Sim-to-real transfer for visual reinforcement learning of deformable object manipulation for robot-assisted surgery

Automation holds the potential to assist surgeons in robotic interventions, shifting their mental work load from visuomotor control to high level decision making. Reinforcement learning has shown promising results in learning complex visuomotor policies, especially in simulation environments where m...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Scheikl, Paul Maria (VerfasserIn) , Tagliabue, Eleonora (VerfasserIn) , Gyenes, Balázs (VerfasserIn) , Wagner, Martin (VerfasserIn) , Dall'Alba, Diego (VerfasserIn) , Fiorini, Paolo (VerfasserIn) , Mathis-Ullrich, Franziska (VerfasserIn)
Dokumenttyp: Article (Journal)
Sprache:Englisch
Veröffentlicht: February 2023
In: IEEE Robotics and automation letters
Year: 2023, Jahrgang: 8, Heft: 2, Pages: 560-567
ISSN:2377-3766
DOI:10.1109/LRA.2022.3227873
Online-Zugang:Verlag, lizenzpflichtig, Volltext: https://doi.org/10.1109/LRA.2022.3227873
Volltext
Verfasserangaben:Paul Maria Scheikl, Eleonora Tagliabue, Balázs Gyenes, Martin Wagner, Diego Dall'Alba, Paolo Fiorini, and Franziska Mathis-Ullrich

