Image processing apparatus and computer program product

1. An image processing apparatus comprising:

a spectral-characteristic estimating unit that estimates, based on a pixel value of a pixel of a stained sample image, a spectral characteristic value of each wavelength at a corresponding point on a stained sample corresponding to the pixel, the stained sample image being obtained by imaging the stained sample that is stained with a plurality of dyes;

a weight setting unit that sets a weight value of each wavelength based on the spectral characteristic value of each wavelength estimated by the spectral-characteristic estimating unit; and

a weighted dye-amount estimating unit that estimates an amount of dye at the corresponding point on the stained sample as a weighted dye amount based on the spectral characteristic value of each wavelength estimated by the spectral-characteristic estimating unit using the weight value of each wavelength.

2. The image processing apparatus according to claim 1 , wherein the weight setting unit sets the spectral characteristic value of each wavelength estimated by the spectral-characteristic estimating unit as the weight value of each wavelength.

3. The image processing apparatus according to claim 1 , wherein the weight setting unit compares the spectral characteristic value of each wavelength estimated by the spectral-characteristic estimating unit with a predetermined threshold, and sets the weight value of each wavelength according to the value relative to the threshold.

4. The image processing apparatus according to claim 1 , wherein the weight setting unit calculates a predetermined converted value from the spectral characteristic value of each wavelength estimated by the spectral-characteristic estimating unit, and sets the converted value as the weight value of each wavelength.

5. The image processing apparatus according to claim 4 , wherein

the spectral characteristic value of each wavelength is spectral transmittance, and

the weight setting value calculates an inverse of absorbance of each wavelength that is obtained by performing logarithmic conversion on the spectral characteristic value of each wavelength as the converted value.

6. The image processing apparatus according to claim 4 , wherein

the spectral characteristic value of each wavelength is spectral transmittance, and

the weight setting value calculates a product of a square root of the spectral characteristic value of each wavelength and an inverse of absorbance of each wavelength that is obtained by performing logarithmic conversion on the spectral characteristic value of each wavelength, as the converted value.

7. The image processing apparatus according to claim 1 , further comprising:

a dye-amount estimating unit that estimates an amount of dye of the stained sample based on the spectral characteristic value of each wavelength estimated by the spectral-characteristic estimating unit;

a spectral-characteristic synthesizing unit that calculates a synthesized spectral-characteristic value of each wavelength by synthesizing a spectral characteristic value based on the amount of dye estimated by the dye-amount estimating unit; and

a difference spectral-characteristic calculating unit that calculates a difference spectral-characteristic value of each wavelength by subtracting the synthesized spectral-characteristic value of each wavelength from the spectral characteristic value of each wavelength estimated by the spectral-characteristic estimating unit, wherein

the weight setting unit sets the weight value of each wavelength based on the difference spectral-characteristic value of each wavelength calculated by the difference spectral-characteristic calculating unit.

8. The image processing unit according to claim 7 , wherein the weight setting unit sets the weight value such that a weight value set for a wavelength at which the difference spectral-characteristic value is large is smaller than a weight value set for a wavelength at which the difference spectra-property value is small.

9. The image processing apparatus according to claim 1 , further comprising:

a weighted spectral-characteristic synthesizing unit that calculates synthesized spectral characteristic value of each wavelength as a weighted synthesized-spectral characteristic value of each wavelength, by synthesizing a spectral characteristic value based on the weighted dye amount estimated by the weighted dye-amount estimating unit; and

an image synthesizing unit that synthesizes an RGB image based on the weighted synthesized-spectral characteristic value of each wavelength.

10. The image processing apparatus according to claim 9 , wherein

the spectral-characteristic estimating unit estimates the spectral characteristic value of each wavelength for all pixels constituting the stained sample image,

the weighted dye-amount estimating unit that estimates the amount of dye at the corresponding point on the stained sample corresponding to respective pixels on the stained sample image as the weighted dye amount, based on the spectral characteristic value of each wavelength estimated for all pixels constituting the stained sample image,

the weighted spectral-characteristic synthesizing unit calculates the weighted synthesized-spectral characteristic value of each wavelength at each pixel of the stained sample image, based on the weighted dye amount at the corresponding point on the stained sample corresponding to each pixel of the stained sample image, and

the image synthesizing unit synthesizes the RGB image by converting the weighted synthesized-spectral characteristic value of each wavelength at each pixel of the stained sample image into an RGB value.

11. The image processing apparatus according to claim 9 , further comprising an image-display control unit that displays the RGB image synthesized by the image synthesizing unit on a display unit.

12. The image processing apparatus according to claim 1 , wherein the dyes include either of hematoxylin, eosin, eosin that stains red corpuscles, and a coloring matter of red corpuscles that are not stained.

