The introduction of Robot-Assisted Minimally Invasive Surgery (RAMIS) has changed the way certain diseases are treated, offering numerous advantages to both patients and surgeons. However, the latter immediately experienced the need to acquire new manual skills to control such tools, as well as feeling the need to adapt their surgical expertise. Many studies have shown that, in this context, virtual reality simulators are a highly effective tool, bringing advantages such as adaptability and consistency of procedures while avoiding the ethical problems associated with traditional training methods. In addition, over the years, many attempts have been made to develop a standardized training curriculum with the goal of maximizing the learning process by combining dexterity and the development of skills more related to surgery, but no common decision has been reached. In this context, this work of thesis has the goal of developing a virtual reality training simulation platform and some learning tasks for the acquiring and the improvement of selected dexterity skills. The console of the da Vinci Research Kit (dVRK) was employed to control the simulated robotic arms. The simulation is realized through Unity3D, which allows the development of simulated scenes, including both rigid and soft body interaction, with an excellent graphic rendering. An experimental study is also proposed involving 9 users with less than one year of previous experience in robotic teleoperation, who have undergone one training session per each of the two tasks taken into account. Specifically, those two tasks aim to improve the maneuverability of the components of the surgical robot and the growth of soft skills. Regarding the first task, the performances have been evaluated in terms of Execution time and Task completion. The analysis has shown a significant enhancement at the end of the training session, thus in the last two repetitions, compared to the results achieved at the beginning. For what concern the second task, instead, there have been analyzed the Execution time and Trajectory error. Unlike the analysis described above, in this case, no statistical significance has been demonstrated, even though the learning curves show better outcomes after the practice. Further analysis should be done involving a bigger sample size, even with clinicians, to assess the validity of training better. To this end, it could be useful to validate the efficiency over multiple sessions. Finally, the virtual reality simulator developed could be enriched with more tasks to investigate other proper movements that characterize RAMIS procedures.
L’introduzione della chirurgia robotica mininvasiva (Robotic Assisted Minimally Invasive Surgery, RAMIS) ha rivoluzionato il modo di trattare molte patologie, apportando benefici sia ai pazienti che ai chirurghi. Ciò nonostante, questi ultimi hanno da subito riscontrato la necessità di acquisire nuove abilità manuali per controllare questi strumenti, oltre che il bisogno di adattare la propria esperienza nell’ambito chirurgico. Molti studi hanno dimostrato come i simulatori di realtà virtuale siano uno strumento efficace in questo contesto, portando benefici quali l’adattabilità e la ripetibilità delle procedure, evitando i problemi etici legati ai metodi di addestramento tradizionali. Inoltre, nel corso degli anni si è cercato di sviluppare una strategia standard di allenamento che massimizzasse l’apprendimento, combinando lo sviluppo di abilità manuali e di competenze più prettamente legate alle procedure chirurgiche, ma non si è giunti ad una conclusione comune. A questo proposito, il lavoro di tesi presentato ha lo scopo di sviluppare un ambiente di simulazione di realtà virtuale e alcuni task per l’acquisizione e il perfezionamento di abilità manuali fondamentali nella chirurgia robotica. In particolare, è stata utilizzata la console del da Vinci Research Kit (dVRK) per il controllo dei bracci robotici simulati, mentre il software impiegato è stato Unity3D. Quest'ultimo consente la creazione di scene di simulazione con un’ottima resa grafica, includendo la possibilità di fare interagire sia tessuti rigidi che deformabili. È stato condotto uno studio sperimentale coinvolgendo 9 utenti con meno di un anno di esperienza nell'ambito della teleoperazione di sistemi robotici, i quali sono stati sottoposti ad una sessione di allenamento per ciascuno dei due task presi in considerazione. Nello specifico, i due task coinvolti nello studio mirano al miglioramento della manovrabilità delle componenti del robot chirurgico e allo sviluppo di competenze trasversali. Per quanto riguarda il primo task, le prestazioni sono state valutate in termini di Tempo di esecuzione e Completamento del task. L'analisi ha mostrato un miglioramento significativo alla fine della sessione di addestramento in relazione ai risultati registrati all'inizio. In merito al secondo task, invece, sono stati analizzati il Tempo di esecuzione e l'Errore di traiettoria. Al contrario di quanto descritto precedentemente, in questo caso non è stata riscontrata rilevanza statistica, anche se le curve di apprendimento relative ad entrambe le metriche hanno mostrato risultati migliori dopo l'allenamento. Ulteriori analisi potrebbero essere condotte con un numero maggiore di utenti, includendo anche medici, per meglio dimostrare la validità del training. A questo proposito, potrebbe essere utile valutare l'efficienza su più sessioni ripetute. Inoltre, il simulatore di realtà virtuale sviluppato potrebbe essere arricchito con nuovi task per valutare altri movimenti tipici delle procedure di chirurgia robotica mininvasiva.
Development of a virtual reality simulation framework for training in robotic-assisted surgery
Gloria, Martina
2021/2022
Abstract
The introduction of Robot-Assisted Minimally Invasive Surgery (RAMIS) has changed the way certain diseases are treated, offering numerous advantages to both patients and surgeons. However, the latter immediately experienced the need to acquire new manual skills to control such tools, as well as feeling the need to adapt their surgical expertise. Many studies have shown that, in this context, virtual reality simulators are a highly effective tool, bringing advantages such as adaptability and consistency of procedures while avoiding the ethical problems associated with traditional training methods. In addition, over the years, many attempts have been made to develop a standardized training curriculum with the goal of maximizing the learning process by combining dexterity and the development of skills more related to surgery, but no common decision has been reached. In this context, this work of thesis has the goal of developing a virtual reality training simulation platform and some learning tasks for the acquiring and the improvement of selected dexterity skills. The console of the da Vinci Research Kit (dVRK) was employed to control the simulated robotic arms. The simulation is realized through Unity3D, which allows the development of simulated scenes, including both rigid and soft body interaction, with an excellent graphic rendering. An experimental study is also proposed involving 9 users with less than one year of previous experience in robotic teleoperation, who have undergone one training session per each of the two tasks taken into account. Specifically, those two tasks aim to improve the maneuverability of the components of the surgical robot and the growth of soft skills. Regarding the first task, the performances have been evaluated in terms of Execution time and Task completion. The analysis has shown a significant enhancement at the end of the training session, thus in the last two repetitions, compared to the results achieved at the beginning. For what concern the second task, instead, there have been analyzed the Execution time and Trajectory error. Unlike the analysis described above, in this case, no statistical significance has been demonstrated, even though the learning curves show better outcomes after the practice. Further analysis should be done involving a bigger sample size, even with clinicians, to assess the validity of training better. To this end, it could be useful to validate the efficiency over multiple sessions. Finally, the virtual reality simulator developed could be enriched with more tasks to investigate other proper movements that characterize RAMIS procedures.File | Dimensione | Formato | |
---|---|---|---|
2022_07_Gloria.pdf
solo utenti autorizzati dal 05/07/2023
Dimensione
13.28 MB
Formato
Adobe PDF
|
13.28 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/190135