Tesi etd-03102025-105512
Link copiato negli appunti
Tipo di tesi
Dottorato
Autore
FAORO, GIOVANNI
URN
etd-03102025-105512
Titolo
Advanced methods for ultrasound-based minimally invasive procedures
Settore scientifico disciplinare
ING-INF/06
Corso di studi
Istituto di Biorobotica - PHD IN BIOROBOTICA
Commissione
relatore Prof.ssa MENCIASSI, ARIANNA
Presidente Prof. CAPINERI, LORENZO
Membro Prof.ssa GIANNAROU, STAMATIA
Presidente Prof. CAPINERI, LORENZO
Membro Prof.ssa GIANNAROU, STAMATIA
Parole chiave
- robotic ultrasound
- image analysis
- deep learning
- anatomical reconstruction
- monitoring
Data inizio appello
30/05/2025;
Disponibilità
parziale
Riassunto analitico
The last few decades have witnessed an exponential growth in medical technologies, from the introduction and refinement of different medical imaging techniques to the application of robotics and artificial intelligence to surgery. The favorable interplay between these technologies could be disruptive for the medical field, facilitating and improving current standards and further enabling the definition of new ones. As an example, artificial intelligence is now widely studied as a powerful tool to improve medical image analysis, reducing the burden on clinicians. Robots have been applied in several scenarios of the surgical practice standardizing the clinical outcome. Furtermore, they could help improving and standardizing medical imaging, by otpimizing, e.g., probe-tissue contact during echography and avoiding operator musculoskeletal disorders due to substained non ergonomic postures. Viceversa, medical imaging represents an optimal source of data for both artificial intelligence training and robotic systems guidance and awareness. Robotic ultrasound (RUS) platforms are a clear example in this context, yet despite they have been studied for decades the technical complexities of these technologies and the strict requirements of the surgical field are still hampering the adoption of RUS approaches. Among the factors hindering the wide use of these platforms, one of the main problems is their lacking awareness.
This thesis addresses the development of a novel RUS platform incrementally improving awareness, adding useful information to effectively perform diagnostic and interventional procedures. From sensorization to therapy monitoring, passing through anatomical reconstruction and target tracking, high autonomy levels are reached. This thesis reports a set of works that investagate the development and improvement of these fundamental technological blocks that stand at the foundation of robust, autonomous and aware robotic platforms. Major attention is paied to cardiovascular surgeries as these deal with the leading cause of death worlwide, thus clearly representing a branch in which precise and safe automation could play an important role in the medical field given the increasing world population.
A novel RUS platform able to reconstruct 3D vascular models, plan catheter insertion, intra operatively register such plan and automatically execute it is presented. Furthermore, a robust ultrasound thermometry method is developed potentially enabling plaque removal through catheter tip heating, a novel and alternative approach to standard stenting. To demonstrate the generalization ability of some of the proposed methods, parallelisms are drawn with a robotic focused ultrasound, a non-invasive therapeutic technology that could potentially revolutionize the treatment approach of various medical conditions.
This thesis addresses the development of a novel RUS platform incrementally improving awareness, adding useful information to effectively perform diagnostic and interventional procedures. From sensorization to therapy monitoring, passing through anatomical reconstruction and target tracking, high autonomy levels are reached. This thesis reports a set of works that investagate the development and improvement of these fundamental technological blocks that stand at the foundation of robust, autonomous and aware robotic platforms. Major attention is paied to cardiovascular surgeries as these deal with the leading cause of death worlwide, thus clearly representing a branch in which precise and safe automation could play an important role in the medical field given the increasing world population.
A novel RUS platform able to reconstruct 3D vascular models, plan catheter insertion, intra operatively register such plan and automatically execute it is presented. Furthermore, a robust ultrasound thermometry method is developed potentially enabling plaque removal through catheter tip heating, a novel and alternative approach to standard stenting. To demonstrate the generalization ability of some of the proposed methods, parallelisms are drawn with a robotic focused ultrasound, a non-invasive therapeutic technology that could potentially revolutionize the treatment approach of various medical conditions.
File
| Nome file | Dimensione |
|---|---|
Ci sono 1 file riservati su richiesta dell'autore. |
|