Techniques for generating a representative image and radiographic interpretation information for a case
View Patent ↗A technique for generating a representative image representing a case and radiographic interpretation information for each case includes calculating wavelet features of a plurality of images that have been taken and stored in the past. The calculated wavelet features and extracted keywords are stored in association with the stored images. The stored images are classified on the basis of the extracted keywords to generate a plurality of groups. For each of the generated groups, a centroid vector of wavelet feature-based feature vectors of respective images corresponding to the keywords is calculated and a spatial distance between the calculated centroid vector and each of the wavelet feature-based feature vectors is calculated. For each of the groups, the image for which the calculated spatial distance is the shortest and the radiographic interpretation information associated with the image is stored as a representative image of that group.
1. An apparatus for generating a representative image representing a case and radiographic interpretation information for each case from medical images based on past cases, comprising:
a wavelet feature calculation unit configured to calculate wavelet features of a plurality of images that have been taken and stored in the past;
a keyword extraction unit configured to extract a keyword included in radiographic interpretation information for each of the stored images;
an information storage unit configured to store the calculated wavelet features and the extracted keywords in association with the respective stored images;
a group generation unit configured to classify the stored images on the basis of the extracted keywords to generate a plurality of groups;
a centroid vector calculation unit configured to calculate, for each of the generated groups, a centroid vector of wavelet feature-based feature vectors of respective images corresponding to the keywords included in that group;
a spatial distance calculation unit configured to calculate, for each of the groups, a spatial distance between the calculated centroid vector and each of the wavelet feature-based feature vectors of the respective images corresponding to the keywords included in that group; and
a representative image storage unit configured to store, for each of the groups, the image for which the calculated spatial distance is the shortest and the radiographic interpretation information associated with the image as a representative image of that group.
2. The apparatus of claim 1 , wherein the wavelet feature calculation unit calculates two-dimensional Gabor wavelet features.
3. The apparatus of claim 1 , further comprising:
a frequency distribution vector calculation unit for calculating frequency distribution vectors for all images by calculating M said wavelet features for each image and binarizing the respective wavelet features for conversion into an M-dimensional bit string, wherein the spatial distance calculation unit calculates the spatial distance as an angle between the centroid vector and each of the calculated frequency distribution vectors, and wherein M is a natural number of 2 or greater.
4. A computer program product, comprising:
a computer-readable storage device; and
program code embodied on the computer-readable storage device, wherein the program code, when executed by a processor, configures the processor to:
calculate wavelet features of a plurality of images that have been taken and stored in the past;
extract a keyword included in radiographic interpretation information for each of the stored images;
store the calculated wavelet features and the extracted keywords in association with the respective stored images;
classify the stored images on the basis of the extracted keywords to generate a plurality of groups;
calculate, for each of the generated groups, a centroid vector of wavelet feature-based feature vectors of respective images corresponding to the keywords included in that group;
calculate, for each of the groups, a spatial distance between the calculated centroid vector and each of the wavelet feature-based feature vectors of the respective images corresponding to the keywords included in that group; and
store, for each of the groups, the image in which the calculated spatial distance is the shortest and the radiographic interpretation information associated with the image as a representative image of that group.
5. The computer program of claim 4 , wherein the wavelet features are two-dimensional Gabor wavelet features.
6. The computer program of claim 4 , wherein the program code, when executed by the processor, further configures the processor to:
calculate frequency distribution vectors for all images by calculating M said wavelet features for each image and binarizing the respective wavelet features for conversion into an M-dimensional bit string, wherein M is a natural number of 2 or greater; and
calculate the spatial distance as an angle between the centroid vector and each of the calculated frequency distribution vectors.