The introduction of robots in the medical field has revolutionized the way of performing surgery, increasing benefits for both patients and surgeons. Nevertheless, Robot-Assisted Minimally Invasive Surgery (RAMIS) introduces new control modalities of the operating instruments. Practical skills training represents a key step for the preparation of robotic surgeons. The capability of Virtual Reality (VR) simulators to create a safe and standardized environment for training, as well as the possibility to provide performance parameters, has led to a great increase of their exploitation in the surgical training process. Virtual reality simulators also open attractive opportunities in terms of augmentation, a concept which has already been explored in rehabilitation and sport training. Augmentation has proven to be an indispensable component to optimize motor learning, and it comprises the concepts of guidance and feedback. Augmented guidance is typically exploited as a way to teach the trainee the ideal movement or the procedure to be learnt, while augmented feedback represents a useful means for error correction. For the purpose of this study a virtual reality needle-driving task was integrated with augmentation in the form of guidance and feedback during the same training session. First guidance was exploited to teach the trainee how to carry out the task and was then substituted by feedback for improving the gesture’s accuracy. As first objective, the effect on motor learning of this novel combination of guidance and feedback was investigated. Moreover, these augmentation modalities were explored over the visual and haptic domains, with the additional aim of identifying the most appropriate sensory modality for practical skills acquisition. To address the purpose of this research, the master console of a da Vinci Research Kit (dVRK) (Intuitive Surgical Inc., Sunnyvale) was exploited together with a virtual reality simulation environment: Assisted Teleoperation with Augmented Reality (ATAR). 32 non-medical participants were involved in a user study. They were randomly subdivided into four groups: one group received no training augmentation, while the others underwent visual, haptic and visuo-haptic augmentation, respectively. Users’ performance was assessed with means of objective metrics involving Phase Completion capability and Time-Accuracy, as well as subjective metrics derived from a post-experimental survey. Results showed a significant improvement, initially introduced by guidance, in the Phase Completion capability of all experimental groups against control. The introduction of the augmented guidance also led to an enhancement of the performance over the Time-Accuracy metrics. The subsequent application of feedback didn't seem to cause a comparable improvement in the objective metrics. According to the accuracy analysis the experimental groups outperformed control at the end of training, while no statistically significant differences were found across sensory domains. For what concerns the execution time, it presented significant lower values in groups involving haptics at evaluation. Interesting results were also derived from the post-experimental survey. Users belonging to the visual group assessed a higher increase in the cognitive overload due to the introduction of the augmentation. In addition, the majority of participants from the visuo-haptic group selected the haptic modality as the most helpful in the error minimization. From these results the potentiality of the haptic augmentation was derived. Further analysis on long term learning and skill retention could better investigate this hypothesis. In addition, the interaction between augmentation modalities could be analysed more thoroughly with means of different training session of subsequent feedback and guidance, as well as protocols involving only guidance or only feedback.
L'introduzione dei robot in medicina ha rivoluzionato il modo di eseguire le operazioni chirurgiche, portando a benefici sia per i pazienti che per i chirurghi. Tuttavia, nonostante gli innumerevoli vantaggi che comporta, la chirurgia robotica (Robot-Assisted Minimally Invasive Surgery, RAMIS) introduce anche nuove e complesse modalità di controllo degli strumenti chirurgici. Di conseguenza, l’acquisizione di abilità pratiche rappresenta una caratteristica fondamentale per la preparazione dei chirurghi robotici. I simulatori di Realtà Virtuale (VR) sono emersi negli ultimi anni come promettente alternativa all’esperienza diretta in sala, con i vantaggi di ripetibilità delle procedure e assenza del danno potenziale al paziente. Inoltre, tali simulatori introducono la possibilità di fornire aumentazione durante il training di abilità chirurgiche. L'aumentazione si è dimostrata una componente indispensabile nell'apprendimento motorio e comprende i concetti di guida e feedback. La guida aumentata viene tipicamente sfruttata come mezzo per insegnare al tirocinante il movimento ideale o la procedura da apprendere, mentre il feedback aumentato rappresenta uno strumento utile per la correzione dell’errore. Ai fini di questo studio, un task in realtà virtuale è stato integrato con tecniche di aumentazione sotto forma di guida e feedback durante la stessa sessione di training. L'introduzione della guida ha avuto la finalità di insegnare all’utente i movimenti che caratterizzano il task. Una volta acquisita la procedura da parte dell'utente, la guida è stata quindi sostituita con il feedback per migliorare l'accuratezza dei gesti. Come primo obiettivo del lavoro, sono stati analizzati i vantaggi in termini di apprendimento motorio di questa innovativa combinazione di guida e feedback. Inoltre, tali modalità di aumentazione sono state testate del dominio sensoriale visivo e aptico. E' stato quindi definito un secondo scopo della ricerca, che consiste nell'identificare la modalità sensoriale più appropriata per l'acquisizione di abilità pratiche. Ai fini di questo lavoro, la master console di un da Vinci Research Kit (dVRK) (Intuitive Surgical Inc., Sunnyvale) è stata sfruttata insieme a un ambiente di simulazione in realtà virtuale: ATAR. 32 partecipanti sono stati coinvolti in uno studio sperimentale che li ha suddivisi in modo casuale in quattro gruppi: un gruppo non ha ricevuto alcuna aumentazione durante il training, mentre gli altri