3D DEFORMABLE OBJECT EXTRACTION AND TRACKING IN MEDICAL IMAGING
The aim of this research line is to provide advanced tools for medical image analysis like image-guided assisted surgery. In particular, object extraction of 3D deformable bodies is a key task in this context, as it allows to measure and visualize 3D anatomical structures and delineate regions of interest. To this end, real-time data coregistration, segmentation and tracking are critical issues that are investigated.
AUGMENTED REALITY PLATFORMS FOR ROBOT-ASSISTED LAPAROSCOPIC SURGERY
The idea is to create augmented reality platforms that endow the surgeon with extra abilities integrated in the same control console. Examples of abilities are: endoscope control based on eyes tracking and customized images superimposed to the 3D rendering of the scene; qualitative force feedback provided in a visual augmented reality environment.
BRAIN COMPUTER INTERFACES IN ROBOTICS
The goal of this activitiy is to explore the use EEG-based brain-computer interface (BCI) to command robotic prostheses and to monitor humans' congitive/emotional state during Human-Robot-Interaction (HRI). To this aim, a first investigation of the use of Weightless Neural Networks for the classification of EEG signals for robotic prosthetic control has been provided.
COGNITIVE COMPUTING AND HEALTHCARE
Cognitive computing requires roles and unique and unusual skills such as: machine learning engineers, natural language processing developers, and data scientists. This research aims to develop cognitive applications that can interact with people and/or things (machines and/or other computers).
CYBER SECURITY FOR eHEALTH and ROBOTICS
Nowadays, the deployment of social and assistive robots has brought great convenience to people’s lives, especially in the health-care domain, where two main challenges have to be faced: i) securing robots from hackers' attacks which could violate people privacy or – even worse – threaten humans’ life from both physical and emotional/social point of views; ii) providing secure storage and exchange of eHealth data, protection and control over personal data. The aim of this task is to explore cutting-edge solutions to cope with the afore mentioned issues.
COMPUTER-GUIDED IMPLANT DENTISTRY
The developments in computer-aided planning have drastically modified the interventional possibilities in implant dentistry, changing the traditional invasive surgical protocol, with a flapless procedure eventual immediate loading. Although computer-guided implant dentistry is an upcoming technology with the potential for more predictive and less invasive implant placement, its performance has to be critically evaluated, because it is already in clinical practice. The aims of this research is to reduce the time required for surgery, to increase the accuracy of the implant positioning using a flapless approach, or in any case one that is minimally invasive, and at the same time to improve the post-operative treatment of the patient through the use of computer-designed surgical guides, developed with a combination of virtual models and mechanical systems.
CONTROL FOR AUTOMATIC TASKS
The aim is to develop autonomous abilities for robotically-assisted surgical robots. In particular we intend to develop algorithms for autonomous or semi-autonomous suturing by using vision-based tissue tracking techniques and force sensing.
DESIGN FOR ADDITIVE MANUFACTURING IN BIOMEDICAL FIELD
The aim of this research line is to design specific 3D CAD models to be produced through Additive Manufacturing (AM) techniques, useful: a) for the surgical planning, b) to improve the accuracy of the surgery, using customized surgical guides, c) to define the internal architecture of scaffolds for the tissue engineering. Crucial in the success of these applications is, in fact, the ability of the designer to design for AM and maximize the possibilities that these techniques can offer: thin walls, cellular structures, free-form channels in the part, integration of functions and personalisation.
DESIGN, MODELING AND PROTOTYPING OF MECHANISMS TO GRIPPING TASKS
The aim of this task is to develop mechanical systems oriented to gripping tasks. This research is crucial in many fields such as prosthetics or humanoid robots, but also surgical tools. The mechanism represents the foundation of all these devices, so it is necessary that its designing and modelling are performed together with the prototyping to identify the best solutions.
EMBEDDED MICROPROCESSOR FOR DATA-ACQUISITION AND CONTROL
Control and monitoring of experimental facilities, detectors and laboratory equipment requires tackling a complex mix of tasks. In a typical application, the control node is asked to acquire analog sensors, to set and read back digital parameters, to drive servo and to assert alarms: all such features fit nicely into a microprocessor/microcontroller-based design. The processor can be selected among a large variety of devices, to match the needed computational bandwidth. We have developed versatile and expandable platforms for deep embedded applications in control and monitoring of detectors, sensors, and complex research equipment. We have studied and implemented in hardware architectures capable to perform resilient, unattended operations in a harsh and potentially hostile environment.
