Int J CARS (2011) 6 (Suppl 1):S90–S93 DOI 10.1007/s11548-011-0577-1
CARDIOVASCULAR IMAGING AND SURGERY
Advantages of using MRI for minimally invasive cardiac surgery M. Li1, D. Mazilu1, K. Horvath1 1 National Institutes of Health, National Heart, Lung and Blood Institute, Bethesda, USA Keywords Beating heart cardiac surgery Real-time MRI MRI guidance Purpose MRI provides excellent visualization particularly in its ability to provide high-resolution images of blood filled structures. Vascular as well as soft tissue visualization can easily be performed simultaneously. The capability of real-time interactive MRI (rtMRI) to quickly change imaging and display parameters, seamlessly tracking a device; allows the physician to smoothly perform cardiac and vascular interventions. With a new generation of imaging systems (1.5T Magnetom Espree, Siemens Medical Solutions, Munich, Germany) that include a wider bore (70 cm) and shorter cylinder (120 cm), surgical access to the patient within the magnet becomes feasible. Using transapical aortic valve implantation (AVI) as an example, we have demonstrated how MRI can be effectively used before, during and immediately after a beating heart cardiac surgical procedure and the advantages of using MRI for minimally invasive cardiac surgery. Methods Pre-operatively, standard MR sequences are performed to obtain the orientation of the heart, evaluate ventricular and valve function, and locate the native valve annulus and the origin of the coronary arteries. Pre-operative imaging also allows establishment of scanning planes to be used for real-time imaging during valve implantation. Three imaging planes were prescribed for simultaneous real-time imaging to guide implantation. Two of these planes are positioned to provide long-axis views of the left ventricle, showing the right coronary artery and left anterior descending coronary artery origins, respectively. The other plane provided an axial view of the aortic valve. The coronary ostia and aortic annulus location are digitally marked. These digital markers remain visible at all times in the 3D rendering for anatomic reference. Intra-operatively, rtMRI provides visualization of the progress of the procedure. The rtMRI interactive system consists of an interactive user interface, operating room large screen display, specialized pulse sequences, and customized image reconstruction software. With this system, multiple parallel and oblique slices can be obtained in rapid succession and can be simultaneously displayed in a 3D rendering. Image slices can be repositioned and added or omitted as needed. The MRI tissue contrast can be interactively channeled by toggling saturation pulses off/on to highlight selected objects. The surgeon views the rtMRI on a projection screen while manipulating the deployment device within the animal in the magnet. He is in contact with the scanner operator via headphones and a microphone to request changes in the imaging planes as needed. MR microcoils are used to tracking the location and orientation of the interventional devices. These MR microcoils can be highlighted with different colors in the MR images.
123
Post-operatively, in addition to anatomic confirmation of adequate placement of the prosthetic valve in relation to the aortic annulus and the coronary arteries, functional assessment of the valve and left ventricle was also obtained with MRI. Results The sets of rtMRI parameters for each of the sequences to perform intracardiac procedures were determined. These include the sequences for pre-surgical planning and post-op evaluation, as well as the sequence for intervention. The sequence used for rtMRI is a steadystate free precession (SSFP) sequence with following scanning parameter: TR = 3.23 ms; TE = 1.62 ms; bandwidth = 800 Hz/pixel; flip angle = 45°; slice thickness = 6.0 mm; FOV = 340mm 9 255 mm; matrix = 192 9 108, with partial phase-encode acquisition. The imaging frame rate is further increased using variable rate view sharing and TSENSE. We have successfully implanted self-expanding aortic prostheses via transapical approach under rtMRI guidance on 23 animals. The time to implant the valve was 60±14 s. This includes the time to correctly orient the prosthetic valve with respect to the native annulus and the coronary ostia. This is more expedient than the results reported with typical fluoro/echo guidance. It is easier to avoid blocking the ostia and deploy the valve on the native aortic annulus under rtMRI as these anatomic features are constantly in view. Due to improved resolution with rtMRI, ventricular unloading by rapid pacing or cardiopulmonary bypass is not required. Additionally, in the case under fluoro/echo guidance, multiple injections of contrast are required to locate the coronaries and approximate the position of the annulus. As a result, radiation and contrast toxicities can occur with fluoro that are not an issue with MRI. Active and passive markers on the devices and the images helped to position and place the valve. The MR microcoils are a superb indicator of the valve orientation in MRI during the procedure. The digital markers placed on the images identify key landmarks and provide surgical references. Post-placement gated CINE MRI revealed excellent myocardial function after valve implantation in both long- and short-axis views. The phase-contrast CINE MR images confirmed good systolic flow with excellent valve leaflet opening and no evidence of turbulence, diastolic regurgitant flow, or paravalvular leak. The perfusion results confirmed adequacy of blood flow at the tissue level, indicating proper valve positioning with respect to the coronary ostia. These results were further confirmed by echocardiograms. Conclusion rtMRI provides excellent visualization for intraoperative guidance of AVI on the beating heart; it certainly can assist the surgeon and improve the success of the operation in appropriate patients. Active markers help to identify the position and orientation of the devices in MRI images. MRI-guided surgery also allows direct functional assessments to be made before, during, and immediately after cardiac procedure that are not obtainable by conventional imaging alone. Expansion of rtMRI guidance to facilitate other types of cardiac surgical procedures, including mitral, pulmonary, and tricuspid valve replacements or repairs, should be considered to minimize trauma and enhance patient benefit.
