US 20040049115 A1
One embodiment of the present invention discloses a computerized method of facilitating cardiac intervention, comprising inputting patient data, creating a computerized interactive model of a diseased heart based on the patient data, wherein the model comprises structural elements, simulating at least one proposed cardiac intervention treatment by adding or deleting structural elements to the model, and determining the effects of the proposed cardiac simulation upon the entire model. The simulating step may be repeated to allow the user to determine an optimal cardiac intervention. Additionally, a template may be created from the model to use as a guide during the cardiac intervention.
1. A computerized method of facilitating cardiac intervention, comprising:
inputting patient data,
creating a computerized interactive model of a diseased heart based on the patient data, wherein the model comprises structural elements,
simulating at least one proposed cardiac intervention treatment by adding or deleting structural elements to the model, and
determining the effects of the proposed cardiac simulation upon the entire model.
2. The method of
3. The method of
4. A method of facilitating cardiac intervention, comprising:
obtaining patient data,
preoperatively determining a diseased portion of the left ventricle,
creating a template to identify the diseased portion of the left ventricle, and
using the template in cardiac intervention procedures.
5. The method of
6. The method of
7. The method of
8. A computerized method of facilitating cardiac intervention, comprising:
inputting patient data,
creating a computerized interactive model of the heart from the patient data, wherein the model is comprised of structural elements, and
manipulating a portion of the structural elements to determine the effects of the manipulation on other structural elements.
 This invention claims the benefit of and incorporates by reference the following U.S. provisional patent applications: Serial No. 60/288,173 filed on Apr. 30, 2001, Serial No. 60/307,218 filed on Jul. 20, 2001, Serial No. 60/318,024 filed on Sep. 7, 2001, and Serial No.60/357,559 filed on Feb. 15, 2002.
 This invention relates generally to systems for identifying the structural elements that contribute to the cardiac performance of an individual patient through the use of imaging methods, and in particular to a computerized system and method for facilitating cardiac intervention methods.
 The circulatory system of a human works as a closed system where the effects of one part of the system are felt by all other parts of the system. For example, if a person's blood pressure rises then there is a corresponding pressure decrease in the venous system, the decrease is much smaller than the increase in the arterial side because of the fact that venous vasculature is more compliant than the arterial vasculature. Within the circulatory system the key component is the heart. Any change to any component of the heart will have an effect felt throughout the entire system.
 The function of a heart in an animal is primarily to deliver life-supporting oxygenated blood to tissue throughout the body. This function is accomplished in four stages, each relating to a particular chamber of the heart. Initially deoxygenated blood is received in the right auricle of the heart. This deoxygenated blood is pumped by the right ventricle of the heart to the lungs where the blood is oxygenated. The oxygenated blood is initially received in the left auricle of the heart and ultimately pumped by the left ventricle of the heart throughout the body. It can be seen that the left ventricular chamber of the heart is of particular importance in this process as it pumps the oxygenated blood through the aortic valve and ultimately throughout the entire vascular system.
 A myocardial infarction (heart attack) will not affect two people in the same manner. The extent of the damage due to the infarction will be based on many factors, such as; location of the infarction, extent of collateral flow in the blockage area, health of the heart prior to infarction, etc. The unique damage will have a corresponding unique effect on the entire cardiac system. The infarction damage in one patient may be isolated to a small section of the ventricle wall. In another person the infarction may involve not only the ventricle wall but also the septum. In still another person the infarction might involve the papillary muscles. Over time these unique damages will cause the heart to respond in different ways to compensate for the damage in an attempt to keep the system operating as best it can.
 Various treatments are currently employed to repair, replace or mitigate the effects of damaged components of the heart. Some of these treatments involve grafting new arteries onto blocked arteries, repairing or replacing valves, reconstructing a dilated left ventricle, administering medication treatment or implanting mechanical devices. All these treatments apply standard repairs to unique problems with a minimum of analysis as to what the optimum intervention should involve. None of the current procedures involve analyzing the performance of the cardiac system after the treatment to see what effect the treatment has had on the entire system. For example, a patient with blocked arteries may undergo the placement of 5-6 grafts on his heart due solely to a short visual inspection of angiographic films that show some stenosis of the arteries of the heart. No analysis is performed to see if placing 3-4 grafts will achieve the same perfusion of the myocardium as the 5-6 grafts. It is simply a situation where the doctor decides that more is better, which may not be true. Placing 5-6 grafts requires more surgical time, longer pump runs and incisions into numerous areas of the body to recover the needed grafts. This increases morbidity to the patient and may contribute to death of the patient who cannot tolerate the additional stress of a longer, more invasive procedure. On some patients the extra grafts may be needed, since collateral flow or flow from other arteries is not sufficient to perfuse the entire myocardium, while on other patients the grafts are not needed, since sufficient flows will be generated from fewer grafts. Currently, the doctor has no way of knowing if the total number of grafts that he put in was necessary.
 A similar procedure is used to place stents in a vessel. Stents are placed in vessels based on an assessment of blockage and ability to access the obstructed area. No method of analysis is performed to determine the effects of placing a stent, or to analyze how many stents should be placed where, or to determine if stenting produces a better result than bypassing.
 The current process for repairing and replacing valves also heavily rely on the doctor's knowledge and intuition. There is no precise way to determine how much a valve or structural component needs to change or what the effect of that change will be, The current procedure for determining if the correct repair was made is to complete the repair, remove the patient from cardiopulmonary bypass and let the heart start beating. When the heart's performance reaches a normal range, an echocardiography is taken of the valve to ensure that it is not regurgitant. If the repair left some regurgitation, then the patient must go back on cardiopulmonary bypass, the heart must be stopped again, reopened and a repair made to the repair or replacement. The checking procedure is repeated after the second repair to ensure that the procedure has been correctly done. This procedure subjects the patient to unnecessary risks by exposing them to longer than necessary bypass runs and reperfusion injuries each time the heart is weaned of cardioplegia, and takes up valuable operating room and staff time. The repairs to repairs scenario for valve procedures is not uncommon. Additionally this assessment method only assesses one factor related to the performance of the valve and ventricle, regurgitation. A doctor may perform a procedure, which corrects the existing problem, but creates another problem or diminishes the performance of the ventricle. The doctor has little if any way of know if he compromised ventricle performance, since current analytical tools only look for flow across the valve. Methods to identify and evaluate the positioning of the valve apparatus and the attached tissue and their combined performance are nonexistent.
