Summary of Research Projects, Russell H. Taylor

This page is still under construction.  You can find more information at the link for the NSF Engineering Research Center for Computer-Integrated Surgical Systems and Technology.  Additional information may be found in several talks that I have given lately.

• A recent TUTORIAL on medical robotics given at the 1999 Symposium on Medical Image Computing and Computer-Assisted Interventions (MICCAI 99) gives an overview of medical robotics and draws on many peoples' work, with a heavy emphasis on work from our group.
• Some of our work on microsurgical augmentation is discussed in a MICCAI 99 presentation by my colleague Pat Jensen, given at MICCAI 99 and in a paper that appeared in that conference.  A revised version will appear in the Millenium Edition of The International Journal of Robotics Research.

 

The slide at left shows two of the systems that my group developed while I was at IBM Research.  The one on the top is a coarse-fine robotic system for inspecting large and delicate printed wiring boards in IBM mainframe computers.  The fine manipulators incorporated endpoint sensing and were capable of micron-scale positioning. 

The system at the bottom (discussed below) is a robotic system for hip replacement surgery.

The central insight behind much of our current work is that both of these systems couple information to action to significantly improve a difficult and delicate process.  Many of the same factors (improved precision, consistency, enabling new processes, reduced error rates, and process learning) that made information-driven machines such a dominant force in advanced manufacturing can have the same effect on surgery.

The material below focuses mainly on the robotics aspects of our current work, which produces the most showy pictures.  However, the computational, modeling, and planning aspects are also of great interest.  I'll try to include more of this material as I get the time to do it.

My earliest experiences with medical robotics involved the development of a robotic system for computer-integrated total hip replacement surgery.  This was a collaborative effort between IBM Research and surgeons  and researchers at the University of California at Davis. The resulting system, which was named "Robodoc^®", by Brent Mittelstadt (then a graduate student at UC Davis), was subsequently developed clincially by a company founded by one of the surgeons, Dr. Howard Paul,.  Our research group still maintains close contact with this company, Integrated Surgical Systems, Inc. of Davis, California.  Some current research projects related to this activity include x-ray fluoroscopy registration and guidance for revision surgery and development of new image-guidance methods and interfaces using the Neuromate^{® robot that ISS has donated to the CISST ERC.}

Our current work on fluoroscopic guidance for robots is tied to several applications and systems.  One activity (illustated at left) is related to our activity with ISS on robotic hip surgery.  This work isaimed at development of novel and robust methods for locating bones with respect to a robot by using only information present in x-ray fluoroscopic images, without relying on external navigation or extensive precalibration.  This work is discussed in a recent MICCAI paper, authored by Yao et al.

Another current activity tied to image guidance of robots is work directed at precise percutaneous delivery of therapeutic agents directly to cancers or other lesions.  This work is a major focus of the CISST ERC, and we are developing many systems for use with biplanar x-rays (left), CT, Ultrasound, and (eventually) MRI.  We are also collaborating with our CISST colleagues at Harvard's Brigham and Women's hospital to develop a common system infrastructure and application shell that will support both the JHU systems and the MRI-compatible robot being incorporated into the MRT "double donut" manipulator at BWH.

The work at left is discussed further in several MICCAI papers, including one by Bzostek, et. al.   Currently, we are mostly working on phantoms and ex-vivo organs, but we have also

I am currently working closely with several colleagues, including Dan Stoianovici (JHU Urology & Mech. Engineering Depts.), Louis Whitcomb (Mechanical Engineering), Patrick Jensen (Ophtholmology) and our students, to develop a modular family of robotic systems for a variety of surgical applications.  The photograph at left shows two of these components in a minimally invasive kidney stone removal.  The first component (called PAKY) is a radiolucent needle driver.  The second component (currently called simply the RCM) is a novel 2-link remote-center-of-motion  robot  that provides a natural "fulcrum" for reorienting the needle. 

Additional details of system at left can be found in  papers and on Dan Stoianovici's web page.

This photograph shows a novel system for microsurgery aumentation.  The mechanical parts of the system include the RCM linkage discussed above, together with commercial XYZ stages, a novel compact instrument carrier developed by Aaron Barnes, and  a force sensing handle.  We are exploring a variety of microsurgical applications, including vitreoretinal surgery, neurosurgery, ENT, spine surgery, and microvascular surgery.  The leader for much of this activity is Pat Jensen, who is the microsurgery assistant testbed leader for the CISST ERC

Initially, the system will be used simply to provide "steady hand" augmentation to provide tremor-free manipulation of delicate tissues at about 10 micron scale.  However, the longer term goal is to incorporate real time image processing, integration of sophisticated modeling and sensing of local anatomy and tissues, and "surgical macros" to provide a true interactive surgical assistant. 

.