scholarly journals Automated Summarisation of SDOCT Volumes using Deep Learning: Transfer Learning vs de novo Trained Networks

2018 ◽  
Author(s):  
Bhavna J. Antony ◽  
Stefan Maetschke ◽  
Rahil Garnavi
Keyword(s):  
Deep Learning ◽  
Transfer Learning ◽  
De Novo ◽  
Imaging Modality ◽  
Age Related Macular Degeneration ◽  
Learning Approaches ◽  
Non Invasive ◽  
Age Related ◽  
Volumetric Images ◽  
Small Set

AbstractSpectral-domain optical coherence tomography (SDOCT) is a non-invasive imaging modality that generates high-resolution volumetric images. This modality finds widespread usage in ophthalmology for the diagnosis and management of various ocular conditions. The volumes generated can contain 200 or more B-scans. Manual inspection of such large quantity of scans is time consuming and error prone in most clinical settings. Here, we present a method for the generation of visual summaries of SDOCT volumes, wherein a small set of B-scans that highlight the most clinically relevant features in a volume are extracted. The method was trained and evaluated on data acquired from age-related macular degeneration patients, and “relevance” was defined as the presence of visibly discernible structural abnormalities. The summarisation system consists of a detection module, where relevant B-scans are extracted from the volume, and a set of rules that determines which B-scans are included in the visual summary. Two deep learning approaches are presented and compared for the classification of B-scans - transfer learning and de novo learning. Both approaches performed comparably with AUCs of 0.97 and 0.96, respectively, obtained on an independent test set. The de novo network, however, was 98% smaller than the transfer learning approach, and had a run-time that was also significantly shorter.

scholarly journals Deep learning is effective for the classification of OCT images of normal versus Age-related Macular Degeneration

2016 ◽  
Author(s):  
Cecilia S Lee ◽  
Doug M Baughman ◽  
Aaron Y Lee
Keyword(s):  
Deep Learning ◽  
Macular Degeneration ◽  
Imaging Modality ◽  
Age Related Macular Degeneration ◽  
Automated Image Analysis ◽  
Clinical Endpoints ◽  
Patient Level ◽  
Age Related ◽  
Peak Sensitivity ◽  
Automated Screening

Objective: The advent of Electronic Medical Records (EMR) with large electronic imaging databases along with advances in deep neural networks with machine learning has provided a unique opportunity to achieve milestones in automated image analysis. Optical coherence tomography (OCT) is the most commonly obtained imaging modality in ophthalmology and represents a dense and rich dataset when combined with labels derived from the EMR. We sought to determine if deep learning could be utilized to distinguish normal OCT images from images from patients with Age-related Macular Degeneration (AMD). Design: EMR and OCT database study Subjects: Normal and AMD patients who had a macular OCT. Methods: Automated extraction of an OCT imaging database was performed and linked to clinical endpoints from the EMR. OCT macula scans were obtained by Heidelberg Spectralis, and each OCT scan was linked to EMR clinical endpoints extracted from EPIC. The central 11 images were selected from each OCT scan of two cohorts of patients: normal and AMD. Cross-validation was performed using a random subset of patients. Receiver operator curves (ROC) were constructed at an independent image level, macular OCT level, and patient level. Main outcome measure: Area under the ROC. Results: Of a recent extraction of 2.6 million OCT images linked to clinical datapoints from the EMR, 52,690 normal macular OCT images and 48,312 AMD macular OCT images were selected. A deep neural network was trained to categorize images as either normal or AMD. At the image level, we achieved an area under the ROC of 92.78% with an accuracy of 87.63%. At the macula level, we achieved an area under the ROC of 93.83% with an accuracy of 88.98%. At a patient level, we achieved an area under the ROC of 97.45% with an accuracy of 93.45%. Peak sensitivity and specificity with optimal cutoffs were 92.64% and 93.69% respectively. Conclusions: Deep learning techniques achieve high accuracy and is effective as a new image classification technique. These findings have important implications in utilizing OCT in automated screening and the development of computer aided diagnosis tools in the future.