 Similarly, there is no method to determine when to replace or repair a valve. This is left to the judgment of the doctor based on review of two dimensional echocardiography studies. Doctors who are unfamiliar with repair techniques may opt for replacement when repair is not only possible but also when repair is the best course of action for the patient. The replacement will be done without knowing what effect it will have on the other elements of the mitral valve apparatus, left ventricle, left atrium and the overall functioning of the heart. A replacement that attaches the chordae tendinae to the new valve may have a much different effect on the ventricle than a replacement that excludes the chordae tendinae. Currently there is no method to assist the doctor in making this assessment. Repairs are typically undertaken to shorten the chordae and annulus without knowing what effect the repairs will have on the entire apparatus. The current solution is experimentation on the patient in real time; make the repair and let the heart beat to see what the repair has done.
 What is needed therefore is a reliable method and apparatus to allow a doctor to determine which elements of the heart are not contributing to, or are decrementing from, the performance of the heart, and a method and apparatus to allow the doctor to simulate the treatment on a portion of those elements and see the effect the treatment has on the other elements and the heart as a whole.
 One embodiment of the present invention discloses a computerized system and method of facilitating cardiac intervention, comprising inputting patient data, creating a computerized interactive model of a diseased heart based on the patient data, wherein the model comprises structural elements, simulating at least one proposed cardiac intervention treatment by adding or deleting structural elements to the model, and determining the effects of the proposed cardiac simulation upon the entire model. The simulating step may be repeated to allow the user to determine an optimal cardiac intervention. Additionally, a template may be created from the model to use as a guide during the cardiac intervention.
 Some embodiments are directed to the preoperative analysis of a patient's heart condition and computer assisted manipulation preoperatively of the patient's heart to perform procedures such as coronary artery bypass grafting, stent placement, surgical ventricular repair, valve repair and replacement and numerous other procedures some involving the use of implantable devices.
FIG. 1 is flowchart illustrating one embodiment of the present invention.
FIG. 2 is flowchart illustrating one embodiment of the present invention.
FIG. 3 is flowchart illustrating one embodiment of the present invention.
FIG. 4 is flowchart illustrating one embodiment of the present invention.
FIG. 5 is flowchart illustrating one embodiment of the present invention.
FIG. 5A: Illustrates sectional views of MRI and Echocardiography (long axis).
FIG. 5B: Illustrates sectional views of MRI and Echocardiography (short axis).
FIG. 6: Illustrates the heart in various stages of the cardiac cycle.
FIG. 7: Illustrates a comparison of systole and diastole images of the ventricle to show effect of wall thickening
FIG. 8: Illustrates a comparison of systole and diastole images to determine border zone between akinetic and functional tissue.
FIG. 9: Illustrates a model created from MRI images.
FIG. 10: Illustrates a model with a finite element grid.
FIG. 11: Illustrates the different structural elements of the heart.
FIG. 12: Illustrates the interactive features of the apparatus.
FIG. 12A: Illustrates making an incision into the heart.
FIG. 12B: Illustrates placing sutures and opening the incision in the ventricle.
FIG. 12C: Illustrates the sizing and shaping device in the ventricle.
FIG. 13: Illustrates different types of surgical manipulations.
FIG. 13A: Illustrates a Fontan Stitch—creation of neck for placement of the patch (FIG. 13L shows with patch).
FIG. 13B: Illustrates suture placement to imbricate stretched tissue.
FIG. 13C: Illustrates anastomosis
FIG. 13D: Illustrates Mitral Valve with insufficiency and Valve after it is corrected with annuloplasty ring.
FIG. 13E: Illustrates placement of a Myocor splint.
FIG. 13F: Illustrates completed bypass grafts.
FIG. 13G: Illustrates a mechanical heart valve.
FIG. 13H: Illustrates an Acorn Corcap.
FIG. 13I: Illustrates a linear closure of an opening.
FIG. 13J: Illustrates a buttress suture.
FIG. 13K: Illustrates reforming the ventricle to give a new volume.
FIG. 13L: Illustrates placement of a patch to close an opening in the ventricle.
FIG. 13M: Illustrates tightening the mitral annulus with a suture.
FIG. 13N: Illustrates replacing an aortic valve.
FIG. 13O: Illustrates repairing a mitral valve—excising a portion and regrafting the leaflets.
FIG. 14: Graphs of physiological functions of the heart.
FIG. 14A: Illustrates a Frank-Starling curve.
FIG. 14B: Illustrates pressure volume loops.
FIG. 15: Illustrates hemodynamic model.
FIG. 15A: Illustrates outputs from a hemodynamic model of the heart and circulatory system by Professor Ying Sun, of The University of Rhode Island.
FIG. 16: Illustrates a mesh that has anatomical landmarks of the heart and the location of the diseased tissue superimposed on it.
FIG. 17: Illustrates a sizing and shaping device with the location of the diseased area of the ventricle marked on its surface.
FIG. 18: Illustrates a pre-cut shape to allow the doctor to identify on the heart the diseased tissue.
FIG. 19: Illustrates various potential patches of different sizes and shapes to seal an opening in a ventricle.
FIG. 20: Illustrates a patch that has apical shape.
FIG. 21: Illustrates a patch with fibers that have strength in one axis different from the strength in the other axis.
 The methods and apparatus of various embodiments of the present invention will be described generally with reference to the drawings for the purpose of illustrating the present preferred embodiments of the invention only and not for purposes of limiting the same. The illustrated embodiments address the ability of the doctor to accurately assess the effects of cardiac disease on an individual patient and to use an appropriate treatment to restore the cardiac system to its optimal or best acceptable condition. In one embodiment, this is accomplished by using an analytical tool that takes images of the patient's own heart as in FIGS. 5A, 5B and other data and converts them to a multidimensional finite element model, such as illustrated in FIG. 10, which may interact and respond to other models or a set of models. The model of the heart FIG. 10 may also be connected to a model of the circulatory system and a model of the cardiac system. These models may simulate the performance of the heart and its effect on the circulatory system. This is a procedure that determines the appropriate areas to be repaired or replaced or otherwise medically treated for each individual patient uniquely, and determines the effects that the treatment may have the structural element treated, the other structural elements of the heart and on the entire heart.
