined main elements, which is obtained by imaging a specimen stained with a plurality of pigments.

  For biological tissue specimens, especially pathological specimens, block specimens obtained by organ excision and specimens obtained by needle biopsy are sliced to a thickness of several microns and then magnified using a microscope to obtain various findings. Is widely practiced. In particular, transmission observation using an optical microscope is one of the most popular observation methods because the equipment is relatively inexpensive and easy to handle, and has been performed for a long time. In this case, since the sliced biological specimen hardly absorbs and scatters light and is nearly colorless and transparent, it is general to stain with a dye prior to observation.

  Various dyeing methods have been proposed, and the total number thereof reaches 100 or more. Particularly, regarding pathological specimens, hematoxylin-eosin staining using two of blue-violet hematoxylin and red eosin as pigments ( Hereinafter, “H & E staining” is used as a standard.

  Hematoxylin is a natural substance collected from plants and itself has no dyeability. However, its oxide, hematin, is a basophilic dye and binds to a negatively charged substance. Since deoxyribonucleic acid (DNA) contained in the cell nucleus is negatively charged by a phosphate group contained as a constituent element, it binds to hematin and is stained blue-violet. As described above, it is not hematoxylin that has a staining property but hematin, which is an oxide thereof. However, since it is common to use hematoxylin as a name of a pigment, the following is followed.

  Eosin is an acidophilic dye that binds to positively charged substances. Whether amino acids or proteins are charged positively or negatively is affected by the pH environment, and the tendency to be positively charged under acidic conditions becomes stronger. For this reason, acetic acid may be added to the eosin solution. Proteins contained in the cytoplasm are stained from red to light red by binding to eosin.

  In a specimen (stained specimen) after H & E staining, cell nuclei, bone tissue, etc. are stained blue-purple, and cytoplasm, connective tissue, erythrocytes, etc. are stained red, so that they can be easily visually recognized. As a result, the observer can grasp the size and positional relationship of the elements constituting the tissue such as the cell nucleus and can morphologically determine the state of the stained specimen.

  The observation of the stained specimen is performed not only by visual observation by an observer but also by displaying the stained specimen on a display screen of an external device after performing multiband imaging. FIG. 16 is a diagram illustrating an example of a multiband image obtained by imaging a stained specimen. The observer discriminates and classifies tissue elements in pathological specimens such as cell nuclei from his / her own judgment from a stained specimen or a multiband image obtained by imaging this stained specimen, and size, positional relationship, existence ratio of multiple tissue elements, etc. The pathological diagnosis is performed by determining the degree of disease progression.

  In addition, in order to accurately perform such pathological diagnosis and reduce the burden on the observer, attempts have been made to automatically classify main elements such as tissue elements included in a stained specimen (for example, patents). Reference 1). According to Patent Document 1, first, a relative amount of a dye fixed to each point of a specimen of a pathological specimen corresponding to each pixel of the multiband image is estimated from a multiband image of a pathological specimen stained with H & E. . Then, the estimated pigment amounts of hematoxylin and eosin are plotted on the pigment amount plane to classify each pixel of the multiband image. At this time, a sample whose content is clear is used as teacher information, and the boundaries of clusters in the pigment amount plane are determined in advance. The distribution position on the dye amount plane of the dye amount estimated for each pixel of the multiband image changes depending on which main element the image position belongs to, and the dye at the image position belonging to the same main element Since the distribution positions of the amounts are relatively close, each pixel of the multiband image can be classified for each main element using the estimated pigment amount.

JP 2005-331394 A

  However, when each image position of the multiband image is classified for each main element using the technique of Patent Document 1, since the teacher information is used in the classification process, it is possible to use another sample such as a standard stained sample. It was necessary to obtain sample points representing each main element in advance. Moreover, when the stained state of the stained specimen and the specimen from which the sample points have been acquired are greatly different, there is a problem that the teacher information is not adapted and accurate classification cannot be performed.

  On the other hand, it is possible to classify each pixel of a multiband image without using teacher information, but iteratively repeats the classification process while correcting the classification result based on a predetermined determination criterion to search for a better classification Since it is a law, the time required for the classification process increases.

  The present invention has been made in view of the above-described conventional problems, and each image position of a sample image obtained by imaging a sample stained with a plurality of pigments, for each main element to which each image position belongs, An object of the present invention is to provide an image processing apparatus and an image processing program that can be reliably classified in a short time.

  In order to solve the above-described problems and achieve the object, an image processing apparatus according to the present invention is a specimen including at least one of a plurality of predetermined main elements, and is stained with a plurality of dyes In an image processing apparatus that processes a sample image obtained by imaging a sample, a dye amount estimation unit that estimates a first dye amount, a second dye amount, and a third dye amount at each image position of the sample image, and the dye Based on each dye amount at each image position estimated by the amount estimating means, a conversion means for converting each image position into a dye amount space, and a dye amount threshold value for each dye amount in the dye amount space are determined. A pigment amount threshold value determining means, a space dividing means for dividing the dye amount space into a plurality of partial regions by a plane determined by each dye amount threshold value determined by the dye amount threshold value determining means, Accordance division result by dividing means, characterized in that it comprises a classification means for classifying each image position for each of the key elements.

