Semantic denoising autoencoders for retinal optical coherence tomography

Show simple item record

dc.identifier.uri http://dx.doi.org/10.15488/10274
dc.identifier.uri https://www.repo.uni-hannover.de/handle/123456789/10347
dc.contributor.author Laves, Max-Heinrich
dc.contributor.author Ihler, Sontje
dc.contributor.author Kahrs, Lüder Alexander
dc.contributor.author Ortmaier, Tobias
dc.date.accessioned 2020-12-08T15:27:06Z
dc.date.available 2020-12-08T15:27:06Z
dc.date.issued 2019
dc.identifier.citation Laves, M.-H.; Ihler, S.; Kahrs, L.A.; Ortmaier, T.: Semantic denoising autoencoders for retinal optical coherence tomography. In: Proceedings of SPIE 11078 (2019), 1107818. DOI: https://doi.org/10.1117/12.2526936
dc.description.abstract Noise in speckle-prone optical coherence tomography tends to obfuscate important details necessary for medical diagnosis. In this paper, a denoising approach that preserves disease characteristics on retinal optical coherence tomography images in ophthalmology is presented. We propose semantic denoising autoencoders, which combine a convolutional denoising autoencoder with a priorly trained ResNet image classifier as regularizer during training. This promotes the perceptibility of delicate details in the denoised images that are important for diagnosis and filters out only informationless background noise. With our approach, higher peak signal-to-noise ratios with PSNR = 31.0 dB and higher classification performance of F1 = 0.92 can be achieved for denoised images compared to state-of-the-art denoising. It is shown that semantically regularized autoencoders are capable of denoising retinal OCT images without blurring details of diseases. eng
dc.language.iso eng
dc.publisher Bellingham : SPIE
dc.relation.ispartofseries Proceedings of SPIE 11078 (2019)
dc.rights Es gilt deutsches Urheberrecht. Das Dokument darf zum eigenen Gebrauch kostenfrei genutzt, aber nicht im Internet bereitgestellt oder an Außenstehende weitergegeben werden. Dieser Beitrag ist aufgrund einer (DFG-geförderten) Allianz- bzw. Nationallizenz frei zugänglich.
dc.subject Computer-aided diagnosis eng
dc.subject Image classication eng
dc.subject Image enhancement eng
dc.subject Machine learning eng
dc.subject.ddc 620 | Ingenieurwissenschaften und Maschinenbau ger
dc.title Semantic denoising autoencoders for retinal optical coherence tomography eng
dc.type conferenceObject
dc.type article
dc.type Text
dc.relation.doi https://doi.org/10.1117/12.2526936
dc.bibliographicCitation.volume 11078
dc.bibliographicCitation.firstPage 1107818
dc.description.version publishedVersion
tib.accessRights frei zug�nglich


Files in this item

This item appears in the following Collection(s):

Show simple item record

 

Search the repository


Browse

My Account

Usage Statistics