 One embodiment of the present invention is a system and method for capturing the geometry of the heart and its components using imaging technologies such as, but not limited to, MRI imaging, echocardiography, or PET. Turning now to FIG. 1, in step 10 patient data is acquired. Some other factors and information that may be captured inlcude:
 a. Myocardial stiffness
 b. Ventricle wall thickness
 c. Heart rate
 d. Ventricle wall tension
 e. Right and left ventricle volumes
 f. Mitral Valve Annulus
 g. Chordae Tendinaea
 h. Papillary Muscles
 i. Mitral Valve Leaflets
 j. Ventricle Endocardium Border
 k. Ventricle Epicardium Border
 I. Aortic valve annulus
 m. Aortic valve cusps
 n. Tricuspid valve apparatus
 o. Pulmonary valve apparatus
 p. Ventricle wall thickness
 q. Ventricles areas of akinesia
 r. Ventricle areas of dyskinesia
 s. Ventricle areas of asynergy
 t. Ventricle preload
 u. Ventricle filling pressure
 v. Heart's arterial system
 w. Heart's flow through the arterial system
 x. Heart's venous system
 y. Left and right atrium volumes
 z. Left and right atrium wall thickness
 Some or all of these factors may be used to create a multi-dimensional finite element computer model of the heart (step 11). One example of a multi-dimensional model is a three-dimensional model that displays not only the three dimensions of the geometry of the heart but may also depict this geometry as it changes over time. Another dimension may be physiological factors, for example the heart produces a hormone B-type natriuretic peptide in reaction to increased wall stress. This production of the hormone could serve as another dimension to the model. Software producing the model may run on a personal computer type of machine or it may run at a central location or it may be processed at one location and delivered to another location. Such a four dimensional model may allow the doctor to visually inspect the status of all the elements of the heart. This model may be used to determine a variety of information, either pre-treatment, during the treatment or post-treatment, including, but not limited to:
 a. The areas of the mitral, aortic, tricuspid or pulmonary valves that may need to be repaired or replaced and what affect each repair may have on the other components.
 b. What vessels are blocked and may need to be grafted, where to graft and what effect the revascularized muscle may have on the other components.
 c. What areas of the ventricle are akinetic, dyskinetic or hibernating, to show what areas may be excluded during ventricular restoration and what effect the exclusion may have on the other components and aspects of the ventricle and heart.
 d. How this patient's heart may respond to medication treatment.
 e. The effects of placement of a corecap restraining device, Myosplint shape changing device or other device on the outside of the ventricle may effect the heart.
 f. The effects of chordae length adjustment or papillary base relocation may effect the heart.
 g. The effects of placement of any ventricular assist device may have on the heart.
 h. The vessels that are blocked and may need to be stented, where to stent and what affect the revascularized muscle may have on the other components.
 The model may then allow the doctor to select a treatment option (12) and allow the doctor to manipulate the image and model (13). The model may then analyze what effects his virtual treatment may likely have on the cardiac geometry (14), calculates predicted outcomes based on physiology and hemodynamics (16) and displays the potential clinical outcomes to the doctor (17). The potential outcomes display may be but are not limited to the following:
 a. The estimated performance of the valves and ventricle after the procedure; i.e. regurgitation, reduced flow across the valves, ejection fraction etc.
 b. The flow through the grafts or stents and what areas of the myocardium the grafts or stents may perfuse.
 c. The volume and contractile state of the ventricle after excluding tissue.
 d. The positioning and performance of the valve apparatuses after reconstruction of the ventricle.
 e. The effects that a drug or combination of drugs may have on the entire heart.
 The doctor may then be able to select the displayed intervention (18) or decide to try another treatment or modify the current intervention (19) and the cycle may repeat itself. When the doctor accepts the potential clinical outcomes, the model may then produce the specifications for the intervention (20). These specifications may lead to the development of a template or tools or devices to guide the doctor in translating the virtual intervention on the model to the actual intervention on the heart (21). In some cases templates, tools or devices may not be needed to perform the intervention and specifications such as the length of a chordae tendinae may be sufficient output from the model to allow the doctor to perform the intervention. Additional devices may be generated from the models to help the doctor implement the procedure that the model may have predicted to provide the best outcome. Furthermore, the use of some or all of above listed factors may be used to evaluate post-treatment the condition of the patient. A database of surgical pathologies, treatments and outcomes may be gathered, maintained and analyzed to further refine the treatment of cardiac diseases and disorders.
FIG. 1 describes a method and apparatus for performing cardiac valve correction planning. The procedure acquires imaging information, such as, but not limited to, MRI, PET or echocardiography imaging data of the patient's ventricle (10). These imaging systems are common in most hospitals and the leading manufacturers of these systems are General Electric, Siemens and Phillips. Other information such as but not limited to stiffness, wall thickness, heart rate, wall tension, right ventricle volume, valve apparatus locations and epicardium and endocardium borders may be needed to convert the data to a multi dimensional model of the heart.
 The imaging data is often acquired as sectional views (FIGS. 5A, 5B), one way of combining these sectional views and converting them into a model may be done by overlaying the sectional view on a XY grid. FIG. 9 shows diastole picture of the heart in long axis along plane M with the grid superimposed. The points of intersection of endocardium (and epicardium) with the grid are identified in XY coordinates. Similarly XY coordinates are identified of all the planes. Since the angular relationship between each plane is known (angle θ, in FIG. 5A), all the data points can be converted into XYZ coordinates. The boundary layer generated by connecting the internal point Pi1,2,3 . . . defines the endocardial boundary, and the boundary layer generated by connecting the external points PO1,2,3 . . . defines the epicardial boundary. This defines the heart in a three dimensional space. Once the three dimensional model is created the time frame of the cardiac over which all the images were made can be added to show the heart move in time during its cardiac cycle (four dimensional model).