  Further, an image processing program according to the present invention is a computer that processes a sample image obtained by imaging a sample that includes at least one of a plurality of predetermined main elements and is stained with a plurality of pigments. , A dye amount estimating step for estimating a first dye amount, a second dye amount and a third dye amount at each image position of the sample image, and each dye at each image position estimated by the dye amount estimating step Based on the amount, determined by a conversion step for converting each image position into a dye amount space, a dye amount threshold determination step for determining a dye amount threshold for each dye amount in the dye amount space, and a dye amount threshold determination step A space dividing step of dividing the dye amount space into a plurality of partial regions according to a plane determined by each of the dye amount thresholds, and the space dividing step Accordance split result, characterized in that to perform the classification step, classifying the respective image position for each of the key elements.

  According to the present invention, each image position of the sample image to be classified can be classified for each main element according to the distribution in the pigment amount space, and classification can be performed without being affected by the staining state. In addition, the classification process can be realized in a shorter time than the classification process by the conventional clustering.

  DESCRIPTION OF EMBODIMENTS Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings. In the present embodiment, the tissue position of each cell nucleus, cytoplasm, and red blood cell, and an H & E-stained pathological sample including at least one of the four main elements of the background are imaged, and the image position of the captured multiband image A case will be described in which the classification is performed by the category corresponding to each main element. The number and type of main elements are examples and can be set as appropriate. For example, in addition to cell nuclei, cytoplasm, and erythrocytes, tissue elements such as bone tissue stained with hematoxylin and connective tissue stained with eosin may be classified.

  First, the configuration of the image processing apparatus according to the present embodiment will be described. FIG. 1 is a block diagram illustrating a functional configuration of the image processing apparatus. In the present embodiment, the image processing apparatus 10 includes an image acquisition unit 110, an operation unit 120, a display unit 130, a storage unit 140, an image processing unit 150, and a control unit 160 that controls each unit of the apparatus. .

  The image acquisition unit 110 captures an H & E-stained pathological specimen to be classified (hereinafter referred to as “target stained specimen”) and acquires a 6-band multiband image. FIG. 2 is a diagram illustrating a configuration of the image acquisition unit 110. As shown in FIG. 2, the image acquisition unit 110 includes an RGB camera 111 including an imaging device such as a CCD, a sample holding unit 113 on which the target stained sample S is placed, and a target stained sample S on the sample holding unit 113. Illuminating unit 115 for transmitting and illuminating, an optical system 117 for condensing and imaging the transmitted light from the target stained specimen S, a filter unit 119 for limiting the wavelength band of the imaged light to a predetermined range, and the like.

  The RGB camera 111 is widely used in digital cameras and the like, and has RGB color filters arranged in a mosaic pattern on a monochrome image sensor. The RGB camera 111 is installed so that the center of the image to be captured is positioned on the optical axis of the illumination light. FIG. 3 is a diagram schematically illustrating an example of the arrangement of color filters. In this case, each pixel can image only one of R, G, and B components, but by using neighboring pixel values, the insufficient R, G, and B components are interpolated. This technique is disclosed in, for example, Japanese Patent No. 3510037. If a 3CCD type camera is used, R, G and B components in each pixel can be acquired from the beginning. In this embodiment, any imaging method may be used. In the following, it is assumed that R, G, and B components can be acquired in each pixel of an image captured by the RGB camera 111.

  The filter unit 119 includes two optical filters 1191 a and 1191 b each having different spectral transmittance characteristics, and these are held by a rotary optical filter switching unit 1193. FIG. 4A is a diagram illustrating a spectral transmittance characteristic of one optical filter 1191a, and FIG. 4B is a diagram illustrating a spectral transmittance characteristic of the other optical filter 1191b. For example, first, first imaging is performed using the optical filter 1191a. Next, the optical filter to be used is switched to the optical filter 1191b by the rotation of the optical filter switching unit 1193, and the second imaging is performed using the optical filter 1191b. By the first imaging and the second imaging, a 3-band image is obtained, and a 6-band multiband image is obtained by combining both results. The number of optical filters is not limited to two, and two or more optical filters can be used. The acquired multiband image is output to the control unit 160 and held in the storage unit 140 as a target specimen image.

  In this image acquisition unit 110, the illumination light emitted from the illumination unit 115 passes through the target stained sample S placed on the sample holding unit 113. Then, the transmitted light that has passed through the target stained specimen S passes through the optical system 117 and the optical filters 1191a and 1191b, and then forms an image on the image sensor of the RGB camera 111. The filter unit 119 including the optical filters 1191 a and 1191 b may be installed at any position on the optical path from the illumination unit 115 to the RGB camera 111. FIG. 5 shows an example of the spectral sensitivity of each of the R, G, and B bands when the illumination light from the illumination unit 115 is imaged by the RGB camera 111 via the optical system 117.

  The operation unit 120 is realized by, for example, a keyboard, a mouse, a touch panel, various switches, and the like, and outputs an operation signal corresponding to the operation input to the control unit 160.

  The display unit 130 is realized by a display device such as an LCD or an ELD, and displays various screens based on a display signal input from the control unit 160.

  The storage unit 140 is realized by various IC memories such as ROM and RAM such as flash memory that can be updated and stored, information storage media such as a built-in or data communication terminal, a hard disk connected by a data communication terminal, a CD-ROM, and a reading device thereof. A program relating to the operation of the image processing apparatus 10, a program for realizing various functions provided in the image processing apparatus 10, data relating to execution of these programs, and the like are stored. Also stored is an image processing program 141 for classifying each image position of the target specimen image obtained by imaging the stained specimen for each main element to which each image position belongs.