Int J CARS (2011) 6 (Suppl 1):S90–S93 Prototype system for developing MRI-guided and robot assisted minimally invasive intracardiac procedures E. Yeniaras1, J. Lamaury1, Y. Hedayati1, N.V. Sternberg1 and N.V. Tsekos1 1 Medical Robotics Laboratory, University of Houston, Houston, TX, USA Keywords Robotic surgery Heart phantom Virtual reality Real-time image guidance Purpose The goal of this work is to develop different hardware and software elements pertinent to magnetic resonance imaging (MRI) guided and robot-assisted intracardiac procedures, and integrate them into a modular platform suitable for further development and testing in the laboratory and at the MR suite. This prototype system, which has been partly inspired by our previous conceptual design [1], includes appropriate tools and dedicated software modules that operate synergistically for planning an operation, controlling a semi-autonomous robot based on MRI data and adjusting on-the-fly the image acquisition parameters of the scanner to better suit the particular conditions of the intervention as it evolves. Moreover, the practicality of a specific actuated heart phantom is also investigated in this work for future in vitro experiments. Methods The system consists of three major components: (1) a computational core, (2) a 7-degrees-of-freedom (7-DOF) robotic manipulator, and (3) an 18-DOF MR-compatible actuated heart phantom. Figure 1 depicts the main architecture and the data flow within the system. Computational core, which is connected to both ‘‘MR Scanner’’ and ‘‘Data Server’’ for image acquisition via TCP/IP protocol, has dedicated modules not only for MRI based guidance but also for controlling actuated devices within the system. The visual information about the area of operation (AoP) is also provided to human– machine-information-interface (HIMI) by computational core [2]. The robotic manipulator was specifically designed for intracardiac procedures and has two parts: a 5-DOF extrathoracic unit (ExtraU) that resides outside the patient to provide access to the apex through an intracostally placed port and a 2-DOF intracardiac unit (IntraU) that maneuvers inside the beating heart toward the operational target point. ExtraU actively compensates heart motion and provides actuation to the IntraU [3]. The functionality of the prototype system was first tested virtually for transapical aortic valve replacement (AVR) in beating heart [4]. Cine MRI datasets (with true fast imaging, steady state precession pulse sequence) were used for surgical planning, while on-the-fly
Fig. 1 The core architecture of the prototype system with data flow is depicted
S91 guidance was performed with real-time MR slices with repetition time of 48.4 ms collected from healthy subjects (n=10). Computational core generated dynamic trajectories from the apex to the aortic annulus for simulating a prosthetic valve deployment with transapical approach. For performing in vitro experiments, the system incorporates an MR-compatible, computer controlled and actuated phantom that mimics the motion and dimensions of certain anatomical landmarks in human heart [5]. The structure and kinematics of this phantom replicates the motion of the trocar/apex, access tube and aortic annulus, since through those structures the manipulator should maneuver in a transapical AVR. With 18 DOFs the cardiac phantom continuously positions three tubular structures (each representing one of the above listed anatomical landmarks). The current version of the phantom has an access tube with a diameter of 9 mm, while the diameters of the aorta and the trocar are 23 mm and 12 mm severally. Results Our image guidance methodology showed that a cylindrical access tube can be defined inside the left ventricle for safe deployment of a catheter-like device. For 10 subjects, the average base diameter of this virtual tube was 9 mm in systole and 22 mm in diastole respectively. Transapical AVR was successfully simulated for several different combinations of initial heart phases when deployment starts, robot actuation speeds and device dimensions. It was also shown that it was possible to design an MR-compatible robotic manipulator and a dynamic heart phantom for simulating the semi-autonomous deployment in the beating heart. Conclusion Recent studies presented herein illustrate the practicality of