sono stati sottoposti rispettivamente a un'aumentazione visiva, aptica e visuo-aptica. Le prestazioni degli utenti sono state valutate tramite metriche oggettive che coinvolgono la capacità di Completamento della Fase e il Tempo-Accuratezza, ma anche tramite metriche soggettive derivate da un'indagine post-sperimentale. I risultati hanno mostrato un miglioramento significativo rispetto al controllo, inizialmente introdotto dalla guida, nella capacità di completamento del task in tutti i gruppi sperimentali. L'introduzione della guida aumentata ha avuto effetti positivi in termini di performance anche sulla metrica di Tempo-Accuratezza. La successiva applicazione del feedback non ha sembrato portare a un miglioramento comparabile nelle metriche di valutazione oggetive. Dall'analisi di accuratezza è emerso come i gruppi sperimentali abbiano superato il controllo in quanto a performance alla fine del training, mentre non sono state riscontrate differenze statisticamente significative tra i domini sensoriali. Per quanto riguarda il tempo di esecuzione, sono stati ottenuti valori significativamente più bassi nei gruppi che coinvolgono l’aptica. Informazioni rilevanti sono state inoltre ricavate dal questionario post-sperimentale, da cui è risultato che gli utenti del gruppo visivo hanno stimato un maggiore incremento del carico cognitivo a causa dell'introduzione dell'aumentazione. Inoltre, la maggior parte degli utenti del gruppo visuo-aptico ha selezionato la modalità aptica come più utile ai fini della minimizzazione dell'errore. Da questi risultati, è stata quindi derivata la potenzialità dell'aumentazione aptica ai fini dell'apprendimento motorio. Ulteriori indagini sul training a lungo termine e sul mantenimento delle abilità potrebbero essere utili nel validare questa ipotesi. Inoltre, l'interazione tra modalità di aumentazione potrebbe essere analizzata in maniera più approfondita per mezzo di sessioni di training che coinvolgano l'applicazione di feedback seguito da guida, ma anche di protocolli specifici che coinvolgano una sola delle due modalità.
Multi-sensory augmentation for simulation-based training in robot assisted surgery
Galli de Paratesi, Chiara
2019/2020
Abstract
The introduction of robots in the medical field has revolutionized the way of performing surgery, increasing benefits for both patients and surgeons. Nevertheless, Robot-Assisted Minimally Invasive Surgery (RAMIS) introduces new control modalities of the operating instruments. Practical skills training represents a key step for the preparation of robotic surgeons. The capability of Virtual Reality (VR) simulators to create a safe and standardized environment for training, as well as the possibility to provide performance parameters, has led to a great increase of their exploitation in the surgical training process. Virtual reality simulators also open attractive opportunities in terms of augmentation, a concept which has already been explored in rehabilitation and sport training. Augmentation has proven to be an indispensable component to optimize motor learning, and it comprises the concepts of guidance and feedback. Augmented guidance is typically exploited as a way to teach the trainee the ideal movement or the procedure to be learnt, while augmented feedback represents a useful means for error correction. For the purpose of this study a virtual reality needle-driving task was integrated with augmentation in the form of guidance and feedback during the same training session. First guidance was exploited to teach the trainee how to carry out the task and was then substituted by feedback for improving the gesture’s accuracy. As first objective, the effect on motor learning of this novel combination of guidance and feedback was investigated. Moreover, these augmentation modalities were explored over the visual and haptic domains, with the additional aim of identifying the most appropriate sensory modality for practical skills acquisition. To address the purpose of this research, the master console of a da Vinci Research Kit (dVRK) (Intuitive Surgical Inc., Sunnyvale) was exploited together with a virtual reality simulation environment: Assisted Teleoperation with Augmented Reality (ATAR). 32 non-medical participants were involved in a user study. They were randomly subdivided into four groups: one group received no training augmentation, while the others underwent visual, haptic and visuo-haptic augmentation, respectively. Users’ performance was assessed with means of objective metrics involving Phase Completion capability and Time-Accuracy, as well as subjective metrics derived from a post-experimental survey. Results showed a significant improvement, initially introduced by guidance, in the Phase Completion capability of all experimental groups against control. The introduction of the augmented guidance also led to an enhancement of the performance over the Time-Accuracy metrics. The subsequent application of feedback didn't seem to cause a comparable improvement in the objective metrics. According to the accuracy analysis the experimental groups outperformed control at the end of training, while no statistically significant differences were found across sensory domains. For what concerns the execution time, it presented significant lower values in groups involving haptics at evaluation. Interesting results were also derived from the post-experimental survey. Users belonging to the visual group assessed a higher increase in the cognitive overload due to the introduction of the augmentation. In addition, the majority of participants from the visuo-haptic group selected the haptic modality as the most helpful in the error minimization. From these results the potentiality of the haptic augmentation was derived. Further analysis on long term learning and skill retention could better investigate this hypothesis. In addition, the interaction between augmentation modalities could be analysed more thoroughly with means of different training session of subsequent feedback and guidance, as well as protocols involving only guidance or only feedback.File | Dimensione | Formato | |
---|---|---|---|
2020_12_GallideParatesi.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: Testo della tesi
Dimensione
24.77 MB
Formato
Adobe PDF
|
24.77 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/10589/175450