Deep Lesion Graphs in the Wild: Relationship Learning and Organization of Significant Radiology Image Findings in a Diverse Large-scale Lesion Database

Abstract

Radiologists in their daily work routinely find and annotate significant abnormalities on a large number of radiology images. Such abnormalities, or lesions, have collected over years and stored in hospitals' picture archiving and communication systems. However, they are basically unsorted and lack semantic annotations like type and location. In this paper, we aim to organize and explore them by learning a deep feature representation for each lesion. A large-scale and comprehensive dataset, DeepLesion, is introduced for this task. DeepLesion contains bounding boxes and size measurements of over 32K lesions. To model their similarity relationship, we leverage multiple supervision information including types, self-supervised location coordinates and sizes. They require little manual annotation effort but describe useful attributes of the lesions. Then, a triplet network is utilized to learn lesion embeddings with a sequential sampling strategy to depict their hierarchical similarity structure. Experiments show promising qualitative and quantitative results on lesion retrieval, clustering, and classification. The learned embeddings can be further employed to build a lesion graph for various clinically useful applications. We propose algorithms for intra-patient lesion matching and missing annotation mining. Experimental results validate their effectiveness.

Accepted by CVPR2018. DeepLesion url added

Cited by in corpus (10)

• Generative Adversarial Network in Medical Imaging: A Review
• Momentum Contrastive Learning for Few-Shot COVID-19 Diagnosis from Chest CT Images
• Reinventing 2D Convolutions for 3D Images
• DAN-Net: Dual-Domain Adaptive-Scaling Non-local Network for CT Metal Artifact Reduction
• Adaptive Convolutional Dictionary Network for CT Metal Artifact Reduction
• Detecting Outliers with Foreign Patch Interpolation
• Orientation-Shared Convolution Representation for CT Metal Artifact Learning
• Learning from Multiple Datasets with Heterogeneous and Partial Labels for Universal Lesion Detection in CT
• Accurate Weakly-Supervised Deep Lesion Segmentation using Large-Scale Clinical Annotations: Slice-Propagated 3D Mask Generation from 2D RECIST
• CT Image Enhancement Using Stacked Generative Adversarial Networks and Transfer Learning for Lesion Segmentation Improvement