Weakly Supervised Medical Diagnosis and Localization from Multiple Resolutions

Abstract

Diagnostic imaging often requires the simultaneous identification of a multitude of findings of varied size and appearance. Beyond global indication of said findings, the prediction and display of localization information improves trust in and understanding of results when augmenting clinical workflow. Medical training data rarely includes more than global image-level labels as segmentations are time-consuming and expensive to collect. We introduce an approach to managing these practical constraints by applying a novel architecture which learns at multiple resolutions while generating saliency maps with weak supervision. Further, we parameterize the Log-Sum-Exp pooling function with a learnable lower-bounded adaptation (LSE-LBA) to build in a sharpness prior and better handle localizing abnormalities of different sizes using only image-level labels. Applying this approach to interpreting chest x-rays, we set the state of the art on 9 abnormalities in the NIH's CXR14 dataset while generating saliency maps with the highest resolution to date.

submitted to ECCV 2018

References in corpus (15)

• Adam: A Method for Stochastic Optimization
• Deep Residual Learning for Image Recognition
• OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks
• CheXNet: Radiologist-Level Pneumonia Detection on Chest X-Rays with Deep Learning
• Object Detectors Emerge in Deep Scene CNNs
• Attention-based Deep Multiple Instance Learning
• Classifying and Segmenting Microscopy Images Using Convolutional Multiple Instance Learning
• Fully Convolutional Multi-Class Multiple Instance Learning
• Diagnose like a Radiologist: Attention Guided Convolutional Neural Network for Thorax Disease Classification
• Wider or Deeper: Revisiting the ResNet Model for Visual Recognition
• Learning Deep Features for Discriminative Localization
• Learning to diagnose from scratch by exploiting dependencies among labels
• Classification and Disease Localization in Histopathology Using Only Global Labels: A Weakly-Supervised Approach
• The One Hundred Layers Tiramisu: Fully Convolutional DenseNets for Semantic Segmentation
• Learning to recognize Abnormalities in Chest X-Rays with Location-Aware Dense Networks

Cited by in corpus (22)

• Learning Hierarchical Attention for Weakly-supervised Chest X-Ray Abnormality Localization and Diagnosis
• Weakly Supervised Deep Learning for Thoracic Disease Classification and Localization on Chest X-rays
• ChestNet: A Deep Neural Network for Classification of Thoracic Diseases on Chest Radiography
• Machine-Learning-Based Multiple Abnormality Prediction with Large-Scale Chest Computed Tomography Volumes
• GraphXCOVID: Explainable Deep Graph Diffusion Pseudo-Labelling for Identifying COVID-19 on Chest X-rays
• Learning Inductive Attention Guidance for Partially Supervised Pancreatic Ductal Adenocarcinoma Prediction
• Multi-task Learning for Chest X-ray Abnormality Classification on Noisy Labels
• ThoraX-PriorNet: A Novel Attention-Based Architecture Using Anatomical Prior Probability Maps for Thoracic Disease Classification
• An interpretable classifier for high-resolution breast cancer screening images utilizing weakly supervised localization
• Energy Models for Better Pseudo-Labels: Improving Semi-Supervised Classification with the 1-Laplacian Graph Energy
• Automatic Diagnosis of Pneumothorax from Chest Radiographs: A Systematic Literature Review
• A New Window Loss Function for Bone Fracture Detection and Localization in X-ray Images with Point-based Annotation
• Globally-Aware Multiple Instance Classifier for Breast Cancer Screening
• Dynamic Routing on Deep Neural Network for Thoracic Disease Classification and Sensitive Area Localization
• Many-to-One Distribution Learning and K-Nearest Neighbor Smoothing for Thoracic Disease Identification
• Weighing Features of Lung and Heart Regions for Thoracic Disease Classification
• High resolution weakly supervised localization architectures for medical images
• Learning to estimate label uncertainty for automatic radiology report parsing
• Weakly supervised training of pixel resolution segmentation models on whole slide images
• Reciprocal Landmark Detection and Tracking with Extremely Few Annotations
• Learning Visual Context by Comparison
• Self-Guided Multiple Instance Learning for Weakly Supervised Disease Classification and Localization in Chest Radiographs