Periodic Reporting for period 1 - FAROS (Functionally Accurate Robotic Surgery)
Période du rapport: 2021-01-01 au 2022-06-30
A key motivation for introducing robots in the operating room has been their ability to deliver superhuman performance. Surgical robots are able to execute highly precise gestures that rapidly and stably position instruments. However, while expert surgeons know from experience which gestures and instrument positions are opportune, robots have a hard time to figure out what motions are appropriate. This is because the surgical action itself alters the anatomy in a drastic and hard to predict manner. This makes that any pre-operatively planned trajectory becomes quickly obsolete. Even if executed with the greatest precision, the plan, being outdated, would not be functional anymore. Surgeons are trained to accommodate their gestures to this variability. Aside from prior experience, surgeons rely on all their sensory channels drawing also from tactile and auditory cues to adapt to the situation and adjust their motion. FAROS develops sensing and control methods to endow surgical robots with similar sensing, reasoning and decision-making capability. The goal is to introduce functionally accurate robotic behaviour. The robot motion that is commanded is accurate in the sense that is adapted to all sorts of variations and specifically geared at establishing goals that maximally achieve the targeted functionality. We do not target precise geometric locations here, but rather intermediate goals that are relevant from a clinical viewpoint such as removal of specific tissue types, avoiding damage to critical structures, establishing a firm connection with bone. By exploiting more and alternative sensors than humans can, FAROS technology will be able to deliver functional accuracy beyond human capabilities. E.g. in situations where humans are deprived from their native senses, e.g. when working deep inside structures (such as vertebrae) under visual obstruction or facing disturbed haptic feel, FAROS will be able to access sensors that look or sense through tissue and assess relations that are hard to discern otherwise. Pedicle screw placement (PSP) and endoscopic lumbar discectomy (ELD) are selected as use cases to explore the accuracy levels that can be reached with this new approach to robotic surgery.
Within the first reporting period the consortium made solid progress in setting up a common software-hardware platform for supporting research on Functionally Accurate Robotic surgery. The common platform supports a broad range of sensory inputs that are accessible for running real-time decision-making. A modular bi-manual collaborative robotic platform was established (and duplicated) to tackle and support improved PSP and/or ELD. As far as we are aware, this is the first surgical robotics platform maximally exploiting real-time control capabilities offered by ROS2 and by ROS2-OROCOS. Through use of the task specification language, eTaSL, complex robotic behaviours can be conveniently programmed by specifying the behaviour as sets of constraints. A detailed work flow analysis was conducted by the clinical experts which allowed us to reason about the ideal and desirable robot behaviours across the different phases of the procedures. Researchers made great progress in integrating sensors and gathering data during a first series of 3 integration weeks and data-collection campaigns. An encompassing data-sharing agreement was established allowing maximal data-exchange among partners. The involvement of clinical experts in these activities allowed for interpretation of acquired data and through life participation in validation experiments, allowed for fast feedback loops and for incorporating insights in the behaviour/functionalities that are being developed. Apart from a solid publication output, a few “firsts” can be reported: first use of real-time HSI based visualisation to guide an endoscopic cadaver experiment, first cadaver experiment under guidance of a co-manipulation robot suggesting significant benefits in instrument management, first smart drilling system prototype integrating a set of non-visual sensors (electrical impedance, accelerometers, drilling torque, drilling speed, and vibro-acoustics), first development of a behaviour-based framework based on an eTaSL-OROCOS-ROS2 for surgery, first PSP data collection in phantom, animal and human cadaver with newly integrated sensors, and a first development of a method to identify PSP functional parameters only on the dorsal aspect of the spine. Extensive experimentation on relevant models led to a unique and extensive (1.04 TB) data-set that will be used next to optimise the accuracy of robot-assisted PSP and ELD interventions.
FAROS will advance control of surgical robots endowing them with the capability to reach unprecedented levels of functional accuracy. FAROS explores a broad range of non-radiative and non-visual departing from the conventional routes that rely on fluoroscopy and/or camera systems prone to radiation and/or line-of-sight issues.
