Registration and motion compensation of a needle placement robot for CT-guided spinal procedures Sheng Xu *a , Kevin Cleary b , Dan Stoianovici c , and Gabor Fichtinger a a Engineering Research Center, Johns Hopkins University, Baltimore, MD, USA 21218; b Imaging Science and Information Systems (ISIS) Center, Department of Radiology, Georgetown University Medical Center, Washington, DC, USA 20007; c James Buchanan Brady Urological Institute, Johns Hopkins Medical Institutions, Baltimore, MD, USA 21224 ABSTRACT Computed tomography (CT) guided needle placement is an established practice in the medical field. The efficacy of these procedures is related to the accuracy of needle placement. Current free-hand techniques have limitations in accuracy, which is often affected by the patient motion. In response to these problems and as a testbed for future developments, we propose a robotically assisted needle placement system consisting of a mobile CT scanner, a needle insertion robot, and an optical localizer. This paper presents the overall system concept and concentrates on the system registration and compensation of the patient motion. Accuracy results using an abdominal phantom are also presented. Keywords: Image guidance, medical imaging, medical robotics, needle placement, registration, tracking 1. INTRODUCTION 1.1 Clinical significance Spinal disorders are the fastest growing musculoskeletal subspecialty. It is estimated that over 70% of the population experiences significant low-back pain during their lives. Recent advances in medical imaging have propelled minimally invasive image-guided biopsy and local therapies into public attention 1 . Computed tomography (CT) guided nerve blocks and facet joint injections have proven to be safe and effective methods to treat spinal pain 2 . Currently, percutaneous (through the skin) placement of needles into the spine is performed freehand. Based on CT or fluoroscopy, the physician identifies the skin entry point and the target, thus defining the desired needle trajectory. The physician then aligns the needle by hand and advances it towards the target gradually while checking the position of the needle by re-imaging as necessary. The main problem is that the physician has limitations in accuracy when initially lining up the needle and subsequently staying on course. In addition, motion of a lumbar vertebral body due to breathing alone is up to 1.3 mm (peak to peak), and surgeon-induced motion up to ten times greater 3 . When the physician releases the needle, the needle can drift or tilt away from the desired path. In response to these problems, we propose integrating intra-operative CT imaging with a medical robot for precise placement of the needle. The workflow of the current manual procedure is practically identical to the steps followed by our robotic system. This parallelism offers a unique opportunity for gradual transition from a manual procedure to a fully robotic intervention. While experienced physicians can complete these procedures without difficulty, there is a need for precise and consistent aiming and delivery of the needle. The longevity of pain relief is thought to be associated with the spatial accuracy of needle placement. The system can also serve as a testbed for the precise robotically guided needle placement systems of the future. 1.2 Prior technical developments The history of medical robotics dates back to 1985, when Kwoh applied a PUMA robot to orient a needle for biopsy of * sheng@cs.jhu.edu; phone 1 410 516-4318; fax 1 410 516-5553 Medical Imaging 2005: Visualization, Image-Guided Procedures, and Display, edited by Robert L. Galloway, Jr., Kevin R. Cleary, Proc. of SPIE Vol. 5744 (SPIE, Bellingham, WA, 2005) · 1605-7422/05/$15 · doi: 10.1117/12.594731 661