Pregled bibliografske jedinice broj: 1035967
Adaptive Thresholding for Single Click Left Atrial Appendage Segmentation
Adaptive Thresholding for Single Click Left Atrial Appendage Segmentation // Proceedings ELMAR-2019 / Muštra, Mario ; Vuković, Josip ; Zovko-Cihlar, Branka (ur.).
Zadar, 2019. str. 35-38 doi:10.1109/ELMAR.2019.8918651 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 1035967 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Adaptive Thresholding for Single Click Left Atrial Appendage Segmentation
Autori
Vdovjak, Krešimir ; Leventić, Hrvoje ; Habijan, Marija ; Galić, Irena
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
Proceedings ELMAR-2019
/ Muštra, Mario ; Vuković, Josip ; Zovko-Cihlar, Branka - Zadar, 2019, 35-38
ISBN
978-1-7281-2182-6
Skup
61st International Symposium Electronics in Marine (ELMAR 2019)
Mjesto i datum
Zadar, Hrvatska, 23.09.2019. - 25.09.2019
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
Adaptive thresholding ; CCTA ; LAA ; Left atrial appendage segmentation
Sažetak
We present a novel automatic adaptive thresholdingmethod which, when combined with our previously publishedsemi-automatic LAA segmentation method, provides a frame-work for the accurate LAA segmentation requiring minimal userinteraction. The method proposed in this paper automaticallyperforms the adaptive thresholding step, thus the frameworknow requires only a single click inside the LAA to perform thesegmentation.Performance of the proposed method is evaluated on 17 CCTAdatasets manually segmented by two medical experts. The LAAsegmentation results using the proposed adaptive thresholdingmethod are compared to the segmentation results achieved witha manually selected threshold. The proposed method results ina comparable DICE overlap to the segmentation results with amanually selected threshold, while removing the subjectivity ofthe manual selection.
Izvorni jezik
Engleski
Znanstvena područja
Računarstvo
POVEZANOST RADA
Ustanove:
Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek
Profili:
Marija Habijan
(autor)
Hrvoje Leventić
(autor)
Irena Galić
(autor)
Krešimir Vdovjak
(autor)