Method and system for patient specific planning of cardiac therapies on preoperative clinical data and medical images

1. A method for patient-specific cardiac therapy planning, comprising:

generating a patient-specific anatomical model of left and right ventricles from medical image data of a patient;

generating a patient-specific computational heart model based on the patient-specific anatomical model of the left and right ventricles and patient-specific clinical data, wherein generating a patient-specific computational heart model based on the patient-specific anatomical model of the left and right ventricles and patient-specific clinical data comprises:

determining patient-specific parameters of the computational heart model based on the patient-specific anatomical model of the left and right ventricles and the patient-specific clinical data by:

simulating heart function using the computational heart model, and

adjusting parameters of the computational heart model to control simulated clinical parameters resulting from the simulation of heart function using the computational heart model to match corresponding measured clinical parameters for the patient; and

simulating cardiac resynchronization therapy (CRT) at one or more anatomical locations using the patient-specific computational heart model by, wherein the computational heart model includes a cardiac electrophysiology module that simulates cardiac electrophysiology, a cardiac biomechanical module, and a cardiac boundary conditions module and the step of simulating CRT at one or more anatomical locations using the patient-specific computational heart model comprises:

simulating heart function with the patient-specific computational heart model with a stimulated current introduced at the one or more anatomical locations in the cardiac electrophysiological module of the computational heart model.

2. The method of claim 1 , wherein the step of generating a patient-specific anatomical model of left and right ventricles from medical image data of a patient comprises:

detecting a patient-specific left ventricle model and a patient-specific right ventricle model in the medical image data; and

fusing the left ventricle model and the right ventricle model into a single bi-ventricular volumetric mesh.

3. The method of claim 2 , wherein the step of generating a patient-specific anatomical model of left and right ventricles from medical image data of a patient further comprises:

mapping spatial information corresponding to at least one of scars, grey zones, or fibrosis onto a tetrahedral representation of bi-ventricular volumetric mesh.

4. The method of claim 2 , wherein the step of generating a patient-specific anatomical model of left and right ventricles from medical image data of a patient further comprises:

generating a model of fiber orientation based on the bi-ventricular volumetric mesh.

5. The method of claim 4 , wherein the step of generating a model of fiber orientation based on the bi-ventricular volumetric mesh comprises:

determining a constant orientation for fibers on the epicardium and endocardium between a base plane and an apex of the heart;

determining fibers around the mitral, tricuspid, and pulmonary valves to have a circumferential orientation and fibers around the left ventricle outflow tract to have a longitudinal orientation;

performing geodesic interpolation of fiber orientations for fibers on the endocardium and epicardium between the base plane and the valves; and

calculating orientations of fibers across the myocardium using linear interpolation.

6. The method of claim 1 , wherein the electrophysiological module comprises a phenomenological model that models the change in trans-membrane potential as a summation of inward current, outward current, stimulated current, and a diffusion term.

7. The method of claim 6 , wherein the electrophysiological module comprises a first phenomenological model to model trans-membrane potential in the Purkinje fibers coupled with a second phenomenological model to model trans-membrane potential in the myocardium.

8. The method of claim 1 , wherein the step of adjusting the parameters of the computational heart model to control simulated clinical parameters resulting from the simulation of heart function using the computational heart model to match corresponding measure clinical parameters for the patient comprises:

adjusting a tissue diffusivity parameter of the electrophysiological module to align a simulated QRS with a QRS of the patient measured using ECG.

9. The method of claim 1 , wherein the cardiac biomechanical module comprises a passive component that models an elasticity of the myocardium tissue and an active component that models the active contraction of the muscle in response to an action potential.

10. The method of claim 9 , wherein the passive component comprises a co-rotational elasticity model.

11. The method of claim 9 , wherein the passive component comprises a hyper-elastic model.

12. The method of claim 9 , wherein the active component comprises a multi-scale model of myocyte contraction.

13. The method of claim 9 , wherein the step of adjusting the parameters of the computational heart model to control simulated clinical parameters resulting from the simulation of heart function using the computational heart model to match corresponding measure clinical parameters for the patient comprises:

adjusting a tissue stiffness parameter of the passive component of the cardiac biomechanical module to minimize differences between simulated cardiac motion and observed cardiac motion in the medical image data; and

adjusting an active contraction strength parameter of the active component of the cardiac biomechanical module to minimize differences between simulated cardiac motion and observed cardiac motion in the medical image data and to minimize differences between a simulated ejection fraction and a measured ejection fraction for the patient.

14. The method of claim 1 , wherein the cardiac boundary conditions module models hemodynamics parameters using lumped parameters within each of the left ventricle and the right ventricle, models arterial pressure using a lumped Windkessel model, models atrial pressure using a lumped model, and models the cardiac phases as sequential states.

