Image processing device, computer storage medium, and method for detecting and displaying nucleated target cells
A device comprises: a nucleus-candidate-region extracting section that extracts, from a captured image obtained by image-capturing a sample piece including a target cell having a nucleus, a nucleus candidate region corresponding to the nucleus; a basic-probability-information acquiring section that acquires, for each of a plurality of determination subject regions determined on the basis of the nucleus candidate region extracted by the nucleus-candidate-region extracting section, basic probability information indicating probability that an image in the determination subject region is an image of the target cell, on the basis of a feature amount of the image of the determination subject region; and a probability-information calculating section that calculates probability information indicating probability that an image in a display subject region corresponding to the nucleus candidate region is the image of the target cell, on the basis of the basic probability information acquired for each of the plurality of determination subject regions.
1. An image processing device comprising
a nucleus-candidate-region extracting unit that extracts, from a captured image obtained by image-capturing a sample piece including a target cell having a nucleus, a nucleus candidate region corresponding to the nucleus;
a basic-probability-information acquiring section that acquires, for each of a plurality of determination subject regions determined on the basis of the nucleus candidate region extracted by the nucleus-candidate-region extracting unit, basic probability information indicating probability that an image in the determination subject region is an image of the target cell, on the basis of a feature amount of the image of the determination subject region; and
a probability-information calculating section that calculates probability information indicating probability that an image in a display subject region corresponding to the nucleus candidate region is the image of the target cell, on the basis of the basic probability information acquired for each of the plurality of determination subject regions,
wherein the probability-information calculating section
calculates, as the probability information indicating the probability that the image in the display subject region corresponding to the nucleus candidate region is the image of the target cell, an increasing function value that uses the basic probability information acquired for each of the plurality of determination regions as a variable, or an increasing function value that uses the basic probability information acquired for any of the plurality of determination regions and the number of the plurality of determination subject regions as variables.
2. The image processing device according to claim 1 ,
wherein the basic-probability-information acquiring section
acquires, for each nucleus candidate region extracted by the nucleus-candidate-region extracting unit, the basic probability information for each of the plurality of determination subject regions determined on the basis of the nucleus candidate region,
wherein the probability-information calculating section
calculates, for each nucleus candidate region extracted by the nucleus-candidate-region extracting unit, the probability information indicating the probability that the image in the display subject region corresponding to the nucleus candidate region is the image of the target cell, and
wherein the image processing device further includes
a display section that displays an image of each display subject region in a descending order of the probability indicated by the probability information.
3. The image processing device according to claim 1 ,
wherein the basic-probability-information acquiring section
acquires a value obtained by inputting a feature amount of the image of the determination subject region to an identifier generated by a learning algorithm, as the basic probability information indicating the probability that the image in the determination subject region is the image of the target cell.
4. The image processing device according to claim 1 , further comprising:
a setting unit that sets a display target region corresponding to the nucleus candidate region, on the basis of a plurality of determination subject regions determined on the basis of the nucleus candidate region.
5. The image processing device according to claim 4 ,
wherein the setting unit
sets a determination subject region from which basic probability information having maximum probability among the plurality of determination subject regions determined on the basis of the nucleus candidate region, as a display target region corresponding to the nucleus candidate region.
6. The image processing device according to claim 1 ,
wherein the probability-information calculating section
calculates a sum of the basic probability information acquired for each of the plurality of determination subject regions, as the probability information indicating the probability that the image in the display subject region corresponding to the nucleus candidate region is the image of the target cell.
7. A non-transitory computer readable storage medium storing a program causing a computer to execute a process, the process comprising:
extracting, from a captured image obtained by image-capturing a sample piece including a target cell having a nucleus, a nucleus candidate region corresponding to the nucleus;
acquiring, for each of a plurality of determination subject regions determined on the basis of the extracted nucleus candidate region, basic probability information indicating probability that an image in the determination subject region is an image of the target cell, on the basis of a feature amount of the image of the determination subject region; and
calculating probability information indicating probability that an image in a display subject region corresponding to the nucleus candidate region is the image of the target cell, on the basis of the basic probability information acquired for each of the plurality of determination subject regions,
wherein
the calculating calculates, as the probability information indicating the probability that the image in the display subject region corresponding to the nucleus candidate region is the image of the target cell, an increasing function value that uses the basic probability information acquired for each of the plurality of determination regions as a variable, or an increasing function value that uses the basic probability information acquired for any of the plurality of determination regions and the number of the plurality of determination subject regions as variables.
8. An image processing method comprising:
extracting, from a captured image obtained by image-capturing a sample piece including a target cell having a nucleus, a nucleus candidate region corresponding to the nucleus;
acquiring, for each of a plurality of determination subject regions determined on the basis of the extracted nucleus candidate region, basic probability information indicating probability that an image in the determination subject region is an image of the target cell, on the basis of a feature amount of the image of the determination subject region; and
calculating probability information indicating probability that an image in a display subject region corresponding to the nucleus candidate region is the image of the target cell, on the basis of the basic probability information acquired for each of the plurality of determination subject regions,
wherein
the calculating calculates, as the probability information indicating the probability that the image in the display subject region corresponding to the nucleus candidate region is the image of the target cell, an increasing function value that uses the basic probability information acquired for each of the plurality of determination regions as a variable, or an increasing function value that uses the basic probability information acquired for any of the plurality of determination regions and the number of the plurality of determination subject regions as variables.