  The image processing unit 150 is realized by hardware such as a CPU. The image processing unit 150 includes a spectrum estimation unit 151, a pigment amount estimation unit 152, a pigment amount space distribution diagram generation unit 153, a pigment amount threshold value determination unit 154, a space division unit 155, a class classification unit 156, A label image generation unit 157. The spectrum estimation unit 151 estimates the spectral characteristics of each image position of the target sample image based on the pixel value of the target sample image. The pigment amount estimation unit 152 stains the pigment amount and cytoplasm of hematoxylin that stains the cell nucleus at each image position of the target specimen image based on the reference spectral characteristics of the hematoxylin and eosin pigments used for staining the target staining specimen. The amount of eosin dye and the amount of eosin dyed erythrocytes are estimated. The dye amount space distribution map generation unit 153 converts each image position of the target specimen image into a dye amount space having the dye amount of each dye as a dye amount characteristic axis based on each estimated dye amount. Generate a spatial map. The dye amount threshold value determination unit 154 determines a dye amount threshold value for each dye amount based on the generated dye amount space distribution diagram. The space dividing unit 155 divides the dye amount space into a plurality of partial regions by a plane determined by the determined dye amount threshold value of each dye amount. The class classification unit 156 classifies each image position of the target sample image for each main element according to the division result. The label image generation unit 157 generates a display image (main element label image) from the target specimen image according to the classification result.

  The control unit 160 is realized by hardware such as a CPU. The control unit 160 is connected to each unit constituting the image processing apparatus 10 based on the operation signal input from the operation unit 120, the image data input from the image acquisition unit 110, the program and data stored in the storage unit 140, and the like. Instruction, data transfer, and the like, and overall operation of the image processing apparatus 10 is controlled. In addition, the control unit 160 controls the operation of the image acquisition unit 110 to acquire the target sample image, and the main element label image generated by the label image generation unit 157 is displayed on the display unit 130. A label image display control unit 163 that performs display control.

  FIG. 6 is a flowchart showing a processing procedure in the image processing apparatus 10. Note that the processing described here is realized by the operation of each unit of the image processing apparatus 10 in accordance with the image processing program 141 stored in the storage unit 14.

  That is, first, the sample image acquisition control unit 161 controls the operation of the image acquisition unit 110 to perform multiband imaging of the target stained sample and acquire the target sample image (step S11).

  Subsequently, the spectrum estimation unit 151 estimates the spectral characteristics of the target stained sample based on the pixel value of the target sample image acquired in step S11 (step S13). Specifically, the spectrum estimation unit 151 estimates the spectral transmittance characteristic at each image position (each sample point of the target stained specimen corresponding to each pixel) based on the pixel value of each pixel. As an estimation method, for example, Wiener estimation is used, and the spectral transmittance at each image position is estimated from the target sample image. For example, “Color Correction of Pathological Images Based on Dye Amount Quantification” (OPTICAL REVIEW, Vol. 12, No. 4, 2005, p.293-300) ”can be used.

Ｆは光学フィルタ１１９１ａ，１１９１ｂの分光透過率、ＳはＲＧＢカメラ１１１の分光感度特性、Ｅは照明の分光放射特性、Ｒ_SSは標本の分光透過率の自己相関行列、Ｒ_NNは撮像に使用するＲＧＢカメラ１１１のノイズの自己相関行列をそれぞれ表す。ただし式（１）は、波長方向に離散化し、行列を用いて複数のバンドに関する式を集約して表したものであるため、バンドを表す変数ｂが陽に記述されていない。また、波長λに関する積分は行列の積に置き換えられている。例えば、波長方向のサンプル点数をＤ、バンド数をＢとすれば（ここではＢ＝６）、Ｆは、Ｂ行Ｄ列の行列である。Ｓは、Ｄ行Ｄ列の対角行列であり、対角要素が、波長λにおけるＲＧＢカメラ１１１の分光感度ｓ（λ）に対応している。同様に、Ｅは、Ｄ行Ｄ列の対角行列であり、対角要素が、波長λにおける照明の分光放射特性ｅ（λ）に対応している。Ｒ_SSはＤ行Ｄ列の行列であり、Ｒ_NNはＢ行Ｂ列の行列である。また、Ｇ（ｘ）はＢ行１列の行列、Ｔ＾（ｘ）はＤ行１列の行列となる。 Here, a method for estimating spectral transmittance using Wiener estimation will be described. In the Wiener estimation, an estimated value T ^ (x) of the spectral transmittance at this image position is estimated from the matrix representation G (x) of the pixel value at the point x of the target sample image according to the following equation (1). T ^ indicates that a hat (^) representing an estimated value is attached on T.

F is the spectral transmittance of the optical filters 1191a and 1191b, S is the spectral sensitivity characteristic of the RGB camera 111, E is the spectral radiation characteristic of illumination, R _SS is the autocorrelation matrix of the spectral transmittance of the specimen, and R _NN is used for imaging The noise autocorrelation matrix of the RGB camera 111 is shown respectively. However, since the expression (1) is discretized in the wavelength direction and expressions related to a plurality of bands are aggregated using a matrix, the variable b representing the band is not explicitly described. In addition, the integration with respect to the wavelength λ is replaced with a matrix product. For example, if the number of sample points in the wavelength direction is D and the number of bands is B (B = 6 here), F is a matrix of B rows and D columns. S is a diagonal matrix of D rows and D columns, and the diagonal elements correspond to the spectral sensitivity s (λ) of the RGB camera 111 at the wavelength λ. Similarly, E is a diagonal matrix of D rows and D columns, and the diagonal elements correspond to the spectral radiation characteristic e (λ) of illumination at the wavelength λ. R _SS is a matrix of D rows and D columns, and R _NN is a matrix of B rows and B columns. G (x) is a matrix of B rows and 1 column, and T ^ (x) is a matrix of D rows and 1 column.