15. The method of claim 14 , wherein the cardiac boundary conditions module models ventricular pressure in isovolumetric phases using correction and prediction model that corrects a current pressure estimate at a current time step such that endocardial volume is preserved and predicts a next pressure estimate at the next time step.

16. The method of claim 14 , wherein the cardiac boundary conditions module models an effect of arteries and atria on ventricular motion using a base stiffness model.

17. The method of claim 14 , wherein the cardiac boundary conditions module models effects of neighboring organs and the pericardium on ventricular motion using a pericardium constraint model.

18. The method of claim 1 , wherein the simulation of the CRT using the patient-specific heart model calculates changes in cardiac parameters after the simulated CRT, and changed cardiac parameters are used as predictors for therapy planning.

19. An apparatus for patient-specific cardiac therapy planning, comprising:

a processor; and

a memory storing computer program instructions, which when executed by the processor cause the process to perform operations comprising:

generating a patient-specific anatomical model of left and right ventricles from medical image data of a patient;

generating a patient-specific computational heart model based on the patient-specific anatomical model of the left and right ventricles and patient-specific clinical data, wherein generating a patient-specific computational heart model based on the patient-specific anatomical model of the left and right ventricles and patient-specific clinical data comprises:

determining patient-specific parameters of the computational heart model based on the patient-specific anatomical model of the left and right ventricles and the patient-specific clinical data by:

simulating heart function using the computational heart model, and

adjusting parameters of the computational heart model to control simulated clinical parameters resulting from the simulation of heart function using the computational heart model to match corresponding measured clinical parameters for the patient; and

simulating CRT at one or more anatomical locations using the patient-specific computational heart model, wherein the computational heart model includes a cardiac electrophysiology module that simulates cardiac electrophysiology, a cardiac biomechanical module, and a cardiac boundary conditions module and simulating CRT at one or more anatomical locations using the patient-specific computational heart model comprises:

simulating heart function with the patient-specific computational heart model with a stimulated current introduced at the one or more anatomical locations in the cardiac electrophysiological module of the computational heart model.

20. The apparatus of claim 19 , wherein generating a patient-specific anatomical model of left and right ventricles from medical image data of a patient comprises:

detecting a patient-specific left ventricle model and a patient-specific right ventricle model in the medical image data; and

fusing the left ventricle model and the right ventricle model into a single bi-ventricular volumetric mesh.

21. The apparatus of claim 20 , wherein generating a patient-specific anatomical model of left and right ventricles from medical image data of a patient further comprises:

generating a model of fiber orientation based on the bi-ventricular volumetric mesh.

22. A non-transitory computer readable medium storing computer program instructions for patient-specific cardiac therapy planning, the computer program instructions when executed by a processor causing the processor to perform operations comprising:

generating a patient-specific anatomical model of left and right ventricles from medical image data of a patient;

generating a patient-specific computational heart model based on the patient-specific anatomical model of the left and right ventricles and patient-specific clinical data, wherein generating a patient-specific computational heart model based on the patient-specific anatomical model of the left and right ventricles and patient-specific clinical data comprises:

determining patient-specific parameters of the computational heart model based on the patient-specific anatomical model of the left and right ventricles and the patient-specific clinical data by:

simulating heart function using the computational heart model, and

adjusting parameters of the computational heart model to control simulated clinical parameters resulting from the simulation of heart function using the computational heart model to match corresponding measured clinical parameters for the patient; and

simulating cardiac resynchronization therapy (CRT) at one or more anatomical locations using the patient-specific computational heart model, wherein the computational heart model includes a cardiac electrophysiology module that simulates cardiac electrophysiology, a cardiac biomechanical module, and a cardiac boundary conditions module and the simulating CRT at one or more anatomical locations using the patient-specific computational heart model comprises:

simulating heart function with the patient-specific computational heart model with a stimulated current introduced at the one or more anatomical locations in the cardiac electrophysiological module of the computational heart model.

23. The non-transitory computer readable medium of claim 22 , wherein the operation of generating a patient-specific anatomical model of left and right ventricles from medical image data of a patient comprises:

detecting a patient-specific left ventricle model and a patient-specific right ventricle model in the medical image data; and

fusing the left ventricle model and the right ventricle model into a single bi-ventricular volumetric mesh.

24. The non-transitory computer readable medium of claim 23 , wherein the operation of generating a patient-specific anatomical model of left and right ventricles from medical image data of a patient further comprises:

generating a model of fiber orientation based on the bi-ventricular volumetric mesh.