Abnormal shadow detecting system
View Patent ↗In an abnormal shadow detecting system, a characteristic value on the shape and/or the density of a prospective area of a micro calcification shadow set in an image of an object is extracted on the basis of image data representing the image of the object, and whether a micro calcification shadow really exists in the prospective area is determined on the basis of the calculated characteristic value.
1. An abnormal shadow detecting system comprising:
a characteristic value calculating means which calculates a characteristic value on the shape and/or the density of a prospective area of a micro calcification shadow set in an image of an object on the basis of image data representing the image of the object, wherein the characteristic value comprises at least one of
a first threshold value at the time at which the number of prospective points of micro calcification extracted from the prospective area reaches a predetermined number,
the difference between the first threshold value and a second threshold value at the time at which a first prospective point is extracted,
the total volume of the predetermined number of prospective points extracted by the use of the first threshold value,
the dispersion in size of the predetermined number of prospective points extracted by the use of the first threshold value,
the dispersion in density of the predetermined number of prospective points extracted by the use of the first threshold value,
the sum-square of the high frequency component signals of the part of the image of the object after subjected to FFT corresponding to a central region of the prospective areas,
a first frequency at which one-dimensional response in the frequency space after the FFT becomes not larger than a predetermined value for the first time,
the coefficient of variation of the dispersion in lengths in a plurality of directions of the prospective point at the center of the prospective area from its center of gravity, and
the maximum length of the prospective point at the center of the prospective area; and
a discriminating means which discriminates whether a micro calcification shadow really exists in the prospective area on the basis of the calculated characteristic value.
2. An abnormal shadow detecting system comprising:
a characteristic value calculating means which calculates a characteristic value on the shape and/or the density of a prospective area of a micro calcification shadow set in an image of an object on the basis of image data representing the image of the object; and
a discriminating means which discriminates whether a micro calcification shadow really exists in the prospective area on the basis of the calculated characteristic value, wherein the prospective area includes a common micro calcification shadow, which is high in density, and an isolated micro calcification shadow, which is low in density and the discriminating means is arranged to discriminate whether a micro calcification shadow really exists in the prospective area on the basis of different criteria for discriminating between when the prospective area is of common micro calcification, which is high in density, and when the prospective area is of isolated micro calcification, which is low in density.
3. An abnormal shadow detecting system comprising:
a prospective point extracting means which extracts a plurality of prospective points representing a prospective micro calcification shadow in an image of a breast on the basis of image data representing the image of the breast,
a prospective area detecting means which detects a prospective micro calcification area including extracted prospective points on the basis of the image data,
a corrected count calculating means which calculates a corrected count of each of the extracted prospective points for counting the number of the extracted prospective points differently weighting the extracted prospective points on the basis of dispersion in size and/or density of all the prospective points in a predetermined area including the extracted prospective point, and
a discriminating means which calculates the proportion of the number of the prospective points included in a prospective area as counted on the corrected count of each prospective point to the number of the prospective points included in the overall image as counted on the corrected count of each prospective point and discriminates whether a micro calcification shadow really exists in the prospective area on the basis of the proportion.
4. An abnormal shadow detecting system as defined in claim 3 in which the corrected count for each prospective point is larger as the dispersion in size and/or density of all the prospective points in a predetermined area including the prospective point increases, and smaller as the dispersion in size and/or density decreases, and
the discriminating means is arranged to determine that a micro calcification shadow really exists in the prospective area when the proportion is not smaller than a predetermined value.
5. An abnormal shadow detecting system as defined in claim 1 in which the prospective area includes a common micro calcification shadow, which is high in density, and an isolated micro calcification shadow, which is low in density and the discriminating means is arranged to discriminate whether a micro calcification shadow really exists in the prospective area on the basis of different criteria for discriminating between when the prospective area is of common micro calcification, which is high in density, and when the prospective area is of isolated micro calcification, which is low in density.
6. An abnormal shadow detecting system as defined in claim 2 in which the discriminating means discriminates on the basis of the likelihood ratio based on Mahalanobis distances.
7. An abnormal shadow detecting system as defined in claim 2 wherein the discriminating means discriminates by the use of a neural network.
8. An abnormal shadow detecting system comprising:
a characteristic value calculating means which calculates a plurality of characteristic values on the shape and/or the density of a prospective area of a micro calcification shadow set in an image of an object on the basis of image data representing the image of the object, and
a discriminating means which discriminates whether a micro calcification shadow really exists in the prospective area on the basis of the plurality of calculated characteristic values,
wherein the plurality of characteristic values comprises more than one of
a first threshold value at the time at which the number of prospective points of micro calcification extracted from the prospective area reaches a predetermined number,
the difference between the first threshold value and a second threshold value at the time at which a first prospective point is extracted,
the total volume of the predetermined number of prospective points extracted by the use of the first threshold value,
the dispersion in size of the predetermined number of prospective points extracted by the use of the first threshold value,
the dispersion in density of the predetermined number of prospective points extracted by the use of the first threshold value,
the sum-square of the high frequency component signals of the part of the image of the object after subjected to FFT corresponding to a central region of the prospective areas,
a first frequency at which one-dimensional response in the frequency space after the FFT becomes not larger than a predetermined value for the first time,
the coefficient of variation of the dispersion in lengths in a plurality of directions of the prospective point at the center of the prospective area from its center of gravity, and
the maximum length of the prospective point at the center of the prospective area.
9. An abnormal shadow detecting system comprising:
a characteristic value calculating means which calculates a characteristic value on the shave and/or the density of a prospective area of a micro calcification shadow set in an image of an object on the basis of image data representing the image of the object, and
a discriminating means which discriminates whether a micro calcification shadow really exists in the prospective area on the basis of the calculated characteristic value,
wherein the prospective area includes a common micro calcification shadow, which is high in density, and an isolated micro calcification shadow, which is low in density and the discriminating means is arranged to discriminate whether a micro calcification shadow really exists in the prospective area on the basis of different criteria for discriminating between when the prospective area is of common micro calcification, which is high in density, and when the prospective area is of isolated micro calcification, which is low in density.
10. An abnormal shadow detecting system as defined in claim 9 in which the discriminating means discriminates on the basis of the likelihood ratio based on Mahalanobis distances.
11. An abnormal shadow detecting system as defined in claim 9 wherein the discriminating means discriminates by the use of a neural network.
12. An abnormal shadow detecting system comprising:
a characteristic value calculating means which calculates a plurality of characteristic values on the shape and/or the density of a prospective area of a micro calcification shadow set in an image of an object on the basis of image data representing the image of the object, and
a discriminating means which discriminates whether a micro calcification shadow really exists in the prospective area on the basis of the plurality of calculated characteristic values,
wherein the characteristic value calculating means differently weights the plurality of characteristic values on the basis of dispersion in size and/or density of prospective points in a predetermined area including the prospective points.