Mutually improved endoscopic image synthesis and landmark detection in unpaired image-to-image translation

The CycleGAN framework allows for unsupervised image-to-image translation of unpaired data. In a scenario of surgical training on a physical surgical simulator, this method can be used to transform endoscopic images of phantoms into images which more closely resemble the intra-operative appearance o...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Sharan, Lalith (VerfasserIn) , Romano, Gabriele (VerfasserIn) , Köhler, Sven (VerfasserIn) , Kelm, Halvar (VerfasserIn) , Karck, Matthias (VerfasserIn) , De Simone, Raffaele (VerfasserIn) , Engelhardt, Sandy (VerfasserIn)
Dokumenttyp: Article (Journal)
Sprache:Englisch
Veröffentlicht: January 2022
In: IEEE journal of biomedical and health informatics
Year: 2022, Jahrgang: 26, Heft: 1, Pages: 127-138
ISSN:2168-2208
Online-Zugang: Volltext
Verfasserangaben:Lalith Sharan, Gabriele Romano, Sven Koehler, Halvar Kelm, Matthias Karck, Raffaele De Simone and Sandy Engelhardt

MARC

LEADER 00000caa a2200000 c 4500
001 1798141248
003 DE-627
005 20220820163309.0
007 cr uuu---uuuuu
008 220407s2022 xx |||||o 00| ||eng c
024 7 |a 10.1109/JBHI.2021.3099858  |2 doi 
035 |a (DE-627)1798141248 
035 |a (DE-599)KXP1798141248 
035 |a (OCoLC)1341458836 
040 |a DE-627  |b ger  |c DE-627  |e rda 
041 |a eng 
084 |a 33  |2 sdnb 
100 1 |a Sharan, Lalith  |d 1993-  |e VerfasserIn  |0 (DE-588)1255069554  |0 (DE-627)1798142104  |4 aut 
245 1 0 |a Mutually improved endoscopic image synthesis and landmark detection in unpaired image-to-image translation  |c Lalith Sharan, Gabriele Romano, Sven Koehler, Halvar Kelm, Matthias Karck, Raffaele De Simone and Sandy Engelhardt 
264 1 |c January 2022 
300 |a 11 
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 07.04.2022 
520 |a The CycleGAN framework allows for unsupervised image-to-image translation of unpaired data. In a scenario of surgical training on a physical surgical simulator, this method can be used to transform endoscopic images of phantoms into images which more closely resemble the intra-operative appearance of the same surgical target structure. This can be viewed as a novel augmented reality approach, which we coined Hyperrealism in previous work. In this use case, it is of paramount importance to display objects like needles, sutures or instruments consistent in both domains while altering the style to a more tissue-like appearance. Segmentation of these objects would allow for a direct transfer, however, contouring of these, partly tiny and thin foreground objects is cumbersome and perhaps inaccurate. Instead, we propose to use landmark detection on the points when sutures pass into the tissue. This objective is directly incorporated into a CycleGAN framework by treating the performance of pre-trained detector models as an additional optimization goal. We show that a task defined on these sparse landmark labels improves consistency of synthesis by the generator network in both domains. Comparing a baseline CycleGAN architecture to our proposed extension (DetCycleGAN), mean precision (PPV) improved by $+61.32$, mean sensitivity (TPR) by $+37.91$, and mean $F_1$ score by $+0.4743$. Furthermore, it could be shown that by dataset fusion, generated intra-operative images can be leveraged as additional training data for the detection network itself. 
650 4 |a CycleGAN 
650 4 |a Generative adversarial networks 
650 4 |a landmark detection 
650 4 |a landmark localization 
650 4 |a Maintenance engineering 
650 4 |a mitral valve repair 
650 4 |a Semantics 
650 4 |a Surgery 
650 4 |a surgical simulation 
650 4 |a surgical training 
650 4 |a Task analysis 
650 4 |a Training 
650 4 |a Valves 
700 1 |a Romano, Gabriele  |d 1991-  |e VerfasserIn  |0 (DE-588)1204102082  |0 (DE-627)168924464X  |4 aut 
700 1 |a Köhler, Sven  |e VerfasserIn  |0 (DE-588)124712830X  |0 (DE-627)1780697864  |4 aut 
700 1 |a Kelm, Halvar  |e VerfasserIn  |0 (DE-588)1255070277  |0 (DE-627)179814350X  |4 aut 
700 1 |a Karck, Matthias  |d 1961-  |e VerfasserIn  |0 (DE-588)130833525  |0 (DE-627)505909340  |0 (DE-576)298366169  |4 aut 
700 1 |a De Simone, Raffaele  |d 1958-  |e VerfasserIn  |0 (DE-588)1075518466  |0 (DE-627)833385674  |0 (DE-576)443621667  |4 aut 
