Distributional Shifts in Automated Diabetic Retinopathy Screening

  title={Distributional Shifts in Automated Diabetic Retinopathy Screening},
  author={Jay Nandy and Wynne Hsu and Mong Li Lee},
Deep learning-based models are developed to automatically detect if a retina image is `referable' in diabetic retinopathy (DR) screening. However, their classification accuracy degrades as the input images distributionally shift from their training distribution. Further, even if the input is not a retina image, a standard DR classifier produces a high confident prediction that the image is `referable'. Our paper presents a Dirichlet Prior Network-based framework to address this issue. It… Expand

Figures and Tables from this paper


An Incremental Feature Extraction Framework for Referable Diabetic Retinopathy Detection
  • Jay Nandy, W. Hsu, M. Lee
  • Computer Science
  • 2016 IEEE 28th International Conference on Tools with Artificial Intelligence (ICTAI)
  • 2016
This paper first learns a universal Gaussian mixture model (GMM) from a small set of annotated images and applies this universal GMM as the prior belief to learn an adaptive GMM for individual images to capture the characteristics of referable versus non-referable images. Expand
An ensemble deep learning based approach for red lesion detection in fundus images
This paper proposes a novel method for red lesion detection based on combining both deep learned and domain knowledge that reported the highest performance on a per-lesion basis on DIARETDB1 and e-ophtha, and for screening and need for referral on MESSIDOR compared to a second human expert. Expand
Building Trust in Deep Learning System towards Automated Disease Detection
It is proposed to use uncertainty estimates of the deep learning system’s prediction to know when to accept or to disregard its prediction, and the effectiveness of using such estimates in a real-life application for the screening of diabetic retinopathy is evaluated. Expand
Deriving Probabilistic SVM Kernels From Flexible Statistical Mixture Models and its Application to Retinal Images Classification
The developed hybrid model is introduced in this paper as an effective SVM kernel able to incorporate prior knowledge about the nature of data involved in the problem at hand and, therefore, permits a good data discrimination. Expand
Development and Validation of a Deep Learning System for Diabetic Retinopathy and Related Eye Diseases Using Retinal Images From Multiethnic Populations With Diabetes
In this evaluation of retinal images from multiethnic cohorts of patients with diabetes, the DLS had high sensitivity and specificity for identifying diabetic retinopathy and related eye diseases. Expand
LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop
This work proposes to amplify human effort through a partially automated labeling scheme, leveraging deep learning with humans in the loop, and constructs a new image dataset, LSUN, which contains around one million labeled images for each of 10 scene categories and 20 object categories. Expand
Very Deep Convolutional Networks for Large-Scale Image Recognition
This work investigates the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting using an architecture with very small convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers. Expand
EyePACS: An Adaptable Telemedicine System for Diabetic Retinopathy Screening
The use of license-free Web-based software, standard interfaces, and flexible protocols has allowed primary care providers to adopt retinopathy screening with minimal effort and resources. Expand
Uncertainty in Deep Learning
This work develops tools to obtain practical uncertainty estimates in deep learning, casting recent deep learning tools as Bayesian models without changing either the models or the optimisation, and develops the theory for such tools. Expand
Deep Anomaly Detection with Outlier Exposure
In extensive experiments on natural language processing and small- and large-scale vision tasks, it is found that Outlier Exposure significantly improves detection performance and that cutting-edge generative models trained on CIFar-10 may assign higher likelihoods to SVHN images than to CIFAR-10 images; OE is used to mitigate this issue. Expand