13. The image processing apparatus according to claim 1 , wherein estimation of the weighted dye amount by the weighted dye-amount estimating unit is performed using Lambert-Beer law.

14. The image processing apparatus according to claim 1 , wherein the spectral characteristic value includes spectral transmittance, spectral reflectivity, and absorbance.

15. An image processing unit comprising:

a spectral-characteristic estimating unit that estimates, based on a pixel value of a pixel of a stained sample image, a spectral characteristic value of each wavelength at a corresponding point on the stained sample corresponding to the pixel, the stained sample image being obtained by imaging a stained sample that is stained with a plurality of dyes;

a weight setting unit that sets a weight value of each wavelength based on the spectral characteristic value of each wavelength corresponding to the pixel value of the pixel according to correlation between the pixel value and the spectral characteristic value of each wavelength, the correlation set in advance; and

a weighted dye-amount estimating unit that estimates an amount of dye at the corresponding point on the stained sample as a weighted dye amount based on the spectral characteristic value of each wavelength estimated by the spectral-characteristic estimating unit using the weight value of each wavelength.

16. The image processing apparatus according to claim 15 , further comprising:

a weighted spectral-characteristic synthesizing unit that calculates synthesized spectral characteristic value of each wavelength as a weighted synthesized-spectral characteristic value of each wavelength, by synthesizing a spectral characteristic value based on the weighted dye amount estimated by the weighted dye-amount estimating unit; and

an image synthesizing unit that synthesizes an RGB image based on the weighted synthesized-spectral characteristic value of each wavelength.

17. The image processing apparatus according to claim 16 , wherein

the spectral-characteristic estimating unit estimates the spectral characteristic value of each wavelength for all pixels constituting the stained sample image,

the weighted dye-amount estimating unit that estimates the amount of dye at the corresponding point on the stained sample corresponding to respective pixels on the stained sample image as the weighted dye amount, based on the spectral characteristic value of each wavelength estimated for all pixels constituting the stained sample image,

the weighted spectral-characteristic synthesizing unit calculates the weighted synthesized-spectral characteristic value of each wavelength at each pixel of the stained sample image, based on the weighted dye amount at the corresponding point on the stained sample corresponding to each pixel of the stained sample image, and

the image synthesizing unit synthesizes the RGB image by converting the weighted synthesized-spectral characteristic value of each wavelength at each pixel of the stained sample image into an RGB value.

18. The image processing apparatus according to claim 16 , further comprising an image-display control unit that displays the RGB image synthesized by the image synthesizing unit on a display unit.

19. The image processing apparatus according to claim 15 , wherein the dyes include either of hematoxylin, eosin, eosin that stains red corpuscles, and a coloring matter of red corpuscles that are not stained.

20. The image processing apparatus according to claim 15 , wherein estimation of the weighted dye amount by the weighted dye-amount estimating unit is performed using Lambert-Beer law.

21. The image processing apparatus according to claim 15 , wherein the spectral characteristic value includes spectral transmittance, spectral reflectivity, and absorbance.

22. An image processing unit comprising:

a spectral-characteristic estimating unit that estimates, based on a pixel value of a pixel of a stained sample image, a spectral characteristic value of each wavelength at a corresponding point on the stained sample corresponding to the pixel, the stained sample image being obtained by imaging a stained sample that is stained with a plurality of dyes;

a weight setting unit that sets a weight value of each wavelength using either one of single-stain spectral-characteristic values of each color that are measured for the single-stained samples stained respectively by the dyes; and

a weighted dye-amount estimating unit that estimates an amount of dye at the corresponding point on the stained sample as a weighted dye amount based on the spectral characteristic value of each wavelength estimated by the spectral-characteristic estimating unit using the weight value of each wavelength.

23. The image processing apparatus according to claim 22 , further comprising:

a dye-amount estimating unit that estimates an amount of dye of the stained sample based on the spectral characteristic value of each wavelength estimated by the spectral-characteristic estimating unit; and

a single-stain spectral-characteristic selecting unit that identifies a dye that is dominant at the corresponding point on the stained sample based on the amount of dye estimated by the dye-amount estimating unit, and selects either one of the single-stain spectral-characteristic values of each color according to the identified dye.

24. The image processing apparatus according to claim 22 , further comprising a single-stain spectral-characteristic selecting unit that identifies a dye that is dominant at the corresponding point on the stained sample based on the spectral characteristic value of each wavelength estimated by the spectral-characteristic estimating unit, and selects either one of the single-stain spectral-characteristic values of each color according to the identified dye.

25. The image processing apparatus according to claim 22 , further comprising a single-stain spectral-characteristic selecting unit that identifies a dye that is dominant at the corresponding point on the stained sample based on the pixel value of the stained sample image, and selects either one of the single-stain spectral-characteristic values of each color according to the identified dye.