ETHICAL, LEGAL AND SOCIAL ASPECTS (ELSA) OF ROBOTICS AND ARTIFICIAL INTELLIGENCE
The aim of this research line is to analyze the social impact of new robotic and AI technologies, to evaluate their ethical and legal implications, and to suggest, on the basis of these principled assessments, suitable research (and more broadly societal) policies. Of special interest are ELSA which concern robotic systems interacting with humans in the workplace, at home, and in other human habitats and activities.
FORMAL METHODS FOR MODELLING MEDICAL GUIDELINES FOR ROBOTIC SURGERY
All approaches for robotic-assisted surgery should be based on a formal and checkable representation of the medical guidelines on which to conduct the aforementioned surgery. Thanks to a formal representation of medical guidelines for robotic surgery, it will be possible to identify the best surgical paths to complete a given intervention in fast and safe way, by means of some typical computer science techniques such as the model checking verification. In this way, we will be confident that a robotic surgeon will be always able to face any unexpected situation, by changing opportunely its behaviour and minimising any kind of risk for the patient.The main aim of this research topic is to investigate on the design and development of the best formal techniques for modelling medical guidelines in robotic surgery.
GAIT ANALYSIS FOR EVALUATING MOTOR PERFORMANCE OF ATHLETES WITH DISABILITIES
The goal is to develop a reliable classification tests for athletes with intellectual disabilities to bring discipline back into the international rowing and to allow him to address the sport and then competitions adequately. For this purpose, the gait analysis will be used to measure motor performance in different sport specialities.
HAPTIC INTERFACES FOR AUTONOMOUS ROBOT-ASSISTED SURGERY
The main goal of this research is to design the future generation of robotic surgeons enhanced with innovative haptic interfaces. These interfaces are capable of reproducing the kinaesthetic and the tactile aspects of the human sense and providing robots with a plethora of additional sensor information, which will improve dramatically the performance of robot-assisted surgery in terms of maximisation of medical benefits and minimisation of risks. This research activity will investigate about the design of efficient artificial intelligent techniques for controlling haptic devices by learning of human tasks in surgery activities.
HIGH-SPEED, LOW-LATENCY SERIAL LINKS BASED ON RECONFIGURABLE HARDWARE
High-speed serial links play a key role in data transfer in digital systems. Serial transmissions allow to overcome skew-related problems between the data and clock lines and achieve a much higher throughput (nearly 10 Gbps w.r.t nearly 100 Mbps). Serialization also reduces the devices pin count and the number of traces on printed circuit boards, simplifying the system. Reconfigurable digital devices, such as Field Programmable Gate Arrays (FPGAs), offer serial IO capabilities up to a 28 Gbps in latest generation static RAM based (SRAM-based) devices. Our research group has several years of experience in the design of serial links for fast data transfers based on reconfigurable hardware. Our expertise includes designs with minimal transfer latency for real-time applications. We have also experience in the design and testing of high-speed serial links FPGAs for the deployment in radiation areas. We investigate novel methodologies to harden designs implemented in SRAM-based FPGAs against radiation-induced single event upsets.
INTERACTIONS BETWEEN BIOMOLECULES DRIVEN BY FEMTOSECOND UV LASER PULSES
The high average power delivered by this source has been exploited to achieve strong UV photon absorption in solution of living cells as well as antibodies. In the former case the absorption leads to DNA-protein and protein-protein cross-link, whereas in the latter system the UV absorption by the tryptophan residues gives rise to the disulfide bridges breaking thereby producing a reactive site that eventually allows an effective tethering of the antibody on gold surfaces. While all these interactions benefit from the relatively high average power, in view of the thermal phenomena arising in the nanosecond time domain, the high fluence laser system, such as nanosecond lasers, would be detrimental for the study of large molecules. Hence, our laser system results to be particularly suitable for inducing electronic excitation in complex media.
