Robotic-assisted minimally invasive surgery

Minimally invasive surgery (MIS) powered by robotic systems can greatly improve surgeon's skills. Dedicated instruments that enter in the human body through tiny holes can be tele-controlled by the surgeon through haptic devices, enhancing perception and execution of surgery tasks. Touching points generate forces based on object parameters (such as stiffness, impedance and friction) that are fed back to the human by haptic devices. Haptic feedback provides realism to contact interactions, distinguishing not only different objects, but also free-space to contact (and vice-versa) transitions. Scaling on-line tactile feedback can magnify or reduce haptic telepresence whenever needed, which is a key functionality for advanced MIS systems. Addressing surgical kinematic constraints in the control design enables manipulation techniques to focus only in the task space, increasing surgeons' dexterity. Guaranteed stability in contact with stiff objects is an important milestone (e.g. touching another instrument or a bone) which demands advanced control techniques, being a major obstacle for practical applications. Nowadays, robotic-assisted MIS does not include haptic feedback in the main control loop to avoid instability problems. In this project we plan to develop a robotic-assisted MIS setup with haptic telepresence, which has not been achieved so far by any commercial system. The advantages over traditional surgical practices include: 1) Augmented/scaled reality (e.g. motion and force augmentation or scaling for microsurgery); 2) Better comfort for the surgeon; 3) Real-time integration of intra-operative data (e.g. image-guided motion and force-controlled motion); 4) Accurate path following; 5) Enhanced mobility. Extra degrees of freedom inside the body can be controlled by the surgeon; 6) Compensation of physiological motions and surgical constraints; 7) Compensation for surgeon's hand tremor; 8) Less pain and trauma and shorter recovery time; 9) Tele-surgery; 10)Training, learning and mentoring using virtual models; The work plan is the following. After installing basic hardware and software, we will perform model identification of the medical robot, needed for advanced compliant motion control. Several haptic tele-manipulation architectures will be designed and tested to connect the master station (controlled by the surgeon) to the medical robot. The physical limits of the system setup (e.g. time delay, bandwidth, sampling time, noise and input saturation) have to be mapped into an optimal haptic feeling, coded into the control architecture. Safety issues cannot be neglected, since humans are in the workspace. Therefore, the overall architecture has to guarantee fault tolerance to sensor failure, high-fidelity haptic feedback and robustness to model errors. Experimental tests in organic objects will be carried out, paving the way to real surgical applications.