Structural Visual Guidance Attention Networks In Retinopathy Of Prematurity
Convolutional neural networks (CNNs) have shown great performance in medical diagnostic applications. However, because their black-box nature, clinicians are reluctant to trust CNN diagnostic outcomes. Incorporating visual attention capabilities in CNNs enhances interpretability by highlighting regi...
Saved in:
Published in | 2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI) pp. 353 - 357 |
---|---|
Main Authors | , , , , , , , , , |
Format | Conference Proceeding |
Language | English |
Published |
IEEE
13.04.2021
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Abstract | Convolutional neural networks (CNNs) have shown great performance in medical diagnostic applications. However, because their black-box nature, clinicians are reluctant to trust CNN diagnostic outcomes. Incorporating visual attention capabilities in CNNs enhances interpretability by highlighting regions in the images that CNNs utilize for prediction. Clinicians can often provide domain knowledge on relevant features: e.g., to diagnose retinopathy of prematurity (ROP), structural information such as tortuosity of vessels aid clinicians in diagnosing ROP. We propose a Structural Visual Guidance Attention Networks (SVGA-Net) method, that leverages structural domain knowledge to guide visual attention in CNNs. Experiments on a dataset of 5512 posterior retinal images, taken using a wide-angle fundus camera, show that SVGA-Net achieves 0.987 and 0.979 AUC to predict plus and normal categories, respectively. SVGA-Net consistently results in higher AUC compared to visual attention CNNs without guidance, baseline CNNs, and CNNs with structured masks. |
---|---|
AbstractList | Convolutional neural networks (CNNs) have shown great performance in medical diagnostic applications. However, because their black-box nature, clinicians are reluctant to trust CNN diagnostic outcomes. Incorporating visual attention capabilities in CNNs enhances interpretability by highlighting regions in the images that CNNs utilize for prediction. Clinicians can often provide domain knowledge on relevant features: e.g., to diagnose retinopathy of prematurity (ROP), structural information such as tortuosity of vessels aid clinicians in diagnosing ROP. We propose a Structural Visual Guidance Attention Networks (SVGA-Net) method, that leverages structural domain knowledge to guide visual attention in CNNs. Experiments on a dataset of 5512 posterior retinal images, taken using a wide-angle fundus camera, show that SVGA-Net achieves 0.987 and 0.979 AUC to predict plus and normal categories, respectively. SVGA-Net consistently results in higher AUC compared to visual attention CNNs without guidance, baseline CNNs, and CNNs with structured masks. |
Author | Yildiz, I. Campbell, J. P. Yildiz, V. Ostmo, S. Chiang, M. F. Dy, J. Kalpathy-Cramer, J. Ioannidis, S. Tian, P. Erdogmus, D. |
Author_xml | – sequence: 1 givenname: V. surname: Yildiz fullname: Yildiz, V. organization: Northeastern University,Cognitive Systems Laboratory,ECE Department,MA,USA – sequence: 2 givenname: S. surname: Ioannidis fullname: Ioannidis, S. organization: Northeastern University,Cognitive Systems Laboratory,ECE Department,MA,USA – sequence: 3 givenname: I. surname: Yildiz fullname: Yildiz, I. organization: Northeastern University,Cognitive Systems Laboratory,ECE Department,MA,USA – sequence: 4 givenname: P. surname: Tian fullname: Tian, P. organization: Northeastern University,Cognitive Systems Laboratory,ECE Department,MA,USA – sequence: 5 givenname: J. P. surname: Campbell fullname: Campbell, J. P. organization: Oregon Health & Science University,Department of Ophthalmology,OR,USA – sequence: 6 givenname: S. surname: Ostmo fullname: Ostmo, S. organization: Oregon Health & Science University,Department of Ophthalmology,OR,USA – sequence: 7 givenname: J. surname: Kalpathy-Cramer fullname: Kalpathy-Cramer, J. organization: Massachusetts General Hospital,Department of Radiology,MA,USA – sequence: 8 givenname: M. F. surname: Chiang fullname: Chiang, M. F. organization: Oregon Health & Science University,Department of Ophthalmology,OR,USA – sequence: 9 givenname: D. surname: Erdogmus fullname: Erdogmus, D. organization: Northeastern University,Cognitive Systems Laboratory,ECE Department,MA,USA – sequence: 10 givenname: J. surname: Dy fullname: Dy, J. organization: Northeastern University,Cognitive Systems Laboratory,ECE Department,MA,USA |
