Current Work in Medical Robotics

Eve Coste-Maniere can be contacted here : evereve@mac.com

I am a research scientist at INRIA Sophia Antipolis where I lead the Chir Medical Robotics team (Chirurgie, Informatique & Robotique).

The central objective of my work is the integration of robotic techniques and image processing in surgery to make the planning and execution of surgical interventions more accurate and less invasive. The optimization of the performances of surgical procedures is based on safe and real-time software integration for medical robots. The following main underlying steps of robotic surgery integration are identified and investigated:

• the modelling and visualisation of deformable anatomical entities: here geometric approaches to computer vision are combined to medical image analysis and processing, to computational geometry. The objective is to provide a patient dependant dynamic model.
• the planning and simulation of robotic interventions using the above mentioned models. With the patient's pre-operative model, we formulate the needs of the surgeon and the characteristics of the robot as mathematical criteria, in order to optimize the settings of the intervention. Then we automatically reproduce expected surgeons' movements and guaranty their feasibility. We also simulate the intervention in real-time, paying particular attention to potential collisions between the robotic arms. in vivo trials are currently conducted.
• the secure real time execution of robotic procedures with augmented reality. This steps introduces medical imaging and modeling in the operating theatre. Results of the two previous phases are used to augment the reality perceived by the surgeon to guide his gesture.

  As studied during my PhD work, the core challenge is here to develop a safe control architecture, with a set of tools for implementation and analysis to ease the integration of all the building blocks required in a surgical robotics application with predictable performance. It relies on the definition of a formal theoretical framework and associated tools that allow the specification, the validation, and the implementation of generic robot and vision control schemes (as investigated in the ORCCAD system).

  In this context, the definition of a generic and efficient mission programming method, the Maestro language (cf Nicolas Turro's work), for robotic systems has been proposed. It is based on the use of formal methods (e.g Esterel Synchronous Programming Language (Meije research group)) to fully benefit from their mathematical foundations, gateway to formal verifications.

Today, a coronary bypass procedure has been chosen as a test bed for this integration. This work is performed in close cooperation with two cardiac surgery teams lead by Alain Carpentier (Hôpital Européen Georges Pompidou, France) and Friedrich Mohr and Volkmar Falk (Herzzentrum Leipzig, Germany). The tele-operated robot da Vinci has been chosen from outset.

Industrial collaborations include Intuitive Surgical, Inc (USA), and General Electrics. Our main international collaborations includes the National Institute of Health with the Johns Hopkins University, and Stanford University. Also, the team participates in projects funded by the French ministry of research "télémédecine", the National Network in Communication Research (RNRT) for the VTHD program, and the National Network of Technology and Health (RNTS) through the EndoXiRob project.