INTRABODY NANONETWORKS FOR HUMAN AUGMENTATION AND MONITORING
Nanoscale devices, implanted inside the human body and interconnected in a network, could be used for radically new medical diagnosis and treatment techniques. In this scenario, the research efforts are currently devoted to map the physiological mechanisms underlying the human neuronal activities into communication engineering system models.
MODEL BUILDING AND DATA ANALYSIS, HIGH PERFORMANCE AND DISTRIBUTED COMPUTING, NETWORKING AND MONITORING FOR HEALTHCARE AND ROBOTICS APPLICATIONS
Main topics of the research activities: a) Development of technologies to optimize high performance/high throughput computing and big data access and preservation, through grid and cloud paradigms for healthcare and robotics applications; b) Statistical analysis of experimental data; c) Object Oriented Software design and development in the context of large projects; d) Detailed simulation of the effects of radiation in matter and the response of measurement devices.
MOTIVATIONS FOR A COMMON RESEARCH ENVIRONMENT
Theory of electromagnetic fields supports several application of interest in advanced medical equipment, provide rationale for monitoring and remote sensing of the human body, guide noninvasive techniques for some medical therapies, describe fundamentals of interaction of human bodies with electrical equipment.
MULTIFUNCTIONAL ROBOTIC HANDS: DESIGN AND CONTROL
This research is devoted to the development and the employment of artificial hands in all those fields where the use of anthropomorphic prehensile devices makes the difference, such as humanoid robots, prosthetics, surgical tools. The innovation will be linked not only to the mechanical design and integrated sensor systems but also to novel planning and control algorithms generating intelligent and effective actions. For the hand design, research starts from underactuated design solutions based on the concept of postural synergies and goes toward innovative solutions. On one hand, these are aimed at simplification of the manipulation device itself, for improving its affordability, in terms of weight and size, without reducing the functionalities. On the other, they are aimed at simplification of the control. About control and planning algorithms, one objective is that of merging vision and touch information and integrating them in reasoning to execute complex manipulation tasks. Furthermore, methods to learn from human data and to map synergies to artificial hands are investigated. Learning and control strategies based on neural networks, reinforcement learning and sensory-motor synergies are investigated and tested on anthropomorphic robotic hands.
NEW SENSING DEVICES
The aim is to develop new sensors for surgical robotic tools to increase the information feedback that arrives to the surgeon. In particular, we intend to develop effective sensing solutions to measure the forces exchanged between surgical tools and tissues and new class of sensors able to acquire chemical, vital information from the field. Possible candidate are FO sensors, and, in particular, FBG based sensors, or recently-synthesized families of organic conjugated compounds.
NEW SURGICAL INSTRUMENTS FOR TELEROBOTIC SURGERY
We aim at designing new surgical instruments or improving existing instruments for telerobotic surgery systems. The close collaboration of engineers and clinicians will be crucial to analyze the surgical workflow, derive the clinical need, conceive the mechanical design and test the new instruments.
OPEN-SOURCE ENGINES FOR SURGERY REALISTIC SIMULATION
The aim is that of improving existing open-sources engines to obtain realistic simulation of typical surgical scenarios that can be used for training in robotic surgery and for designing of new strategies for autonomous control of particular tasks such as robotic suturing. In this context, dense environment reconstruction, tracking and modeling of deformable objects are investigated. Another important issue is the integration of friction models and collision detection algorithms to handle the collision between a soft body (deformable organ) and a rigid body (spatula) as well as their dynamic interaction.
REAL TIME IMAGE PROCESSING
The interaction with a complex environment and the drive towards augmented reality, require the implementation of always more complex image and video processing algorithms. Higher video resolution, multiple cameras, and the image analyses conducted using various video sensing techniques, require an enormous processing capability to obtain real time results. The research activity wants to implement the steps of the processing chains that require large computing capabilities on dedicated integrated circuits (Field Programmable Gate Arrays, Application Specific Integrated Circuits, Digital Signal Processors) that provide the computing power while satisfying the weight, power, and cost constraints of the system.
