Machine learning for tissue labeling segmentation

1. A computer implemented method for directed machine learning comprising:

receiving a plurality of features including an intensity data feature and an location data feature of an image;

condensing the intensity data feature and the location data feature into a feature vector;

processing the common feature vector by each of a plurality of classifiers, each classifier trained for a respective trained class among a plurality of classes;

outputting, from each classifier, a value of the feature vector corresponding to the respective trained class; and

assigning the feature vector a label according to a comparison of the values of the classifiers, wherein the label corresponds to a classifier having a selected value from among the values of the classifiers, and wherein the assignment produces a segmentation of the image.

2. The computer implemented method of claim 1 , wherein the classifiers execute a binary support vector machine with radial basis functions trained to identify the feature vector as either tissue or not tissue.

3. The computer implemented method of claim 1 , wherein the classifiers execute a Gaussian mixture classification for selecting the selected value from among the values of the classifiers.

4. The computer implemented method of claim 1 , wherein the classifiers execute a Bayesian Network based classification for selecting the selected value from among the values of the classifiers.

5. The computer implemented method of claim 1 , further comprising training each classifier according to a predetermined weighting of each feature, wherein the assignment of the feature to the label implicitly includes the predetermined weighting of the classifier outputting the selected value, wherein different classifiers have different predetermined weightings.

6. The computer implemented method of claim 1 , wherein condensing further comprises condensing a spatial feature into the feature vector.

7. A computer readable medium embodying instructions executable by a processor to perform a method for directed machine learning, the method steps comprising:

receiving a plurality of features including an intensity data feature and an location data feature of an image;

condensing the intensity data feature and the location data feature into a feature vector;

processing the common feature vector by each of a plurality of classifiers, each classifier trained for a respective trained class among a plurality of classes;

outputting, from each classifier, a value of the feature vector corresponding to the respective trained class; and

assigning the feature vector a label according to a comparison of the values of the classifiers, wherein the label corresponds to a classifier having a selected value from among the values of the classifiers, and wherein the assignment produces a segmentation of the image.

8. The computer readable medium of claim 7 , wherein the classifiers execute a binary support vector machine with radial basis functions trained to identify the feature vector as either tissue or not tissue.

9. The computer readable medium of claim 7 , wherein the classifiers execute a Gaussian mixture classification from among the values of the classifiers.

10. The computer readable medium of claim 7 , wherein the classifiers execute a Bayesian Network based classification from among the values of the classifiers.

11. The computer readable medium of claim 7 , further comprising training each classifier according to a predetermined weighting of each feature, wherein the assignment of the feature to the label implicitly includes the predetermined weighting of the classifier outputting the selected value, wherein different classifiers have different predetermined weightings.

12. The computer readable medium of claim 7 , wherein condensing further comprises condensing a spatial feature into the feature vector.

13. A computer implemented method for directed machine learning comprising:

receiving features including intensity data and location data of an image;

determining intensity probabilities and spatial probabilities for the intensity data and the location data, respectively;

condensing the intensity probabilities and the spatial probabilities into a feature vector;

processing the feature vector by a plurality of classifiers, each classifier trained for a respective trained class among a plurality of classes;

outputting, from each classifier, a probability of the feature vector belong to the respective trained class; and

assigning the feature vector a label according to the probabilities of the classifiers, wherein the assignment produces a segmentation of the image.

14. The computer implemented method of claim 13 , wherein the classifiers execute a binary support vector machine with radial basis functions trained to identify the feature vector as either tissue or not tissue.

15. The computer implemented method of claim 13 , wherein the classifiers execute a Gaussian mixture classification for selecting a probability from among the probabilities output by the classifiers.

16. The computer implemented method of claim 13 , wherein the classifiers execute a Bayesian Network based classification for selecting a probability from among the probabilities output by the classifiers.

17. The computer implemented method of claim 13 , further comprising training each classifier according to a predetermined weighting of each feature, wherein the assignment of the feature to the label implicitly includes the predetermined weighting of the classifier outputting a selected probability from among the probabilities output by the classifiers, wherein different classifiers have different predetermined weightings.

18. The computer implemented method of claim 13 , wherein condensing further comprises condensing a spatial feature into the feature vector.