Method and apparatus for planning computer-aided diagnosis
The invention provides a method and apparatus for classifying a region of interest in imaging data, the method comprising: calculating a feature vector for at least one region of interest in the imaging data, said feature vector including features of a first modality; projecting the feature vector for the at least one region of interest in the imaging data using a decision function to generate a classification, wherein the decision function is based on classified feature vectors including features of a first modality and features of a second modality; estimating the confidence of the classification if the feature vector is enhanced with features of the second modality.
1. Method for classifying a region of interest in imaging data, the method comprising:
providing imaging data measured using a first modality, said imaging data excluding data measured using a second modality;
calculating a feature vector for at least one region of interest in the imaging data, said feature vector including features calculated from the provided imaging data;
projecting the feature vector for the at least one region of interest in the imaging data using a decision function to generate a classification, wherein the decision function is based on classified feature vectors including features of the first modality and features of at least the second modality;
calculating a confidence value of the classification; and
estimating an increase in the confidence value of the classification if the feature vector were enhanced with features of the second modality calculated from imaging data measured using the second modality.
2. The method according to claim 1 , further comprising determining an outcome value based on outcome values associated with the classified feature vectors.
3. The method according to claim 1 , further comprising marginalizing the decision function for features of the second modality.
4. The method according to claim 3 , further comprising marginalizing the decision function for features of a third modality.
5. The method according to claim 4 , further comprising selecting one of the second modality and the third modality.
6. The method according to claim 4 , wherein the decision function has been trained using the respective set of classified feature vectors to project a feature vector to a classification.
7. The method according to claim 6 , wherein the decision function is based on one of a support vector machine (SVM), a decision tree, or a boosted stump.
8. The method according to claim 7 , wherein the imaging data represents a human lung or a human breast.
9. The method according to claim 8 , wherein the imaging data is a computer tomography (CT) image.
10. The method according to claim 9 , wherein the first, second, or third modality is characterized by a type of contrasting agent used in obtaining the imaging data.
11. The method according to claim 9 , wherein the first, second, or third modality is characterized by a type of equipment used in obtaining the imaging data.
12. The method according to claim 1 , further comprising determining and presenting a recommendation for obtaining second modality features.
13. The method according to claim 12 , wherein the decision function is based on one of a support vector machine (SVM), a decision tree, or a boosted stump.
14. Computation platform comprising a computation device, wherein the computation device is adapted for:
receiving imaging data measured using a first modality, said imaging data excluding data measured using a second modality;
calculating a feature vector for at least one region of interest in the imaging data, said feature vector including features calculated from the received imaging data;
projecting the feature vector for the at least one region of interest in the imaging data using a decision function to generate a classification, wherein the decision function is based on classified feature vectors including features of the first modality and features of at least the second modality;
calculating a confidence value of the classification; and
estimating an increase in the confidence value of the classification if the feature vector were enhanced with features of the second modality calculated from imaging data measured using the second modality.
15. The platform according to claim 14 , further adapted for determining an outcome value based on outcome values associated with the classified feature vectors.
16. Non-transitory computer readable medium comprising computer instructions for implementing the method according to claim 1 .