scholarly journals Deep-learning based retinal fluid segmentation in optical coherence tomography images using a cascade of ENets

2021 ◽  
Author(s):  
Sophie Loizillon ◽  
Cédric Meurée ◽  
Camille Breuil ◽  
Timothée Faucon ◽  
Arnaud Lambert
Keyword(s):  
Optical Coherence Tomography ◽  
Imaging Modality ◽  
Region Of Interest ◽  
Random Forest Classifier ◽  
Age Related Macular Degeneration ◽  
Optical Coherence ◽  
Deep Convolutional Neural Networks ◽  
Non Invasive ◽  
Age Related ◽  
Inner Limiting Membrane

Optical coherence tomography (OCT) is a non-invasive, painless and reproducible examination which allows ophthalmologists to visualize retinal layers. This imaging modality is useful to detect diseases such as diabetic macular edema (DME) or age related macular degeneration (AMD), which are associated with fluid accumulations. In this paper, a cascade of deep convolutional neural networks is proposed using ENets for the segmentation of fluid accumulations in OCT B-Scans. After denoising the B-Scans, a first ENet extracts the region of interest (ROI) between the inner limiting membrane (ILM) and the Bruch's membrane (BM), whereas the second ENet segments the fluid in the ROI. A random forest classifier was applied on the segmented fluid regions to reject false positive. Our framework was trained on three different datasets with several diseases such as diabetic retinopathy (DR) and AMD. Our method achieves an average Dice Score for fluid segmentation of 0.80, 0.83 and 0.83 on the UMN DME, UMN AMD and Kermany datasets respectively.

scholarly journals Optical Coherence Tomography Angiography

2018 ◽  
Vol 11 (2) ◽  
pp. 91
Author(s):  
Manish Nagpal ◽  
Gujarat India ◽  
◽  
Keyword(s):  
Optical Coherence Tomography ◽  
Optical Coherence Tomography Angiography ◽  
Imaging Modality ◽  
Venous Occlusion ◽  
Age Related Macular Degeneration ◽  
Routine Practice ◽  
Optical Coherence ◽  
Non Invasive ◽  
Age Related ◽  
Clinical Scenarios

Optical coherence tomography angiography (OCTA) is a new revolutionary non-invasive imaging modality, built on the platform of optical coherence tomography (OCT). This technique works on the principle of ‘decorrelation’ and is still a nascent technology in its infancy with tremendous potential applicability for diagnosing retinal and choroidal vascular diseases. Its non-invasive nature, and the ability to generate images of retinal and choroidal vasculature, allows it to replace and/or supplement the current angiographic gold standards, fluorescein angiography (FA) and indocyanine green angiography (ICGA), if not in all but certainly in most retinal and choroidal pathologies. Still, there exists a major challenge in terms of its wide-scale availability, equipment and processing techniques, presence of artifacts, limitations of imaging capability, and lack of common vocabulary among retinal specialists for interpretation. In this review we intend to describe this novel technique by highlighting its key features, and comparing it with FA and ICGA. We will also discuss its applicability in various clinical scenarios such as diabetic retinopathy, age-related macular degeneration, retinal venous occlusion, choroiditis, and in routine practice. Further studies are needed to more definitively determine OCTA’s utility in the clinical setting and to establish if this technology may offer a non-invasive option of visualizing the retinal vasculature in detail.

scholarly journals Automatic detection of 39 fundus diseases and conditions in retinal photographs using deep neural networks

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Ling-Ping Cen ◽  
Jie Ji ◽  
Jian-Wei Lin ◽  
Si-Tong Ju ◽  
Hong-Jie Lin ◽  
...  
Keyword(s):  
Deep Learning ◽  
High Efficiency ◽  
Characteristic Curve ◽  
Weighted Average ◽  
Age Related Macular Degeneration ◽  
Learning Platform ◽  
Age Related ◽  
Public Data ◽  
Primary Test ◽  
Retinal Fundus

