RETOUCH: The Retinal OCT Fluid Detection and Segmentation Benchmark and Challenge

Bogunović, Hrvoje; Venhuizen, Freerk G.; Klimscha, Sophie; Apostolopoulos, Stefanos; Bab–Hadiashar, Alireza; Bağcı, Ulaş; Beg, Mirza Faisal; Bekalo, Loza; Chen, Qiang; Ciller, Carlos; Gopinath, Kaundinya; Gostar, Amirali Khodadadian; Jeon, Kiwan; Ji, Zexuan; Kang, Sung Ho; Koozekanani, Dara D.; Lu, Donghuan; Morley, Dustin; Parhi, Keshab K.; Park, Hyoung Suk; Rashno, Abdolreza; Sarunic, Marinko V.; Shaikh, Saad; Sivaswamy, Jayanthi; Tennakoon, Ruwan; Yadav, Shivin; Zanet, Sandro De; Waldstein, Sebastian M.; Gerendas, Bianca S.; Klaver, Caroline C W; Sánchez, Clara I.; Schmidt‐Erfurth, Ursula

doi:10.1109/tmi.2019.2901398

Cited by 182 publications

(147 citation statements)

References 61 publications

Supporting

Mentioning

131

Contrasting

Unclassified

Order By: Relevance

“…12 More recently, and underlining the interest of this particular application, an on-line challenge was organized stemming from a medical imaging conference that assimilated and compared a number of different deep learning approaches for segmenting three retinal fluid types in OCT data; intra-and sub-retinal fluid (SRF) and pigment epithelial detachment (PED). 13 This task maps nicely to the work we present in the following and is returned to in the discussion section of this paper.…”

Section: Introductionmentioning

confidence: 56%

Automated Deep Learning-based Multi-class Fluid Segmentation in Swept-Source Optical Coherence Tomography Images

Oakley¹,

Sodhi

Russakoff³

et al. 2020

Preprint

View full text Add to dashboard Cite

PurposeTo evaluate the performance of a deep learning-based, fully automated, multi-class, macular fluid segmentation algorithm relative to expert annotations in a heterogeneous population of confirmed wet age-related macular degeneration (wAMD) subjects.MethodsTwenty-two swept-source optical coherence tomography (SS-OCT) volumes of the macula from 22 from different individuals with wAMD were manually annotated by two expert graders. These results were compared using cross-validation (CV) to automated segmentations using a deep learning-based algorithm encoding spatial information about retinal tissue as an additional input to the network. The algorithm detects and delineates fluid regions in the OCT data, differentiating between intra- and sub-retinal fluid (IRF, SRF), as well as fluid resulting from in serous pigment epithelial detachments (PED). Standard metrics for fluid detection and quantification were used to evaluate performance.ResultsThe per slice receiver operating characteristic (ROC) area under the curves (AUCs) for each of these fluid types were 0.90, 0.94 and 0.94 for IRF, SRF and PED, respectively. Per volume results were 0.94 and 0.88 for IRF and PED (SRF being present in all cases). The correlation of fluid volume between the expert graders and the algorithm were 0.99 for IRF, 0.99 for SRF and 0.82 for PED.ConclusionsAutomated, deep learning-based segmentation is able to accurately detect and quantify different macular fluid types in SS-OCT data on par with expert graders.

show abstract

Section: Introductionmentioning

confidence: 56%

Automated Deep Learning-based Multi-class Fluid Segmentation in Swept-Source Optical Coherence Tomography Images

Oakley¹,

Sodhi

Russakoff³

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Deep learning methods for segmentation of the total retinal segmentation, 7 retinal fluids like IRC, SRF 8 or PED 9 have been proposed. Furthermore, the RETOUCH challenge was recently organized to measure the performance of state-of-the-art methods for the detection and segmentation of retinal fluids in OCT. 10 In this work, for the first time, the total retina volume as well as PED is segmented in SELF-OCT image data using a deep learning-based approach. The segmentation approach builds on our preliminary work 11,12 and consists of a CNN that segments the total retina as well as PEDs in three-dimensional scans of the SELF-OCT and is based on the popular U-Net architecture.…”

Section: Purposementioning

confidence: 99%

Segmentation of retinal low-cost optical coherence tomography images using deep learning

Kepp

Sudkamp

Burchard

et al. 2020

Medical Imaging 2020: Computer-Aided Diagnosis

View full text Add to dashboard Cite

The treatment of age-related macular degeneration (AMD) requires continuous eye exams using optical coherence tomography (OCT). The need for treatment is determined by the presence or change of disease-specific OCTbased biomarkers. Therefore, the monitoring frequency has a significant influence on the success of AMD therapy. However, the monitoring frequency of current treatment schemes is not individually adapted to the patient and therefore often insufficient. While a higher monitoring frequency would have a positive effect on the success of treatment, in practice it can only be achieved with a home monitoring solution. One of the key requirements of a home monitoring OCT system is a computer-aided diagnosis to automatically detect and quantify pathological changes using specific OCT-based biomarkers. In this paper, for the first time, retinal scans of a novel self-examination low-cost full-field OCT (SELF-OCT) are segmented using a deep learningbased approach. A convolutional neural network (CNN) is utilized to segment the total retina as well as pigment epithelial detachments (PED). It is shown that the CNN-based approach can segment the retina with high accuracy, whereas the segmentation of the PED proves to be challenging. In addition, a convolutional denoising autoencoder (CDAE) refines the CNN prediction, which has previously learned retinal shape information. It is shown that the CDAE refinement can correct segmentation errors caused by artifacts in the OCT image.

show abstract

“…Through a "fair and meaningful comparison of algorithms" (the "head-to-head" aspect of challenges), the community may understand which methods are most (and least) suited to address a specific task. [12][13][14] It should be noted that this benefit is achieved only if publications subsequent to the challenge clearly report relevant details of the top-performing (and least-performing) methods. 15 Furthermore, conclusions drawn from the relative performance of participating algorithms are only reliable to the extent that the individual algorithms were designed, implemented, and executed in a manner compatible with the challenge dataset.…”

Section: Rebuttal: Samuel G Armato Phdmentioning

confidence: 99%

Biomedical image analysis challenges should be considered as an academic exercise, not an instrument that will move the field forward in a real, practical way

2020

View full text Add to dashboard Cite

RETOUCH: The Retinal OCT Fluid Detection and Segmentation Benchmark and Challenge

Cited by 182 publications

References 61 publications

Automated Deep Learning-based Multi-class Fluid Segmentation in Swept-Source Optical Coherence Tomography Images

Automated Deep Learning-based Multi-class Fluid Segmentation in Swept-Source Optical Coherence Tomography Images

Segmentation of retinal low-cost optical coherence tomography images using deep learning

Biomedical image analysis challenges should be considered as an academic exercise, not an instrument that will move the field forward in a real, practical way

Contact Info

Product

Resources

About