 Once the multi dimensional object is defined, it can be converted to elements of a finite element model and a finite element mesh that represent the heart and its components to create the model (FIG. 10). Some of the components of the heart that may be identified as different structural elements of a finite element model are listed below but the apparatus and method is not limited to these components;
 These elements may have different structural properties. The structural properties of myocardium and other cardiac structures can be obtained from various sources in literature like, the properties of the ventricle myocardium can be found in, J. M. Guccione et. al., “Passive Material Properties of Intact Ventricular Myocardium Determined from a Cylindrical Model, Journal of Biomechanical Engineering Vol. 113, February 1991. Once all the structures are geometrically defined and structural properties are known, a finite element model can be created. The general creation of finite element models is well known in the art. A method of converting a defined object to a finite element mesh is describes in U.S. Pat. No. 5,892,515, and is hereby incorporated by reference. “Finite element analysis” is a mathematical approach to solving large (complex) problems. Generally the subject is segmented or discretized into many pieces that have closed form solutions. That is, each piece is definable by a linear equation, and hence is a “finite element. Collectively, the linear equations of the pieces form a system of equations that are simultaneously solvable. Computer programs for simulating finite element analysis in various applications exist; for example, design engineers use finite modeling programs. Typically many thousands of elements are created to model a subject object and in particular three-dimensional objects. For each element, there is geometric information such as an (x, y, z) coordinate at a point in the element, an element type, material property, stress value, displacement, thermal value, etc. Such information is definable by linear equations for the elements. To that end, finite analysis is employed to model the subject object. Examples of finite modeling programs include: ABAQUS by Hibbitt, Karlsson, and Sorensen, Inc. of Pawtucket, R.I., ANSYS by Swanson Analysis Systems Inc. of Houston, Pa: SUPERTAB by Structural Dynamics Research corp. of Ohio; and PATRAN by PDA Engineering of Costa Mesa, Calif.
 Once a finite element model has been created, an image of the heart and some of its structural elements may appear on a monitor to allow the doctor to interact with the model. An image as illustrated in FIG. 10 may be displayed along with relevant data on the state of the heart for example left ventricle volume, blood pressure, ejection fraction, heart rate. The image may be interactively connected to the model (11) to allow the doctor to simulate the effects of the treatment before it is administered. For example, a pull down menu that is commonly used in many software applications like word processing software or CAD software can be accessed to select the type of treatment desired (12) surgical ventricular repair, bypass grafting, mitral valve repair etc. For example, a doctor can select the mitral valve option to shorten the chordae tendinae or tighten the mitral annulus. In the chordae tendinae example, the model may separate the chordae elements from the entire model and present it to the doctor, to allow the doctor to interact with the elements. This interaction can come in various forms. A pull down menu standard to most software programs could present the doctor with a list of options, such as selecting the type of scalpel to use, the type of suture material etc. The physical characteristics of these implements can be entered into a database (22) that the model can access. Once the doctor has selected the implement to use a box or another pull down menu can appear asking for further information on how to use the implement. For example, with a scalpel the box will ask the doctor how long and how deep he wants to make the incision. The doctor will then be asked to identify by click with a mouse or stylus the start and end points of the incision. When these steps are complete an incision may appear on the model corresponding to the input of the doctor and sized appropriately for the heart according to the characteristics of myocardium etc. that are built into the finite element model (14). Methods to model the physical properties of the heart exist to create the manipulation portion of the model. A method to create a finite element model of the heart is written about by K. D. Costa et. al., “A Three-Dimensional Finite Element Method for Large Elastic Deformations of Ventricular Myocardium: I-Cylindrical and Spherical Polar Coordinates, Journal of Biomechanical Engineering, November 1996, Vol. 118 pp. 452463. The physical properties of the elements of the heart on which to base the finite element equations for the structural elements can be found in, Hunter P. J., et. al., “Modeling the mechanical properties of cardiac muscle”, Progress in Biophysics & Molecular Biology 69 (1998) pp. 289-331. Modeling the diseased areas of the left ventricle has been described in Rez Mazhari, et. al., “Integrative Models for Understanding the Structural Basis of Regional Mechanical Dysfunction in Ischemic Myocardium”, Annals of Biomedical Engineering, Vol. 28, pp. 979-2000. The properties of the ventricle myocardium can be found in, J. M. Guccione et. al., “Passive Material Properties of Intact Ventricular Myocardium Determined from a Cylindrical Model, Journal of Biomechanical Engineering Vol. 113, February 1991. Once the doctor has shortened the chordae the model presents the image of the new shorter element and presents an image of the other elements with the effect that the shortening of the chordae has had on them along with clinical outcomes (16)(17). The doctor may save the results of the first intervention and repeat the procedure in a different manner (19) to compare the outcomes of different interventions. The doctor can then select the optimal outcomes (18) and perform the procedure in that manner. Optimal outcomes may be based on a variety of cardiac performance parameters. They may include, ejection fraction, end systolic volume, stroke volume index, cardiac output, mitral regurgitation, pulmonary artery pressure, mean arterial pressure, percentage of asynergy etc. Optimal outcomes are very doctor dependent, some doctors may prefer higher ejection fraction and may be willing to tolerate slight mitral regurgitation. Other doctors will tolerate no mitral regurgitation and accept a lower ejection fraction to achieve no regurgitation through the mitral valve. When the doctor is satisfied that the intervention is the optimal possible for this patient, he accepts the intervention and the model will produce specifications to assist the doctor in performing the intervention (20). In this example the specifications may be simply a display of the final length of the chordae. In more complicated procedures the specifications may result in the production of patient specific devices, which will assist the doctor with translating the virtual intervention to an actual intervention on the patient. The patient specific devices can be simple variations to the existing devices like customized annuloplasty ring or they can be more complex devices like prosthetic mitral apparatus. With the information provided by the model the doctor can proceed with the intervention assured that the result will be the optimal possible (21).
