Matches in SemOpenAlex for { <https://semopenalex.org/work/W2899013683> ?p ?o ?g. }
- W2899013683 abstract "Acute stroke lesion segmentation and prediction tasks are of great clinicalinterest as they can help doctors make better informed time-critical treatmentdecisions. Automatic segmentation of these lesions is a complex task due totheir heterogeneous appearance, dynamic evolution and inter-patientdifferences. Typically, acute stroke lesion tasks are approached with methodsdeveloped for chronic stroke or other brain lesions. However, thepathophysiology and anatomy of acute stroke establishes an inherently differentproblem that needs special consideration. In this work, we propose a novel deeplearning architecture specially designed for acute stroke tasks that involveapproximating complex non-linear functions with reduced data. Within ourstrategy, class imbalance is tackled using a hybrid strategy based onstate-of-the-art train sampling strategies designed for other brain lesionrelated tasks, which is more suited to the anatomy and pathophysiology of acutestroke lesions. The proposed method is evaluated on three unrelated publicinternational challenge datasets (ISLES) without any dataset specifichyper-parameter tuning. These involve the tasks of sub-acute stroke lesionsegmentation, acute stroke penumbra estimation and chronic extent predictionfrom acute MR images. The performance of the proposed architecture is analysedboth against similar deep learning architectures from chronic stroke andrelated biomedical tasks and also by submitting the segmented test images forblind online evaluation on each of the challenges. When compared with the restof submitted strategies, our method achieves top-rank performance among thebest submitted entries in all the three challenges, showing its capability todeal with different unrelated tasks without hyper-parameter tuning. In order topromote the reproducibility of our results, a public version of the proposedmethod has been released." @default.
- W2899013683 created "2018-11-09" @default.
- W2899013683 creator A5027168396 @default.
- W2899013683 creator A5034174740 @default.
- W2899013683 creator A5060062918 @default.
- W2899013683 creator A5064897887 @default.
- W2899013683 creator A5077881500 @default.
- W2899013683 creator A5085592048 @default.
- W2899013683 date "2018-10-31" @default.
- W2899013683 modified "2023-10-18" @default.
- W2899013683 title "SUNet: a deep learning architecture for acute stroke lesion segmentation and outcome prediction in multimodal MRI." @default.
- W2899013683 cites W1677182931 @default.
- W2899013683 cites W1686810756 @default.
- W2899013683 cites W1743074346 @default.
- W2899013683 cites W1884191083 @default.
- W2899013683 cites W1901129140 @default.
- W2899013683 cites W1930624869 @default.
- W2899013683 cites W1987869189 @default.
- W2899013683 cites W2000269560 @default.
- W2899013683 cites W2129869438 @default.
- W2899013683 cites W2194775991 @default.
- W2899013683 cites W2217077692 @default.
- W2899013683 cites W2301358467 @default.
- W2899013683 cites W2464708700 @default.
- W2899013683 cites W2484736472 @default.
- W2899013683 cites W2626711511 @default.
- W2899013683 cites W2758131076 @default.
- W2899013683 cites W2784209709 @default.
- W2899013683 cites W2919115771 @default.
- W2899013683 cites W2963076262 @default.
- W2899013683 cites W2963881378 @default.
- W2899013683 cites W2312104448 @default.
- W2899013683 hasPublicationYear "2018" @default.
- W2899013683 type Work @default.
- W2899013683 sameAs 2899013683 @default.
- W2899013683 citedByCount "2" @default.
- W2899013683 countsByYear W28990136832019 @default.
- W2899013683 countsByYear W28990136832020 @default.
- W2899013683 crossrefType "posted-content" @default.
- W2899013683 hasAuthorship W2899013683A5027168396 @default.
- W2899013683 hasAuthorship W2899013683A5034174740 @default.
- W2899013683 hasAuthorship W2899013683A5060062918 @default.
- W2899013683 hasAuthorship W2899013683A5064897887 @default.
- W2899013683 hasAuthorship W2899013683A5077881500 @default.
- W2899013683 hasAuthorship W2899013683A5085592048 @default.
- W2899013683 hasConcept C108583219 @default.
- W2899013683 hasConcept C119857082 @default.
- W2899013683 hasConcept C126322002 @default.
- W2899013683 hasConcept C127413603 @default.
- W2899013683 hasConcept C142724271 @default.
- W2899013683 hasConcept C154945302 @default.
- W2899013683 hasConcept C162324750 @default.
- W2899013683 hasConcept C187736073 @default.
- W2899013683 hasConcept C2776572282 @default.
- W2899013683 hasConcept C2780451532 @default.
- W2899013683 hasConcept C2780577055 @default.
- W2899013683 hasConcept C2780645631 @default.
- W2899013683 hasConcept C2781156865 @default.
- W2899013683 hasConcept C3020166492 @default.
- W2899013683 hasConcept C41008148 @default.
- W2899013683 hasConcept C541997718 @default.
- W2899013683 hasConcept C71924100 @default.
- W2899013683 hasConcept C78519656 @default.
- W2899013683 hasConcept C89600930 @default.
- W2899013683 hasConcept C99508421 @default.
- W2899013683 hasConceptScore W2899013683C108583219 @default.
- W2899013683 hasConceptScore W2899013683C119857082 @default.
- W2899013683 hasConceptScore W2899013683C126322002 @default.
- W2899013683 hasConceptScore W2899013683C127413603 @default.
- W2899013683 hasConceptScore W2899013683C142724271 @default.
- W2899013683 hasConceptScore W2899013683C154945302 @default.
- W2899013683 hasConceptScore W2899013683C162324750 @default.
- W2899013683 hasConceptScore W2899013683C187736073 @default.
- W2899013683 hasConceptScore W2899013683C2776572282 @default.
- W2899013683 hasConceptScore W2899013683C2780451532 @default.
- W2899013683 hasConceptScore W2899013683C2780577055 @default.
- W2899013683 hasConceptScore W2899013683C2780645631 @default.
- W2899013683 hasConceptScore W2899013683C2781156865 @default.
- W2899013683 hasConceptScore W2899013683C3020166492 @default.
- W2899013683 hasConceptScore W2899013683C41008148 @default.
- W2899013683 hasConceptScore W2899013683C541997718 @default.
- W2899013683 hasConceptScore W2899013683C71924100 @default.
- W2899013683 hasConceptScore W2899013683C78519656 @default.
- W2899013683 hasConceptScore W2899013683C89600930 @default.
- W2899013683 hasConceptScore W2899013683C99508421 @default.
- W2899013683 hasLocation W28990136831 @default.
- W2899013683 hasOpenAccess W2899013683 @default.
- W2899013683 hasPrimaryLocation W28990136831 @default.
- W2899013683 hasRelatedWork W1901129140 @default.
- W2899013683 hasRelatedWork W2163605009 @default.
- W2899013683 hasRelatedWork W2217077692 @default.
- W2899013683 hasRelatedWork W2592929672 @default.
- W2899013683 hasRelatedWork W2795142235 @default.
- W2899013683 hasRelatedWork W2912084136 @default.
- W2899013683 hasRelatedWork W2912802107 @default.
- W2899013683 hasRelatedWork W2915335211 @default.
- W2899013683 hasRelatedWork W2979365272 @default.
- W2899013683 hasRelatedWork W3021149250 @default.
- W2899013683 hasRelatedWork W3022713997 @default.
- W2899013683 hasRelatedWork W3081850693 @default.