Training artificial intelligence models for radiation therapy

1. A method comprising:

iteratively training, by a server, an artificial intelligence model, wherein with each iteration the server:

identifies a radiation therapy treatment attribute; and

when an epsilon value does not satisfy a threshold corresponding to a likelihood of exploration or exploitation training of the artificial intelligence model, the server:

generates, using an existing policy applied to the identified radiation therapy treatment attribute, a first predicted radiation therapy treatment attribute;

generates, using a predefined computer model, a second predicted radiation therapy treatment attribute; and

trains the artificial intelligence model using one of the first predicted radiation therapy treatment attribute or the second predicted radiation therapy treatment attribute that is associated with a higher reward, such that the artificial intelligence model is configured to predict a third predicted radiation therapy treatment attribute in accordance with its corresponding reward.

2. The method of claim 1 , wherein the server iteratively repeats training the artificial intelligence model until the existing policy satisfies an accuracy threshold.

3. The method of claim 1 , wherein the epsilon value is received from a system administrator.

4. The method of claim 1 , wherein the server further trains the existing policy based on the predefined computer model using a supervised training method.

5. The method of claim 1 , wherein the server revises the epsilon value, such that a likelihood of generation of the first or second predicted radiation therapy treatment attribute is higher or lower than generation of a random radiation therapy treatment attribute.

6. The method of claim 1 , wherein the predefined computer model is specific to a clinic.

7. The method of claim 1 , wherein the predefined computer model is configured to optimize a dose-histogram volume calculation.

8. A system comprising:

one or more processors; and

a non-transitory memory to store computer code instructions, the computer code instructions when executed cause the one or more processors to:

iteratively train an artificial intelligence model, wherein with each iteration the processor;

identifies a radiation therapy treatment attribute; and

when an epsilon value does not satisfy a threshold corresponding to a likelihood of exploration or exploitation training of the artificial intelligence model, the one or more processors:

generate, using an existing policy applied to the identified radiation therapy treatment attribute, a first predicted radiation therapy treatment attribute;

generate, using a predefined computer model, a second predicted radiation therapy treatment attribute; and

train the artificial intelligence model using one of the first predicted radiation therapy treatment attribute or the second predicted radiation therapy treatment attribute that is associated with a higher reward, such that the artificial intelligence model is configured to predict a third predicted radiation therapy treatment attribute in accordance with its corresponding reward.

9. The system of claim 8 , wherein the one or more processors iteratively repeat training the artificial intelligence model until the existing policy satisfies an accuracy threshold.

10. The system of claim 8 , wherein the epsilon value is received from a system administrator.

11. The system of claim 8 , wherein the one or more processors further train the existing policy based on the predefined computer model using a supervised training method.

12. The system of claim 8 , wherein the one or more processors revise the epsilon value, such that a likelihood of generation of the first or second predicted radiation therapy treatment attribute is higher or lower than generation of a random radiation therapy treatment attribute.

13. The system of claim 8 , wherein the predefined computer model is specific to a clinic.

14. The system of claim 8 , wherein the predefined computer model is configured to optimize a dose-histogram volume calculation.

15. A system comprising:

a server in communication with an artificial intelligence model, the server configured to:

iteratively train the artificial intelligence model, wherein with each iteration the server is configured to:

identifies a radiation therapy treatment attribute; and

when an epsilon value does not satisfy a threshold corresponding to a likelihood of exploration or exploitation training of the artificial intelligence model, the server:

generates, using an existing policy applied to the identified radiation therapy treatment attribute, a first predicted radiation therapy treatment attribute;

generates, using a predefined computer model, a second predicted radiation therapy treatment attribute; and

trains the artificial intelligence model using one of the first predicted radiation therapy treatment attribute or the second predicted radiation therapy treatment attribute that is associated with a higher reward, such that the artificial intelligence model is configured to predict a third predicted radiation therapy treatment attribute in accordance with its corresponding reward.

16. The system of claim 15 , wherein the server is further configured to:

iteratively repeat training the artificial intelligence model until the existing policy satisfies an accuracy threshold.

17. The system of claim 15 , wherein the epsilon value is received from a system administrator.

18. The system of claim 15 , wherein the server is further configured to:

train the existing policy based on the predefined computer model using a supervised training method.

19. The system of claim 15 , wherein the server is further configured to:

revise the epsilon value, such that a likelihood of generation of the first or second predicted radiation therapy treatment attribute is higher or lower than generation of a random radiation therapy treatment attribute.

20. The system of claim 15 , wherein the predefined computer model is specific to a clinic.