Federico Barbagli

Stanford University, Palo Alto, California, United States

Are you Federico Barbagli?

Claim your profile

Publications (69)30.93 Total impact

  • 07/2014;
  • 07/2014;
  • [Show abstract] [Hide abstract]
    ABSTRACT: We present a method for real-time, freehand 3D ultrasound (3D-US) reconstruction of moving anatomy, with specific application towards guiding the catheter ablation procedure in the left atrium. Using an intracardiac echo (ICE) catheter with a pose (position/orientation) sensor mounted to its tip, we continually mosaic 2D-ICE images of a left atrium phantom model to form a 3D-US volume. Our mosaicing strategy employs a probabilistic framework based on simultaneous localization and mapping (SLAM), a technique commonly used in mobile robotics for creating maps of unexplored environments. The measured ICE catheter tip pose provides an initial estimate for compounding 2D-ICE image data into the 3D-US volume. However, we simultaneously consider the overlap-consistency shared between 2D-ICE images and the 3D-US volume, computing a "corrected" tip pose if need be to ensure spatially-consistent reconstruction. This allows us to compensate for anatomic movement and sensor drift that would otherwise cause motion artifacts in the 3D-US volume. Our approach incorporates 2D-ICE data immediately after acquisition, allowing us to continuously update the registration parameters linking sensor coordinates to 3D-US coordinates. This, in turn, enables real-time localization and display of sensorized therapeutic catheters within the 3D-US volume for facilitating procedural guidance.
    Ultrasound in medicine & biology 12/2011; 37(12):2037-54. DOI:10.1016/j.ultrasmedbio.2011.08.007 · 2.10 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we extend the concept of the contrast sensitivity function—used to evaluate video projectors—to the evaluation of haptic devices. We propose using human observers to determine if vibrations rendered using a given haptic device are accompanied by artifacts detectable to humans. This determination produces a performance measure that carries particular relevance to applications involving texture rendering. For cases in which a device produces detectable artifacts, we have developed a protocol that localizes deficiencies in device design and/or hardware implementation. In this paper, we present results from human vibration detection experiments carried out using three commercial haptic devices and one high performance voice coil motor. We found that all three commercial devices produced perceptible artifacts when rendering vibrations near human detection thresholds. Our protocol allowed us to pinpoint the deficiencies, however, and we were able to show that minor modifications to the haptic hardware were sufficient to make these devices well suited for rendering vibrations, and by extension, the vibratory components of textures. We generalize our findings to provide quantitative design guidelines that ensure the ability of haptic devices to proficiently render the vibratory components of textures.
    IEEE Transactions on Haptics 03/2011; 4:134-146. DOI:10.1109/TOH.2011.5 · 2.03 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: We present a method for catheter localization in the left atrium based on the unscented particle filter (UPF), a Monte Carlo method employed in stochastic state estimation. Using an intracardiac echo (ICE) ultrasound catheter, we acquire ultrasound images of the atrium from multiple configurations and iteratively determine the catheter’s pose with respect to anatomy. At each time step, the catheter’s change in pose is determined using either a six-degree-of-freedom electromagnetic pose sensor or a robotic guide catheter whose kinematics serve as a pseudo-pose measurement. Sensor and kinematic model uncertainties are explicitly considered when computing the localization estimate. Acquired ultrasound images are compared with simulated ultrasound images based on segmented computed tomography (CT) or magnetic resonance (MR) data of the left atrium. The results of these comparisons are used to refine the localization estimate. After considering less than 30 seconds’ worth of ICE data, our algorithm converges to an accurate pose estimate. Furthermore, our algorithm is robust to sensor drift and kinematic model errors, as well as gradual, unmodeled movements in the anatomy. Such problems typically complicate traditional image-based localization.
    The International Journal of Robotics Research 01/2010; 29:643-665. DOI:10.1177/0278364909105332 · 2.50 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The catheter ablation procedure is a minimally invasive surgery used to treat atrial fibrillation. Difficulty visualizing the catheter inside the left atrium anatomy has led to lengthy procedure times and limited success rates. In this paper, we present a set of algorithms for reconstructing 3D ultrasound data of the left atrium in real-time, with an emphasis on automatic tissue classification for improved clarity surrounding regions of interest. Using an intracardiac echo (ICE) ultrasound catheter, we collect 2D-ICE images of a left atrium phantom from multiple configurations and iteratively compound the acquired data into a 3D-ICE volume. We introduce two new methods for compounding overlapping US data-occupancy-likelihood and response-grid compounding-which automatically classify voxels as "occupied" or "clear," and mitigate reconstruction artifacts caused by signal dropout. Finally, we use the results of an ICE-to-CT registration algorithm to devise a response-likelihood weighting scheme, which assigns weights to US signals based on the likelihood that they correspond to tissue-reflections. Our algorithms successfully reconstruct a 3D-ICE volume of the left atrium with voxels classified as "occupied" or "clear," even within difficult-to-image regions like the pulmonary vein openings. We are robust to dropout artifact that plagues a subset of the 2D-ICE images, and our weighting scheme assists in filtering out spurious data attributed to ghost-signals from multi-path reflections. By automatically classifying tissue, our algorithm precludes the need for thresholding, a process that is difficult to automate without subjective input. Our hope is to use this result towards developing 3D ultrasound segmentation algorithms in the future.
