Method and system for clinical decision support with local and remote analytics

1. A method for non-invasive assessment of coronary artery disease (CAD) of a patient, comprising:

receiving medical image data of a patient;

extracting a clinical indicator of CAD from the medical image data;

automatically selecting a processing strategy for assessing CAD of the patient using joint on-site and off-site processing using a trained machine learning based intelligent artificial agent based on the clinical indicator of CAD, a purpose for assessing the CAD, and clinical requirements for a current clinical scenario;

performing non-invasive assessment of the CAD of the patient based on the medical image data of the patient using the joint on-site and off-site processing according to the selected processing strategy;

determining whether results of the non-invasive assessment of the CAD of the patient using the on-site processing and results of the non-invasive assessment of the CAD of the patient using the off-site processing are consistent; and

in response to determining that the results of the non-invasive assessment of the CAD of the patient using the on-site processing and the results of the non-invasive assessment of the CAD of the patient using the off-site processing are consistent, determining a final assessment of the CAD of the patient by combining the results of the non-invasive assessment of the CAD of the patient using the on-site processing and the results of the non-invasive assessment of the CAD of the patient using the off-site processing.

2. The method of claim 1 , wherein automatically selecting a processing strategy for assessing CAD of the patient using joint on-site and off-site processing using a trained machine learning based intelligent artificial agent based on the clinical indicator of CAD, a purpose for assessing the CAD, and clinical requirements for a current clinical scenario comprises:

automatically selecting the processing strategy for assessing the CAD of the patient using the joint on-site and off-site processing based on the clinical requirements including one or more of time requirements, fidelity requirements, computing resources currently available on-site, or a type of the medical image data.

3. The method of claim 1 , wherein:

extracting a clinical indicator of CAD from the medical image data comprises:

performing calcium scoring in the medical image data of the patient using a trained machine learning based classifier; and

automatically selecting a processing strategy for assessing CAD of the patient using joint on-site and off-site processing using a trained machine learning based intelligent artificial agent based on the clinical indicator of CAD, a purpose for assessing the CAD, and clinical requirements for a current clinical scenario comprises:

automatically selecting the processing strategy for assessing CAD of the patient using the joint on-site and off-site processing based on the calcium scoring in the medical image data of the patient.

4. The method of claim 1 , wherein automatically selecting a processing strategy for assessing CAD of the patient using joint on-site and off-site processing using a trained machine learning based intelligent artificial agent based on the clinical indicator of CAD, a purpose for assessing the CAD, and clinical requirements for a current clinical scenario comprises:

automatically selecting the processing strategy using a trained deep neural network (DNN) based on the clinical indicator of CAD input to the trained DNN, wherein the trained DNN is trained using deep reinforcement learning based on a database of patient-specific medical image data for various patients and corresponding outcomes or clinical histories of the various patients.

5. The method of claim 4 , wherein the clinical indicator is a calcium scoring of the medical image data of the patient.

6. The method of claim 1 , wherein the on-site processing utilizes a machine learning based CAD assessment algorithm and the off-site processing utilizes a CAD assessment algorithm based on computational modeling.

7. The method of claim 1 , wherein the on-site processing utilizes a first computational CAD assessment algorithm and the off-site processing utilizes a second computational CAD assessment algorithm with a higher fidelity computational model of coronary artery blood flow than the first computational CAD assessment algorithm.

8. The method of claim 1 , wherein performing non-invasive assessment of the CAD of the patient based on the medical image data of the patient using the joint on-site and off-site processing according to the selected processing strategy comprises:

performing one or more operations of a CAD assessment algorithm using the on-site processing and one or more operations of the CAD assessment algorithm using the off-site processing.

9. The method of claim 1 , further comprising:

automatically determining whether additional processing is necessary for assessing the CAD of the patient; and

in response to a determination that additional processing is necessary for assessing the CAD of the patient, automatically selecting an updated processing strategy for assessing the CAD of the patient using one of on-site processing, off-site processing, or the joint on-site and off-site processing based on the clinical requirements for the current clinical scenario and the final assessment of CAD performed using the previous processing strategy.

10. An apparatus for non-invasive assessment of coronary artery disease (CAD) of a patient, comprising:

means for receiving medical image data of a patient;

means for extracting a clinical indicator of CAD from the medical image data;

means for automatically selecting a processing strategy for assessing CAD of the patient using joint on-site and off-site processing using a trained machine learning based intelligent artificial agent based on the clinical indicator of CAD, a purpose for assessing the CAD, and clinical requirements for a current clinical scenario;

means for performing non-invasive assessment of the CAD of the patient based on the medical image data of the patient using the joint on-site and off-site processing according to the selected processing strategy;

means for determining whether results of the non-invasive assessment of the CAD of the patient using the on-site processing and results of the non-invasive assessment of the CAD of the patient using the off-site processing are consistent; and

means for determining a final assessment of the CAD of the patient by combining the results of the non-invasive assessment of the CAD of the patient using the on-site processing and the results of the non-invasive assessment of the CAD of the patient using the off-site processing in response to determining that the results of the non-invasive assessment of the CAD of the patient using the on-site processing and the results of the non-invasive assessment of the CAD of the patient using the off-site processing are consistent.

