Systems and methods for predicting perfusion deficits from physiological, anatomical, and patient characteristics

1. A computer-implemented method for estimating patient-specific blood perfusion to guide diagnosis or treatment of cardiovascular disease, the method comprising:

receiving measured data of a characteristic of perfusion of blood through tissues of a plurality of individuals;

receiving measured data of one or more anatomical or physiological parameters for each of the plurality of individuals;

generating a mapping of the one or more anatomical or physiological parameters to the characteristic of perfusion of blood, by training a machine learning algorithm to learn the mapping using a training set comprising the measured data of the characteristic of perfusion of blood and the measured data of the one or more anatomical or physiological parameters;

generating, from image data of a patient, a patient-specific vessel model of a vessel of the patient and a patient-specific tissue model of a target tissue perfused by the vessel;

extracting, from the patient-specific vessel model or the patient-specific tissue model, one or more patient-specific values of the one or more anatomical or physiological parameters;

computing an estimate of the characteristic of perfusion of blood through the patient-specific tissue model using the mapping, by inputting the one or more patient-specific values of the one or more anatomical or physiological parameters into the trained machine learning algorithm to obtain the estimate as an output of the trained machine learning algorithm; and

generating a virtual scan comprising the computed estimate of the characteristic of perfusion of blood.

2. The computer implemented method of claim 1 , wherein the one or more anatomical or physiological parameters include an estimated or measured blood flow to a plurality of regions of a vessel or tissue at one or more physiological states.

3. The computer implemented method of claim 1 , wherein the one or more anatomical or physiological parameters include one or more anatomical geometries of: vessel size, vessel shape, vessel tortuosity, vessel length, vessel thickness, estimated territories of the perfusion of blood within a tissue or vascular network, or a combination thereof.

4. The computer implemented method of claim 1 , wherein

the one or more patient-specific values of the one or more anatomical or physiological parameters characterize the patient in one or more physiological states, and

the one or more physiological states include one or more of: a resting patient state, a hyperemic state, an exercise state, a postprandial state, a gravitational state, an emotional state, a state of hypertension, a medicated state, or a combination thereof.

5. The computer implemented method of claim 1 , wherein the characteristic of perfusion of blood includes one or more of a fractional flow reserve, flow magnitude, flow direction, or a combination thereof.

6. The computer implemented method of claim 1 , wherein the one or more anatomical or physiological parameters include one or more image characteristics of a target tissue or a vessel in one or more physiological states, including one or more of local average intensities, texture characteristics, standard image, or a combination thereof.

7. The computer implemented method of claim 1 , wherein the one or more anatomical or physiological parameters include one or more of: patient characteristics, target tissue disease characteristics, electromechanical measurements, or a combination thereof.

8. The computer implemented method of claim 1 , wherein the computing the estimate of the characteristic of perfusion of blood through the patient-specific tissue model includes comparing the blood flow in the target tissue at one or more physiological states.

9. The computer implemented method of claim 1 , wherein the patient-specific vessel model and the patient-specific tissue model includes one or more of:

a coronary vasculature and the myocardium;

a cerebral vasculature and the brain;

a peripheral vasculature and muscle;

a hepatic vasculature and a liver;

a renal vasculature and a kidney;

a visceral vasculature and a bowel;

or any target organ and a vasculature supplying blood to said target organ.

10. The computer implemented method of claim 1 , further comprising:

adjusting the one or more patient-specific values of the one or more physiological parameters based on the estimate of the characteristic of perfusion of blood; and

simulating a blood flow characteristic using the estimate and the adjusted one or more patient-specific values of the one or more anatomical or physiological parameters.

11. The computer implemented method of claim 1 , further comprising:

receiving one or more desired perfusion characteristics at one or more locations in the target tissue;

comparing the estimate of the characteristic of perfusion of blood at the one or more locations in the target tissue with the one or more desired perfusion characteristics at the one or more locations; and

changing one or more of the one or more patient-specific values of the one or more anatomical or physiological parameters or the patient-specific tissue model, based on the comparison.

12. The computer implemented method of claim 1 , wherein

the virtual scan comprises the computed estimate of the characteristic perfusion of blood overlaid on an image of the target tissue to thereby simulate a SPECT or PET scan in one or more physiological states, and

the method further comprises outputting the one or more virtual perfusion images to an electronic storage medium.

13. The method of claim 1 , further comprising,

adjusting the one or more patient-specific values of the one or more anatomical or physiological parameters based on the estimate of the characteristic of the perfusion of blood; and

restoring blood flow through the patient-specific tissue model using the adjusted one or more patient-specific values of the one or more anatomical or physiological parameters.