700 1 |a Engelhardt, Sandy  |d 1987-  |e VerfasserIn  |0 (DE-588)1122674465  |0 (DE-627)876003080  |0 (DE-576)481436049  |4 aut 
773 0 8 |i Enthalten in  |a Institute of Electrical and Electronics Engineers  |t IEEE journal of biomedical and health informatics  |d New York, NY : IEEE, 2013  |g 26(2022), 1, Seite 127-138  |h Online-Ressource  |w (DE-627)728472279  |w (DE-600)2687987-6  |w (DE-576)375637281  |x 2168-2208  |7 nnas 
773 1 8 |g volume:26  |g year:2022  |g number:1  |g pages:127-138  |g extent:11  |a Mutually improved endoscopic image synthesis and landmark detection in unpaired image-to-image translation 
951 |a AR 
992 |a 20220407 
993 |a Article 
994 |a 2022 
998 |g 1122674465  |a Engelhardt, Sandy  |m 1122674465:Engelhardt, Sandy  |d 910000  |d 910100  |e 910000PE1122674465  |e 910100PE1122674465  |k 0/910000/  |k 1/910000/910100/  |p 7  |y j 
998 |g 1075518466  |a De Simone, Raffaele  |m 1075518466:De Simone, Raffaele  |d 910000  |d 910200  |d 50000  |e 910000PD1075518466  |e 910200PD1075518466  |e 50000PD1075518466  |k 0/910000/  |k 1/910000/910200/  |k 0/50000/  |p 6 
998 |g 130833525  |a Karck, Matthias  |m 130833525:Karck, Matthias  |d 910000  |d 910200  |e 910000PK130833525  |e 910200PK130833525  |k 0/910000/  |k 1/910000/910200/  |p 5 
998 |g 1255070277  |a Kelm, Halvar  |m 1255070277:Kelm, Halvar  |p 4 
998 |g 124712830X  |a Köhler, Sven  |m 124712830X:Köhler, Sven  |d 910000  |d 910100  |e 910000PK124712830X  |e 910100PK124712830X  |k 0/910000/  |k 1/910000/910100/  |p 3 
998 |g 1204102082  |a Romano, Gabriele  |m 1204102082:Romano, Gabriele  |d 910000  |d 910200  |e 910000PR1204102082  |e 910200PR1204102082  |k 0/910000/  |k 1/910000/910200/  |p 2 
998 |g 1255069554  |a Sharan, Lalith  |m 1255069554:Sharan, Lalith  |d 910000  |d 910100  |e 910000PS1255069554  |e 910100PS1255069554  |k 0/910000/  |k 1/910000/910100/  |p 1  |x j 
999 |a KXP-PPN1798141248  |e 4113496064 
BIB |a Y 
SER |a journal 
JSO |a {"physDesc":[{"extent":"11 S."}],"recId":"1798141248","note":["Gesehen am 07.04.2022"],"language":["eng"],"type":{"bibl":"article-journal","media":"Online-Ressource"},"title":[{"title":"Mutually improved endoscopic image synthesis and landmark detection in unpaired image-to-image translation","title_sort":"Mutually improved endoscopic image synthesis and landmark detection in unpaired image-to-image translation"}],"relHost":[{"name":{"displayForm":["Institute of Electrical and Electronics Engineers"]},"pubHistory":["17.2013 -"],"titleAlt":[{"title":"Biomedical and health informatics"}],"corporate":[{"display":"Institute of Electrical and Electronics Engineers","role":"aut"}],"id":{"zdb":["2687987-6"],"doi":["10.1109/JBHI.6221020"],"issn":["2168-2208"],"eki":["728472279"]},"part":{"year":"2022","volume":"26","pages":"127-138","issue":"1","extent":"11","text":"26(2022), 1, Seite 127-138"},"origin":[{"publisherPlace":"New York, NY","dateIssuedKey":"2013","dateIssuedDisp":"2013-","publisher":"IEEE"}],"title":[{"title":"IEEE journal of biomedical and health informatics","title_sort":"IEEE journal of biomedical and health informatics"}],"language":["eng"],"type":{"media":"Online-Ressource","bibl":"periodical"},"note":["Gesehen am 14.03.2023"],"recId":"728472279","physDesc":[{"extent":"Online-Ressource"}],"disp":"Institute of Electrical and Electronics EngineersIEEE journal of biomedical and health informatics"}],"person":[{"given":"Lalith","role":"aut","family":"Sharan","display":"Sharan, Lalith"},{"display":"Romano, Gabriele","family":"Romano","role":"aut","given":"Gabriele"},{"family":"Köhler","role":"aut","given":"Sven","display":"Köhler, Sven"},{"family":"Kelm","given":"Halvar","role":"aut","display":"Kelm, Halvar"},{"display":"Karck, Matthias","family":"Karck","role":"aut","given":"Matthias"},{"given":"Raffaele","role":"aut","family":"De Simone","display":"De Simone, Raffaele"},{"display":"Engelhardt, Sandy","family":"Engelhardt","role":"aut","given":"Sandy"}],"origin":[{"dateIssuedDisp":"January 2022","dateIssuedKey":"2022"}],"id":{"eki":["1798141248"],"doi":["10.1109/JBHI.2021.3099858"]},"name":{"displayForm":["Lalith Sharan, Gabriele Romano, Sven Koehler, Halvar Kelm, Matthias Karck, Raffaele De Simone and Sandy Engelhardt"]}} 
SRT |a SHARANLALIMUTUALLYIM2022