 A separate model or models may be able to determine the clinical outcomes of the procedure. For example the physiological and hemodynamic conditions of the heart may be modeled. The physiological properties of the heart are well understood, the Frank-Starling curve and the law of Laplace etc., and are written about in numerous publications to include Hurst et. al., Hurst's The Heart, McGraw-Hill, 1998, excerpts are FIGS. 14A, 14B. Frank-Starling curve varies from heart to heart based on various factors, like contractility, wall stress, sphericity index, diseased state etc. The curve that best matches a given patient can be obtained by comparing the patient specific characteristics to those of other patients in a CHF database 33 (FIG. 2).
 A hemodynamic model has been developed and published by Professor Ying Sun, et. al., “A comprehensive model for right-left heart interaction under the influence of pericardium and baroreflex, The American Journal of Physiology, 1997, pp. H1499-H 514, FIGS. 15A, 15B. These two models may interact with the finite element model to show the doctor what effect his interaction has had on the other elements and the whole heart. The physiological models may vary from very simple such as an equation of a curve of Stroke Volume vs. End Diastolic Volume as in the Frank-Starling curve, to much more complicated computational biology models. The hemodynamic models may also vary from simple models of the pressure drop vs. flow relationship to complex computational flow dynamics like the one published by Makhijani et. al. “Three-dimensional coupled fluid—Structure simulation of pericardial bioprosthetic aortic valve function” ASAIO Journal 1997; 43:M387-M392.
 As another example, the placement of an annuloplasty ring may be simulated to show its effect on the annulus, connected tissue and ventricle, FIG. 13D. The patient's heart will be imaged (10) and the image converted to a finite element model (11). The software may allow the doctor to select the type of treatment desired (12) and access a database to select the device to be used (22), in this case an annuloplasty ring. The model may then display to the doctor the mitral valve and allow him to instruct the model on where to place the ring, which suture to use in securing the ring, how much tension to put on the sutures, distance between each bite etc (13). The model may then apply this intervention to the mitral valve annulus and the other elements of the mitral valve and the other components of the ventricle and the heart as a whole (14). Other data, if necessary, is pulled into the equation (15) if needed. The model may recreate the image on the monitor to show the doctor the effects of his interaction (16). The potential clinical outcomes (17) may be determined by the model through interaction with the physiological and hemodynamic models FIGS. 14, 15. This simulation may show the ring's effect on the size and orientation of the annulus as well as the effect the ring may have on the connected tissue, i.e. does it affect the length of the chordae tendinae, shape of the ventricle, etc. The model may be analyzed to show the surface area of the opening of the shortened annulus, how much flow may come through that opening and how the change in flow may affect the ventricle. The model may predict if there is a mitral valve prolapse. A database of medical devices FIGS. 13C, 13D, 13E, 13G, 13H can be created and accessed to allow the simulation of these devices. These devices can be tested for physical properties and these physical properties encoded into a finite element model, as has been done for elements of the heart described above. The finite element models for the devices are stored in the database (22) and accessed by the doctor by selecting the object by its common name. For example, prosthetic valves and prosthetic valve apparatus (mechanical and bioprosthetic) may be called upon to place different artificial valves into the heart, and then the performance of the heart with the different valves may be assessed to select the correct valve for this patient. The model might also give estimated values of post-surgery performance of the heart. It might display estimated ejection fraction, regurgitation, sphericity of ventricle, volume of the ventricle, percentage of shortening on the long and short axis, and maximum and minimum flows across the valves, tension in chordae etc. In some instances it is likely that off the shelf devices do not provide optimum results. For example the annuloplasty rings comes in various sizes, it is likely that for a given patient when a smaller size is used it may end up creating more than acceptable tension in the chordae, while going to next size up can lead to mitral insufficiency. In such situation the model may come up with a specification for the ring that falls between those two sizes, which offers the best possible outcome for that patient.
FIG. 2 depicts another method and apparatus for treating cardiac disease. Images of the ventricle are taken (23) and a finite element mesh model is created of the ventricle and of the structural elements as described previously (24). The doctor chooses a treatment option (25), surgical ventricular repair. The doctor using pull down menus, or another standard interactive means chooses the implements that are need to perform this procedure (26). The doctor performs the treatment by interacting with the image and the model (27). Interacting with the model, the doctor selects a scalpel, identifies where and how to incise the ventricle. He then identifies the tissue he wants to exclude and places a Fontan stitch FIG. 13A. When the doctor excludes tissue the model eliminates the sections of the finite model that correspond to this area from the calculations of the ventricle parameters and outcomes. The model may keep these elements solely as graphical depictions. The model may try various degrees of volume reduction of the ventricle FIG. 13K and changes in the shape of the ventricle. The finite element model calculates this change in shape of the ventricle and also calculates how this change has affected the other structural elements of the ventricle and the heart. As the model reshapes the ventricle to make it smaller, it may show the effect this has on the other structures like the mitral apparatus. The model may show the new location of the papillary muscles, new angle of the chordae tendinae to the mitral annulus, etc. The finite element model may use known methods described previously to calculate the reaction of different structural elements to changes in another element. For example, the geometric alterations may in turn have effects on various other cardiac performance characteristics i.e. smaller ventricles may have lower wall stress and can result in better contractility. The model can then prompt the user to choose a patch to cover the opening that may be left in the ventricle and to reinforce the septum, FIG. 13L. If the opening in the ventricle is small, less than 3 centimeters, the model may tell the user to close the opening in the ventricle without a patch. The user can identify the suture placement locations as described previously and specify the amount of tension to be placed on the sutures. The model may depict the opening being closed with these sutures. The model accomplishes this by taking the boundary layers at the edge of the opening and moving them towards each other. When the boundary layers meet, the model recalculates the finite element model shapes that should depict this closure area. For example if the finite element model is made of triangles the triangles on the boundary layer may be smaller than the average triangle in the model. When the two smaller triangles on the boundary layers meet at the closure line the smaller triangles may be combined into one average sized triangle. The finite element model may then interact with the outcomes predictor (30). The outcomes predictor may be composed of but not limited to a hemodynamic model FIG. 15 (30 a), a physiological model FIG. 14 (30 b) and other calculations (30 c). These models may interact until the physiological and hemodynamic models are within tolerances of know physiological and hemodynamic constraints, or balanced (32). The acceptance criteria may be SVI (stroke volume index) to be between 22 to 50 ml/mt2 and that PAP (pulmonary artery pressure) to be within 10 to 25 mmhg, and ejection fraction to be above 30% and ESVI (end systolic volume index) to be between 25 and 60 ml/mt2. If after 50 attempts, for example, the models cannot become balanced; the model may ask the doctor to alter his intervention. Once the models are balanced, the model may display the ventricle with the new shape and volume to the doctor along with potential clinical outcomes such as ejection fraction, mitral regurgitation etc. (34)(35). The doctor can then accept these clinical outcomes (36) or return to the original model and image (37) and try a new treatment or modify the initial treatment. The doctor may perform multiple iterations of the procedure and compare clinical outcomes to determine which procedure is optimal for the patient (36). When the doctor accepts the intervention that is optimal for the patient, the model may then create specifications to help the doctor translate the simulated intervention to an actual procedure (38). The model may determine the size, shape and volume of the ventricle desired and a unique shaping and sizing device may be created for the patient from this information to assist the doctor in performing the procedure (39).