The spectral transmittance F of the optical filters 1191a and 1191b, the spectral sensitivity characteristic S of the RGB camera 111, and the spectral radiation characteristic E of illumination are measured in advance using a spectrometer after selecting the equipment to be used. Here, the spectral transmittance of the optical system 117 is approximated to 1.0. However, when the deviation from the approximate value 1.0 cannot be allowed, the spectral transmittance of the optical system 117 is also measured in advance. The spectral radiation characteristic E of illumination may be multiplied. Further, the autocorrelation matrix R _SS of the spectral transmittance of the sample and the autocorrelation matrix R _NN of the imaging noise of the RGB camera 111 are also measured in advance. The R _SS is obtained by preparing a typical specimen stained with H & E, and measuring the spectral transmittance of a plurality of points with a spectrometer to obtain an autocorrelation matrix. In order to improve the statistical accuracy, it is better to measure about 100 points without deviation in the sample. On the other hand, R _NN acquires a multiband image by the image acquisition unit 110 without a sample, obtains a variance of pixel values for each band of the obtained 6-band multiband image, and uses this as a diagonal component. It is obtained by generating a matrix. The obtained spectral transmittance estimated value (hereinafter referred to as “spectral transmittance data”) T ^ (x) is stored in the storage unit 140.

  Subsequently, the dye amount estimation unit 152 estimates the dye amount of the target stained specimen based on the spectral transmittance data estimated in step S13 (step S15). Specifically, the dye amount estimation unit 152 is based on the spectral transmittance data at each image position of the target sample image, and the dye fixed at each image position (each sample point of the target stained sample corresponding to each pixel). Estimate the amount. There are three types of pigments to be estimated: hematoxylin that stains the cell nucleus as the first pigment amount, eosin that stains the cytoplasm as the second pigment amount, and eosin that stains the red blood cells as the third pigment amount. is there. Here, hematoxylin is abbreviated as dye H, eosin dyed cytoplasm is abbreviated as dye E, and eosin dyed erythrocytes is abbreviated as dye R. Strictly speaking, erythrocytes have their own unique color even without staining, and after H & E staining, the color of erythrocytes and the color of eosin changed in the staining process are superimposed. Observed. For this reason, the combination of both is called dye R.

ここでｋ_H（λ），ｋ_E（λ），ｋ_R（λ）は、それぞれ色素Ｈ、色素Ｅ、色素Ｒに対応したｋ（λ）を表す。またｄ_H，ｄ_E，ｄ_Rは、マルチバンド画像の各画像位置における色素Ｈ、色素Ｅ、色素Ｒの仮想的な厚さを表す。本来色素は、標本中に分散して存在するため、厚さという概念は正確ではないが、標本が単一の色素で染色されていると仮定した場合と比較して、どの程度の量の色素が存在しているかを表す相対的な色素量の指標となる。すなわち、ｄ_H，ｄ_E，ｄ_Rはそれぞれ色素Ｈ、色素Ｅ、色素Ｒの色素量を表しているといえる。なお、ｋ_H（λ），ｋ_E（λ），ｋ_R（λ）は、単一の色素で染色した標本を予め用意し、その分光透過率を分光計で測定することによって、Lambert-Beer則から容易に求めることができる。 When the target stained specimen that has been subjected to H & E staining is stained with three types of dyes, dye H, dye E, and dye R, the following equation (2) is established at each wavelength λ according to the Lambert-Beer rule.

Here, k _H (λ), k _E (λ), and k _R (λ) represent k (λ) corresponding to the dye H, the dye E, and the dye R, respectively. D _H , d _E , and d _R represent the virtual thicknesses of the dye H, the dye E, and the dye R at each image position of the multiband image. Since the dye is inherently dispersed in the specimen, the concept of thickness is not accurate, but how much dye is compared to the assumption that the specimen is stained with a single dye. It is an indicator of the relative amount of pigment that indicates whether or not the selenium is present. That is, it can be said that d _H , d _E , and d _R represent the dye amounts of the dye H, the dye E, and the dye R, respectively. K _H (λ), k _E (λ), and k _R (λ) are prepared in advance by preparing a specimen stained with a single dye, and measuring the spectral transmittance with a spectrometer. It can be easily obtained from the law.

式（３）において未知変数はｄ_H，ｄ_E，ｄ_Rの３つであるから、少なくとも３つの異なる波長λに対して式（３）を連立させれば、これらを解くことができる。より精度を高めるために、４つ以上の異なる波長λに対して式（３）を連立させ、重回帰分析を行ってもよい。 When the element corresponding to the wavelength λ of the estimated spectral transmittance data T ^ (x) is t ^ (x, λ) and the logarithm of both sides of the equation (5) is taken, the following equation (3) is obtained.