11. The apparatus of claim 10 , wherein the means for automatically selecting a processing strategy for assessing CAD of the patient using joint on-site and off-site processing using a trained machine learning based intelligent artificial agent based on the clinical indicator of CAD, a purpose for assessing the CAD, and clinical requirements for a current clinical scenario comprises:

means for automatically selecting the processing strategy for assessing the CAD of the patient using the joint on-site and off-site processing based on the clinical requirements including one or more of time requirements, fidelity requirements, computing resources currently available on-site, or a type of the medical image data.

12. The apparatus of claim 10 , further comprising:

means for automatically determining whether additional processing is necessary for assessing the CAD of the patient; and

means for automatically selecting an updated processing strategy for assessing the CAD of the patient using one of on-site processing, off-site processing, or the joint on-site and off-site processing based on the clinical requirements for the current clinical scenario and the final assessment of CAD performed using the previous processing strategy in response to a determination that additional processing is necessary for assessing the CAD of the patient.

13. A non-transitory computer readable medium storing computer program instructions for non-invasive assessment of coronary artery disease (CAD) of a patient, the computer program instructions when executed by a processor cause the processor to perform operations comprising:

receiving medical image data of a patient;

extracting a clinical indicator of CAD from the medical image data;

automatically selecting a processing strategy for assessing CAD of the patient using joint on-site and off-site processing using a trained machine learning based intelligent artificial agent based on the clinical indicator of CAD, a purpose for assessing the CAD, and clinical requirements for a current clinical scenario;

performing non-invasive assessment of the CAD of the patient based on the medical image data of the patient using the joint on-site and off-site processing according to the selected processing strategy;

determining whether results of the non-invasive assessment of the CAD of the patient using the on-site processing and results of the non-invasive assessment of the CAD of the patient using the off-site processing are consistent; and

in response to determining that the results of the non-invasive assessment of the CAD of the patient using the on-site processing and the results of the non-invasive assessment of the CAD of the patient using the off-site processing are consistent, determining a final assessment of the CAD of the patient by combining the results of the non-invasive assessment of the CAD of the patient using the on-site processing and the results of the non-invasive assessment of the CAD of the patient using the off-site processing.

14. The non-transitory computer readable medium of claim 13 , wherein automatically selecting a processing strategy for assessing CAD of the patient using joint on-site and off-site processing using a trained machine learning based intelligent artificial agent based on the clinical indicator of CAD, a purpose for assessing the CAD, and clinical requirements for a current clinical scenario comprises:

automatically selecting the processing strategy for assessing the CAD of the patient using the joint on-site and off-site processing based on the clinical requirements including one or more of time requirements, fidelity requirements, computing resources currently available on-site, or a type of the medical image data.

15. The non-transitory computer readable medium of claim 13 , wherein:

extracting a clinical indicator of CAD from the medical image data comprises:

performing calcium scoring in the medical image data of the patient using a trained machine learning based classifier; and

automatically selecting a processing strategy for assessing CAD of the patient using joint on-site and off-site processing using a trained machine learning based intelligent artificial agent based on the clinical indicator of CAD, a purpose for assessing the CAD, and clinical requirements for a current clinical scenario comprises:

automatically selecting the processing strategy for assessing CAD of the patient using the joint on-site and off-site processing based on the calcium scoring in the medical image data of the patient.

16. The non-transitory computer readable medium of claim 13 , wherein performing non-invasive assessment of the CAD of the patient based on the medical image data of the patient using the joint on-site and off-site processing according to the selected processing strategy comprises:

performing one or more operations of a CAD assessment algorithm using the on-site processing and one or more operations of the CAD assessment algorithm using the off-site processing.

17. The non-transitory computer readable medium of claim 13 , wherein the operations further comprise:

automatically determining whether additional processing is necessary for assessing the CAD of the patient; and

in response to a determination that additional processing is necessary for assessing the CAD of the patient, automatically selecting an updated processing strategy for assessing the CAD of the patient using one of on-site processing, off-site processing, or the joint on-site and off-site processing based on the clinical requirements for the current clinical scenario and the final assessment of CAD performed using the previous processing strategy.