Following advances can already be reported :
● in Da Silva et al., 2022, we experimentally investigated the possibility of finely controlling the force applied by the robot to the environment based on the robot's built-in joint impedance controller.
● in Suter et al. 2022, we proposed PSP planning method based on the dorsal aspect of the spine, which can be further applied on the 3D intraoperative US reconstruction that has been developed.
● in Da Silva et al. 2022, we demonstrated the feasibility of automatic bone breach prevention based on bio-electrical conductivity sensing using previously developed breach detection algorithms and the newly developed equipment within FAROS. The robotic drilling setup stopped within a -0.6 / +1.5mm interval from the bone/canal interface in more than 100 ex-vivo experiments.
● in Gruijthuijsen et al. 2022, we introduced and demonstrated a framework for semantically rich surgical scope control. The proposed platform allows the surgeon to perform a bi-manual coordination and navigation tasks while the robotic arm autonomously performs scope positioning. To the best of our knowledge, this work is the first to report how to construct an autonomous instrument tracking system that allows for solo-surgery using only the endoscope as a sensor to track the surgical tools. A user study confirmed the suitability of the whole robotic system for a standardised bi-manual coordination and navigation surgical task.
● We further made progress in establishing a new approach to ELD under collaborative guidance. The method was tested on an ex-vivo model and received great interest by the participating clinicians.
In consultation with the industrial partner, clinicians and the External Advisory Board, FAROS moves forward to maximize the impact of its research efforts. In the second reporting period we will further work on optimizing our experimental plan and processing our unique data-sets to establish key relations and models that translate towards functionally accurate motion and decision-making. Advances in robotic PSP, minimal invasive PSP and MIS ELD are targeted. While targeting critical spinal applications, the methods to improve functional robotic accuracy are believed to be of broad use also in other surgical fields. Through our active dissemination strategy we aim to broadly grow awareness of the importance of this new approach to robotic surgery.
Following advances can already be reported :
● in Da Silva et al., 2022, we experimentally investigated the possibility of finely controlling the force applied by the robot to the environment based on the robot's built-in joint impedance controller.
● in Suter et al. 2022, we proposed PSP planning method based on the dorsal aspect of the spine, which can be further applied on the 3D intraoperative US reconstruction that has been developed.
● in Da Silva et al. 2022, we demonstrated the feasibility of automatic bone breach prevention based on bio-electrical conductivity sensing using previously developed breach detection algorithms and the newly developed equipment within FAROS. The robotic drilling setup stopped within a -0.6 / +1.5mm interval from the bone/canal interface in more than 100 ex-vivo experiments.
● in Gruijthuijsen et al. 2022, we introduced and demonstrated a framework for semantically rich surgical scope control. The proposed platform allows the surgeon to perform a bi-manual coordination and navigation tasks while the robotic arm autonomously performs scope positioning. To the best of our knowledge, this work is the first to report how to construct an autonomous instrument tracking system that allows for solo-surgery using only the endoscope as a sensor to track the surgical tools. A user study confirmed the suitability of the whole robotic system for a standardised bi-manual coordination and navigation surgical task.
● We further made progress in establishing a new approach to ELD under collaborative guidance. The method was tested on an ex-vivo model and received great interest by the participating clinicians.
In consultation with the industrial partner, clinicians and the External Advisory Board, FAROS moves forward to maximize the impact of its research efforts. In the second reporting period we will further work on optimizing our experimental plan and processing our unique data-sets to establish key relations and models that translate towards functionally accurate motion and decision-making. Advances in robotic PSP, minimal invasive PSP and MIS ELD are targeted. While targeting critical spinal applications, the methods to improve functional robotic accuracy are believed to be of broad use also in other surgical fields. Through our active dissemination strategy we aim to broadly grow awareness of the importance of this new approach to robotic surgery.