The concepts and benefits of hands-on robotic surgery and active-constraint robotics are introduced. The argument is made for systems to be cost effective and simple in order that they can be justified for a large range of surgical procedures. The case is made for robotic systems to have a clear justification, with benefits compared to those from cheaper navigation systems. The need to have robust systems, that require little surgical training and no technical presence in the operating room, is also discussed. An active constraint medical robot, the Acrobot System, is described together with its use in a prospective randomized controlled trial of unicondylar knee arthroplasty (UKA), comparing the performance of the Acrobot System with conventional surgery. Twenty-eight patients awaiting UKA were randomly allocated to have the operation performed conventionally or with the assistance of the Acrobot. The results of the trial are presented together with a discussion of the need for measures of accuracy to be introduced so that the efficacy of the robotic surgery can be immediately identified, rather than having to wait for a number of years before long-term clinical improvements can be demonstrated.

A "hands-on" robotic system for total knee replacement (TKR) surgery is presented. A computer tomography-based preoperative planning software is used to accurately plan the procedure. Intraoperatively, the surgeon guides a small special-purpose robot, called Acrobot, which is mounted on a gross positioning device. The Acrobot uses active constraint control, which constrains the motion to a predefined region, and thus allows the surgeon to safely cut the knee bones to fit a TKR prosthesis with high precision. A noninvasive anatomical registration method is described. The system has been successfully used in seven clinical trials with encouraging results.

A special-purpose robot system has been developed for neurosurgery. It contains control mechanisms developed from flight-simulator experience to allow precise motion and force-control using very low-force input. The robot constrains the tool motion to a safe region using methods developed for robotic surgery. These regions are defined pre-operatively using MRI segmentation techniques developed with the help of a simulator, ROBO-SIM. The intraoperative target location is tracked in real-time using ultrasound transducers. By applying updates using the ultrasound probe, safety is maintained by compensating for brain-shift. The overall system is being evaluated on phantoms.