Enhanced personalized evaluation of coronary artery disease using an integration of multiple medical imaging techniques

1. A method for providing a personalized evaluation of coronary artery disease (CAD) for a patient, the method comprising:

acquiring one or more non-invasive images depicting a patient's coronary arteries;

extracting a first set of features of interest from the one or more non-invasive images;

applying a first machine learning model to the first set of features of interest to yield a prediction of one or more coronary measures of interest at a certain location in the coronary arteries;

acquiring one or more invasive images depicting the patient's coronary arteries, wherein the invasive images depict regions of interest selected based on the prediction of one or more coronary measures of interest;

extracting a second set of features of interest from the one or more invasive images;

combining the first set of features of interest and the second set of features of interest to yield a combined set of features of interest; and

applying a second machine learning model to the combined set of features of interest to yield an enhanced prediction of the one or more coronary measures of interest accounting for full coronary circulation.

2. The method of claim 1 , wherein the non-invasive images comprise Computed Tomography Angiography (CCTA) images and the invasive images comprise X-ray Angiography (XA) images.

3. The method of claim 1 , wherein the combined set of features of interest further includes the prediction of the one or more coronary measures of interest.

4. The method of claim 1 , wherein the first set of features are extracted directly from the one or more non-invasive images.

5. The method of claim 1 , further comprising:

generating a geometric model of the patient's coronary arteries using the one or more non-invasive images,

wherein the plurality of features are extracted from the geometric model.

6. The method of claim 1 , wherein the one or more coronary measures of interest comprise a measurement of Fractional Flow Reserve (FFR).

7. The method of claim 1 , wherein the one or more coronary measures of interest comprise a measurement of one or more of instantaneous wave-free ratio (IFR), ratio of resting distal pressure to aortic pressure (rest Pd/Pa), basal stenosis resistance (BSR), hyperemic stenosis resistance (HSR), and index of microcirculatory resistance (IMR).

8. The method of claim 1 , further comprising:

determining whether the enhanced prediction of the one or more coronary measures of interest is in a gray zone of a hybrid decision making strategy; and

if the one or more coronary measures of interest is in the gray zone of the hybrid decision making strategy, performing an invasive measurement of the one or more coronary measures of interest to acquire invasive measurement data.

9. The method of claim 8 , further comprising:

retraining the machine learning model based on the invasive measurement data and the combined set of features of interest.

10. The method of claim 8 , wherein the hybrid decision making strategy is based on a decision of performing iFR measurements or FFR measurements.

11. The method of claim 8 , wherein the invasive measurement is an FFR measurement.

12. A computer-implemented method for providing a personalized evaluation of CAD for a patient, the method comprising:

acquiring one or more non-invasive images depicting a patient's coronary arteries;

extracting a first set of features of interest from the one or more non-invasive images;

applying a first machine learning model to the first set of features of interest to yield a prediction of one or more coronary measures of interest at a certain location in the coronary arteries;

acquiring one or more invasive images depicting the patient's coronary arteries, wherein the invasive images depict regions of interest selected based on the prediction of one or more coronary measures of interest;

extracting a second set of features of interest from the one or more invasive images; and

performing (i) a correction of at least a portion of the first set of features of interest using the second set of features of interest or (ii) a correction of at least a portion of the second set of features of interest using the first set of features of interest to yield a corrected set of features of interest;

applying a second machine learning model to the corrected set of features of interest to yield an updated prediction of the one or more coronary measures of interest accounting for full coronary circulation.

13. The method of claim 12 , further comprising:

acquiring an invasive measurement of the one or more coronary measures of interest; and

performing additional corrections on the corrected set of features of interest based on the invasive measurement.

14. The method of claim 12 , wherein the non-invasive images comprise CCTA images and the invasive images comprise XA images.

15. The method of claim 12 , wherein the first set of features of interest includes the prediction of the one or more coronary measures of interest.

16. A computer-implemented method for providing a personalized evaluation of CAD for a patient, the method comprising:

acquiring one or more non-invasive images depicting the patient's coronary arteries;

applying a first machine learning model to a first set of features of interest extracted from the one or more non-invasive images to yield a first prediction of one or more coronary measures of interest at a certain location in the coronary arteries;

determining whether the first prediction of the one or more coronary measures of interest indicates functionally significant CAD; and

if the first prediction of coronary measures of interest indicates functionally significant CAD, then scheduling the patient for invasive imaging of the patient's coronary arteries;

acquiring one or more invasive images depicting the patient's coronary arteries, wherein the invasive images depict regions of interest selected based on the prediction of one or more coronary measures of interest; and

applying a second machine learning model to a second set of features of interest extracted from the one or more invasive images to yield a second prediction of the one or more coronary measures of interest accounting for full coronary circulation.

17. The method of claim 16 , further comprising:

using the first prediction and the second prediction of the one or more coronary measures of interest to make one or more treatment decisions.

18. The method of claim 16 , wherein the non-invasive images comprise CCTA images and the invasive images comprise XA images.