 A method to make a custom sizing and shaping device is by generating a 3D CAD file (DXF or STL formats) that has the outline of the interior of the ventricle and load this file into a CNC milling machine. This machine may take the file and create a 3-Dimensional mandrel from the file. This mandrel can then be dipped in a number of solutions such as plastisol and urethane to form a pliable balloon like object that can be taken off the mandrel. A cap of similar material can be added to the top and a tube for filling the shaping and sizing device with fluid can be added, FIG. 17. The best solution to reconstruct the ventricle may require the use of a patch to reinforce the septum and/or close a hole remaining in the ventricle. The model may be able to show the doctor what shape patch may be needed to perform this task and a specially constructed patch may be made for this patient. A method to manufacture this custom patch could be to purchase cardiovascular patches currently sold by Boston Scientific/Meadox, or W. L. Gore, for examples. The model can generate a CAD file defining the shape of the opening in the ventricle. The shape of the opening can be printed and used as a template. The template could be placed on the patch and the patch cut to the shape and then sterilized FIG. 19. The model may lead to other tools that help the doctor implement the solution that the model has created like a patch with an apex etc FIG. 20.
 Other way of doing an SVR procedure is to start of with a desired volume of the ventricle and selecting a ventricle sizer. The model may interact with the computational model of the ventricle sizer. The operational steps are similar to those mentioned in the earlier paragraph, except that the ventricle is formed over the ventricle sizer. The output of the model in this case can be a patient specific unique shaped patch that is needed to perform the intervention.
 Similarly, the model can interact with finite element models of many currently marketed devices such as but not limited to the Myocor Inc Myosplint FIG. 13E, the Acorn Inc Corcap FIG. 13H or biventricular pacing from either Medtronic or Guidant may be made into a model. In each case the model may produce outcomes of the intervention with these devices. If the doctor likes the outcomes then specifications can be produced in order to transfer the results of virtual surgery to real surgery. In some instances specific tools or devices can be generated. The doctor takes these tools, devices and or specifications and conducts the procedure (39)
 This method and apparatus may also be used in an automatic mode FIG. 4. A doctor could simply input a desired outcome or outcomes such as a defined ejection fraction range, ventricle volume range etc. The software may then run numerous iterations of all the different types of treatments and produce expected treatment options that meet defined criteria for that particular patient. The results may be ranked to allow the doctor to select the best treatment with the best outcome. The software may also just run and supply the best possible outcome without any input from a doctor besides the required data to run the software. The software may again present the doctor with expected outcomes prioritized. The software may report to the doctor that the desired outcome from a specific treatment is not possible and thereby force the doctor to reconsider his selection criteria options.
 Before treatment, in order to determine which areas of the heart may need to be repaired or replaced, the patient may undergo an imaging procedure such as an MRI scan, PET scan or an Echocardiography scan to determine the location and condition of the components of the heart (10). The patient's current ventricular anatomical landmarks may be determined by manually tracing the epicardium and endocardium or it may be done by automated border detection software, which may quickly outline the location of different structures within the ventricle from the imaging data. This scan, with the borders delineated, is converted into a multi dimensional picture of the heart and may include all valves, arterial and venous structures of the heart (11). Parts of the valve apparatus, which may not fully appear with the automated border detection software (chordae tendinae) for example, may be quickly hand traced to complete the four dimensional dataset. The multi-dimensional image may also show regurgitation across the valves using different color gradients to show severity, as is currently done with echocardiography.
 Post-treatment imaging such as MRI, PET and echocardiography scanning of the above listed measurement points may show the doctor how well the patient has done in treatment. The images of the patient's heart before treatment, the models depiction of the treated heart with performance characteristics can all be saved in a database. The doctor can compare the actual data with the predicted and determine how to improve his technique to achieve the theoretical best results. Long-term follow up is enhanced when current images of the heart can be compared to pre- and post-treatment images of the heart. These images can be analytically compared for small changes in the heart's geometry and alignment. If small changes are detected early, less invasive measures can be taken to stop or slow the progression of the abnormality. The surgeons can also use this database to pull up data on past patients who may have similar characteristics as the current patient under consideration, and compare his current treatment options to the past ones. Such methods may further contribute to improvement of techniques.