integrating appropriate computational tools to facilitate the MRI-based volumetric (3D) image guidance. Although it was designed for in vitro experiments, the phantom could be used also for extensive studies of developing imaging methods, robotic manipulators, and methods for practicing as well as for training. Aside from being a part of an ongoing study of developing a complete image-guided intracardiac surgical system, the main motivation for the implementation of such a device originates from the certain benefits to clinicians involved in the development of MR-guided robotic surgeries. We envision that the system will be applicable for in vivo testing upon its completion, allowing us to expand the approach to other sophisticated surgeries. References [1] Yeniaras, E., Lamaury, J., Deng, Z., Tsekos, N.V.: ‘‘Towards A New Cyber-Physical System for MRI-Guided and RobotAssisted Cardiac Procedures,’’ in Proceedings of 10th IEEE International Conference on Information Technology and Applications in Biomedicine (ITAB 2010), 2010, pp. 1–5. [2] Yeniaras, E., Deng, Z., Syed, M.A., Davies, M.G., Tsekos, N.V.: ‘‘A Novel Virtual Reality Environment for Preoperative Planning and Simulation of Image Guided Intracardiac Surgeries with Robotic Manipulators,’’ in Stud Health Technol Inform 163 (2011), IOS Press, pp. 716–722. [3] Yeniaras, E., Lamaury, J., Navkar, N.V., Shah, D.J., Chin, K., Deng, Z., Tsekos, N.V.: ‘‘Magnetic Resonance Based Control of a Robotic Manipulator for Interventions in the Beating Heart,’’ in Proceedings of 2011 IEEE International Conference on Robotics and Automation (ICRA 2011), 2011, to appear. [4] McRae, M.E., Rodger, M., Bailey, B.A.: Transcatheter and transapical aortic valve replacement. Crit Care Nurse 29 (2009) 22–37; quiz 38. [5] Sternberg, N.V., Hedayati, Y., Yeniaras, E., Christoforou, E., Tsekos, N.V.: ‘‘Design of an actuated phantom to mimic the motion of cardiac landmarks for the study of image-guided intracardiac interventions,’’ in Proceedings of 2010 IEEE International Conference on Robotics and Biomimetics (ROBIO 2010), 2010, pp. 856–861.
123
S92 Intraoperative assistance system for minimally invasive transapical aortic valve implantation based on 2D fluoroscopic image guidance M.E. Karar1, T. Noack2, D. Holzhey2, V. Falk3, F.-W. Mohr2, O. Burgert1 1 University of Leipzig, Innvoation Center Computer Assisted Surgery (ICCAS), Leipzig, Germany 2 University Hospital Leipzig for Heart Surgery, Heart Center Leipzig, Leipzig, Germany 3 University Hospital Zurich, Division of Heart and Vascular Surgery, Zurich, Switzerland Keywords Aortic valve replacement Biomedical image processing Image-guided surgery X-ray fluoroscopy Purpose Transapical aortic valve implantation (TAVI) is a minimally invasive technique to be applied for elderly and inoperable patients with severe aortic stenosis, avoiding sternotomy and cardiopulmonary bypass [1]. In the TAVI, a stented valve bioprosthesis that is temporarily crimped upon a balloon catheter, is inserted through the apex into the aortic root via a left anterolateral mini-thoracotomy. After reaching the correct position of implantation, the stented valve prosthesis is deployed by an inflatable balloon to reach its final diameter, thus fixing the prosthesis to the aortic wall. Once deployed, the prosthesis can not be repositioned. Therefore, the correct placement of stented aortic valve prosthesis is crucial under routinely angiographic and fluoroscopic X-ray imaging with a C-arm system. To solve the fluoroscopic guidance problems associated with current TAVI, we present a new assistance system to guide the TAVI intraoperatively. The developed system supports the physician in finding an appropriate placement for the aortic valve prosthesis by automatically defining the safe area of valve implantation on live 2D fluoroscopic images. Methods For assisting the TAVI, the intraoperative assistance system is connected with X-ray angiographic and fluoroscopic imaging system as depicted in Fig. 1. This developed system can automatically define the appropriate placement of the prosthesis as follows. Overlay of projected 3D aortic root model and valve landmarks A 3D model of the aortic root is provided by automatic segmentation of the aorta and detection of anatomical valve landmarks which are coronary ostia, commisures and three lowest points of the leaflets in interventional C-arm CT images (syngo DynaCT, Siemens AG, Healthcare Sector, Forchheim, Germany) [2, 3]. The 3D model of