14. A system for estimating patient-specific blood perfusion to guide diagnosis or treatment of cardiovascular disease, the system comprising:

a data storage device storing instructions for determining patient-specific characteristics of the perfusion of blood; and

a processor configured to execute the instructions to perform a method including the steps of:

receiving measured data of a characteristic of perfusion of blood through tissues of a plurality of individuals;

receiving measured data of one or more anatomical or physiological parameters for each of the plurality of individuals;

generating a mapping of the one or more anatomical or physiological parameters to the characteristic of perfusion of blood, by training a machine learning algorithm to learn the mapping using a training set comprising the measured data of the characteristic of perfusion of blood and the measured data of the one or more anatomical or physiological parameters;

generating, from image data of a patient, a patient-specific vessel model of a vessel of the patient and a patient-specific tissue model of a target tissue perfused by the vessel;

extracting, from the patient-specific vessel model or the patient-specific tissue model, one or more patient-specific values of the one or more anatomical or physiological parameters;

computing an estimate of the characteristic of perfusion of blood through the patient-specific tissue model using the mapping, by inputting the one or more patient-specific values of the one or more anatomical or physiological parameters into the trained machine learning algorithm to obtain the estimate as an output of the trained machine learning algorithm; and

generating a virtual scan comprising the computed estimate of the characteristic of perfusion of blood.

15. The system of claim 14 , wherein the one or more anatomical or physiological parameters include, at one or more physiological states, one or more of the following:

an estimated or measured blood flow to a plurality of regions of a vessel or tissue;

anatomical characteristics;

image characteristics;

patient characteristics;

target tissue disease characteristics;

electromechanical measurements; or

a combination thereof.

16. The system of claim 14 , wherein

the one or more patient-specific values of the one or more anatomical or physiological parameters characterize the patient in one or more physiological states, and

the one or more physiological states include, one or more of: a resting patient state, a hyperemic state, an exercise state, a postprandial state, a gravitational state, an emotional state, a state of hypertension, a medicated state, or a combination thereof.

17. The system of claim 14 , wherein the characteristic of the perfusion of blood include one or more of: fractional flow reserve, flow magnitude, flow direction, or a combination thereof.

18. The system of claim 14 , wherein the computing the estimate of the a characteristic of perfusion of blood through the patient-specific tissue model includes comparing the blood flow in the target tissue at one or more physiological states.

19. The system of claim 14 , wherein the patient-specific vessel model and the patient-specific tissue model includes one or more of:

a coronary vasculature and the myocardium;

a cerebral vasculature and the brain;

a peripheral vasculature and muscle;

a hepatic vasculature and a liver;

a renal vasculature and a kidney;

a visceral vasculature and a bowel;

or any target organ and a vasculature supplying blood to said target organ.

20. The system of claim 14 , wherein the method further comprises:

adjusting the one or more patient-specific values of the one or more anatomical or physiological parameters based on the estimate of the characteristic of the perfusion of blood; and

simulating a blood flow characteristic using the estimate of the characteristic of blood perfusion and the adjusted one or more patient-specific values of the one or more anatomical or physiological parameters.

21. A non-transitory computer readable medium for performing a method on a computer system containing computer-executable programming instructions for estimating patient-specific blood perfusion to guide diagnosis or treatment of cardiovascular disease, the method comprising:

receiving measured data of a characteristic of perfusion of blood through tissues of a plurality of individuals;

receiving measured data of one or more anatomical or physiological parameters for each of the plurality of individuals;

generating a mapping of the one or more anatomical or physiological parameters to the characteristic of perfusion of blood, by training a machine learning algorithm to learn the mapping using a training set comprising the measured data of the characteristic of perfusion of blood and the measured data of the one or more anatomical or physiological parameters;

generating, from image data of a patient, a patient-specific vessel model of a vessel of the patient and a patient-specific tissue model of a target tissue perfused by the vessel;

extracting, from the patient-specific vessel model or the patient-specific tissue model, one or more patient-specific values of the one or more anatomical or physiological parameters;

computing an estimate of the characteristic of perfusion of blood through the patient-specific tissue model using the mapping, by inputting the one or more patient-specific values of the one or more anatomical or physiological parameters into the trained machine learning algorithm to obtain the estimate as an output of the trained machine learning algorithm; and

generating a virtual scan comprising the computed estimate of the characteristic of perfusion of blood.