    International Journal of Computer Assisted Radiology and Surgery 09/2009; 4(5):425-37. DOI:10.1007/s11548-009-0354-6 · 1.66 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Human vibrotactile detection experiments were used to compare temporal sinusoids displayed on three commercial haptic devices to a high-fidelity linear voice-coil actuator. The three commercial haptic devices we used span the cost spectrum, supposing that cost of a device is correlated with the fidelity of its virtual textures. This turned out not to be the case. The results indicated that none of the three haptic devices we tested were able to render perceptually distortion-free, periodically regular vibrations at detection threshold levels. Further investigation into the electrical and mechanical device properties that limited the performance of these devices revealed that D/A resolution, amplifier non-linearity and stiction were the primary sources of signal corruption.
    World Haptics 2009 - Third Joint EuroHaptics conference and Symposium on Haptic Interfaces for Virtual Environment and Teleoperator Systems, Salt Lake City, UT, USA, 18-20 March 2009; 01/2009
  • [Show abstract] [Hide abstract]
    ABSTRACT: We present a method for registering real-time ultrasound of the left atrium to an outdated, anatomic surface mesh model, whose shape differs from that of the anatomy. Using an intracardiac echo (ICE) catheter with mounted 6DOF electromagnetic position/orientation sensor (EPS), we acquire images of the left atrium and determine where the ICE catheter must be positioned relative to the surface mesh to generate similar, "virtual" ICE images. Further, we use an affine warping model to infer how the shape of the surface mesh differs from that of the atrium. Our registration and warping algorithm allows us to display EPS-sensorized catheters inside the surface mesh, facilitating guidance for left atrial procedures. By solving for the atrium-to-mesh warping parameters, we ensure that tissue contact in the anatomy is properly displayed as tissue contact in the mesh. After considering less than thirty seconds worth of ICE data, we are able to accurately localize EPS measurements within the surface mesh, despite surface mesh warpings of up to +/-20% along and about the principal axes of the left atrium. Further, because our estimation framework is iterative and continuous, our accuracy improves as new data is acquired.
    Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 01/2009; 2009:5567-70. DOI:10.1109/IEMBS.2009.5333739
  • [Show abstract] [Hide abstract]
    ABSTRACT: To describe a novel flexible robotic system for performing retrograde intrarenal surgery. Remote robotic flexible ureterorenoscopy was performed bilaterally in 5 acute swine (10 kidneys). A novel 14F robotic catheter system, which manipulated a passive optical fiberscope mounted on a remote catheter manipulator was used. The technical feasibility, efficiency, and reproducibility of accessing all calices were assessed. Additionally, laser lithotripsy of calculi and laser ablation of renal papillae were performed. The robotic catheter system could be introduced de novo in eight ureters; two ureters required balloon dilation. The ureteroscope could be successfully manipulated remotely into 83 (98%) of the 85 calices. The time required to inspect all calices within a given kidney decreased with experience from 15 minutes in the first kidney to 49 seconds in the last (mean 4.6 minutes). On a visual analog scale (1, worst to 10, best), the reproducibility of caliceal access was rated at 8, and instrument tip stability was rated at 10. A renal pelvic perforation constituted the solitary complication. Histologic examination of the ureter showed changes consistent with acute dilation without areas of necrosis. A novel robotic catheter system is described for performing retrograde ureterorenoscopy. The potential advantages compared with conventional manual flexible ureterorenoscopy include an increased range of motion, instrument stability, and improved ergonomics. Ongoing refinement is likely to expand the role of this technology in retrograde intrarenal surgery in the near future.
    Urology 08/2008; 72(1):42-6. DOI:10.1016/j.urology.2008.01.076 · 2.13 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The increasing demand for virtual reality applications in several scientific disciplines feeds new research perspectives dealing with robotics, automation, and computer science. In this context, one of the topics is the design of advanced force-feedback devices allowing not only kinesthetic interaction with virtual objects but also locomotion and navigation inside virtual worlds. This has the main advantage to stimulate human vestibular apparatus, thus increasing the overall realism of simulation. Particularly, this paper deals with mobile haptic interfaces (MHIs), built by combining standard force-feedback devices with mobile platforms. We investigated which factors may affect the transparency of this kind of devices, identifying in mobile robot dynamics a possible cause of loss of transparency. Hence, in this paper, we present a method to analyze dynamic performance of an MHI and some basic guidelines to design controller in order to meet desired specifications. Experimental validation of the theoretical results is reported.