AbstractRetinal fundus diseases can lead to irreversible visual impairment without timely diagnoses and appropriate treatments. Single disease-based deep learning algorithms had been developed for the detection of diabetic retinopathy, age-related macular degeneration, and glaucoma. Here, we developed a deep learning platform (DLP) capable of detecting multiple common referable fundus diseases and conditions (39 classes) by using 249,620 fundus images marked with 275,543 labels from heterogenous sources. Our DLP achieved a frequency-weighted average F1 score of 0.923, sensitivity of 0.978, specificity of 0.996 and area under the receiver operating characteristic curve (AUC) of 0.9984 for multi-label classification in the primary test dataset and reached the average level of retina specialists. External multihospital test, public data test and tele-reading application also showed high efficiency for multiple retinal diseases and conditions detection. These results indicate that our DLP can be applied for retinal fundus disease triage, especially in remote areas around the world.

Updates in deep learning research in ophthalmology

2021 ◽  
Vol 135 (20) ◽  
pp. 2357-2376
Author(s):  
Wei Yan Ng ◽  
Shihao Zhang ◽  
Zhaoran Wang ◽  
Charles Jit Teng Ong ◽  
Dinesh V. Gunasekeran ◽  
...  
Keyword(s):  
Deep Learning ◽  
Critical Role ◽  
Disease Classification ◽  
Age Related Macular Degeneration ◽  
Security And Privacy ◽  
Generative Adversarial Networks ◽  
End User ◽  
Ethical Assessment ◽  
Age Related ◽  
Patient Consent

Abstract Ophthalmology has been one of the early adopters of artificial intelligence (AI) within the medical field. Deep learning (DL), in particular, has garnered significant attention due to the availability of large amounts of data and digitized ocular images. Currently, AI in Ophthalmology is mainly focused on improving disease classification and supporting decision-making when treating ophthalmic diseases such as diabetic retinopathy, age-related macular degeneration (AMD), glaucoma and retinopathy of prematurity (ROP). However, most of the DL systems (DLSs) developed thus far remain in the research stage and only a handful are able to achieve clinical translation. This phenomenon is due to a combination of factors including concerns over security and privacy, poor generalizability, trust and explainability issues, unfavorable end-user perceptions and uncertain economic value. Overcoming this challenge would require a combination approach. Firstly, emerging techniques such as federated learning (FL), generative adversarial networks (GANs), autonomous AI and blockchain will be playing an increasingly critical role to enhance privacy, collaboration and DLS performance. Next, compliance to reporting and regulatory guidelines, such as CONSORT-AI and STARD-AI, will be required to in order to improve transparency, minimize abuse and ensure reproducibility. Thirdly, frameworks will be required to obtain patient consent, perform ethical assessment and evaluate end-user perception. Lastly, proper health economic assessment (HEA) must be performed to provide financial visibility during the early phases of DLS development. This is necessary to manage resources prudently and guide the development of DLS.

scholarly journals Improving Efficiency in Separating Blood Vessels from Retinal Images with Deep Learning Techniques

2019 ◽  
Vol 8 (2S11) ◽  
pp. 3637-3640
Keyword(s):  
Diabetic Retinopathy ◽  
Deep Learning ◽  
Macular Degeneration ◽  
Blood Vessels ◽  
Age Related Macular Degeneration ◽  
Retinal Images ◽  
Retinal Vessels ◽  
Age Related ◽  
Learning Techniques ◽  
Retinal Pathologies