BookMark | eNotj99KwzAchaMouM09gSB5gdb88q_J5Rw6i8OJU29H2qQY3dKRpkjf3oI7N9_V-Thnii5CGxxCt0ByAKLvyu19yRUFyCmhkGvOmFJwhqYgpeBAuYRzNAHNRaa4oFdo3nXfZEzBOSN8gp63KfZ16qPZ40_f9SNWvbcm1A4vUnIh-TbgF5d-2_jT4TLgN5d8aI8mfQ140-DX6A5m7Ps0XKPLxuw7Nz9xhj4eH96XT9l6syqXi3XmKWEpqyqjKRNWiZoxa5woqNTGSlUQUHUthBz3WdoI2whDpNLE6IJIahhIa0XFZujm3-udc7tj9AcTh93pO_sD8RVQTA |
ContentType | Conference Proceeding |
DBID | 6IE 6IL CBEJK RIE RIL |
DOI | 10.1109/ISBI48211.2021.9433881 |
DatabaseName | IEEE Electronic Library (IEL) Conference Proceedings IEEE Proceedings Order Plan All Online (POP All Online) 1998-present by volume IEEE Xplore All Conference Proceedings IEEE Xplore IEEE Proceedings Order Plans (POP All) 1998-Present |
DatabaseTitleList | |
Database_xml | – sequence: 1 dbid: RIE name: IEEE Xplore url: https://proxy.k.utb.cz/login?url=https://ieeexplore.ieee.org/ sourceTypes: Publisher |
DeliveryMethod | fulltext_linktorsrc |
Discipline | Engineering |
EISBN | 1665412461 9781665412469 |
EISSN | 1945-8452 |
EndPage | 357 |
ExternalDocumentID | 9433881 |
Genre | orig-research |
GroupedDBID | 23N 6IE 6IF 6IK 6IL 6IN AAJGR ABLEC ADZIZ ALMA_UNASSIGNED_HOLDINGS BEFXN BFFAM BGNUA BKEBE BPEOZ CBEJK CHZPO IEGSK IPLJI M43 OCL RIE RIL RNS |
ID | FETCH-LOGICAL-i203t-bba9235d85c33dae57269ad687018cc556000d2f5df5a06890a97062a316dd5b3 |
IEDL.DBID | RIE |
IngestDate | Wed Jun 26 19:28:59 EDT 2024 |
IsPeerReviewed | false |
IsScholarly | true |
Language | English |
LinkModel | DirectLink |
MergedId | FETCHMERGED-LOGICAL-i203t-bba9235d85c33dae57269ad687018cc556000d2f5df5a06890a97062a316dd5b3 |
PageCount | 5 |
ParticipantIDs | ieee_primary_9433881 |
PublicationCentury | 2000 |
PublicationDate | 2021-April-13 |
PublicationDateYYYYMMDD | 2021-04-13 |
PublicationDate_xml | – month: 04 year: 2021 text: 2021-April-13 day: 13 |
PublicationDecade | 2020 |
PublicationTitle | 2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI) |
PublicationTitleAbbrev | ISBI |
PublicationYear | 2021 |
Publisher | IEEE |
Publisher_xml | – name: IEEE |
SSID | ssj0000744304 |
Score | 2.210479 |
Snippet | Convolutional neural networks (CNNs) have shown great performance in medical diagnostic applications. However, because their black-box nature, clinicians are... |
SourceID | ieee |
SourceType | Publisher |
StartPage | 353 |
SubjectTerms | Attention Cameras CNN Convolutional neural networks Interpretability Knowledge engineering Medical diagnosis Retina Retinopathy ROP Visualization |
Title | Structural Visual Guidance Attention Networks In Retinopathy Of Prematurity |
URI | https://ieeexplore.ieee.org/document/9433881 |
hasFullText | 1 |
inHoldings | 1 |
isFullTextHit | |
isPrint | |
link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjV07T8MwELZKJ1h4tIi3PDDiNHFiJxkBUSioBVGKulV2zpYqpBSpyQC_nnMSykMMTI4sOYnuZH139n3fEXIqlA9GW8tkbDMWxSZgCoGF6TgFhfClU-7IycORvJlEt1MxbZGzFRfGGFMVnxnPPVZ3-bDISndU1ksjTKgcz3ot8XnN1VqdpyAURpiaNyTgwE97g_HFIEowwcEskAdes_hHF5UKRPqbZPj5-bp25MUrC-1l77-UGf_7f1uk-0XXow8rINomLZPvkI1vSoMdcjeudGKdxgZ9ni9LHK7LOTiX0_OiqGse6aiuCV_SQU4fHRd64RoWv9F7i-934q5Vq7sumfSvni5vWNNHgc25HxZMa4VhnIBEZGEIyoiYy1SBxK0aJFkmXNDjA7cCLLpOJqmv0tiXXIWBBBA63CXtfJGbPUIlBFwJmzmVwshqriE2JgoBgyzgCdh90nFmmb3WUhmzxiIHf08fknXnGnc5E4RHpI2WMMeI8YU-qZz7AVyUp5g |
link.rule.ids | 310,311,786,790,795,796,802,23958,23959,25170,27958,55109 |
linkProvider | IEEE |
linkToHtml | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjV07T8MwELYqGICFR4t4k4GRpIkTO8kIiNLQB4i2qFtl52ypQkqRmgzw6zknoTzEwBQrVizrTtF3Z9_3HSEXTLigpNY2D3VqB6HybIHAYsswBoHwJWNqyMmDIe9OgvspmzbI5YoLo5Qqi8-UY4blXT4s0sIclbXjABMqw7NeR5x344qttTpRQTAMMDmvacA4305G10kQYYqDeSD1nPrzH31UShjpbJPB5waq6pEXp8ilk77_0mb87w53SOuLsGc9rqBolzRUtke2vmkNNklvVCrFGpUN63m-LPBxV8zBON26yvOq6tEaVlXhSyvJrCfDhl6YlsVv1oPG9Y28a9nsrkUmndvxTdeuOynYc-r6uS2lwECOQcRS3wehWEh5LIDjz-pFacpM2OMC1Qw0Oo9HsSvi0OVU-B4HYNLfJ2vZIlMHxOLgUcF0anQKAy2phFCpwAcMs4BGoA9J05hl9lqJZcxqixz9_fqcbHTHg_6snwx7x2TTuMlc1Xj-CVlDq6hTRPxcnpWO_gBk-Kru |
openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=proceeding&rft.title=2021+IEEE+18th+International+Symposium+on+Biomedical+Imaging+%28ISBI%29&rft.atitle=Structural+Visual+Guidance+Attention+Networks+In+Retinopathy+Of+Prematurity&rft.au=Yildiz%2C+V.&rft.au=Ioannidis%2C+S.&rft.au=Yildiz%2C+I.&rft.au=Tian%2C+P.&rft.date=2021-04-13&rft.pub=IEEE&rft.eissn=1945-8452&rft.spage=353&rft.epage=357&rft_id=info:doi/10.1109%2FISBI48211.2021.9433881&rft.externalDocID=9433881 |