Disruptive, image-guided smart solutions.Hands-free, Scale-free, Stress-free.www.spectrorural.com

IA, Data Science & Digital Health.

Research in the domain of Radiomics (Quantitative Medicine) exploring machine learning models for quantitative Radiology image analysis as a decision support tool in the context of prognostic prediction and computer-aided diagnosis of Idiopathic Lung Diseases.Keywords: Artificial Intelligence, Radiomics, Medical Image Analysis

This project aimed to minimize radiation dose delivered to patient and to medical staff under robot-driven Interventional Radiology (image-guided) procedures. By assisting robot-driven 3D needle planning under space-constrained scenario, we target the development of new clinical applications, which are not currently feasible by simply using existent clinical approaches in interventional radiology.I also support research grant writing and management of consortium involving European… Show more This project aimed to minimize radiation dose delivered to patient and to medical staff under robot-driven Interventional Radiology (image-guided) procedures. By assisting robot-driven 3D needle planning under space-constrained scenario, we target the development of new clinical applications, which are not currently feasible by simply using existent clinical approaches in interventional radiology.I also support research grant writing and management of consortium involving European universities and companies in the Interventional Robotics domain. Key-words: Computer Vision, Interventional Radiology Robots, Medical Image Analysis. Show less

My PhD work supported robot-driven minimally invased procedures using slender tools (semi-rigid needles), often used for tissue ablation, biopsies, delivery of medication, anesthesia, etc.A novel Physically-based model was devised to compensate for needle deflection (and tissue deformation) during its interaction with soft tissues. A predictive image-guided model was the core of the development of a needle path planning approach with needle steering and entry-point assistance.We… Show more My PhD work supported robot-driven minimally invased procedures using slender tools (semi-rigid needles), often used for tissue ablation, biopsies, delivery of medication, anesthesia, etc.A novel Physically-based model was devised to compensate for needle deflection (and tissue deformation) during its interaction with soft tissues. A predictive image-guided model was the core of the development of a needle path planning approach with needle steering and entry-point assistance.We develeloped a 3D navigation tool with assistance resources based on Augmented Reality. A modular teleoperation architecture was designed to assist Interventional Radiology applications in ARCS (Abdomino-pelvic Robotic-driven semi-rigid needle insertion performed in CT or MR image-guided Scenario). The system was built in C/C++ on top of CamiTK (Computer-Assisted Medical Intervention Toolkit) and OpenCV libraries. Results obtained from in-vitro experiments using different robot platforms (LPR and Raven-II) showed potential of our method to be placed between the best in the literature in terms of performance and robustness of needle targeting. Moreover, currently, it is the only approach providing a solution for teleoperated needle rotation, translation and insertion into ARCS. Show less

Automatic detection and tracking of robotic-driven flexible needles (nitinol) under ultrasound and video images. The tool aimed to provide position and orientation of slender tools during c-like (nonlinear) trajectories. It was integrated into a visual servoing system to steer needle using duty cycle technique. The tool was implemented in C, using OpenCV under Linux.

Development of Computer-Aided Diagnosis System in Radiology. R&D in partnership with FORP - School of Odontology of Ribeirão Preto, University of Sao Paulo - Brazil. The image analysis and registration tool was developed in C language and OpenCV library.

Development of tools for web-based geographical databases and big data handling. Developed in C/C#.

Computer vision system for driver´s assistance. Neuromorphic embedded system for pattern recognition of driver´s sleepiness and distraction. Developed in C/C++ and OpenCV.