Retinal vessels ID means to isolate the distinctive retinal configuration issues, either wide or restricted from fundus picture foundation, for example, optic circle, macula, and unusual sores. Retinal vessels recognizable proof investigations are drawing in increasingly more consideration today because of pivotal data contained in structure which is helpful for the identification and analysis of an assortment of retinal pathologies included yet not restricted to: Diabetic Retinopathy (DR), glaucoma, hypertension, and Age-related Macular Degeneration (AMD). With the advancement of right around two decades, the inventive methodologies applying PC supported systems for portioning retinal vessels winding up increasingly significant and coming nearer. Various kinds of retinal vessels segmentation strategies discussed by using Deep Learning methods. At that point, the pre-processing activities and the best in class strategies for retinal vessels distinguishing proof are presented.

scholarly journals Autophagy Stimulus Promotes Early HuR Protein Activation and p62/SQSTM1 Protein Synthesis in ARPE-19 Cells by Triggering Erk1/2, p38MAPK, and JNK Kinase Pathways

2018 ◽  
Vol 2018 ◽  
pp. 1-15 ◽  
Author(s):  
Nicoletta Marchesi ◽  
Natthakan Thongon ◽  
Alessia Pascale ◽  
Alessandro Provenzani ◽  
Ali Koskela ◽  
...  
Keyword(s):  
Kinase Inhibitors ◽  
Rna Binding ◽  
De Novo ◽  
Disease Onset ◽  
Age Related Macular Degeneration ◽  
Protein Aggregate ◽  
Altered Expression ◽  
Protein Activation ◽  
Age Related ◽  
Early Effects

RNA-binding protein dysregulation and altered expression of proteins involved in the autophagy/proteasome pathway play a role in many neurodegenerative disease onset/progression, including age-related macular degeneration (AMD). HuR/ELAVL1 is a master regulator of gene expression in human physiopathology. In ARPE-19 cells exposed to the proteasomal inhibitor MG132, HuR positively affects at posttranscriptional level p62 expression, a stress response gene involved in protein aggregate clearance with a role in AMD. Here, we studied the early effects of the proautophagy AICAR + MG132 cotreatment on the HuR-p62 pathway. We treated ARPE-19 cells with Erk1/2, AMPK, p38MAPK, PKC, and JNK kinase inhibitors in the presence of AICAR + MG132 and evaluated HuR localization/phosphorylation and p62 expression. Two-hour AICAR + MG132 induces both HuR cytoplasmic translocation and threonine phosphorylation via the Erk1/2 pathway. In these conditions, p62 mRNA is loaded on polysomes and its translation in de novo protein is favored. Additionally, for the first time, we report that JNK can phosphorylate HuR, however, without modulating its localization. Our study supports HuR’s role as an upstream regulator of p62 expression in ARPE-19 cells, helps to understand better the early events in response to a proautophagy stimulus, and suggests that modulation of the autophagy-regulating kinases as potential therapeutic targets for AMD may be relevant.

scholarly journals Utility of Deep Learning Methods for Referability Classification of Age-Related Macular Degeneration

2018 ◽  
Vol 136 (11) ◽  
pp. 1305 ◽  
Author(s):  
Phillippe Burlina ◽  
Neil Joshi ◽  
Katia D. Pacheco ◽  
David E. Freund ◽  
Jun Kong ◽  
...  
Keyword(s):  
Deep Learning ◽  
Macular Degeneration ◽  
Age Related Macular Degeneration ◽  
Learning Methods ◽  
Age Related

Assessment of CFH and HTRA1 polymorphisms in age-related macular degeneration using classic and machine-learning approaches

2020 ◽  
Vol 41 (6) ◽  
pp. 539-547
Author(s):  
Antonieta Martínez-Velasco ◽  
Andric C. Perez-Ortiz ◽  
Bani Antonio-Aguirre ◽  
Lourdes Martínez-Villaseñor ◽  
Esmeralda Lira-Romero ◽  
...  
Keyword(s):  
Machine Learning ◽  
Macular Degeneration ◽  
Age Related Macular Degeneration ◽  
Learning Approaches ◽  
Age Related
Sign in / Sign up

Export Citation Format

Share Document