Since there are three unknown variables d _H , d _E , and d _R in equation (3), these can be solved if equation (3) is combined for at least three different wavelengths λ. In order to further improve the accuracy, the multiple regression analysis may be performed by simultaneous equations (3) for four or more different wavelengths λ.

As an example, when considering the case where the equation (3) is simultaneously provided for the three wavelengths λ ₁ , λ ₂ , and λ ₃ , the matrix can be expressed as the following equation (4).

このようにして推定された各画像位置における色素Ｈ、色素Ｅ、色素Ｒの各色素量ｄ_H，ｄ_E，ｄ_Rは、記憶部１４０に格納される。 Therefore, d _H , d _E and d _R can be estimated according to the following equation (5).

The dye amounts d _H , d _E , and d _R of the dye H, the dye E, and the dye R at each image position estimated in this way are stored in the storage unit 140.

Subsequently, based on the dye amounts d _H , d _E , and d _R estimated in step S15, the dye amount space distribution map generation unit 153 converts the image positions of the target specimen image to the dye H, the dye E, and the dye R. A dye amount space distribution diagram (hereinafter referred to as “HER”) in which each image position is plotted on this HER dye amount space after being converted into a dye amount space (hereinafter referred to as “HER dye amount space”) with each dye amount as an axis. (Referred to as “pigment amount space distribution diagram”) (step S17). FIG. 7 is a diagram showing an example of a HER dye amount space distribution diagram. In the present embodiment, it is assumed that the HER dye amount space is configured so that the dye H, the dye E, and the dye R are in a positive range (H ≧ 0, E ≧ 0, R ≧ 0). Data relating to the generated HER dye amount spatial distribution map is stored in the storage unit 140.

Subsequently, the dye amount threshold value determination unit 154 determines the dye amount threshold values th _H , th _E , and th for each of the dye amounts of the dye H, the dye E, and the dye R based on the image positions in the converted HER dye amount space. _R is determined (step S19). The dye amount thresholds th _H , th _E , and th _R are thresholds for dividing the distribution for each dye at each image position. In the present embodiment, for example, Otsu's threshold method “Mathematical research on feature extraction in pattern recognition (Noriyuki Otsu, Research Institute of Electronics Technology Research No. 818, 1981)” is used. R threshold values T _H , T _E and T _R are obtained. In Otsu's threshold method, when the distribution of each pigment amount is divided into two sets (classes), the variation within each class is the smallest (minimum intra-class variance) and the classes are farthest apart (maximum inter-class variance). ) To determine the threshold value. Then, according to the following equation (6), the obtained threshold values T _H , T _E , T _R are multiplied by coefficients a, b, c, and the values of the respective threshold values T _H , T _E , T _R are adjusted to give a dye. The quantity threshold values th _H , th _E , and th _R are obtained. The values of the coefficients a, b, and c are each set to “0.5”, for example.
th _H = aT _H
th _E = bT _E
th _R = cT _R (6)
The dye amount threshold values th _H , th _E , and th _R determined in this manner are stored in the storage unit 140.

Note that the method for determining the dye amount thresholds t _H , t _E , t _R is not limited to this, and the dye amount statistics (for example, the average value, median value, most frequent value, etc.) of each dye are used. It may be determined on the basis of a differential histogram method.

Subsequently, the space dividing unit 155 performs space division processing, and divides the HER dye amount space into partial regions by a plane determined by the dye amount thresholds th _H , th _E , and th _{R of} the respective dye amounts determined in step S19 (step S19). S21). 8A to 8 and FIG. 9 are diagrams for explaining the space division processing in step S21.

First, the space dividing unit 155 includes three planes, a plane α that is a first plane, a plane β that is a second plane, and a plane γ that is a third plane, according to the dye amount thresholds th _H , th _E , and th _R. The planes α, β, and γ are determined. FIGS. 8-1 to 3 are diagrams showing three planes α, β, and γ determined based on the dye amount thresholds th _H , th _E , and th _R , respectively. In the present embodiment, as shown in FIG. 8A, the plane α is defined as a plane that passes through the dye amount threshold th _H of the dye amount of the dye _H and is perpendicular to the dye amount characteristic axis of the dye H. Further, as shown in FIG. 8B, the plane β passes through the dye amount threshold th _E of the dye amount of the dye _E and is parallel to the dye amount characteristic axis of the dye R, and the amount of the dye H of the dye H It is defined as a plane determined by the dye amount threshold th _H. The plane β is not limited to the plane determined by the dye amount threshold th H of the dye amount of the dye _H , and passes through the dye amount threshold th _E of the dye amount of the dye _E and is parallel to the dye amount characteristic axis of the dye R. The plane can also be set as appropriate. 8C, the plane γ is defined as a plane that passes through the dye amount threshold th _R of the dye amount of the dye _R and is perpendicular to the dye amount characteristic axis of the dye R. Each plane α, β, γ is expressed by the following equation (7).
Plane α: H = th _H
Plane β: th _E H + th _H E−th _H th _E = 0
Plane γ: R = th _R (7)

Next, the space dividing unit 155 divides the HER dye amount space into the partial regions P _i (i = 1) of the number of partial regions m by the three planes α, β, γ determined by the respective dye amount threshold values th _H , th _E , th _R. , 2,..., M; where m ≦ 8). In the present embodiment, the HER dye amount space is composed of the dye H, the dye E, and the dye R in a positive range (H ≧ 0, E ≧ 0, R ≧ 0). 6 FIG. 9 is a diagram showing six reference partial areas that can be divided by planes α, β, and γ. As shown in FIG. 9, the HER dye amount space has a maximum of six references by each plane α, β, and γ. It can be _divided into partial areas P _{s1 to} P _s6 .