 One of the problems surgeons confront while doing SVR procedure is to determine, the demarcation line between viable and akinetic tissue. For this purpose a non-interactive model, which can show the location of diseased area of the ventricle can be used. The model may show on the image which areas of the ventricle are akinetic or dyskinetic to determine which areas might be excluded during SVR procedure. One method of doing this is to take the images from MRI or echocardiography FIG. 5. These images are a combination of sections of the heart imaged during one cardiac cycle, so that each section contains a complete cycle FIG. 5. These slices are combined to create one image. The images at the end of systole and the end of diastole are then identified, FIG. 7. These images are overlaid by aligning markers that don't move such as the aortic valve annulus and a grid pattern is then superimposed on these images, FIG. 8. Each intersection of the grid that intersects the epicardium and endocardium is identified. The geometric center of the heart is calculated and imaginary lines (rays) are drawn from this center. Two points on each ray are recorded; the points are defined as point of intersection of the ray to the endocardium and epicardial boundary. The distance between these two points gives the wall thickness (d). Wall thickness is calculated on the diastole image dd and on the systole image ds. Normally ds>dd when the heart functions normally, that is because the myocardial wall thickens during systole to create pumping action. If a section of the heart muscle is diseased then ds=dd, meaning that portion of the wall is not thickening, it is referred to as akinetic tissue, it could either dead or non-contributing tissue. All the rays that correspond to akinetic tissue are identified (all rays where ds=dd). The boundary layer of the akinetic area is then established by comparing each of the akinetic rays to its neighboring rays. For any given akinetic ray, if at least one of its neighboring rays is kinetic (ds>dd) then that akinetic ray is the boundary layer ray. Once all rays on the boundary layer are identified, the point of intersection of the boundary layer rays on the endocardial boundary defines the border zone between the viable and akinetic tissue.
 Once the location of the diseased section is identified with respect to other cardiac structures, a 3D CAD file (DXF or STL files) may be generated which shows the location of the border area with respect to a known landmark on the heart. One may then create a template that may match the diseased area and have anatomical landmarks from the heart such as Left Anterior Descending artery or the Atrial ventricular groove to ease alignment of the template to the diseased area. The template may be in a form of a balloon that is patient specific with the same shape and size as the interior of the ventricle, and with border zone marked on it or it can be a like a glove that fits on the outside of the heart with border zone and landmark points marked on it. Such tools may be very helpful in order execute SVR procedure with greater precision.
 Another method to determine the diseased area of the ventricle is to measure the motion of the endocardium towards a centerline of the ventricle. This is popularly referred to as “centerline method” it determines the region of no motion by evaluation how much motion at 90 points of the ventricle the motion differs from the standard motion. In the centerline method any tissue that moves less than 2 standard deviations from a typical movement level of normal heart is considered diseased. This algorithm could be applied in the above-mentioned model to identify the border zone. The model may generate an image using different color gradients to depict the range of lack of motion from the standard. This color grading may give the doctor a precise location for tissue to exclude and may give assurance that the doctor will not exclude any viable tissue. A template showing the status of the myocardium stated above may be provided to the doctor to use as an aid in excluding the tissue. The gradient image may be used for both idiopathic and ischemic cardiomyopathy patient assessment.
 When the tissue is excluded as described in paragraph 22, there may be a hole left in the ventricle that a surgeon will fill. One device that might cover this hole is a patch that could aid in the contraction of the left ventricle. One form of this patch may be made of a fabric that is pretensioned and stretched to fill the hole left in the ventricle. The pretensioning places stress on the fibers, which assist the ventricle in contraction when going back to their relaxed state during systole. Another variation could be that the short axis fibers are of a different strength than the long axis fibers, thus aiding the greater contraction along the short axis FIG. 21. The patch could have the pretensioned fibers only in the center of the patch, decreasing the tension exerted by the patch on the ventricle walls, but still providing some assistance to the ventricle during contraction.
 Another embodiment of current invention, this apparatus and method may be used to plan for bypass or stent interventions FIG. 3. by showing the location and condition of the arterial system of the heart. Imaging of the arterial system with identification of lesions and blockage has been performed for at least 10 years with ventriculograms. This process injects a dye into the aortic root, which supplies the cardiac arterial system with blood. The dye flows through the arterial system with the blood and can be image with X-rays to identify where the constricted points of the arterial vessels are located. The arterial system may then be mapped and a finite element model applied to the system to determine the width of the vessels, location of constrictions etc. to allow the model to predict how much blood is flowing to each portion of the heart. This may be correlated to displays of viable tissue, so that if the patient has had a myocardial infarction and has dead tissue, the doctor will not use the best graftable conduits FIG. 13F to graft to vessels feeding these areas or place stents on these vessels. Or perhaps the doctor will choose to not graft or stent at all. This model may give the doctor the opportunity to place different grafts or stents on different vessels to analyze the perfusion effect on the heart for the different combinations. The grafts or stent models may come from the database of surgical equipment and devices discussed previously (22). The model may then be run to show the doctor the effect that his grafts or stents may likely have on the entire cardiac and circulatory system, so that he may select the best combination of locations for that particular patient.
 This apparatus and method may also be used to show the effects that interventions on the left vent outflow tract and aortic valve may have on other elements and the entire heart. The outflow tract changes position as people age and an acute angle in the left vent outflow tract may contribute to poor performance of the ventricle and/or the aortic valve. The model may show the positioning of the left vent outflow tract and may show the doctor turbulence or restrictions in blood flow through this area. There are many companies that have developed flow dynamic software. One such model was developed CDFRC (Huntsville, Ala.) and published by Makhijani et. al. “Three-dimensional coupled fluid - Structure simulation of pericardial bioprosthetic aortic valve function”, ASAIO Journal 1997; 43:M387-M392. If desired the doctor may virtually manipulate the left vent outflow tract into different positions and then run the models to see which position of the tract provides the best flow dynamics. This may then tell the doctor if he needs to adjust the positioning of the left vent outflow tract and may show the effects that this new position will have on the performance of the heart. Poor performance of the aortic valve can limit the amount of blood that the ventricle can eject. The model may display the aortic valve and allow the doctor to virtually manipulate the valve and assess if the manipulations have increased the performance of the valve and increased the performance of the cardiac system. The doctor can then take the best results and perform those manipulations on the actual valve.
 The method and apparatus can also be used to simulate the effects of drugs on the heart and its components. A database of drugs and their effects can be developed and the doctor can interact with the model by selecting a type of drug and dosage amount. The model can then give the doctor the results of the treatment, whether it has resulted in a change in the geometry of the heart and its components and if the performance of the heart has improved. For example the model can simulate the effects of vasodilators that can diminish the afterload of the heart or the effect of norepinephrine, which increases the contractility of the heart.