SEMANTICS IN MEDICAL ROBOTICS
Medical robotics includes a number of devices used for surgery, medical training, rehabilitation therapy, prosthetics, and assisting people with disabilities. Nowadays, robotic devices are used to replace missing limbs, perform delicate surgical procedures, deliver neurorehabilitation therapy to stroke patients, teach children with learning disabilities, and perform a growing number of other health-related tasks. However, usually concrete models and specific knowledge are not available for objects or events in the medical robot's work environment. Hence, a medical robot system has to rely on more generalized modes of inference to infer the semantic content of the context. Adequate models and knowledge may describe broad categories of objects or events, acquired through training on sets of numerous examples. Knowledge may also be inferred from similarities and correspondences discovered between novel and known cases. As a matter of fact, there is a growing tendency to introduce high-level semantic knowledge into robotic systems. When a medical robot encounters unknown objects in its environment and semantic models are available, the perceptual system can derive knowledge from the relationships established with known objects of a similar typology. Moreover, through semantic modeling of low-level features within a scenario, robots can generate representation of such features at a level of abstraction where logical reasoning methods could be applied for decision making. Furthermore, at such semantic level more than one modality can be merged to complement each other and produce logical inferences. As a result, different cognitive systems have become quite popular among the research community, especially those using deep learning techniques over images and language sources, showing promising results.
SERIOUS GAME TO SUPPORT THE REHABILITATION AND TRAINING OF CHILDREN WITH AUTISM SPECTRUM DISORDER
The serious games are interactive experiences that have the appearance and / or structure of a real game; their aim is to develop certain skills and abilities in the player for the achievement of a final target. One of the most promising applications of serious games is represented by tools for supporting people with Autism Spectrum Disorder (ASD). The goal is to investigate the use serious games for the inclusion of children with ASD and for the development of their personal and professional skills.
SOCIAL AND ASSISTIVE ROBOTICS
The aim is to develop personal robots, as assistive technological tools, that can collaborate with people. Such robots are expected to incrementally learn user preferences and to modify and adapt their behavior accordingly. This adaptation requires learning a model of human behavior and integrating this model into the decision-making algorithm of the robot. Creating robotic systems capable to correctly model and recognize the human behavior and of adapting their behavior to the user is a very critical task, especially in the domain of assistive robotics and when working with vulnerable user populations.
STATISTICAL LEARNING FOR ROBOTICS AND HUMAN AUGMENTATION
Main topics of research: a) Model-building and data analysis in cognitive computing and automation control; b) Development of accurate algorithms for supervised classification and pattern recognition; b) Data processing and editing for missing data imputation and data fusion; c) Smoothing and adaptive automation control using P-splines and non-parametric regression; d) Clustering of time series; e) Probabilistic and non-parametric learning for non-linear dynamical systems.
To remotely access and control real objects in perceived real time, an innovative network infrastructure is required for transporting haptic information. In this scenario, the research efforts are currently focused on the definition of the IEEE P1918.1 Tactile Internet.
ULTRAFAST DYNAMICS OF BIOMOLECULES
The aim of this group is the application of ultrashort laser pulses (~ 100 fs) to several topics of interest in life sciences. The laser system is based on a custom PHAROS source (Light Conversion Ltd) - delivering up to 1.5 mJ at 1030 nm with 2kHz rep rate. The IR pulse can be used to pump a Harmonic Generator stage (HIRO) - with II (515 nm), III (343 nm), and IV (258 nm) harmonic outputs - and/or an Optical Parametric Amplifier (ORPHEUS) - tunable from 0.2 μm to 3 μm. An additional Harmonic Generator is also available to produce harmonics of the oscillator output at 80 MHZ. Various equipments such as delay stages, spectrographs and CCD cameras allow us to perform time resolved fluorescence experiments on molecules of biological interest (5-benzyl-uracil and glucose oxidase are currently studied).
USE OF DA VINCI VIRTUAL REALITY SIMULATOR FOR TRAINING IN ROBOTIC GYNECOLOGICAL SURGERY
The aim of this task is to demonstrate that da Vinci VR simulator is able to fasten the learning curve of doctors with no previous experience in gynecologic laparoscopic surgery.
WEBCASTING OF ROBOTIC GYNECOLOGICAL SURGERY
The aim of this task is to increase the diffusion of robotic gynecological surgery through a webcasting of to the live surgeries with possibility of an interaction of trainees and surgeons