Corresponding main element types are set in advance in the six reference partial areas P _{s1 to} P _s6 . FIG. 10 is a diagram illustrating an example of a correspondence relationship between the reference partial regions P _{s1 to} P _s6 and main elements. According to the example shown in FIG. 10, the reference partial region P _s1 has a cell nucleus, the reference partial region P _s2 has a red blood cell, the reference partial region P _s3 has a cytoplasm, the reference partial region P _s4 has a background, Red blood cells are assigned to the partial areas P _s6 , respectively. In addition, a reject area indicating that no corresponding main element exists is assigned to the reference partial area P _s5 . This is because, from the characteristics of H & E staining, when the image position belongs to the reference partial region P _s5 , it is considered that the amount of the dye includes an estimation error, and thereby, the dye at the image position belonging to the reference partial region P _s5. The quantity can be treated as data with error.

Then, the space dividing unit 155 determines the plane range of each plane α, β, γ. Thereby, the number m of partial regions is determined, and the HER dye amount space is divided into m partial regions P _i (i = 1, 2,..., M; where m ≦ 6). The plane ranges of the planes α, β, and γ are determined according to a user operation, for example. Data relating to the determined plane range of each plane α, β, γ is stored in the storage unit 140. From the correspondence between the reference partial areas P _{s1 to} P _s6 and the main elements shown in FIG. 10 and the correspondence relation between the reference partial areas P _{s1 to} P _s6 and the reject area, the user can determine each plane α, β, γ. Specify the plane area. In this case, the control unit 160, for example, the correspondence between the reference partial regions P _{s1 to} P _s6 shown in FIG. 10 and the main elements, the HER dye amount space distribution diagram generated in step S17, and each determined in step S19. Control is performed to display the dye amount thresholds th _H , th _E , and th _R on the display unit 130, and control is performed to display a notification of a request for specifying the plane range of each plane α, β, γ. FIG. 11 is a diagram illustrating an example of a notification screen for a plane range designation request. On the notification screen W10, input boxes I11 to I16 for receiving inputs of upper and lower limits of the plane ranges of the planes α, β, and γ are arranged. Here, the plane range of the planes α and β is specified by the value of the dye R, and the plane range of the plane γ is specified by the value of the dye H. The user designates a desired plane range of each plane α, β, γ by inputting the upper limit / lower limit of the plane range to be changed via the operation unit 120 into the corresponding input boxes I11 to I16. Further, the designation operation is canceled by pressing the cancel button B11, or the designation operation is confirmed by pressing the decision button B13.

For example, when “th _R ” is input to the input box I13 and the upper limit value of the plane range of the plane β is designated, the plane range of the plane β is determined as R <th _R. Since H ≧ 0, E ≧ 0, and R ≧ 0, the number of partial regions m is determined to be “5”, and as a result, the HER dye amount space has five partial regions P _i (i = 1, 2, 3). , 4, 5). FIG. 12 is a diagram showing a configuration plane and a configuration range of each of the five partial regions P _i (i = 1, 2, 3, 4, 5) in this case, and corresponding main elements. As shown in FIG. 12, the constituent planes of the partial region P ₂ are a plane α and a plane β. This is because the plane range of the plane β is determined as R <th _{R. In} this case, the reference partial areas P _s2 and P _s6 shown in FIG. 9 are combined to form one partial area P _2. It will be divided.

Further, the correspondence between each actually divided partial area P _i and the main element or reject area is determined based on the correspondence between the reference partial areas P _{s1 to} P _s6 and the main element or reject area shown in FIG. However, these correspondence relationships may be changed in accordance with a user operation. For example, in the example of FIG. 12, “red blood cells” are set as the main element corresponding to the partial region P ₂ in which the reference partial regions P _s2 and P _s6 are combined. This is because, as shown in FIG. 10, the main elements assigned to the reference partial regions P _s2 and P _s6 are “red blood cells”, but different types of planes of the planes α, β, and γ determined. When the reference partial areas to which the main elements are assigned are joined, for example, a selection operation for one of the main elements is received and the corresponding main element is set. For the reject area, for example, a user operation such as setting a partial area to which the main element is assigned as the reject area or setting a main element to the partial area to which the reject area is assigned is accepted. For example, in the example of FIG. 12, “cell nucleus” is set in the partial region P ₅ corresponding to the reference partial region P _S5 (see FIG. 10) to which the reject region is assigned.

Subsequently, the class classification unit 156 classifies each image position of the target specimen image for each main element (step S23). Specifically, the class classification unit 156 determines each partial region P _i according to the correspondence between each determined partial region P _i (i = 1, 2,..., M; where m ≦ 6) and main elements. Each image position belonging to _i is assigned to a corresponding main element, and a HER dye amount space classification diagram is generated. Data relating to the generated HER dye amount space classification diagram is stored in the storage unit 140. FIG. 13 is a diagram showing a HER dye amount space classification diagram. As shown in FIG. 13, each image position of the target specimen image converted into the HER dye amount space has five partial regions P _{1 to} P ₅ divided by the determined plane ranges of the respective planes α, β, and γ. Are classified for each main element depending on which partial region belongs to. For example, each pixel position belonging to the partial region P ₁ is classified as “cell nucleus” based on the correspondence between the partial region P _i and the main elements shown in FIG.