Fig. 1 Block diagram of intraoperative fluoroscopic assistance system connected with an interventional C-arm imaging system during the transapical aortic valve implantation
123
Int J CARS (2011) 6 (Suppl 1):S90–S93 aorta and valve landmarks are projected into a 2D fluoroscopic plane using the transformation matrix of C-arm imaging system, in order to perform the overlay of aortic root model and valve landmarks on fluoroscopic images during the intervention. Tracking of aortic valve prosthesis The prosthesis is tracked in the fluoroscopic image sequences by using the template matching approach to estimate the position of aortic valve prosthesis and a shape model to extract the corner points of the prosthesis [4]. To start the prosthesis tracking procedure, an initialization step is performed by manually defining the corner points of the prosthesis in the first image of sequence to provide the required algorithm parameters such as the shape model parameters. Estimation of safe area of valve implantation In this study, the correct placement of aortic valve prosthesis is defined to be one-third to onehalf of prosthesis’s length above and perpendicular to the aortic annulus [1]. Therefore, the assistance system calculates and visualizes the safe area of implantation by defining the normal of valve annulus between the lowest points of leaflets and the coronary ostia, showing the updating position of aortic valve prosthesis until final placement within the clinical accepted margins. Results The developed assistance system has been tested and evaluated on five fluoroscopic image sequences and the related 3D aorta models from the clinical routine of the TAVI. The tested fluoroscopic images are 1024 9 1024 pixels in size. The pixel size is approximately 0.2 mm. In Fig. 2a, a 3D model of the aortic root (yellow meshes) and valve landmarks shown as colored points; namely coronary ostia (red), commisures (green) and lowest points of the leaflets (blue). The maximum position errors are less than 5.0 mm for the aorta model overlay on fluoroscopic images. Figure 2b shows the tracked prosthesis with maximum localization errors less than 1.0 mm which is within the clinical accepted range (B5.0 mm). Figure 2c combines the overlay and tracking results of aorta model and prosthesis, respectively, to align the prosthesis’s length with the estimated safe area of valve implantation. Conclusion A new intraoperative assistance system has been developed for improving the accuracy of transapical aortic valve implantations. The developed system visualizes automatically the safe area of valve implantation as well as the tracked aortic valve prosthesis under live 2D fluoroscopy guidance. For minimizing the position errors of aorta model overlay on fluoroscopic images, automatic registration of the aortic root model and valve landmarks to contrast fluoroscopic images is still under development. Moreover, we are now preparing the hardware setup of developed assistance system for the clinical evaluation.
Fig. 2 Results of developed assistance system: a 3D model of aortic root with anatomical valve landmarks, b tracking of aortic valve prosthesis and c overlay of aorta model (yellow) and tracked prosthesis (pink) into the safe area of valve implantation (white) on a live fluoroscopic image
Int J CARS (2011) 6 (Suppl 1):S90–S93 References [1] Walther T, Dewey T, Borger MA, Kempfert J, Linke A, Becht R, Falk V, Schuler G, Mohr FW, Mack M (2009) Transapical aortic valve implantation: step by step. Ann Thorac Surg 87:276–283. [2] Zheng Y, John M, Liao R, Boese J, Kirschstein U, Georgescu B, Zhou SK, Kempfert J, Walther T, Brockmann G, Comaniciu D (2010) Automatic aorta segmentation and valve landmark detection in C-arm CT: application to aortic valve implantation. Med Image Comput Comput Assist Interv 13:476–483.
S93 [3]
[4]
Kempfert J, Falk V, Schuler G, Linke A, Merk D, Mohr FW, Walther T (2009) Dyna-CT during minimally invasive off-pump transapical aortic valve implantation. Ann Thorac Surg 88:2041. Karar ME, Merk DR, Chalopin C, Walther T, Falk V, Burgert O (2010) Aortic valve prosthesis tracking for transapical aortic valve implantation. Int J Comput Assist Radiol Surg. doi: 10.1007/s11548-010-0533-5.
123