In previous work, we demonstrated that real-time 3D ultrasound is an effective imaging technology for guiding surgeries inside of the heart. In this project, we have developed a 3D ultrasound-guided motion compensation system that tracks heart motion and allows the surgeon to operate on the fast moving structures of the beating heart without risk of damaging them. We focus on the beating heart repair of the mitral valve through mitral annuloplasty. The design of the system is guided by the clinical observation that the rapid motion of the mitral annulus is dominated by translation along a single axis between the left atrium and left ventricle. This allows the use of a simplified 1 DOF motion compensation system that can be used for surgical procedures like the anchor driving found in mitral annuloplasty.

• Surgical Instrument Design
  The robotic motion compensation instrument required for this procedure must have sufficient bandwidth to compensate for the fast-moving structures in the heart, as well as a sufficient force output to perform suturing and stapling tasks on heart tissue. It must also be able to operate within the limited confines of the heart.

  By noting that the motion of the mitral annulus is nearly unixial, we have developed a fast, lightweight, and hand held instrument that meets these surgical requirements (Figure 3). The instrument is capable of delivering surgical anchors to tissue while tracking the fast motion of the mitral annulus [1].

• Real-Time Instrument and Tissue Tracking in 3D Ultrasound
  Automatic, real-time segmentation of a tissue target in 3D ultrasound is challenging due to poor shape definition and the number of computations required to process volumes at 28 Hz (~40 MB/sec). We achieve real-time tissue tracking by making use of the instrument to designate a tissue target in the ultrasound volume, essentially by pointing at it. Tracking is turned into a two-stage problem where the instrument is first segmented in the volume through a GPU-based algorithm [2] and then the target is taken to be the tissue directly in front of the instrument [4].

• Time Delay Compensation
  The acquisition and processing of 3D ultrasound volumes introduces a substantial delay in which time the mitral valve can traverse the majority of its trajectory. Left uncompensated, these delays would lead to collisions between the robotic instrument and annulus that would result in tissue damage. To avoid this outcome, we exploit the near periodicity of the mitral valve trajectory to predict its path and hence compensate for time delay [3].