  Subsequently, the label image generation unit 157 generates a main element label image from the target specimen image based on the classification result of step S23 (step S25). Specifically, the label image generation unit 157 first assigns label information to each pixel position according to the classified main elements. Then, the label image generating unit 157 obtains a main element label image by changing the luminance value of the corresponding target specimen image according to the label information given to each image position. That is, based on the label information given to each image position of the target specimen image, the luminance value is changed for each main element to which the same label information is attached, and each image position of the target specimen image is the same for each label information A main element label image expressed in luminance is generated.

  Then, the label image display control unit 163 performs control to display the main element label image generated in step S25 on the display unit 130 (step S27). FIG. 14 is a diagram showing an example of the main element label image, and shows an example in which the luminance value is changed in the order of cell nucleus, cytoplasm, red blood cell, and background. Although the main element label image is obtained by changing the luminance value based on the classification result, the main element label image may be obtained by changing the display color according to the label information of each image position. Further, the main element label image generated here may be displayed superimposed on the target specimen image. Alternatively, in accordance with the label information of each image position, a boundary line indicating the region of the main element may be displayed on the target sample image, and control for identifying and displaying the region of the main element in the target sample image may be performed.

According to the present embodiment described above, each image position of a specimen image including predetermined main elements and subjected to H & E staining is determined based on the amount of each of the dye H, the dye E, and the dye R at each image position. While converting to the HER dye amount space, the dye amount thresholds th _H , th _R , and th _E are determined for each dye amount, and three planes α, β determined by the determined dye amount thresholds th _H , th _R , and th _E are determined. , Γ can divide the pigment amount space into partial regions. According to this, the dye amount thresholds th _H , th _R , th _E are determined according to the distribution in the dye amount space of each image position of the sample image to be classified, and the determined dye amount thresholds th _H , th _{R are determined.} , Th _E , the pigment amount space can be divided into partial regions. Then, based on the correspondence between each partial region and the main element, each image position can be classified for each main element. Therefore, each image position of the target specimen image can be surely classified for each main element without being affected by the staining state. Further, the processing time can be shortened as compared with the classification processing by the conventional clustering.

  In addition, it is good also as a structure provided with the user interface for designating the reference partial area | region used for a class classification | category from the predetermined reference | standard partial area | regions. In such a configuration, if the reference partial region used for classification is designated, the correspondence between the designated reference partial region and the main element is validated. Then, the image positions belonging to the reference partial area to be used are classified into main elements corresponding to the reference partial area. For example, when the type and number of main elements included in the target stained specimen are known, the user designates the reference partial area to which the main elements are assigned as the reference partial area to be used. According to this, it is possible to classify each image position of the target specimen image for each main element by dividing the HER dye amount space into partial regions according to the type and number of main elements included in the target stained specimen. it can.

  Further, a configuration having a user interface for directly specifying the connection between the reference partial areas may be adopted. If the coupling between the reference partial areas is designated, the HER dye amount space is divided into partial areas by determining the plane ranges of the planes α, β, γ according to the designated reference partial areas.

Further, the threshold values T _H , T _E , T _R and the parameters a, b, c for obtaining the respective dye amount threshold values t _H , t _E , t _R can be adjusted according to the user operation. Also good. For example, the control unit 160 performs control to display the HER dye amount space distribution diagram and the like illustrated in FIG. 7 on the display unit 130, and the threshold values T _H , T _E , T _R and the coefficients a, b, c. Control to display notification of input request. FIG. 15 is a diagram illustrating an example of a notification screen for requesting input of each parameter. In the notification screen W20, input boxes I21 to I25 for receiving input operations for each value are arranged. The user inputs a desired value of each parameter into the corresponding input boxes I21 to I25 via the operation unit 120. In this case, the dye amount threshold value determination unit 154 determines the dye amount threshold values t _H , t _E , and t _R according to the parameter values input in this way.

In addition, a HER dye amount space diagram generated by the dye amount space distribution diagram generation unit 153 illustrated in FIG. 7 and a HER dye amount space classification diagram generated by the class classification unit 156 illustrated in FIG. It may be displayed on the display unit 130. Further, after the class classification unit 156 generates the HER dye amount space classification diagram, the thresholds T _H , T _E , T _R and the coefficients a, b, c for obtaining the dye amount thresholds th _H , th _E , th _R are obtained. It is good also as a structure which can adjust the value of each parameter of according to user operation. In this case, the dye amount thresholds th _H , th _E and th _R are recalculated according to the adjusted values of the parameters, and the HER dye amounts reflecting the obtained dye amount thresholds th _H , th _E and th _R are obtained. It is also possible to perform control to generate the space classification map again and display it on the display unit 130. At this time, by installing a high-speed processor in the image processing apparatus 10, the adjustment result of each parameter can be immediately reflected in the HER dye amount space classification diagram, so that the dye amount threshold th _H is confirmed while checking the class classification result. , Th _E , th _R can be appropriately adjusted to change the planes α, β, γ, and the operability can be greatly improved.

  In the above embodiment, the case where the pathological specimen stained with H & E is observed through transmission has been described. However, the present invention can also be applied to a biological specimen stained using another staining method. The present invention can be similarly applied not only to observation of transmitted light but also to observation of reflected light, fluorescence, and light emission.

