Iterative Augmentation of Visual Evidence for Weakly-Supervised Lesion Localization in Deep Interpretability Frameworks: Application to Color Fundus Images

  title={Iterative Augmentation of Visual Evidence for Weakly-Supervised Lesion Localization in Deep Interpretability Frameworks: Application to Color Fundus Images},
  author={Cristina Gonz{\'a}lez-Gonzalo and Bart Liefers and Bram van Ginneken and Clara I. S{\'a}nchez},
  journal={IEEE Transactions on Medical Imaging},
Interpretability of deep learning (DL) systems is gaining attention in medical imaging to increase experts’ trust in the obtained predictions and facilitate their integration in clinical settings. We propose a deep visualization method to generate interpretability of DL classification tasks in medical imaging by means of visual evidence augmentation. The proposed method iteratively unveils abnormalities based on the prediction of a classifier trained only with image-level labels. For each image… 
Deep Interpretable Classification and Weakly-Supervised Segmentation of Histology Images via Max-Min Uncertainty
High uncertainty is introduced as a criterion to localize non-discriminative regions that do not affect classifier decision, and is described with original Kullback-Leibler (KL) divergence losses evaluating the deviation of posterior predictions from the uniform distribution.
Applications of Artificial Intelligence for the Diagnosis, Prognosis, and Treatment of Age-related Macular Degeneration
The high prevalence of AMD, the effective treatments for a subset of the disease, and the need for better therapies for other subsets of the Disease make the application of artificial intelligence (AI) to AMD particularly important.
Deep Learning Approaches for Automatic Localization in Medical Images
This study presents a short review of DNN implementation for medical images and validates its efficacy on benchmarks, and discusses the challenges associated with the application of the DNN for medical image localization which can drive further studies in identifying potential future developments in the relevant field of study.
Progressive Multi-scale Consistent Network for Multi-class Fundus Lesion Segmentation
A progressive multi-scale consistent network (PMCNet) that integrates the proposed progressive feature fusion (PFF) block and dynamic attention block (DAB) to address the two issues of multi- scale and feature inconsistency in the multi-class segmentation of fundus lesions, which will produce better feature representation in the feature space.
Factibilidad de la utilización de la inteligencia artificial para el cribado de pacientes con COVID-19 en Paraguay
RESUMEN Objetivo. Estudiar la factibilidad de utilización de la inteligencia artificial como método sensible y específico para el cribado de COVID-19 en pacientes con afecciones respiratorias
A new detection model of microaneurysms based on improved FC-DenseNet
A novel model for the detection of MAs in fluorescein fundus angiography (FFA) images based on the improved FC-DenseNet, MAs-FC- DenseNet is proposed, which would be used for mass screening of DR patients.
Negative Evidence Matters in Interpretable Histology Image Classification
of using both positive and negative evidence, the first obtained from a classifier, and the other naturally available in datasets.
Grouping and Decoupling Mechanism for Diabetic Retinopathy Image Grading
This paper designs a novel deep convolutional neural network to classify DR, which also generates attention map indicating the location of lesions, and outperforms other methods on the Messidor dataset.


Improving weakly-supervised lesion localization with iterative saliency map refinement
A novel deep visualization framework for improving weakly-supervised lesion localization by applying an iterative approach where, in each step, the interpretation maps focus on different, less discriminative areas of the images, but still important for the final classification, reaching a more refined localization of abnormalities.
Weakly-supervised localization of diabetic retinopathy lesions in retinal fundus images
It is shown that on retina images different lesion areas fundamental for diabetic retinopathy are detected on an image level with high accuracy, comparable or exceeding supervised methods.
Visual Feature Attribution Using Wasserstein GANs
This paper develops a novel feature attribution technique based on Wasserstein Generative Adversarial Networks (WGAN), which performs substantially better than the state-of-the-art for visual attribution on a synthetic dataset and on real 3D neuroimaging data from patients with mild cognitive impairment and Alzheimer's disease.
Using a Deep Learning Algorithm and Integrated Gradients Explanation to Assist Grading for Diabetic Retinopathy.
Visualizing Deep Learning Models for the Detection of Referable Diabetic Retinopathy and Glaucoma
The findings suggest that this visualization method can highlight traditional regions in disease diagnosis, substantiating the validity of the deep learning models investigated and may promote the clinical adoption of these models.
Automated Grading of Age-Related Macular Degeneration From Color Fundus Images Using Deep Convolutional Neural Networks
It is demonstrated that automated algorithms could play a role that is independent of expert human graders in the current management of AMD and could address the costs of screening or monitoring, access to health care, and the assessment of novel treatments that address the development or progression of AMD.
Zoom-in-Net: Deep Mining Lesions for Diabetic Retinopathy Detection
A convolution neural network based algorithm for simultaneously diagnosing diabetic retinopathy and highlighting suspicious regions and a network termed Zoom-in-Net which mimics the zoom-in process of a clinician to examine the retinal images is proposed.
Deep image mining for diabetic retinopathy screening
Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs.
An algorithm based on deep machine learning had high sensitivity and specificity for detecting referable diabetic retinopathy and diabetic macular edema in retinal fundus photographs from adults with diabetes.