26. The image processing apparatus according to claim 22 , wherein

the single-stain spectral-characteristic values of each color are constituted of a plurality of single-stain spectral-characteristic values of respective wavelength that are measured at a plurality of corresponding points on the single-stained sample stained with a corresponding dye, and

the weight setting unit calculates standard deviation of each wavelength from the single-stain spectral-characteristic value selected by the single-stain spectral-characteristic selecting unit, and sets the weight value of each wavelength based on a result of calculation.

27. The image processing apparatus according to claim 22 , further comprising a single-stain spectral-characteristic measuring unit that measures the single-stain spectral-characteristic values of respective wavelength of the respective single-stained samples by a spectrometer.

28. The image processing apparatus according to claim 22 , further comprising:

a weighted spectral-characteristic synthesizing unit that calculates synthesized spectral characteristic value of each wavelength as a weighted synthesized-spectral characteristic value of each wavelength, by synthesizing a spectral characteristic value based on the weighted dye amount estimated by the weighted dye-amount estimating unit; and

an image synthesizing unit that synthesizes an RGB image based on the weighted synthesized-spectral characteristic value of each wavelength.

29. The image processing apparatus according to claim 28 , wherein

the spectral-characteristic estimating unit estimates the spectral characteristic value of each wavelength for all pixels constituting the stained sample image,

the weighted dye-amount estimating unit that estimates the amount of dye at the corresponding point on the stained sample corresponding to respective pixels on the stained sample image as the weighted dye amount, based on the spectral characteristic value of each wavelength estimated for all pixels constituting the stained sample image,

the weighted spectral-characteristic synthesizing unit calculates the weighted synthesized-spectral characteristic value of each wavelength at each pixel of the stained sample image, based on the weighted dye amount at the corresponding point on the stained sample corresponding to each pixel of the stained sample image, and

the image synthesizing unit synthesizes the RGB image by converting the weighted synthesized-spectral characteristic value of each wavelength at each pixel of the stained sample image into an RGB value.

30. The image processing apparatus according to claim 28 , further comprising an image-display control unit that displays the ROB image synthesized by the image synthesizing unit on a display unit.

31. The image processing apparatus according to claim 22 , wherein the dyes include either of hematoxylin, eosin, eosin that stains red corpuscles, and a coloring matter of red corpuscles that are not stained.

32. The image processing apparatus according to claim 22 , wherein estimation of the weighted dye amount by the weighted dye-amount estimating unit is performed using Lambert-Beer law.

33. The image processing apparatus according to claim 22 , wherein the spectral characteristic value includes spectral transmittance, spectral reflectivity, and absorbance.

34. A computer program product having a computer readable medium including programmed instructions for estimating an amount of dye of a stained sample from a stained sample image obtained by imaging the stained sample stained with a plurality of dyes, wherein the instructions, when executed by a computer, cause the computer to perform:

estimating, based on a pixel value of a pixel constituting the stained sample image, a spectral characteristic value of each wavelength at a corresponding point on the stained sample corresponding to the pixel;

setting a weight value of each wavelength based on the estimated spectral characteristic value of each wavelength; and

estimating an amount of dye at the corresponding point on the stained sample as a weighted dye amount based on the estimated spectral characteristic value of each wavelength using the weight value of each wavelength.

35. A computer program product having a computer readable medium including programmed instructions for estimating an amount of dye of a stained sample from a stained sample image obtained by imaging the stained sample stained with a plurality of dyes, wherein the instructions, when executed by a computer, cause the computer to perform:

estimating, based on a pixel value of a pixel constituting the stained sample image, a spectral characteristic value of each wavelength at a corresponding point on the stained sample corresponding to the pixel;

setting a weight value of each wavelength based on the spectral characteristic value of each wavelength corresponding to the pixel value of the pixel according to correlation between the pixel value and the spectral characteristic value of each wavelength, the correlation set in advance; and

estimating an amount of dye at the corresponding point on the stained sample as a weighted dye amount based on the estimated spectral characteristic value of each wavelength using the weight value of each wavelength.

36. A computer program product having a computer readable medium including programmed instructions for estimating an amount of dye of a stained sample from a stained sample image obtained by imaging the stained sample stained with a plurality of dyes, when executed by a computer, cause the computer to perform:

estimating, based on a pixel value of a pixel constituting the stained sample image, a spectral characteristic value of each wavelength at a corresponding point on the stained sample corresponding to the pixel;

setting a weight value of each wavelength using either one of single-stain spectral-characteristic values of each color that are measured for the single-stained samples stained respectively by the dyes; and

estimating an amount of dye at the corresponding point on the stained sample as a weighted dye amount based on the estimated spectral characteristic value of each wavelength using the weight value of each wavelength.