MARC

LEADER 00000caa a22000002c 4500
001 1859336620
003 DE-627
005 20240307054403.0
007 cr uuu---uuuuu
008 230912s2023 xx |||||o 00| ||eng c
024 7 |a 10.1109/LRA.2022.3227873  |2 doi 
035 |a (DE-627)1859336620 
035 |a (DE-599)KXP1859336620 
035 |a (OCoLC)1425212336 
040 |a DE-627  |b ger  |c DE-627  |e rda 
041 |a eng 
084 |a 33  |2 sdnb 
100 1 |a Scheikl, Paul Maria  |e VerfasserIn  |0 (DE-588)1302319159  |0 (DE-627)1859352022  |4 aut 
245 1 0 |a Sim-to-real transfer for visual reinforcement learning of deformable object manipulation for robot-assisted surgery  |c Paul Maria Scheikl, Eleonora Tagliabue, Balázs Gyenes, Martin Wagner, Diego Dall'Alba, Paolo Fiorini, and Franziska Mathis-Ullrich 
264 1 |c February 2023 
300 |a 8 
336 |a Text  |b txt  |2 rdacontent 
337 |a Computermedien  |b c  |2 rdamedia 
338 |a Online-Ressource  |b cr  |2 rdacarrier 
500 |a Gesehen am 12.09.2023 
520 |a Automation holds the potential to assist surgeons in robotic interventions, shifting their mental work load from visuomotor control to high level decision making. Reinforcement learning has shown promising results in learning complex visuomotor policies, especially in simulation environments where many samples can be collected at low cost. A core challenge is learning policies in simulation that can be deployed in the real world, thereby overcoming the sim-to-real gap. In this letter, we bridge the visual sim-to-real gap with an image-based reinforcement learning pipeline based on pixel-level domain adaptation and demonstrate its effectiveness on an image-based task in deformable object manipulation. We choose a tissue retraction task because of its importance in clinical reality of precise cancer surgery. After training in simulation on domain-translated images, our policy requires no retraining to perform tissue retraction with a 50% success rate on the real robotic system using raw RGB images. Furthermore, our sim-to-real transfer method makes no assumptions on the task itself and requires no paired images. This letter introduces the first successful application of visual sim-to-real transfer for robotic manipulation of deformable objects in the surgical field, which represents a notable step towards the clinical translation of cognitive surgical robotics. 
650 4 |a Computer vision for medical robotics 
650 4 |a Grippers 
650 4 |a Medical robotics 
650 4 |a reinforcement learning 
650 4 |a Robots 
650 4 |a Surgery 
650 4 |a surgical robotics: Laparoscopy 
650 4 |a Task analysis 
650 4 |a Training 
650 4 |a Visualization 
700 1 |a Tagliabue, Eleonora  |e VerfasserIn  |4 aut 
700 1 |a Gyenes, Balázs  |e VerfasserIn  |4 aut 
700 1 |a Wagner, Martin  |d 1988-  |e VerfasserIn  |0 (DE-588)1060231980  |0 (DE-627)799324337  |0 (DE-576)416324282  |4 aut 
700 1 |a Dall'Alba, Diego  |e VerfasserIn  |4 aut 
700 1 |a Fiorini, Paolo  |e VerfasserIn  |4 aut 
700 1 |a Mathis-Ullrich, Franziska  |e VerfasserIn  |4 aut 
773 0 8 |i Enthalten in  |t IEEE Robotics and automation letters  |d New York, N.Y. : Institute of Electrical and Electronics Engineers, 2016  |g 8(2023), 2 vom: Feb., Seite 560-567  |h Online-Ressource  |w (DE-627)846122235  |w (DE-600)2844561-2  |w (DE-576)45442244X  |x 2377-3766  |7 nnas  |a Sim-to-real transfer for visual reinforcement learning of deformable object manipulation for robot-assisted surgery 
773 1 8 |g volume:8  |g year:2023  |g number:2  |g month:02  |g pages:560-567  |g extent:8  |a Sim-to-real transfer for visual reinforcement learning of deformable object manipulation for robot-assisted surgery 
856 4 0 |u https://doi.org/10.1109/LRA.2022.3227873  |x Verlag  |x Resolving-System  |z lizenzpflichtig  |3 Volltext 
951 |a AR 
992 |a 20230912 
993 |a Article 
994 |a 2023 
998 |g 1060231980  |a Wagner, Martin  |m 1060231980:Wagner, Martin  |d 910000  |d 910200  |e 910000PW1060231980  |e 910200PW1060231980  |k 0/910000/  |k 1/910000/910200/  |p 5 
998 |g 1302319159  |a Scheikl, Paul Maria  |m 1302319159:Scheikl, Paul Maria  |d 50000  |e 50000PS1302319159  |k 0/50000/  |p 1  |x j 
999 |a KXP-PPN1859336620  |e 4374793461 
BIB |a Y 
SER |a journal 
JSO |a {"name":{"displayForm":["Paul Maria Scheikl, Eleonora Tagliabue, Balázs Gyenes, Martin Wagner, Diego Dall'Alba, Paolo Fiorini, and Franziska Mathis-Ullrich"]},"person":[{"family":"Scheikl","role":"aut","given":"Paul Maria","display":"Scheikl, Paul Maria"},{"family":"Tagliabue","role":"aut","given":"Eleonora","display":"Tagliabue, Eleonora"},{"family":"Gyenes","given":"Balázs","role":"aut","display":"Gyenes, Balázs"},{"display":"Wagner, Martin","role":"aut","given":"Martin","family":"Wagner"},{"given":"Diego","role":"aut","display":"Dall'Alba, Diego","family":"Dall'Alba"},{"family":"Fiorini","role":"aut","given":"Paolo","display":"Fiorini, Paolo"},{"family":"Mathis-Ullrich","role":"aut","given":"Franziska","display":"Mathis-Ullrich, Franziska"}],"type":{"media":"Online-Ressource","bibl":"article-journal"},"recId":"1859336620","origin":[{"dateIssuedKey":"2023","dateIssuedDisp":"February 2023"}],"relHost":[{"name":{"displayForm":["IEEE Robotics and Automation Society"]},"corporate":[{"display":"Institute of Electrical and Electronics Engineers","role":"isb"},{"role":"isb","display":"IEEE Robotics and Automation Society"}],"note":["Trotz der Ankündigung, eine OA-only-Zeitschrift zu sein, sind derzeit nicht alle Artikel frei zugänglich","Gesehen am 21. März 2016"],"physDesc":[{"extent":"Online-Ressource"}],"disp":"Sim-to-real transfer for visual reinforcement learning of deformable object manipulation for robot-assisted surgeryIEEE Robotics and automation letters","id":{"zdb":["2844561-2"],"issn":["2377-3766"],"eki":["846122235"]},"origin":[{"publisher":"Institute of Electrical and Electronics Engineers","dateIssuedDisp":"Jan 2016-","publisherPlace":"New York, N.Y."}],"recId":"846122235","type":{"bibl":"periodical","media":"Online-Ressource"},"part":{"extent":"8","pages":"560-567","issue":"2","volume":"8","year":"2023","text":"8(2023), 2 vom: Feb., Seite 560-567"},"language":["eng"],"title":[{"title":"IEEE Robotics and automation letters","title_sort":"IEEE Robotics and automation letters"}],"pubHistory":["Vol. 1, Issue 1 (Jan 2016)-"],"titleUni":[{"title":"Robotics and automation letters"}]}],"note":["Gesehen am 12.09.2023"],"physDesc":[{"extent":"8 S."}],"id":{"doi":["10.1109/LRA.2022.3227873"],"eki":["1859336620"]},"language":["eng"],"title":[{"title":"Sim-to-real transfer for visual reinforcement learning of deformable object manipulation for robot-assisted surgery","title_sort":"Sim-to-real transfer for visual reinforcement learning of deformable object manipulation for robot-assisted surgery"}]} 
SRT |a SCHEIKLPAUSIMTOREALT2023