 The method and apparatus may also be used to simulate the placement of mechanical devices in or on the heart to determine the benefits of the devices. The physical and functional characteristics of these devices may be determined through testing and may be reduced to a finite element model. These finite element models may be placed in a database. The doctor will interact with the model by choosing the device by its common name or product name, for example Myosplint or Corecap FIGS. 13E, 13H, he can then direct its placement by methods described above specifying for example, location, attachment means etc. Left ventricular assist devices may also be added to the database. All these mechanical devices may be simulated to show their effects on the whole heart and its components and these effects can then be compared to other less invasive treatments to determine if the increased invasiveness and cost of these devices is warranted by a corresponding increase in the heart's performance.
 The model may be accessed at a central location and the images of pre- and post-treatment images may be stored and categorized by disease type, surgical procedure, outcome, etc. may also be stored at this location. This database may then be used to perform retrospective studies on the efficacy of different procedures and approaches for different disease states and patients. This database and analysis may contribute to the advancement and refinement of models and help improve their probability. The database may also be used to analyze treatments to compare and empirically demonstrate which are the best treatments for certain patients. The database may also allow doctors to compare their results with the database population. The doctor can then see if his selection of and performance of treatment options is better, equal to, or worse than the group as a whole. If he is worse than the group, the surgeon can use the database to help improve his treatment selection making process and his technique.
 In response to these and other problems, an improved apparatus and method is provided for capturing the geometry of the heart and its components using imaging technologies such as, but not limited to, MRI imaging, echocardiography, or PET (10). Using imaging information along with other factors may be used to create a multidimensional finite element computer model of the heart (11). The model may display not only the three dimensions of the geometry of the heart but may also depict this geometry as it changes over time. This model may run on a personal computer type of machine or it may run at a central location or it may be processed at one location and delivered to another location to be run. The multi dimensional model may allow the doctor to visually inspect the status of all the elements of the heart. This model may be used to determine a variety of information, either pre-treatment, during the treatment or post-treatment, including, but not limited to:
 a. The areas of the mitral apparatus, aortic, tricuspid or pulmonary valves that may need to be repaired or replaced and what affect each repair may have on the other components.
 b. What vessels are blocked and may need to be grafted, where to graft and what affect the revascularized muscle may have on the other components.
 c. What areas of the ventricle are akinetic, dyskinetic or hibernating, to show what areas may be excluded during ventricular restoration and what effect the exclusion may have on the other components and aspects of the ventricle and heart.
 d. How this patient's heart may respond to medication treatment.
 e. The effects of placement of an Acorn, Myocor or other device on the outside of the ventricle may affect the heart.
 f. The effects of chordae length adjustment or papillary base relocation may affect the heart.
 g. The effects of placement of any ventricular assist device may have on the heart.
 h. What vessels are blocked and may need to be stented, where to stent and what affect the revascularized muscle may have on the other components.
 The model may then allow the doctor to select a treatment option (12) and allow the doctor to manipulate the image and model (13). The model may then analyze what effects his virtual treatment may likely have on the cardiac system display the potential clinical outcomes to the doctor (14), (16). The potential outcomes display may be but are not limited to the following (17):
 a. The estimated performance of the valves and ventricle after the procedure; i.e. regurgitation, reduced flow across the valves, ejection fraction etc.
 b. The volume and contractile state of the ventricle after excluding tissue.
 f. The positioning and performance of the valve apparatuses after reconstruction of the ventricle.
 The doctor may then be able to select the displayed intervention (18) or decide to try another treatment or modify the current intervention (19) and the cycle may repeat itself. When the doctor accepts the potential clinical outcomes, the model may then produce the specifications for the intervention (20). These specifications may lead to the development of a templates or tools or device to guide the doctor in translating the virtual intervention on the model to the actual intervention on the heart. No template or devices may be needed to perform the intervention and specifications such as the length of a chordae tendinae may be sufficient output from the model to allow the doctor to perform the intervention. Additional devices may be generated from the models to help the doctor implement the procedure that the model may have predicted to provide the best outcome. Such devices can include prosthetic mitral apparatus that is patient specific, or a customized annuloplasty ring etc (21). Furthermore, the use of some or all of above listed factors may be used to evaluate post-treatment the condition of the patient. A database of surgical pathologies, treatments and outcomes may be gathered, maintained and analyzed to further refine the treatment of cardiac diseases and disorders.
 This may determine before the treatment what likely effects his treatment of one or more elements of the heart will have on the other elements, and how to optimize the treatment of each component relative to the other components in order to achieve the best performance of the entire cardiac and circulatory system. The method and apparatus should allow the doctor to simulate numerous interventions and allow him to compare the different simulations, so that he can perform the option that gave him the best outcome. These interventions can include but are not limited to; placement of a Myosplint (Myocor Inc., Maple Groove, Minn.), placement of Corcap restraining device (Acorn cardiovascular Inc, St. Paul, Minn.), valve replacement (St. Jude Medical, St. Paul, Minn.), annuloplasty (Edward Lifesciences, Irvine, Calif.), surgical ventricular restoration (Chase Medical, Richardson, Tex.) stent placement (Medtronic, Minneapolis, Minn.), valve repair (Edward Lifesciences, Irvine, Calif.), bypass grafting, pacing, Biventricular pacing (Medtronic, Minneapolis, Minn.) and ventricle assist device (Abiomed, Danvers, Mass.). Surgical Ventricular Restoration (SVR) can be improved by providing a method and apparatus where a doctor can take an image of the patient's heart or ventricle and create an interactive multidimensional model with structural elements. The doctor can then manipulate the model by deleting, adding or rearranging the structural elements to simulate the SVR procedure. The model may integrate all the manipulations relative to each other and then interact with other models such as but not limited to physiological and hemodynamic models. The interactive multidimensional model may recreate the patient's heart or ventricle based on the manipulations conducted by the doctor and depict the new ventricle or heart and display cardiac performance characteristics and parameters. The doctor can perform this simulation numerous times and then compare the performance characteristics and select the optimal procedure. The model can then produce specifications for the selected procedure from which templates or tools can be created to aid the doctor in translating the virtual procedure to the real procedure.
 It is further understood that other modifications, changes and substitutions are intended in the foregoing disclosure and in some instances some features of the disclosure will be employed without corresponding use of other features. Accordingly, it is appropriate that the application be construed broadly and in a manner consistent with the scope of the disclosure.