메뉴
닫기


Automatic Segmentation of Mitochondria and Endolysosomesin Volumetric Electron Microscopy Data
Year of publication
2020
Author
Manca Žerovnik Mekuča, Ciril Bohak, Samo Hudoklin, Byeong Hak Kim, Rok Romih, Min Young Kim and Matija Marolt
Journal
Computers in Biology and Medicine
volume
119
File
첨부 Automatic_Segmentation_of_Mitochondria_and_Endo_lysosomes_in_Volumetric_EM_Data___Computers_in_Biology_and_Medicine.pdf (16.3M) 8회 다운로드 DATE : 2020-03-16 10:58:28

Automaticsegmentationofintracellularcompartmentsisapowerfultechnique,whichprovidesquantitativedataaboutpresence, spatialdistribution, structureandconsequentlythefunctionofcells. Withtherecentdevelopmentofhighthroughputvolumetricdataacquisitiontechniquesinelectron microscopy(EM),manualsegmentationisbecomingamajorbottleneckoftheprocess. Toaidthe cellresearch,weproposeatechniqueforautomaticsegmentationofmitochondriaandendolysosomes obtainedfromurinarybladderurothelialcellsbythedualbeamEMtechnique. Wepresentanovel publiclyavailablevolumetricEMdataset thefirstofurothelialcells,evaluateseveralstate-of-theartsegmentationmethodsonthenewdatasetandpresentanovelsegmentationpipeline,whichis basedonsuperviseddeeplearningandincludesmechanismsthatreducetheimpactofdependenciesintheinputdata,artefactsandannotationerrors. Weshowthatourapproachoutperformsthe comparedmethodsontheproposeddataset.




[702-701] 1370 Sankyuk-dong, Buk-gu, Daegu, Korea
Tel : +82-53-950-7233 / Fax : +82-53-950-5505
Copyrights ⒞ 2019 Kyungpook National University. All Rights Reserved.