TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Papers/Improving Robustness using Joint Attention Network For Det...

Improving Robustness using Joint Attention Network For Detecting Retinal Degeneration From Optical Coherence Tomography Images

Sharif Amit Kamran, Alireza Tavakkoli, Stewart Lee Zuckerbrod

2020-05-16Retinal OCT Disease Classification
PaperPDFCode(official)

Abstract

Noisy data and the similarity in the ocular appearances caused by different ophthalmic pathologies pose significant challenges for an automated expert system to accurately detect retinal diseases. In addition, the lack of knowledge transferability and the need for unreasonably large datasets limit clinical application of current machine learning systems. To increase robustness, a better understanding of how the retinal subspace deformations lead to various levels of disease severity needs to be utilized for prioritizing disease-specific model details. In this paper we propose the use of disease-specific feature representation as a novel architecture comprised of two joint networks -- one for supervised encoding of disease model and the other for producing attention maps in an unsupervised manner to retain disease specific spatial information. Our experimental results on publicly available datasets show the proposed joint-network significantly improves the accuracy and robustness of state-of-the-art retinal disease classification networks on unseen datasets.

Results

TaskDatasetMetricValueModel
Disease PredictionOCT2017Acc95.6Joint-Attention-Network MobileNet-v2
Disease PredictionOCT2017Acc92.4Joint-Attention-Network ResNet50-v1
Disease PredictionOCT2017Acc77.4Joint-Attention-Network OpticNet-71
Disease PredictionSrinivasan2014Acc100Joint-Attention-Network ResNet50-v1
Disease PredictionSrinivasan2014Acc99.68Joint-Attention-Network OpticNet-71
Disease PredictionSrinivasan2014Acc99.36Joint-Attention-Network MobileNet-v2
Medical DiagnosisOCT2017Acc95.6Joint-Attention-Network MobileNet-v2
Medical DiagnosisOCT2017Acc92.4Joint-Attention-Network ResNet50-v1
Medical DiagnosisOCT2017Acc77.4Joint-Attention-Network OpticNet-71
Medical DiagnosisSrinivasan2014Acc100Joint-Attention-Network ResNet50-v1
Medical DiagnosisSrinivasan2014Acc99.68Joint-Attention-Network OpticNet-71
Medical DiagnosisSrinivasan2014Acc99.36Joint-Attention-Network MobileNet-v2

Related Papers

UniNet: A Contrastive Learning-guided Unified Framework with Feature Selection for Anomaly Detection2025-02-28ViT-2SPN: Vision Transformer-based Dual-Stream Self-Supervised Pretraining Networks for Retinal OCT Classification2025-01-28TINC: Temporally Informed Non-Contrastive Learning for Disease Progression Modeling in Retinal OCT Volumes2022-06-30Demystifying Deep Learning Models for Retinal OCT Disease Classification using Explainable AI2021-11-06Matching the Clinical Reality: Accurate OCT-Based Diagnosis From Few Labels2020-10-23Optic-Net: A Novel Convolutional Neural Network for Diagnosis of Retinal Diseases from Optical Tomography Images2019-10-13Retinal OCT disease classification with variational autoencoder regularization2019-03-23MobileNetV2: Inverted Residuals and Linear Bottlenecks2018-01-13