Dataset Information

Prediction of age-related macular degeneration disease using a sequential deep learning approach on longitudinal SD-OCT imaging biomarkers.

ABSTRACT: We propose a hybrid sequential prediction model called "Deep Sequence", integrating radiomics-engineered imaging features, demographic, and visual factors, with a recursive neural network (RNN) model in the same platform to predict the risk of exudation within a future time-frame in non-exudative AMD eyes. The proposed model provides scores associated with risk of exudation in the short term (within 3 months) and long term (within 21 months), handling challenges related to variability of OCT scan characteristics and the size of the training cohort. We used a retrospective clinical trial dataset that includes 671 AMD fellow eyes with 13,954 observations before any signs of exudation for training and validation in a tenfold cross validation setting. Deep Sequence achieved high performance for the prediction of exudation within 3 months (0.96?±?0.02 AUCROC) and within 21 months (0.97?±?0.02 AUCROC) on cross-validation. Training the proposed model on this clinical trial dataset and testing it on an external real-world clinical dataset showed high performance for the prediction within 3-months (0.82 AUCROC) but a clear decrease in performance for the prediction within 21-months (0.68 AUCROC). While performance differences at longer time intervals may be derived from dataset differences, we believe that the high performance and generalizability achieved in short-term predictions may have a high clinical impact allowing for optimal patient follow-up, adding the possibility of more frequent, detailed screening and tailored treatments for those patients with imminent risk of exudation.

SUBMITTER: Banerjee I

PROVIDER: S-EPMC7508843 | biostudies-literature | 2020 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Prediction of age-related macular degeneration disease using a sequential deep learning approach on longitudinal SD-OCT imaging biomarkers.

Banerjee Imon I de Sisternes Luis L Hallak Joelle A JA Leng Theodore T Osborne Aaron A Rosenfeld Philip J PJ Gregori Giovanni G Durbin Mary M Rubin Daniel D

Scientific reports 20200922 1

We propose a hybrid sequential prediction model called "Deep Sequence", integrating radiomics-engineered imaging features, demographic, and visual factors, with a recursive neural network (RNN) model in the same platform to predict the risk of exudation within a future time-frame in non-exudative AMD eyes. The proposed model provides scores associated with risk of exudation in the short term (within 3 months) and long term (within 21 months), handling challenges related to variability of OCT sca ...[more]

PMID: 32963300

Similar Datasets

Project description:PurposeTo evaluate longitudinally the performance of the Notal Vision Home OCT (NVHO), comprising a spectral-domain OCT device for patient self-imaging at home, telemedicine infrastructure for automated data upload, and deep learning algorithm for automated OCT evaluation. The aims were to study the system's performance in daily image acquisition and automated analysis and to characterize the dynamics of retinal fluid exudation in neovascular age-related macular degeneration (nAMD).DesignPilot prospective, observational longitudinal study.ParticipantsFour individuals (mean age, 73.8 years) with nAMD (one or both eyes) undergoing anti-vascular endothelial growth factor therapy in routine clinical practice.MethodsThe participants performed daily self-imaging at home with the NVHO for 1 month. The macular cube scans were uploaded automatically to the Notal Health Cloud. They underwent evaluation separately by the Notal OCT Analyzer (NOA) and human expert graders for fluid presence, segmentation, and volume.Main outcome measuresDaily self-imaging completion, image quality, acquisition time, agreement between automated and human grading of retinal fluid, and temporal dynamics of fluid volume.ResultsOf 240 self-imaging attempts initiated, the number successfully completed was 211 (87.9%). Of these, 97.6% had satisfactory quality. For fluid presence, the NOA agreed with human grading in 94.7% of cases. From a subset of 24 scans with fluid, for agreement between NOA and human fluid volume measurements, the correlation coefficient was 0.996 and mean absolute difference was 1.5 nl (vs. 0.995 and 1.2 nl, respectively, for interhuman agreement). Graphic plots of fluid volume revealed wide variation in the dynamics of fluid exudation and treatment response.ConclusionsThe participants could perform daily self-imaging at home and generate macular cube scans of satisfactory quality. Automated quantitative OCT analysis achieved high agreement with human grading. Daily self-imaging with automated OCT analysis permitted detailed characterization of the dynamics of fluid exudation and revealed wide variation between eyes. Metrics describing these dynamics may become important disease biomarkers. Home OCT telemedicine systems represent an alternative paradigm of disease monitoring; they may allow highly personalized retreatment decisions, with fewer unnecessary injections and clinic visits.