It is a block diagram which shows the function structure of an image processing apparatus. It is a figure which shows the structure of an image acquisition part. It is a figure which shows typically the example of an arrangement | sequence of a color filter. It is a figure which shows the spectral transmittance characteristic of one optical filter. It is a figure which shows the spectral transmittance characteristic of another optical filter. It is a figure which shows the example of the spectral sensitivity of each band of R, G, B. It is a flowchart which shows the procedure of the process in an image processing apparatus. It is a figure which shows an example of a HER pigment amount space distribution map. It is a figure for demonstrating a space division process. It is a figure for demonstrating a space division process. It is a figure for demonstrating a space division process. It is a figure for demonstrating a space division process. It is a figure which shows an example of the correspondence of a reference | standard partial area | region and a main element. It is a figure which shows an example of the notification screen of the designation | designated request | requirement of a planar range. It is a figure which shows the structure plane of each partial area | region, its structure range, and a corresponding main element. It is a figure which shows a HER pigment amount space classification diagram. It is a figure which shows an example of a main element label image. It is a figure which shows an example of the notification screen of the input request of the parameter for calculating | requiring each pigment amount threshold value. It is a figure which shows an example of the multiband image which imaged the dyeing | staining specimen.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Image processing apparatus 110 Image acquisition part 120 Operation part 130 Display part 140 Storage part 141 Image processing program 150 Image processing part 151 Spectrum estimation part 152 Dye quantity estimation part 153 Dye quantity space distribution map generation part 154 Dye quantity threshold value determination part 155 Space Division unit 156 Class classification unit 157 Label image generation unit 160 Control unit 161 Sample image acquisition control unit 163 Label image display control unit

Claims (13)

In an image processing apparatus that processes a specimen image obtained by imaging a specimen that includes at least one of a plurality of predetermined main elements and is stained with a plurality of pigments,
A dye amount estimating means for estimating a first dye amount, a second dye amount, and a third dye amount at each image position of the sample image;
Conversion means for converting each image position into a dye amount space based on each dye amount at each image position estimated by the dye amount estimating means;
A dye amount threshold value determining means for determining a dye amount threshold value of each dye amount in the dye amount space;
Space dividing means for dividing the dye amount space into a plurality of partial regions by a plane determined by each dye amount threshold value determined by the dye amount threshold value determining means;
Class classification means for classifying each image position for each main element according to the division result by the space division means;
An image processing apparatus comprising:   The space dividing unit includes a first plane that passes through the dye amount threshold of the first dye amount and is perpendicular to the first dye amount characteristic axis, and the dye amount of the second dye amount. A second plane passing through the threshold and parallel to the third dye quantity feature axis; and a third plane passing through the dye quantity threshold of the third dye quantity and perpendicular to the third dye quantity feature axis; The image processing apparatus according to claim 1, wherein the pigment amount space is divided into the plurality of partial regions.   The image processing apparatus according to claim 2, wherein the second plane is determined by a dye amount threshold value of the first dye amount.   The image processing apparatus according to any one of claims 1 to 3, wherein the space dividing unit divides the pigment amount space into partial regions having a plurality of predetermined main elements. Correspondences between the partial regions of the pigment amount space divided according to the pigment amount threshold and the plurality of main elements are preset,
The said class classification | category means classify | categorizes each said image position for every said main element according to the main element corresponding to the partial area to which each said image position belongs, The Claim 1 characterized by the above-mentioned. Image processing apparatus. The conversion means includes distribution map generation means for generating a dye amount space distribution diagram in which the image positions are plotted based on the dye amounts at the image positions.
The image processing according to any one of claims 1 to 5, further comprising a distribution diagram display control unit that performs control to display a pigment amount space distribution diagram generated by the distribution diagram generation unit on a display unit. apparatus.   The image processing apparatus according to claim 1, further comprising: a classification result display control unit that performs control to display a class classification result by the class classification unit on a display unit. Label image generation means for generating a label image by attaching label information corresponding to the classified main element to each image position based on the classification result by the classification means;
Label image display control means for performing control to display a label image generated by the label image generation means on a display unit;
The image processing apparatus according to claim 1, further comprising: The main elements include red blood cells;
The first dye amount, the second dye amount, and the third dye amount estimated by the dye amount estimating means are the respective dye amounts of hematoxylin, eosin, and eosin dyed erythrocytes. The image processing apparatus according to any one of 1 to 8.   The image processing apparatus according to claim 9, wherein the first pigment amount is a pigment amount of hematoxylin.   The image processing apparatus according to claim 9, wherein the second dye amount is an eosin dye amount.   The image processing apparatus according to any one of claims 9 to 11, wherein the third pigment amount is a pigment amount of eosin that stains red blood cells. A sample that includes at least one of a plurality of predetermined main elements and that processes a sample image obtained by imaging a sample stained with a plurality of dyes,
A pigment amount estimation step for estimating a first pigment amount, a second pigment amount, and a third pigment amount at each image position of the sample image;
A conversion step of converting each image position into a dye amount space based on each dye amount at each image position estimated by the dye amount estimation step;
A dye amount threshold value determining step for determining a dye amount threshold value of each dye amount in the dye amount space;
A space dividing step of dividing the dye amount space into a plurality of partial regions by a plane determined by each dye amount threshold determined by the dye amount threshold determining step;
A class classification step of classifying the image positions into the main elements according to the division result of the space division step;
An image processing program for executing

Images