    IEEE Transactions on Robotics 07/2008; DOI:10.1109/TRO.2008.924262 · 2.65 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Haptic interaction in six degrees of freedom is critical to numerous applications, but is still prohibitively complex for realistic environments. This paper presents an approach to rendering six-degree-of-freedom contact among virtual objects using a novel data structure referred to as an implicit sphere tree. This data structure allows an extremely compact representation of volumetric objects and extremely rapid intersection testing among objects, which broadens the scope of virtual environments that can be rendered in six degrees of freedom at interactive rates. We introduce this data structure, along with appropriate techniques for collision detection and haptic rendering, and demonstrate its efficiency in representing and manipulating complex models.
    Haptic interfaces for virtual environment and teleoperator systems, 2008. haptics 2008. symposium on; 04/2008
  • [Show abstract] [Hide abstract]
    ABSTRACT: One of the most important advantages of computer simulators for surgical training is the opportunity they afford for independent learning. However, if the simulator does not provide useful instructional feedback to the user, this advantage is significantly blunted by the need for an instructor to supervise and tutor the trainee while using the simulator. Thus, the incorporation of relevant, intuitive metrics is essential to the development of efficient simulators. Equally as important is the presentation of such metrics to the user in such a way so as to provide constructive feedback that facilitates independent learning and improvement. This paper presents a number of novel metrics for the automated evaluation of surgical technique. The general approach was to take criteria that are intuitive to surgeons and develop ways to quantify them in a simulator. Although many of the concepts behind these metrics have wide application throughout surgery, they have been implemented specifically in the context of a simulation of mastoidectomy. First, the visuohaptic simulator itself is described, followed by the details of a wide variety of metrics designed to assess the user's performance. We present mechanisms for presenting visualizations and other feedback based on these metrics during a virtual procedure. We further describe a novel performance evaluation console that displays metric-based information during an automated debriefing session. Finally, the results of several user studies are reported, providing some preliminary validation of the simulator, the metrics, and the feedback mechanisms. Several machine learning algorithms, including Hidden Markov Models and a Naïve Bayes Classifier, are applied to our simulator data to automatically differentiate users' expertise levels.
    Computer Aided Surgery 04/2008; 13(2):63-81. DOI:10.3109/10929080801957712 · 1.08 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We present a method for registering position and orientation data collected from an electroanatomic mapping system (EMS) to a surface mesh based on segmented Computed Tomography (CT) or Magnetic Resonance (MR) images of the left atrium. Our algorithm is based on the Unscented Particle Filter (UPF) for stochastic state estimation. Using an intracardiac echo (ICE) ultrasound catheter with mounted mapping sensor, we acquire ultrasound images of the atrium from multiple configurations and iteratively determine the catheter's pose with respect to anatomy. After considering less than a minute's worth of ICE data, the algorithm converges to an accurate pose estimate which, in turn, yields the registration parameters transforming EMS coordinates to mesh coordinates. The iterative framework of the UPF allows us to be robust to unmodeled EMS noise and drift, problems which complicate traditional registration methods assuming regularity in image data structure.
  • [Show abstract] [Hide abstract]
    ABSTRACT: We present a sensorless method for localizing a robotic catheter inside the left atrium using intracardiac echo (ICE) ultrasound. As the robotic catheter navigates inside the anatomy, its kinematics provide a rough estimate of change in pose. At the same time, an ICE catheter inserted through the robotic catheter’s lumen acquires images to refine this estimate. Our algorithm is based on the Unscented Particle Filter (UPF) for stochastic state estimation. We iteratively determine the catheter’s pose by comparing newly acquired ICE images to segmented Computed Tomography (CT) images of the left atrium. After considering less than fifteen second’s worth of ICE data, the algorithm converges to an accurate pose estimate despite significant deviations from the kinematic model, and unmodeled movements in the anatomy.
    Experimental Robotics, The Eleventh International Symposium, ISER 2008, July 13-16, 2008, Athens, Greece; 01/2008
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents several new metrics related to bone removal and suctioning technique in the context of a mastoidectomy simulator. The expertise with which decisions as to which regions of bone to remove and which to leave intact is evaluated by building a Naïve Bayes classifier using training data from known experts and novices. Since the bone voxel mesh is very large, and many voxels are always either removed or not removed regardless of expertise, the mutual information was calculated for each voxel and only the most informative voxels used for the classifier. Leave-out-one cross validation showed a high correlation of calculated expert probabilities with scores assigned by instructors. Additional metrics described in this paper include those for assessing smoothness of drill strokes, proper drill burr selection, sufficiency of suctioning, two-handed tool coordination, and application of appropriate force and velocity magnitudes as functions of distance from critical structures.
    Studies in health technology and informatics 02/2007; 125:427-32.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: One of the primary barriers to the acceptance of surgical simulators is that most simulators still require a significant amount of an instructing surgeon's time to evaluate and provide feedback to the students using them. Thus, an important area of research in this field is the development of metrics that can enable a simulator to be an essentially self-contained teaching tool, capable of identifying and explaining the user's weaknesses. However, it is essential that these metrics be validated in able to ensure that the evaluations provided by the "virtual instructor" match those that the real instructor would provide were he/she present. We have previously proposed a number of algorithms for providing automated feedback in the context of a mastoidectomy simulator. In this paper, we present the results of a user study in which we attempted to establish construct validity (with inter-rater reliability) for our simulator itself and to validate our metrics. Fifteen subjects (8 experts, 7 novices) were asked to perform two virtual mastoidectomies. Each virtual procedure was recorded, and two experienced instructing surgeons assigned global scores that were correlated with subjects' experience levels. We then validated our metrics by correlating the scores generated by our algorithms with the instructors' global ratings, as well as with metric-specific sub-scores assigned by one of the instructors.
    Studies in health technology and informatics 02/2007; 125:421-6.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper explores the use of haptic feedback to teach an abstract motor skill that requires recalling a sequence of forces. Participants are guided along a trajectory and are asked to learn a sequence of one- dimensional forces via three paradigms: haptic training, visual training, or combined visuohaptic training. The extent of learning is measured by accuracy of force recall. We find that recall following visuohaptic training is significantly more accurate than recall following visual or haptic training alone, although haptic training alone is inferior to visual training alone. This suggests that in conjunction with visual feedback, haptic training may be an effective tool for teaching sensorimotor skills that have a force- sensitive component to them, such as surgery. We also present a dynamic programming paradigm to align and compare spatiotemporal haptic trajectories.
    Second Joint EuroHaptics Conference and Symposium on Haptic Interfaces for Virtual Environment and Teleoperator Systems (WHC 2007), 22-24 March 2007, Tsukuba, Japan; 01/2007
  • Proceedings of the 21th Computer Assisted Radiology and Surgery CARS; 01/2007
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We present a new tactile display for rendering rotational friction at the fingertips. To inform the design and control of this device, we have conducted an initial set of perceptual experiments to investigate perceptual thresholds for rates of rotational slipping at the fingertips. It was found that hu- mans can discriminate a change in rotational sliding rate of 4.13-28.04 rad/sec from a reference rate of 1.57-25.13 rad/sec. An initial device design and prototype are pre- sented. The device utilizes a small friction disk housed in a thimble, driven remotely from a motor worn on the user's wrist. The device directly integrates with a robotic force feedback arm, preserving the force feedback functionality while adding the tactile information on frictional sliding. Insights from the presented psychophysical experiments and initial device prototype will be utilized in the design of the next generation of rotation friction devices that will be used in multi-finger grasp studies.
    Second Joint EuroHaptics Conference and Symposium on Haptic Interfaces for Virtual Environment and Teleoperator Systems (WHC 2007), 22-24 March 2007, Tsukuba, Japan; 01/2007
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Visual and haptic simulation of bone surgery can support and extend current surgical training techniques. The authors present a system for simulating surgeries involving bone manipulation, such as temporal bone surgery and mandibular surgery, and discuss the automatic computation of surgical performance metrics. Experimental results confirm the system's construct validity.
    IEEE Computer Graphics and Applications 12/2006; 26(6):48-57. DOI:10.1109/MCG.2006.140 · 1.12 Impact Factor

Publication Stats

1k Citations
30.93 Total Impact Points

Institutions

  • 2003–2011
    • Stanford University
      • • Department of Computer Science
      • • Department of Mechanical Engineering
      Palo Alto, California, United States
  • 2008
    • Mountain View College
      Mountain View, California, United States
  • 2003–2006
    • Università degli Studi di Siena
      • Department of Information Engineering and Mathematical
      Siena, Tuscany, Italy
  • 2005
    • Università di Pisa
      Pisa, Tuscany, Italy
  • 1999–2002
    • Scuola Normale Superiore di Pisa
      Pisa, Tuscany, Italy