System and method for target tracking using a quality indicator during radiation therapy

1. A method for training a classifier for determining a quality of target localization from a plurality of images acquired by an imaging device during radiation therapy treatment to a patient, the method comprising:

determining a localization result indicating a location of a target in a plurality of images representative of images acquired in a radiation therapy treatment session;

determining a quality indicator for each localization result, the quality indicator representing a precision level of each localization result;

extracting, by processing circuitry, one or more features from each localization result; and

training the classifier using the localization result, the quality indicator, and one or more of the extracted features, the classifier outputs an indication of the precision level of a new localization result of a newly captured image.

2. The method of claim 1 , further comprising:

selecting a subset of images from the plurality of images to provide a quality measure; and

providing samples having a plurality types of qualities by adjusting the localization result.

3. The method of claim 1 , wherein training the classifier comprises:

training the classifier using a supervised machine learning technique, including:

creating a plurality of subsets of features;

determining primary features based on the plurality of subsets of features; and

training the classifier using the primary features.

4. The method of claim 3 , wherein the primary features include at least one of:

a correlation score;

a similarity measure computed a predetermined distance away from the location of the target;

a minimum orthant score; or

a maximum orthant score.

5. The method of claim 1 , wherein the imaging device includes an ultrasound probe and the plurality of images include ultrasound images.

6. The method of claim 1 , wherein the imaging device includes an MRI device and the plurality of images include MR images.

7. A method for determining a quality of target localization from a plurality of images acquired by an imaging device during radiation therapy treatment to a patient, the method comprising:

determining a localization result indicating a location of a target in a plurality of images acquired during radiation therapy treatment to the patient;

determining a localization quality of the localization result for tracking the location of the target using a classifier, the localization quality representing a precision level of each localization result; and

providing to a user an indication of the localization quality or taking treatment related action based on the localization quality.

8. The method of claim 7 , wherein the indication of the localization quality reflects a determination of at least a normal tracking, a low confidence tracking, and a lost tracking.

9. The method of claim 7 , further comprising:

extracting, using processing circuitry, one or more features from the plurality of images and the location result; and

determining the localization quality using the classifier based on the extracted one or more features.

10. The method of claim 7 , further comprising:

tracking real-time movement of the target during the radiation therapy treatment based on the localization result and the localization quality.

11. The method of claim 7 , wherein the treatment related action includes turning off a radiation beam when the localization quality indicates a lost tracking of the target.

12. The method of claim 7 , further comprising:

updating or retraining the classifier in a clinical environment based on new data.

13. A system for training a classifier for determining a quality of target localization from a plurality of images acquired by an imaging device during radiation therapy treatment to a patient, the system comprising:

a memory for storing computer-executable instructions; and

a processor communicatively coupled to the memory, wherein the computer-executable instructions, when executed by the processor, cause the processor to perform operations including:

determining a localization result indicating a location of a target in a plurality of images representative of images acquired in a radiation therapy treatment session;

determining a quality indicator for each localization result, the quality indicator representing a precision level of each localization result;

extracting one or more features from each localization result; and

training the classifier using the localization result, the quality indicator, and one or more of the extracted features, the classifier outputs an indication of the precision level of a new localization result of a newly captured image.

14. A system for determining a quality of target localization from a plurality of images acquired by an imaging device during radiation therapy treatment to a patient, the system comprising:

a memory for storing computer-executable instructions; and

a processor communicatively coupled to the memory, wherein the computer-executable instructions, when executed by the processor, cause the processor to perform operations including:

determining a localization result indicating a location of a target in a plurality of images acquired during radiation therapy treatment to the patient;

determining a localization quality of the localization result for tracking the location of the target using a classifier, the localization quality representing a precision level of each localization result; and

providing to a user an indication of the localization quality or taking treatment related action based on the localization quality.

15. The system of claim 14 , wherein the indication of the localization quality reflects a determination of at least a normal tracking, a low confidence tracking, and a lost tracking.

16. The system of claim 14 , wherein the operations further comprise:

extracting one or more features from the plurality of images and the location result; and

determining the localization quality using the classifier based on the extracted one or more features.

17. A non-transitory computer-readable medium that stores a set of instructions that is executable by at least one processor of a device to cause the device to perform a method for training a classifier for determining a quality of target localization from a plurality of images acquired by an imaging device during radiation therapy treatment to a patient, the method comprising:

determining a localization result indicating a location of a target in a plurality of images representative of images acquired in a radiation therapy treatment session;

determining a quality indicator for each localization result, the quality indicator representing a precision level of each localization result;

extracting, by the at least one processor, one or more features from each localization result; and

training the classifier using the localization result, the quality indicator, and one or more of the extracted features, the classifier outputs an indication of the precision level of a new localization result of a newly captured image.

18. A non-transitory computer-readable medium that stores a set of instructions that is executable by at least one processor of a device to cause the device to perform a method for determining a quality of target localization from a plurality of images acquired by an imaging device during radiation therapy treatment to a patient, the method comprising:

determining a localization result indicating a location of a target in a plurality of images acquired during radiation therapy treatment to the patient;

determining a localization quality of the localization result for tracking the location of the target using a classifier, the localization quality representing a precision level of each localization result; and

providing to a user an indication of the localization quality or taking treatment related action based on the localization quality.

19. The non-transitory computer-readable medium of claim 18 , wherein the indication of the localization quality reflects a determination of at least a normal tracking, a low confidence tracking, and a lost tracking.

20. The non-transitory computer-readable medium of claim 18 , wherein the method further comprises:

extracting, by the at least one processor, one or more features from the plurality of images and the location result; and

determining the localization quality using the classifier based on the extracted one or more features.