Project description:PurposeTimely diagnosis of eye diseases is paramount to obtaining the best treatment outcomes. OCT and OCT angiography (OCTA) have several advantages that lend themselves to early detection of ocular pathology; furthermore, the techniques produce large, feature-rich data volumes. However, the full clinical potential of both OCT and OCTA is stymied when complex data acquired using the techniques must be manually processed. Here, we propose an automated diagnostic framework based on structural OCT and OCTA data volumes that could substantially support the clinical application of these technologies.DesignCross sectional study.ParticipantsFive hundred twenty-six OCT and OCTA volumes were scanned from the eyes of 91 healthy participants, 161 patients with diabetic retinopathy (DR), 95 patients with age-related macular degeneration (AMD), and 108 patients with glaucoma.MethodsThe diagnosis framework was constructed based on semisequential 3-dimensional (3D) convolutional neural networks. The trained framework classifies combined structural OCT and OCTA scans as normal, DR, AMD, or glaucoma. Fivefold cross-validation was performed, with 60% of the data reserved for training, 20% for validation, and 20% for testing. The training, validation, and test data sets were independent, with no shared patients. For scans diagnosed as DR, AMD, or glaucoma, 3D class activation maps were generated to highlight subregions that were considered important by the framework for automated diagnosis.Main outcome measuresThe area under the curve (AUC) of the receiver operating characteristic curve and quadratic-weighted kappa were used to quantify the diagnostic performance of the framework.ResultsFor the diagnosis of DR, the framework achieved an AUC of 0.95 ± 0.01. For the diagnosis of AMD, the framework achieved an AUC of 0.98 ± 0.01. For the diagnosis of glaucoma, the framework achieved an AUC of 0.91 ± 0.02.ConclusionsDeep learning frameworks can provide reliable, sensitive, interpretable, and fully automated diagnosis of eye diseases.Financial disclosuresProprietary or commercial disclosure may be found after the references.

Project description:ObjectiveAge-related macular degeneration (AMD) is a significant cause of blindness, initially characterized by the accumulation of sub-Retinal pigment epithelium (RPE) deposits, leading to progressive retinal degeneration and, eventually, irreversible vision loss. This study aimed to elucidate the differential expression of transcriptomic information in AMD and normal human RPE choroidal donor eyes and to investigate whether it could be used as a biomarker for AMD.MethodsRPE choroidal tissue samples (46 Normal samples, 38 AMD samples) were obtained from the GEO (GSE29801) database and screened for differentially expressed genes in normal and AMD patients using GEO2R and R to compare the degree of enrichment of differentially expressed genes in the GO, KEGG pathway. Firstly, we used machine learning models (LASSO, SVM algorithm) to screen disease signature genes and compare the differences between these signature genes in GSVA and immune cell infiltration. Secondly, we also performed a cluster analysis to classify AMD patients. We selected the best classification by weighted gene co-expression network analysis (WGCNA) to screen the key modules and modular genes with the strongest association with AMD. Based on the module genes, four machine models, RF, SVM, XGB, and GLM, were constructed to screen the predictive genes and further construct the AMD clinical prediction model. The accuracy of the column line graphs was evaluated using decision and calibration curves.ResultsFirstly, we identified 15 disease signature genes by lasso and SVM algorithms, which were associated with abnormal glucose metabolism and immune cell infiltration. Secondly, we identified 52 modular signature genes by WGCNA analysis. We found that SVM was the optimal machine learning model for AMD and constructed a clinical prediction model for AMD consisting of 5 predictive genes.ConclusionWe constructed a disease signature genome model and an AMD clinical prediction model by LASSO, WGCNA, and four machine models. The disease signature genes are of great reference significance for AMD etiology research. At the same time, the AMD clinical prediction model provides a reference for early clinical detection of AMD and even becomes a future census tool. In conclusion, our discovery of disease signature genes and AMD clinical prediction models may become promising new targets for the targeted treatment of AMD.

Project description:PurposeUltrahigh resolution spectral domain-OCT (UHR SD-OCT) enables in vivo visualization of micrometric structural markers which differentially associate with normal aging versus age-related macular degeneration (AMD). This study explores the hypothesis that UHR SD-OCT can detect and quantify sub-retinal pigment epithelium (RPE) deposits in early AMD, separating AMD pathology from normal aging.DesignProspective cross-sectional study.ParticipantsA total of 53 nonexudative (dry) AMD eyes from 39 patients, and 63 normal eyes from 39 subjects.MethodsClinical UHR SD-OCT scans were performed using a high-density protocol. Exemplary high-resolution histology and transmission electron microscopy images were obtained from archive donor eyes. Three trained readers evaluated and labeled outer retina morphological features, including the appearance of a hyporeflective split within the RPE-RPE basal lamina (RPE-BL)-Bruch's membrane (BrM) complex on UHR brightness (B)-scans. A semi-automatic segmentation algorithm measured the thickness of the RPE-BL-BrM split/hyporeflective band.Main outcome measuresQualitative description of outer retinal morphological changes on UHR SD-OCT B-scans; the proportion of the RPE-BL-BrM complex with visible split (%) and the thickness of the resulting hyporeflective band (μm).ResultsIn young normal eyes, UHR SD-OCT consistently revealed an RPE-BL-BrM split/hyporeflective band. Its visibility and thickness were less in eyes of advanced age. However, the split/hyporeflective band was again visible in early AMD eyes. Both qualitative reading and quantitative thickness measurements showed significantly elevated visibility and thickness of the RPE-BL-BrM split/hyporeflective in early AMD eyes compared to age-matched controls.ConclusionsOur imaging results strongly support the hypothesis that appearance of the RPE-BL-BrM split/hyporeflective band in older subjects is dominated by the BL deposit, an indicator of early AMD well known from histology. Ultrahigh resolution SD-OCT can be used to investigate physiological aging as well as early AMD pathology in clinical imaging studies. Developing quantifiable markers associated with disease pathogenesis and progression can facilitate drug discovery, as well as reduce clinical trial times.Financial disclosuresProprietary or commercial disclosure may be found after the references.

Dataset Information

Prediction of age-related macular degeneration disease using a sequential deep learning approach on longitudinal SD-OCT imaging biomarkers.

Publications

Prediction of age-related macular degeneration disease using a sequential deep learning approach on longitudinal SD-OCT imaging biomarkers.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets